• Author / Uploaded
• Divyesh Patel

• Categories
• Petróleo
• Gas natural
• Refinería de petróleo
• Rodamiento (Mecánico)
• Gas natural licuado

Citation preview

HARNESS THE POWER OF ADVANCED HRSG TECHNOLOGY

The industry leader in Heat Recovery Steam Generators for gas turbines up to 30 MW, RENTECH offers a full range of HRSG systems to meet your toughest project requirements. We custom engineer our crossflow two-drum and waterwall designs to perform superbly in the most demanding applications and operating conditions. We master every detail to deliver elemental power for clients worldwide. HARNESS THE POWER WITH RENTECH.

HEAT RECOVERY STEAM GENERATORS WASTE HEAT BOILERS FIRED PACKAGED WATERTUBE BOILERS SPECIALTY BOILERS

WWW.RENTECHBOILERS.COM

AUGUST 2016 | HydrocarbonProcessing.com

FLUID FLOW AND ROTATING EQUIPMENT

Analyze abnormal operations of an HDS reactor loop with dynamic simulation

PROCESS ENGINEERING Choose the most appropriate modeling approach for reactors Distillation technology—Then and now

MAINTENANCE AND RELIABILITY Design operations-and-maintenance-friendly pressure vessels

HEAT TRANSFER Sampling of heat transfer fluid offsets carbon effects on thermal plant efficiency

Grrrrr. Or as we call it: 784 mbar, 24.4 kg, and 107.3 % Level 6LHPHQV3URFHVV,QVWUXPHQWDWLRQ²PHDVXULQJHYHU\WKLQJWKDWPDWWHUV 3UHVVXUH:HLJKLQJ/HYHODQGPRUH 2XUUDQJHRISURFHVVLQVWUXPHQWVDUHSDFNHGZLWKVWDWHRI WKHDUWWHFKQRORJ\:KHWKHU\RXQHHGWRPHDVXUHSUHVVXUH WHPSHUDWXUHIORZZHLJKWOHYHORUXVHHOHFWURSQHXPDWLF SRVLWLRQHUVRXULQVWUXPHQWVDUHHQJLQHHUHGWRJLYH\RX WKHPRVWSUHFLVHPHDVXUHPHQWSRVVLEOH$QGEHFDXVHZH XQGHUVWDQGDFFXUDWHPHDVXULQJLVWKHNH\WRLQFUHDVLQJ

SODQWHIILFLHQF\DQGLPSURYLQJSURGXFWTXDOLW\ZHRIIHUIXOO\ DXWRPDWHGVROXWLRQVIRUHYHU\VWDJHRIWKHSURFHVV:KLFK PHDQV\RXFDQUHO\RQRXUSDVVLRQIRUSUHFLVLRQWKURXJKRXW \RXUHQWLUHSODQW%HFDXVHZHEHOLHYHHYHU\EXVLQHVVVXFFHVV EHJLQVZLWKJUHDWPHDVXULQJ

VLHPHQVFRPSURFHVVLQVWUXPHQWDWLRQ Select 75 at www.HydrocarbonProcessing.com/RS

AUGUST 2016 | Volume 95 Number 8 HydrocarbonProcessing.com

38

24

10 SPECIAL REPORT: FLUID FLOW AND ROTATING EQUIPMENT 25 Analyze abnormal operations of an HDS reactor loop with dynamic simulation O. Garcia, R. Shipman, C. Tong and R. E. Palmer

29

Extend ethylene plant run length with compressor chemical treatment J. M. Hancock and S. Rodrigues

33

Boost capacity of SRUs with mixing devices for oxygen enrichment M. Rajasekhar, V. D. Thakare, G. Srivardhan, V. K. Jayanti, D. K. R. Nambiar, S. R. Singh and V. Shukla

BONUS REPORT: LNG 39 Cylindrical hull concept improves design for offshore FLNG production L. Odeskaug and S. Mokhatab

PROCESS ENGINEERING AND OPTIMIZATION 45 Choose the most appropriate modeling approach for reactors A. A. Jain and A. Gupta

51

4

MAINTENANCE AND RELIABILITY 55 Design operations-and-maintenance-friendly pressure vessels—Part 1 G. Murti

PROCESS CONTROL AND INSTRUMENTATION 59 Utilize genetic programming to develop new point efficiency correlation N. Kasiri, P. Jouybanpour and M. Reza Ehsani

WATER MANAGEMENT 65 Experience with naphtha in sour water emulsions generated in a fractionator overhead accumulator C. McKnight and B. Rumball

ENVIRONMENT AND SAFETY 70 Understand the sources of oil pollution in water M. Yang

Consider post-design changes to confine a hazardous area S. V. Bapat

HEAT TRANSFER 77 Sample heat transfer fluids to offset carbon effects on thermal plant efficiency C. Wright

GAS PROCESSING SUPPLEMENT GP-1 Technology and Business Information for the Global Gas Processing Industry Cover Image: MAN Diesel & Turbo North America, headquartered in Houston, Texas, provides the full array of MAN diesel and gas engines, turbomachinery and after-sales service support. Photo courtesy of MAN Diesel & Turbo North America Inc.

Industry Perspectives

10

Business Trends

15

Industry Metrics

17

Global Project Data

81

Innovations

83

Marketplace

84

Advertiser Index

85

Events

86

People

COLUMNS 7

Publisher’s Letter

9

Editorial Comment

Distillation—Then and now J. C. Gentry, M. Bhargava and M. J. Binkley

73

DEPARTMENTS

The transition into a new era A historical snapshot of a complex global industry

19

Reliability

21

Viewpoint

Monitor electric motor vibration and optimize motor bearing lubricant application An engineer’s guide to networking

www.HydrocarbonProcessing.com

Industry Perspectives Petchem Tech Forum highlights advances in US petrochemicals In mid-July, Hydrocarbon Processing held its inaugural Petchem Tech Forum. The forum was a two-day event at which leading technical experts and professionals provided the latest advances in petrochemical technologies and techniques. It included sessions on topics such as process control and automation, maintenance and reliability, turnarounds/revamps, plant design, water treatment, analytics and risk management. These are extraordinary times for the petrochemical sector, especially in the US. The US petrochemical industry is in the midst of one of the largest industry expansions to ever occur in North America. Cheap, readily available shale gas has provided chemical producers in the US with low-cost feedstocks, fueling over $135 B in new petrochemical capacity. This investment includes a sharp increase in the construction of ethane cracking and derivatives capacity, ammonia and urea plants, and methanol production. By 2020, the US is forecast to start up nearly 10 MMtpy of new ethylene capacity, including new cracker projects, as well as capacity expansions. Companies such as Chevron Phillips Chemical, Dow, ExxonMobil, Ingleside Ethylene, Formosa Plastics, LyondellBasell and Sasol will be instrumental in adding over 8 MMtpy of new US ethylene capacity by the end of the decade (TABLE 1). This first wave of investment constitutes a total capital expenditure of nearly $20 B. Additional expansion projects are expected to add over 1 MMtpy of additional ethylene capacity by 2018. A second wave of new ethane crackers could add 8 MMtpy by the early 2020s. If built, total capital expenditures could top $50 B. According to the American Chemistry Council, gross exports of US chemical products will more than double from $60 B in 2014 to over $120 B by 2030. Billions of dollars will be invested in the construction of pipelines, storage terminals and export capacity through the end of the decade. Hydrocarbon Processing’s Construction Boxscore Database is tracking more than 100 active petrochemical projects in the US. At present, the US represents approximately 17% of global market share for active petrochemical projects. For more information and coverage on Hydrocarbon Processing’s inaugural Petchem Tech Forum, please visit HydrocarbonProcessing.com. TABLE 1. New ethane crackers under construction in the US

P. O. Box 2608 Houston, Texas 77252-2608, USA Phone: +1 (713) 529-4301 Fax: +1 (713) 520-4433 [email protected]

PUBLISHER

Catherine Watkins [email protected]

EDITOR/ASSOCIATE PUBLISHER

Lee Nichols [email protected]

EDITORIAL Executive Editor Managing Editor Technical Editor Digital Editor Reliability/Equipment Editor Contributing Editor Contributing Editor Contributing Editor

Adrienne Blume Mike Rhodes Bob Andrew Kyle Kornegay Heinz P. Bloch Alissa Leeton ARC Advisory Group Anthony Sofronas

MAGAZINE PRODUCTION / +1 (713) 525-4633 Vice President, Production Manager, Editorial Production Artist/Illustrator Senior Graphic Designer Manager, Advertising Production

Sheryl Stone Angela Bathe Dietrich David Weeks Amanda McLendon-Bass Cheryl Willis

ADVERTISING SALES See Sales Offices, page 84.

CIRCULATION / +1 (713) 520-4440 / [email protected] Manager, Circulation

Alice Murrell

SUBSCRIPTIONS Subscription price (includes both print and digital versions): Print—One year $239, two years $419, three years $539. Digital format—One year $239. Airmail rate outside North America $175 additional a year. Single copies $35, prepaid. Hydrocarbon Processing’s Full Data Access subscription plan is priced at $1,695. This plan provides full access to all information and data Hydrocarbon Processing has to offer. It includes a print or digital version of the magazine, as well as full access to all posted articles (current and archived), process handbooks, the HPI Market Data book, Construction Boxscore Database project updates and more. Because Hydrocarbon Processing is edited specifically to be of greatest value to people working in this specialized business, subscriptions are restricted to those engaged in the hydrocarbon processing industry, or service and supply company personnel connected thereto. Hydrocarbon Processing is indexed by Applied Science & Technology Index, by Chemical Abstracts and by Engineering Index Inc. Microfilm copies available through University Microfilms, International, Ann Arbor, Mich. The full text of Hydrocarbon Processing is also available in electronic versions of the Business Periodicals Index.

ARTICLE REPRINTS If you would like to have a recent article reprinted for an upcoming conference or for use as a marketing tool, contact Foster Printing Company for a price quote. Articles are reprinted on quality stock with advertisements removed; options are available for covers and turnaround times. Our minimum order is a quantity of 100.

For more information about article reprints, call Rhonda Brown with Foster Printing Company at +1 (866) 879-9144 ext. 194 or e-mail [email protected]. Hydrocarbon Processing (ISSN 0018-8190) is published monthly by Gulf Publishing Company, 2 Greenway Plaza, Suite 1020, Houston, Texas 77046. Periodicals postage paid at Houston, Texas, and at additional mailing office. POSTMASTER: Send address changes to Hydrocarbon Processing, P.O. Box 2608, Houston, Texas 77252. Copyright © 2016 by Gulf Publishing Company. All rights reserved. Permission is granted by the copyright owner to libraries and others registered with the Copyright Clearance Center (CCC) to photocopy any articles herein for the base fee of $3 per copy per page. Payment should be sent directly to the CCC, 21 Congress St., Salem, Mass. 01970. Copying for other than personal or internal reference use without express permission is prohibited. Requests for special permission or bulk orders should be addressed to the Editor. ISSN 0018-8190/01.

Company

Capacity

Startup

Chevron Phillips Chemical

1.5 MMtpy

3Q/4Q 2017

ExxonMobil

1.5 MMtpy

2017

Dow Chemical

1.5 MMtpy

2Q 2017

Sasol

1.5 MMtpy

3Q/4Q 2019

Formosa Plastics

1.5 MMtpy

1Q 2017

Publication Agreement Number 40034765

1Q 2017

Other Gulf Publishing Company titles include: Gas Processing, Petroleum Economist and World Oil.

Ingleside Ethylene

544 Mtpy

4 AUGUST 2016 | HydrocarbonProcessing.com

President/CEO CFO Vice President Vice President, Production

John Royall Pamela Harvey Ron Higgins Sheryl Stone Printed in USA

people powered

Through our work, we define opportunity. As an integrated energy and chemicals company, we are solving global energy challenges. Continued expansion in our refining and chemicals activities means we are looking for experienced downstream professionals to join our team. Process engineers, business development specialists, and others are contributing to world-class projects with global impact.

Apply now at www.aramco.jobs/hp Select 82 at www.HydrocarbonProcessing.com/RS

™

Advanced FIBER FILM Contactor Technology

Reduce Costs Improve Efficiencies

Hydrocarbon Treating Made Better • Improve Treating Rates • Reduce Carryover • Reduce Plot Space

Introducing the next generation of FIBER FILM® Contactor technology. FFC Plus™ is an advancement to FIBER FILM technology that makes mercaptan extraction more efficient than ever. The latest innovations to this commercially proven technology allow for more than double the treating capacity while reducing overall costs and plot space. FFC Plus can also be easily retrofitted into existing FIBER FILM units as a drop-in solution that is simple to install during a unit turnaround, requiring less time than a standard cleaning. In addition, turndown can be maintained at current levels. Add capacity for tomorrow without losing operability today.

Select 84 at www.HydrocarbonProcessing.com/RS

P. O. Box 2608, Houston, Texas 77252-2608, USA | Phone: +1 (713) 529-4301 | HydrocarbonProcessing.com

The transition into a new era Dear Reader, These are unprecedented times in the oil and gas industry. New technologies have produced gluts of both oil and natural gas, which have provided cheap feedstocks for the downstream processing industries. The cyclical nature of the oil and gas business has had vastly different effects on each region. To oil-exporting nations, reduced oil prices equate to low government revenues. In turn, little money is available to fund social, industrial or infrastructure projects. To other nations, low oil and natural gas prices have been a boon to the processing industries. They have seen a boost in the construction of additional downstream processing capacity, as well as cheap fuel prices for consumers, which spur consumption. As with all cyclical industries, sometimes change is a necessity to ensure the strength and viability of an organization. Change allows a business to build on its strengths and evolve into an even more exceptional enterprise. Hydrocarbon Processing is undertaking such a transition. As of August 1, I have assumed the role of Publisher for two of Gulf Publishing Company’s exceptional brands, Hydrocarbon Processing and Gas Processing. Most of my life has been spent in and around the oil and gas industry. I was born in Chicago, but was largely raised in the oil patches of Iran, the UAE and Texas. As a result of my upbringing, I was familiar with drill collars and Christmas trees (assemblies on surface and subsea wells) before I had heard of Barbie dolls. For the past 20 years, I have represented Gulf Publishing Company’s Hydrocarbon Processing, World Oil and Gas Processing publications in France, Spain, Germany, Switzerland, Austria, Belgium and the Middle East/North Africa regions. My contributions included display and digital advertising sales, organizing industry forecast presentations and event participation, and conducting marketing seminars for clients. During this time, I have found no other publication as highly regarded in the downstream industry as Hydrocarbon Processing. This is not arrogance or by accident; it is based on audited circulation numbers and years of industry professionals’ testimonials. Hydrocarbon Processing reached this position by applying the highest levels of editorial integrity. It is proof that, although consumer and industry publications are largely giving up on editorial standards, our industry still values and needs a trusted source of information. As we look into the future of our industry and publications, technical content is king. This will never change. We will continue to bring our readers the highest-quality technical and operating articles in the industry. It is what we have been doing since 1922, and will continue to do now and into the future. What has, is and will continue to change is how that information is disseminated. Hydrocarbon Processing wants to engage and listen to the industry to provide editorial and data content in the most useful medium and format possible. We have already begun this change with the unveiling of the new Hydrocarbon Processing website—but it does not stop there. Hydrocarbon Processing is exploring ideas such as podcasts, data products and innovative mobile apps. I want to thank you for your devotion to Hydrocarbon Processing. It is gratifying to hear how our publication, website and newsletters provide interesting and, more importantly, sound technical content to make your work and our industry a better place. I also want to thank all of the advertisers who support this publication. Please share your comments, ideas and news with us; we highly value your feedback. It is a dynamic time in the downstream industry, and I can guarantee that Hydrocarbon Processing will continue to be at the forefront of technology, trends and data intelligence.

Catherine Watkins Publisher, Hydrocarbon Processing and Gas Processing

Hydrocarbon Processing | AUGUST 2016 7

Select 69 at www.HydrocarbonProcessing.com/RS

Editorial Comment

LEE NICHOLS, EDITOR/ASSOCIATE PUBLISHER [email protected]

A historical snapshot of a complex global industry As we move into the second half of 2016, we take stock of new project activity in the global downstream processing industry. After all, a detailed trends analysis on new project construction can provide great insight into the strength or weakness of a country’s or region’s economy. FIG. 1 provides a macro-level analysis on new project construction from 2009 to the present. Since 2009, new project announcements have gradually declined. Besides a spike in 2014, which was a direct effect of the construction boom in the US gas processing/LNG and petrochemical industries, the overall trend shows new project announcements decreasing over the sample period. However, the drop in project announcements does not mean that the sky is falling. While the overall new project counts are lower, many nations are witnessing a boost in project construction. Factors such as population, economic, and social class growth; new technologies that provide better processing economics; and the move to diversify product portfolios are the primary drivers behind capital-intensive downstream infrastructure builds.

The high project numbers in 2009 were attributed to a spike in new refining and petrochemical construction projects in Asia, primarily China. During this time-frame, China was experiencing a refinery building boom. The Middle East was also investing heavily (and continues to do so) in new downstream processing capacity, a deliberate move to diversify product portfolios by increasing participation in the refined and petrochemical products market. Over the past few years, the US has been the leader in new downstream processing capacity, primarily through the construction of new gas processing plants, LNG export terminals and petrochemical complexes. As we look back on nearly a decade of project data, the events that lead to the ups and downs of an industry are quite extraordinary. This is just a snapshot of a complex global industry, and it is intended to provide a brief history of the industry’s past and clues as to where it might be headed. For a detailed breakdown and overview of downstream construction activity, please see this month’s Business Trends on pg. 14.

300 2009 2010 2011 2012 2013 2014 2015 2016

250

200

150

100

50 0 Africa

Asia-Pacific

Canada

Europe

Latin America

Middle East

US

FIG. 1. New project announcements by region, 2009–2016. Source: Hydrocarbon Processing’s Construction Boxscore Database.

INSIDE THIS ISSUE

10 Business Trends.

Hydrocarbon Processing explores the state of the downstream construction and investment sector. The review includes a detailed trends analysis on new and active construction projects around the world, a market share analysis and an overview of each region’s project activity and future outlook.

24 Special Report.

Any problems or failures in the fluid flow systems of an HPI facility will impact the entire plant’s operation and profitability. Significant effort is directed to the proper design, installation, operation and maintenance of fluid-handling facilities. The special report investigates the numerous issues around maximizing plant/process unit uptime and reliability, and eliminating leaks and emissions of process liquids and gases.

55 Reliability.

Useful ideas are presented to ensure that the design and selection of pressure vessels meet functional safety requirements while providing operation-and-maintenance-friendly service to end users for years to come.

65 Water Management.

Syncrude Canada’s attempt to modify an overhead accumulator drum to improve naphtha/sour water separation resulted in the formation of naphtha in sour water emulsion. The rapid response to alleviate this problem through the use of a demulsifier is detailed, along with the development of potential mechanical solutions through the application of computational fluid dynamics and cold flow modeling. Hydrocarbon Processing | AUGUST 2016 9

| Business Trends Capital-intensive investments are being made in every sector of the hydrocarbon processing industry, in every region. The following is an update on the global downstream investment and construction sector. This overview provides a detailed trends analysis on new and active downstream construction projects around the world. This information will help shed some light on the state of the downstream construction sector, as well as provide a detailed overview on each region’s project outlook. Photo: Fluor completed the lift of a 250-ft, 570-ton C2 splitter in late 2015. The column is part of Chevron Phillips Chemical’s ethane cracker project at its Cedar Bayou complex in Baytown, Texas. Photo courtesy of Fluor.

LEE NICHOLS, EDITOR/ASSOCIATE PUBLISHER [email protected]

Business Trends The state of the global downstream construction sector terminal final investment decisions, as LNG developers would be hesitant to invest in heavily capital-intensive terminal construction. The following is an update on the global downstream investment and construction sector. This overview provides a detailed trends analysis on new and active downstream construction projects around the world. The following information was developed using Hydrocarbon Processing’s Construction Boxscore Database. For detailed information on active downstream construction projects, as well as a resource for lead generation, market research, trends analysis and planning, please visit www.constructionboxscore.com. New projects. According to Hydrocarbon Processing’s Construction Boxscore Database, over 1,000 new projects have been announced since 2014. Nearly half of these projects were announced in 2014, and the trend shows a steady decline in new project announcements since that time. FIG. 1 shows a Boxscore Database trend analysis on new project announcements from 2014–2016. Each year runs from July 1 to the end of June. New project announcements have fallen from nearly 490 in 2014, to 320 in 2015, to just over 200 within the past year. This represents a yearly decrease of approximately 35%. Over the past year, regions such as Africa, Asia and Europe have gained new project market shares (FIG. 2). Although the US’ new project market share has decreased over the past year, the region still maintains the greatest total number of new project announcements since 2014. The US has announced more New downstream project announcements by region, 2014–2016

What is the state of downstream processing capacity? What regions and sectors are seeing growth, and which ones are stagnant, or possibly shrinking? These simple questions tend to not have simple answers. In short, capital-intensive investments are being made in every sector of the hydrocarbon processing industry, in every region. These investments are ensuring that global demand for petroleum products will be met in the future. The present and future of the downstream processing industry varies immensely by region. The following information will help shed some light on the state of the downstream construction sector, as well as provide a detailed overview on each region’s project outlook. Along with continued capacity growth in the global refining sector, new technologies are moving the industry toward cleaner, lower-sulfur fuels for transportation. Refiners are investing billions of dollars in new units, upgrades/retrofits and expansions to meet new sulfur and emissions regulations. These investments include sulfur-reduction programs such as Tier 3 in the US and Canada, National V in China and Bharat Stage 6 in India. Clean-fuels investments promote the reduction of carbon monoxide, nitrogen oxide, hydrocarbons and particulate matter in both diesel and gasoline vehicles. Refiners around the world are making necessary investments to produce high-quality fuels that meet Euro 4, Euro 5 and Euro 6 specifications. This will continue to be a major theme through the end of the decade. Over the past few years, the world has witnessed a surge in new petrochemical capacity announcements, led by the AsiaPacific, Middle East and US regions. These regions continue to build up petrochemical capacity to satisfy demand, increase product export market share and increase downstream product portfolios. The global petrochemical sector will continue to see growth through the rest of the decade. However, the petrochemical landscape varies significantly between regions. Many new construction projects remain active, but the outlook for the industry is not as bullish as it was 18 to 24 months ago. This sentiment is reflected in new project announcements over the past few years, which will be discussed later. The world is watching natural gas become the fastest-growing fossil fuel. Growth on both the supply and demand sides has resulted in the announcement of billions of dollars of capital investment. The increase in natural gas usage has caused a surge in LNG trade over the past few years. However, even with natural gas demand increasing around the world, LNG supply capacity is outpacing demand growth. This is leading to a glut of supplies, which should peak in the early 2020s, according to forecasts. The present LNG glut could jeopardize future LNG

180 2014 2015 2016

160 140 120 100 80 60 40 20 0

Africa

Asia-Pacific

Canada

Europe

Latin America Middle East

US

FIG. 1. Total new project announcements by region, 2014–2016. Source: Hydrocarbon Processing’s Construction Boxscore Database. Hydrocarbon Processing | AUGUST 2016 11

Business Trends than 280 projects in the past three years. The overwhelming majority of these projects are within the gas processing/LNG and petrochemical industries. In total, nearly 80% of all new US downstream projects have fallen within these two sectors. This activity includes the construction of new cryogenic and gas processing plants, NGL fractionators, hundreds of miles of new pipeline infrastructure and millions of tpy of new LNG export capacity. On the petrochemical front, the US has announced millions of tpy of new capacity growth in ethane cracking and derivative projects, methanol production and ammonia-urea capacity. The country has also announced a multitude of new refining projects to process lighter crude slates produced from US shale basins, as well as additional secondary units and upgrades to meet new US Tier 3 regulations, which will take effect in 2017. The Asia-Pacific region is a close second to the US and has announced 260 new projects in the past three years. This represents approximately 26% of the total number of new projects announced globally within that same time frame. Nearly half of all new projects announced in the region are located in China and India (FIG. 3). Just as China has seen unprecedented growth over the past decade, India is emerging as the globe’s new oil demand center. The country’s burgeoning demand is providing huge potential for downstream oil and gas growth.

15%, 16%, 17%

4%, 5%, 4% 28%, 32%, 21%

17%, 11%, 11% 3%, 6%, 7%

10%, 6%, 6%

23%, 24%, 34%

FIG. 2. Market share analysis of new project announcements, 2014–2016. Source: Hydrocarbon Processing’s Construction Boxscore Database. Other 4% Vietnam 4% Thailand 4% Taiwan 3% South Korea 5% Singapore 6% Philippines 1% Papua New Guinea 1% Myanmar 1% Malaysia 6% Japan 4% Indonesia 7% India 18% China 30% Bangladesh 1% Australia 5%

FIG. 3. Total new project market share in the Asia-Pacific region, 2014– 2016. Source: Hydrocarbon Processing’s Construction Boxscore Database.

12 AUGUST 2016 | HydrocarbonProcessing.com

India’s new project announcements have increased over 140% since 2014. Within the past year, India accounted for 30% of all new projects announced in the Asia-Pacific region. The Middle East has seen a substantial decrease in new project market share since its wave of new capacity announcements a few years ago. Middle Eastern nations rely heavily on oil export revenues. With the decrease in oil prices, the region’s oil exporting nations have taken a substantial hit in government revenues. This has delayed the implementation of certain downstream projects and resulted in multiple project cancellations. Regardless, the region still maintains a vast amount of capital-intensive projects. Saudi Arabia continues to be the leader in new project announcements in the region. Like most Middle Eastern countries, Saudi Arabia is making a deliberate move to increase its participation in the refined and petrochemicals product market. In doing so, the country is diversifying itself away from its reliance on oil export revenues. Like the Middle East, Latin American nations rely heavily on oil export revenues. The region has seen tremendous petroleum product demand growth over the past decade. To satisfy this demand, Latin America has relied heavily on refined project imports, mainly from the US. The region has announced major refining projects to help curb imports, but the drop in crude oil prices has left little money to fund capacity expansions. In the short term, Latin American nations would rather import refined fuels than invest in major expansions or grassroots facilities. This does not mean that the region is void of new project announcements. Some of the bright spots include new refining and petrochemical projects in Mexico and Peru, the growth in Bolivia’s petrochemical sector, as well as additional LNG regasification capacity in Colombia, Puerto Rico, Chile and Jamaica. In Europe, the majority of new project announcements are located in Eastern Europe, Russia and the Commonwealth of Independent States (CIS). These three areas of Europe represent nearly 70% of all new project announcements in the region since 2014. Major ongoing modernization projects in Russia, as well as expansions, upgrades and grassroots refinery and petrochemical projects in the CIS, have contributed to this activity. Canada’s new project market share has hovered around 4% to 5% for the past couple of years. New project announcements have centered on Canada’s desire to monetize excess natural gas supplies. Nearly 75% of recent project announcements in the region were in the gas processing/LNG sector. Historically, Canada has exported almost all of its excess natural gas to the US by pipeline. However, because of the shale gas boom, the US no longer needs to import significant volumes of natural gas from Canada. This has caused a considerable drop-off in Canadian exports to its main customer, with projects showing substantial decreases over the next 20 years. To offset this financial hit, Canada has announced a multitude of LNG export terminals. The majority of these projects are located on the country’s west coast in British Columbia. An additional halfdozen projects have also been announced on the country’s east coast in New Brunswick, Nova Scotia and Quebec. Finally, Africa has seen an incremental increase in new project market share over the past few years. This increase includes new refinery construction in Algeria and Uganda,

Business Trends Boxscore Database, over $60 B in capital projects have been placed on hold or abandoned within the past year. These projects include capital-intensive projects in every region of the globe. Some of the more notable project cancellations and holds include: • Appalachian Resins ethane cracker (Ohio)—$1.3 B • Ascend Performance Materials Chocolate Bayou PDH plant (Texas)—$1.2 B • Atyrau petrochemical complex (Kazakhstan)—$6.3 B • Barrancabermeja refinery modernization and expansion project (Colombia)—$3.4 B • BASF methanol-to-propylene plant (Texas)—$1.4 B • Binh Dinh refinery (Vietnam)—$22 B • Browse FLNG • CHS Spiritwood fertilizer plant (North Dakota)—$3 B • Douglas Channel LNG (Canada)—$600 MM • Downeast LNG (Maine)—$2 B • Marathon Petroleum’s Garyville refinery upgrade (Louisiana)—$2 B • Moin refinery (Costa Rica)—$1.2 B • SOCAR OGPC project (Azerbaijan)—$8.5 B • Triton LNG (Canada) • Valero’s St. Charles methanol project (Louisiana)— $700 MM. Not all of these projects have been delayed, put on hold or canceled due to low oil prices, although that has been the case with many project cancellations over the past few years. Some of these projects have been shelved due to government sanctions, the inability to secure financing or necessary feedstocks, capital expenditure, re-bids of engineering, procurement and construction quotes, public opposition and/or poor economics.

capacity expansions and grassroots refining and petrochemical facilities in Nigeria and Egypt, and new LNG capacity in countries such as Equatorial Guinea, Egypt, Ghana, Morocco, Tanzania and Mozambique. Active projects. At present, the Boxscore Database is tracking more than 2,100 projects around the world (FIG 4). Nearly 80% of total active projects are within the refining and petrochemical sectors. The Asia-Pacific region still dominates in total number of active projects in all downstream sectors, led by projects in China and India. These two countries represent over half of all active projects in the region. Combined, Europe and the US represent approximately 30% of total active downstream project market share. This activity includes new petrochemical and LNG capacity in the US, as well as additional refining and petrochemical builds in Eastern Europe, Russia and the CIS. A detailed breakdown of total active project market share by region is shown on pg. 14. Approximately 60% of active projects are in the preconstruction stage. A breakdown of active projects by activity level is listed below: • Engineering—18% • Front-end engineering design (FEED)—9% • Planning—26% • Feasibility study—7% • Under construction—40%. Abandons/holds. Although the Boxscore Database is tracking more than 2,100 active projects, market conditions have caused many projects to be moved from active status to delayed, put on hold or abandoned altogether. According to the

35 14

125 13

5

101 80

Canada 109

30

94

65

Europe US

171

75

25

72 34 34

104 66

Refining Petrochemical Gas processing/LNG Other

146

42

28

48

Middle East 31

Latin America

Africa

215

198 102

71

Asia-Pacific

FIG. 4. Total active projects by region and sector, July 2016. Source: Hydrocarbon Processing’s Construction Boxscore Database. Hydrocarbon Processing | AUGUST 2016 13

Business Trends

Breakdown of total active project market share by region, July 2016

Europe

CIS 19% Russia 31% Eastern Europe 17% Western Europe 33%

Source: Hydrocarbon Processing’s Construction Boxscore Database

Africa

Other 33% South Africa 8% Nigeria 21% Egypt 24% Angola 7% Algeria 7%

Asia-Pacific

Other 4% Australia 7% Vietnam 4% Thailand 2% Taiwan 2% South Korea 4% Singapore 5% Philippines 2% Pakistan 4% Malaysia 5% China 30% Japan 2% Indonesia 8% India 21%

Canada

14 AUGUST 2016 | HydrocarbonProcessing.com

Other 17% British Columbia 42% Alberta 41%

Latin America

Middle East

United States

Other 25% Venezuela 7% Peru 13% Mexico 23% Jamaica 7% Bolivia 8% Brazil 17%

Other 8% UAE 10% Turkey 9% Saudi Arabia 20% Oman 10% Kuwait 10% Iraq 17% Iran 16%

PADD 5— West Coast 8% PADD 4— Rocky Mountain 6% PADD 3— Gulf Coast 63% PADD 2—Midwest 15% PADD 1—East Coast 8%

MIKE RHODES, MANAGING EDITOR [email protected]

Industry Metrics

May-16

April-16

Mar.-16

Feb.-16

Jan.-16

Dec.-15

Nov.-15

May-16

April-16

Mar.-16

Feb.-16

Jan.-16

Dec.-15

Nov.-15

June-16

May-16

April-16

Mar.-16

Feb.-16

Jan.-16

Dec.-15

Nov.-15

Oct.-15

Sept.-15

Aug.-15

July-15

June-15

Cracking spread, US$/bbl Cracking spread, US$/bbl

Gasoil Fuel oil

Cracking spread, US$/bbl

June-16

May-16

April-16

Mar.-16

Feb.-16

Jan.-16

Dec.-15

Nov.-15

Oct.-15

Sept.-15

Aug.-15

30

Dubai Urals

20 10 Prem. gasoline Jet/kero

0

Gasoil Fuel oil

June-16

May-16

April-16

Mar.-16

Feb.-16

Jan.-16

Dec.-15

Nov.-15

Oct.-15

Sept.-15

Aug.-15

July-15

June-16

May-16

April-16

Mar.-16

Feb.-16

Jan.-16

Dec.-15

Nov.-15

Oct.-15

Sept.-15

Aug.-15

July-15

May-15

-10 -20

June-15

May-15

Prem. gasoline Jet/kero

Singapore cracking spread vs. Oman, 2015–2016*

Brent dated vs. sour grades (Urals and Dubai) spread, 2015–2016* Light sweet/medium sour crude spread, US$/bbl

0

-10 -20

Source: EIA Short-Term Energy Outlook, June 2016

8 6 4 2 0 -2 -4

10

July-15

2017-Q1

30 20

June-15

2016-Q1

40

May-15

2015-Q1

Stock change and balance, MMbpd

Supply and demand, MMbpd

6 5 4 3 2 1 0 -1 -2 -3

Forecast

Stock change and balance World supply World demand

Prem. gasoline Jet/kero Diesel Fuel oil

Rotterdam cracking spread vs. Brent, 2015–2016*

World liquid fuel supply and demand, MMbpd

2014-Q1

Oct.-15

60 50 40 30 20 10 0 -10 -20 May-15

Oil prices, $/bbl

120 110 100 90 80 70 60 W. Texas Inter. 50 Brent Blend 40 Dubai Fateh 30 Source: DOE 20 M J J A S O N D J F M A M J J A S O N D J F M A M 2014 2015 2016

2013-Q1

Sept.-15

US Gulf cracking spread vs. WTI, 2015–2016*

Selected world oil prices, $/bbl

2012-Q1

Japan Singapore

Oct.-15

60

Production equals US marketed production, wet gas. Source: EIA.

100 98 96 94 92 90 88 86 84 82 2011-Q1

US EU 16

70

Sept.-15

M J J A S O N D J F M A M J J A S O N D J F M A M 2014 2015 2016

80

June-15

20

2 1 0

Aug.-15

Monthly price (Henry Hub) 12-month price avg. Production

July-15

3

40

90

June-15

4

Utilization rates, %

60

100

May-15

5

Aug.-15

June-15

May-15

Global refining utilization rates, 2015–2016*

6 Gas prices, $/Mcf

Production, Bcfd

WTI, US Gulf Brent, Rotterdam Oman, Singapore

5

7

80

0

10

0

US gas production (Bcfd) and prices ($/Mcf) 100

15

July-15

An expanded version of Industry Metrics can be found online at HydrocarbonProcessing.com.

Global refining margins, 2015–2016* 20

Margins, US$/bbl

Asia refining margins fell despite strong regional demand. US gasoline demand continued to rise, and strong export opportunities have strengthened gasoil and fuel oil crack spreads, supporting refinery margins. European gasoline crack spread weakened under oversupply pressure, while the middle and bottom of the barrel recovered amid export opportunities.

* Material published permission of the OPEC Secretariat; copyright 2016; all rights reserved; OPEC Monthly Oil Market Report, July 2016. Hydrocarbon Processing | AUGUST 2016 15

MAXIMIZE RELIABILITY, AVAILABILITY & PROFITABILITY With Ariel API 618 Process Compressors

Select 56 at www.HydrocarbonProcessing.com/RS

LEE NICHOLS, EDITOR/ASSOCIATE PUBLISHER [email protected]

Global Project Data New downstream project announcements have continued to decrease since April. Although the trend in new project announcements is in decline, the downstream processing industry has announced more than 250 projects over the past year. The majority of these projects are located in the Asia-Pacific and US regions. These two regions account for nearly 65% of all new project

announcements since July 2015. In Asia, new project market share is dominated by China and India. This activity includes new capacity builds in the refining and petrochemical sectors, as well as new LNG import terminal construction. In the US, project developers are continuing to boost petrochemical capacity and construction of new LNG export infrastructure.

20

Canada

31 71

Europe 44

US

110

37

Middle East Africa

25

Number of projects by region

Asia-Pacific

Latin America

Number of projects by region, with total capital expenditure greater than $1 B 30 27 22 17

18

27

26 20

18

18

26% Planning

21 15

13

40% Under construction 12

June- July- Aug.- Sept.- Oct.- Nov.- Dec.- Jan.- Feb.- Mar.- April- May- June- July15 15 15 15 15 15 15 16 16 16 16 16 16 16

Boxscore new project announcements, June 2015–present

7% Feasibility study 9% FEED 18% Engineering Breakdown of active projects by activity level

Detailed and up-to-date information for active construction projects in the refining, gas processing and petrochemical industries across the globe | ConstructionBoxscore.com Hydrocarbon Processing | AUGUST 2016 17

ZymeFlow® tough. Refinery and petrochem process unit decontamination.

FASTER. SAFER. PREPARE PLANT EQUIPMENT FOR ENTRY IN

8-12 HOURS

Shorten your turnaround using ZymeFlow® Decon. With 25 years of experience in refining and petrochemical applications, the ZymeFlow® brand of specialty decontamination chemistries and services is the global leader for fast and safe vessel entry. The unique ZymeFlow process simultaneously eliminates benzene, LEL's, hydrogen sulfide, and pyrophoric iron sulfides, dramatically reducing the time spent to prepare vessels for entry and hot work. Furthermore, the process is safe for people, equipment, and wastewater plants. ZymeFlow Decon is active in over 50 countries, performing about 200 specialized decontamination projects each year.

Save Time. Increase Safety. Choose ZymeFlow®.

Tel: +1 281.443.0300 [email protected] | www.zymeflow.com ©2016. United Laboratories International, LLC. All Rights Reserved.

A United Laboratories Internationational Brand

Select 93 at www.HydrocarbonProcessing.com/RS

Reliability

HEINZ P. BLOCH, RELIABILITY/EQUIPMENT EDITOR [email protected]

Monitor electric motor vibration and optimize motor bearing lubricant application Reader feedback is very important to us. From the many questions that make their way to our desks, we can sense the pulse of industry trends, the level of training, professional competence and issues of concern. We usually send off a reply and later rework the more interesting correspondence into one of the columns you are reading now. Keep this in mind as you follow our discourse with a reliability professional employed by an owner-operator with a number of ammonia and urea plants in locations where blinding sandstorms are prone to occur. Commendations for an admirably tabulated case history. To begin with, this reliability professional—a.k.a.

“the reader”—deserves credit for making important observations on a double-ended electric motor. A high-pressure carbamate pump and a booster pump were connected to the motor’s shaft ends. • Each motor bearing housing had one vertical (y) probe and one axial (z) probe. Both are seismic probes that resolve acceleration into velocity. No horizontal (x) probe was utilized. • The motor tripped on high vibration at one of its bearings. Initially, only one vertical (y) probe reached the trip value, but the second one did not. After 30 seconds, both probes reached the trip value of 7.1 mm/ sec and the motor was shut down, exactly as intended. • All bearings are deep-groove style 6317, indicating an 85-mm bore. • The failed motor bearings showed bluish discoloration on shafts and bearing inner races, pointing to a lubrication issue. There was no trace of lubricant. • The original design intent was for these bearings to be lubricated by automatic grease-dispensing devices; however, no such automation was in place. The bearings were last (manually) lubricated in September 2014. No re-greasing was carried out until the bearings failed after about 10 months of operation. • After rebuilding the motor, the axial (z) probe was repositioned to the horizontal (x) location. Our reader inquired about API 670 4th Ed. (2000). This industry standard mentions dual-voting logic, which, the reader believes, is adopted by a majority of end users. He noted that the recently released API 670 5th Ed. (2014) recommends single-voting logic for radial vibrations. Based on in-house experience, his company favors either monitoring radial vibration excursions without trip logic or, more recently, two-out-of-two voting logic. The reader sought our advice on

the best voting logic for radial seismic acceleration/vibration monitoring of motors, and asked us to be mindful of everpresent concerns over plant operational availability and machine reliability priorities. Treat root causes, not symptoms. Our advice was experience-based. To “protect” this motor with one transducer per bearing housing is probably cost-justified because the plant already has all the associated electronic modules. However, the facility’s objection to using just one transducer was stated in follow-up correspondence, confirming concerns about spurious trips shutting down a highly profitable plant. Researching the probability of spurious trips in modern installations would be appropriate. The reader’s recollection of failing transducers may have to be updated. Alternatively, if those in authority demanded two seismic transducers and two-out-of-two voting logic, they might plan to install these in vertical (y) and horizontal (x) directions. Implementing twoout-of-two trip voting logic and installing the two probes in readily accessible locations should be considered. Both may be at convenient angles or located at the 12 o’clock and three o’clock positions. The first excursion should sound an alarm

Logic cards in BN

Logic cards in BN

Y or X Hi

And

Y or X HiHi

Y or X HiHi

And

Y or X Hi

Y

X

Y

Vib probes

X Vib probes

Rotor Drive end

Non-drive end

FIG. 1. A squashed orbit with the X-vibration probe shows much less amplitude than the y-probe, so it is possible that during a vibration excursion the y-probe could be in a trip state (HiHi) while the x-probe is indicating normal. Hydrocarbon Processing | AUGUST 2016 19

Reliability if one of the two readings exceeds 7 mm/sec. Automatic trip activation should be linked to both probes measuring an activity exceeding 7 mm/sec. A special caveat is illustrated in FIG. 1. The example shows a squashed orbit with the x-vibration probe showing much less amplitude than the z-probe. Therefore, it is possible that during a vibration excursion the y-probe could be in a trip state (HiHi) while the x-probe is indicating normal. Over the years, we have found that if one end of a rotor is in distress, the other end should show some change from normal. The probe might not be in a HiHi alarm state on the non-distressed end, but it should at least show a Hi alarm on one of its vibration probes (FIG. 1). It is a bit more costly since more cards must be installed in the monitoring rack, but it is generally worth it. Generalized vibration guidelines for pumps and electric motors. The reader may be well aware that the mo-

tor at issue here was driving two carbamate pumps. Because carbamate services are quite notorious for being installed in applications with inadequate separation of net positive suction head (NPSHa and NPSHr), it would be of interest to closely monitor pump vibration in this instance. NPSHa is a function of the system and must be calculated, whereas NPSHr is a function of the pump and must be provided by the pump manufacturer. Vibration velocity on pump bearing housings is measured, and conscientious operators are asked to take these readings once per shift. This compels operators to be in the field where they can use their senses of smell, hearing, vision and touch to determine deviations from normal. Certain types of motor distress are brought on by issues that are rooted in pump deficiencies. As shown online in TABLE 1, the allowable vibration velocity values are a function of pump style and size. Of course, many potential causes of excessive pump vibration exist,1 and the most prevalent include: 1. Rotor unbalance (new residual impeller/rotor unbalance or unbalance caused by impeller metal removal or wear) 2. Shaft (coupling) misalignment 3. Liquid turbulence due to operation too far below the pump best-efficiency flowrate (BEP) 4. Cavitation due to insufficient NPSH margin 5. Pressure pulsations from impeller vane-casing tongue (cutwater) interaction in high-discharge energy pumps. Once a pump has been determined to have a high “total/ all-pass” vibration level, the next step is to identify the cause. A filtered vibration analysis should be obtained, and the first step in such an analysis will be to capture, and then evaluate, the multiples (harmonics) of pump running speed. Evidence of outdated lubrication technology. This plant

would do well to make reliable bearings and lubrication one of its priority concerns. We recommend they adopt only bestavailable bearing selection and plant-wide automated lubrication strategies. The plant’s top technical and mid-level managers should appreciate why dry sump oil mist has been successfully used by Best-of-Class (BoC) companies for the past 40 years. Unless the bearings are lubricated by oil mist, BoCs disal20 AUGUST 2016 | HydrocarbonProcessing.com

low rolling element bearings for electric motors above 500 hp; Siemens allows oil mist in motors up to 3,000 kW (4,692 hp).2 For those insisting on grease, details on automatically or manually applied grease lubrication are important but will differ with the location and orientation of shields (if any) and drain ports. There is considerable reliability impact, depending on the type of grease. Moreover, certain grease application methods sometimes result in incorrect fill volume, excessive grease pressure (deflecting shields), rust or dust in bearing element paths, and bearing flat spots (in an installed spare pump set) due to shafts not being rotated, to name just a few. Again, proper greasing procedures and lubrication management are far more important than placing/mounting/maintaining more monitors on a rolling element-equipped motor bearing housing. An electric motor with 85-mm bearings is obviously not a small machine. It is maintenance-intensive and may require grease replenishment at least six, and in some cases 16, times per year. If rivet heads pop off in a riveted-cage bearing, the motor sometimes grinds to a halt in mere seconds. We refer the reader to an article describing how BoCs use oil mist on many electric motor bearings (Hydrocarbon Processing, March 1977—fully 39 years ago). An estimated 26,000 electric motors and 150,000 process pumps are presently using dry sump oil mist lubrication, and some of these have not needed bearing replacements in the past 35 years. Why the reader’s company is not availing itself of oil mist lubrication is very difficult to comprehend and not worth speculating. The one sure thing we know about achieving reliability is that it cannot be obtained with business-asusual mindsets. Allow us to zero in on the real problem: The reader is probably only responsible for vibration monitoring tasks. His assignment may be limited in scope and he cannot tell higher management that we believe his company is vulnerable in its use of old lubrication technology. Here is how others solved the dilemma: At least two companies accepted our recommendation to send four or five managers to a three-day offsite update session where experts (without allegiance to either vendors or bosses) candidly briefed them on how BoCs become BoCs. The value of teaching mid-level managers in small groups is far greater than trying to present in-plant seminars to 40 disinterested lower-rung folks. Their response, time and time again, has been, “I hear you, but that is how we do things around here, and I cannot do anything about it.” LITERATURE CITED Bloch, H. P., Petrochemical machinery insights, Elsevier Publishing, Oxford, UK, and Waltham, Massachusetts, 2016. 2 Bloch, H. P. and A. Shamim, Oil mist lubrication: Practical applications, Fairmont Publishing Co., Lilburn, Georgia, 1998. 1

HEINZ P. BLOCH resides in Westminster, Colorado. His professional career commenced in 1962 and included long-term assignments as Exxon Chemical’s regional machinery specialist for the US. He has authored over 650 publications, among them 19 comprehensive books. Mr. Bloch holds BS and MS degrees in mechanical engineering. He is an ASME life fellow and maintains registration as a professional engineer in New Jersey and Texas.

Viewpoint

GOUTAM SHAHANI, Shure-Line Construction, Kenton, Delaware; and CARL RENTSCHLER, Akron, Pennsylvania

An engineer’s guide to networking

GOUTAM SHAHANI is VP of sales and marketing at Shure-Line Construction and former business development manager at Linde Engineering North America. Mr. Shahani has over 30 years of industry experience and specializes in industrial gases for the energy, refining and chemical industries. He holds BS and MS degrees in chemical engineering, as well as an MBA.

CARL RENTSCHLER, P.E., is an engineering consultant specializing in project management, business development, client relationship management and procedure development. He has over 40 years of varied engineering and management experience in the energy and petrochemical industries. He holds a BS degree in civil engineering from Penn State University and a master of engineering degree in civil engineering from Cornell University.

Innumerable articles and training programs have been written on the subject of networking. Among the younger generation, social networking has become a primary means of communicating. So, the question may arise as to why another article on networking is needed. This column’s purpose is to discuss networking in the context of engineers working in the global chemical and refining industry. Typically, engineers tend to network only when they need something, such as a career change. Alternatively, the authors recommend networking as an ongoing activity. In today’s volatile business environment with corporate consolidation, corporate restructuring, plant closings and project delays, it is critical for engineers at every level to have a broadbased, active network of peers, mentors and coaches. The old adage of, “It is not what you know, but who you know,” has never been more true than in this dynamic industry atmosphere. Networking is important for many reasons. It can help engineers stay current with the latest developments, such as market trends, new technologies and engineering tools, and industry activities. Networking with peers to benchmark and learn new skills helps an engineer expand their horizons beyond their specialized work function. This helps people do their job better and is applicable to every function, including sales, marketing, procurement, project management, process engineering and construction management. In some disciplines, there is a greater supply of engineers than demand, and networking is one way to differentiate oneself in a competitive employment market. In other words, networking greatly helps position a person for the next career move consistent with industry trends. Creating a viable network. It is important to consider both the softer, conceptual issues and the hard mechanistic methods. Conceptual issues must be approached

with a sincere interest and concern to help others. The ability to listen and connect disparate people, facts and figures is very helpful. People do not forget a kind word or a helpful act, even after many years. It is indeed a small world in terms of people in the capital-intensive process industry. Networking must be a two-way street, and is not something that should be turned on and off only when something is needed. After working in the industry for several decades and watching it evolve, the authors stress the importance of having a strong foundation, in terms of intent, before launching into the nuts and bolts of developing a network. Some good mechanistic techniques are: • Join Linkedin, Toastmasters or other media sites • Connect with colleagues, ex-employees, college alumni, friends and neighbors • Attend conferences, make presentations and write articles • Participate in specialized trade associations and volunteer for committees. A good network includes people that are involved both internally and externally with an organization or company. A lot of value exists in developing relationships within a company. This is especially true in large multinational organizations. By having good contacts within your own and related functions within an organization, it is possible to learn, exchange information and collaborate. Information can flow horizontally and directly between people instead of needing to go up vertically through silos. This is faster and more effective for all parties concerned. Sustaining networking as an SOP. Maintaining and increasing the network becomes a challenge because everyone is busy. For networking to be effective, it is important to conduct it on a continuous basis. The business world is competitive, and many good performers are losing Hydrocarbon Processing | AUGUST 2016 21

Now in its next generation, this puppy’s pedigree is still best-of-breed.

Our next-generation Model 9900 gas analyzer. More sensitivity to sniff out emissions without H2O or CO2 interference. RM

Nothing tracks UV-absorbable gas species better than our new Model 9900RM analyzer. Even moisture and CO2 won’t throw it off. Whether part of your CEM system or a stand-alone gas monitor, its high intensity line source lamps deliver reliable dry gas analysis with a fixed wavelength, resolution to 0.02 nm and accuracy to 1% of full scale. The rugged Model 9900RM meets RATA testing requirements and can be rack, wall or bench mounted. Its advanced data processing capabilities include data logging, graphing, on-board diagnostics and more. It’s big on connectivity too, supporting Modbus/RS485 and Ethernet/Web interface. From performance to sensitivity to reliability, this puppy’s got it all. Learn more at: www.ametekpi.com

© 2016 by AMETEK. All rights reserved.

Select 73 at www.HydrocarbonProcessing.com/RS

Viewpoint their jobs. It is important to distinguish yourself among the other good performers. Networking is a way to stay current with technical advances and business trends. This can add valuable input to your company and makes you a potential candidate to other organizations. Networking should be standard operating procedure for engineers. Each week, time should be set aside to accomplish it, and goals should be set as to the level of networking to be achieved. This is a case where quality overshadows quantity. Focus on making contact with people and organizations aligned with your business and/or technical field. Remember, it is a two-way street, so always be prepared to offer information or assistance to others. Being altruistic is the watchword for effective networking. You will have little success if you are not willing to do your part. When carried out conscientiously, effective networking is a win for you, your colleagues and your company. One significant hurdle is that many technical people find it challenging to interface and network with others, and therefore do not attempt communications. These challenges include: • Technical people do not see communications skills as essential. Many engineers believe communication is something to be endured to get to the technical “meat.” • Technical people are not expected to be strong communicators. Often, technical people feel that they are counted on for their knowledge, and they rely on others to understand if their communication is subpar. This is not a positive approach to promote networking. • Communications are not considered part of an engineer’s makeup. Some technical people give up on communications because they feel it is not in their personality, or because they are intimidated by the extroverted and gregarious people around them. It is inaccurate to assume that introverts are poor communicators; in truth, introverts tend to be good listeners and have the capability to be strong communicators, if motivated. Several opportunities exist to find help in developing communication and networking skills. Located in nearly ev-

ery city, Toastmasters organizations offer an opportunity to enhance communication skills through peers. Organizations with the sole purpose of networking tend to be open, supportive and non-threatening. If you feel reticent about the idea of networking, seek out support from colleagues or an appropriate organization. A proactive approach. As the global

chemical and refinery industries be-

come increasingly more competitive, professional networking has never been more important. Make networking part of your professional life, just as you make a workout part of your daily regimen. It should become a standard operating procedure, as it is a win for all involved and increases your value to your employer. The payoff may not be immediate, but many positives will develop over time.

Crude oil fouling testing helps keep your heat exchangers running longer, cleaner, and more efficiently. Drawing on decades of experience, we provide companies throughout the world with contract fouling services that include • measuring fouling propensity and anticipated fouling characteristics Rotating • assessing the impact of additives Fouling Unit • estimating fouling thresholds at different temperatures and velocities • evaluating the impact of metallurgy • interpreting chemistry data • assessing primary fouling mechanisms HTRI researchers help design a test plan that best suits your goals. Once testing is complete, you receive a report containing a comprehensive summary of the tests, analysis of the results, mitigation strategies, and recommendations to guide your project.

High Temperature Fouling Unit

High Temperature Fouling Unit 2

HTRI engineers use three state-of-the-art research rigs to provide crude oil fouling testing services to companies worldwide. Contact us at [email protected] to learn how we can help you with your next project.

www.htri.net

“HTRI” and the HTRI logo are trademarks or registered trademarks of Heat Transfer Research, Inc. All rights reserved.

Select 151 at www.HydrocarbonProcessing.com/RS Hydrocarbon Processing April Fouling Ad.indd 1

23

3/11/2016 1:47:30 PM

| Special Report FLUID FLOW AND ROTATING EQUIPMENT Most HPI facilities are continuous processes; therefore, any problems or failures in the fluid flow systems will impact the entire plant’s operation and the company’s profitability. Considerable effort is directed on the proper design, installation, operation and maintenance of fluid handling systems. Compressors and pumps provide the motive force to convey various process liquids and gases. Equally important are the support equipment systems, such as valves, piping and instrumentation, as part of the infrastructure to manage products and intermediate streams. This month’s special report investigates the numerous issues around maximizing plant/process unit uptime and reliability, and eliminating leaks and emissions of process liquids and gases. Siemens’ single-shaft centrifugal compressors with horizontally split casings in the STC-SH series can be used for the majority of process applications, including cracked gas, coker gas, process air and refrigerant duties. Photo courtesy of Siemens.

Special Report

Fluid Flow and Rotating Equipment O. GARCIA, R. SHIPMAN, C. TONG and R. E. PALMER, Wood Group Mustang, Houston, Texas

Analyze abnormal operations of an HDS reactor loop with dynamic simulation Hydrodesulfurization (HDS) units are used in a petroleum refinery to process a variety of feeds to alter composition via the addition of hydrogen (H2 ). Process objectives include reducing the sulfur and nitrogen content for subsequent downstream processing. Often, existing HDS facilities are modified for higher throughput, feed composition changes and/or increased hydrotreating severity. A revamp process study is typically undertaken to identify the changes needed to achieve these new process objectives. For the evaluation of the HDS reactor loop, two abnormal operating conditions must be considered: • Heatwave to the reactor effluent equipment and piping caused by the sudden loss of feed • Settle-out pressure in the reactor loop after the loss of recycle gas flow. Conventional approaches for evaluating the impact on the reactor effluent system of a heatwave can result in overconservatism in a revamp or new design.

While conventional methods for calculating the settle-out pressure exist, if the revamped settle-out pressure

exceeds the set pressure of the pressure safety valve (PSV) protecting the HDS reactor loop, then there is generally no

Reactor effluent air cooler

Legend Reactor feed Reactor effluent Recycle/makeup H2 Wash water

PC FC

FC

Wash water

Separator

LC

LC

FC

Recycle gas compressor Recycle gas purge

Product separation

Hydrogen makeup Sour water FC

Hydrocarbon feed

Charge pump

Feed/effluent exchangers

TC

Fuel gas HDS reactor

Heater

FIG. 1. Typical HDS reactor loop.

TABLE. 1 Peak temperatures calculated for feed/effluent exchanger shells

Exchanger

Feed side steady state inlet temp., °F

Feed side steady state outlet temp., °F

Peak feed side outlet temp., °F

Effluent side steady state inlet temp., °F

Effluent side steady state outlet temp., °F

Peak effluent side inlet temp., °F

HX Shell 1

448

495

542

550

485

550

HX Shell 2

425

448

524

485

450

538

HX Shell 3

398

425

498

450

426

517

HX Shell 4

365

398

472

426

397

491

HX Shell 5

323

365

448

397

361

465

HX Shell 6

270

323

424

361

315

441

HX Shell 7

203

270

397

315

256

418

HX Shell 8

123

203

350

256

185

389

Air cooler

–

–

–

175

131

325

Hydrocarbon Processing | AUGUST 2016 25

agreed-upon method for calculating the relieving rate. Dynamic simulation is the preferred approach for analyzing these two contingencies. Key design information is used to develop the dynamic simulation. This includes equipment design details such as tube/shell size, geometry, nozzle locations and elevations. Centrifugal pumps and compressors are modeled using the performance curves. CV data is used to model control valves, while volumes are used for piping to model the holdup. Heatwave caused by loss of feed. In

the HDS unit, a heatwave begins when the liquid feed flow is lost, resulting in a condition where the heat content in the reactor effluent—which normally transfers to the feed in the feed/effluent exchanger—is not removed. As a result, equipment and piping in contact with the reactor effluent will experience higher-than-normal temperatures. Evaluating the peak temperatures for equipment in the reactor loop with conventional methods may result in very conservative design conditions. A dynamic simulation is used to predict the transient response of the temperatures and pressures for this condition. A dynamic simulation of a typical HDS reactor loop (FIG. 1) is used to demonstrate the heatwave analysis. The hydrocarbon liquid feed is pumped by the charge pumps to the reactor loop pressure, and the hydrocarbon liquid rate is regulated by flow control. Liquid hydrocarbon is combined with the recycle H2

from the recycle gas compressor. The combined feed stream goes through a series of feed/effluent exchangers, where the reactor feed absorbs heat from the reactor effluent before being heated to the reactor inlet temperature in the heater. The reactor feed inlet temperature to the HDS reactor is controlled by adjusting the fuel gas flow to the heater burners. The reactor effluent is cooled in the feed/effluent exchangers and then mixed with wash water before entering the reactor effluent air cooler. The air cooler outlet stream enters the separator, which separates the sour water, liquid hydrocarbon and recycle gas. The liquid hydrocarbon and sour water are pressured out of the HDS reactor loop. The vapor stream is split, with some gas purged out of the HDS reactor loop to maintain H2 purity in the recycle gas. The remaining gas is routed through the recycle gas compressor and mixed with the hydrocarbon liquid feed. The following example is for a heatwave resulting from a local power failure where the reactor feed is lost. It is assumed that normal control responses stop the H2 makeup and wash water flows. It is also assumed that fuel gas to the heater is shut off via safety interlock and that there is no heat input to the feed. The recycle gas compressor continues operating and initiates the heatwave. The dynamic model is run until peak temperatures on the feed/ effluent exchangers are observed and begin to decay. Eight shells are used in series for the feed/effluent exchangers (TABLE 1).

600

60,000 Comp. suction Effluent side 1 Heater Comp. discharge

550 500 450

Separator Effluent side 2 Feed side 2 Relieving rate

Air cooler HDS reactor Feed side 1

50,000

350 30,000

300 250

20,000

200 150

10,000

100 50

0

0 0

1

2

3

4 Time, min.

FIG. 2. Settle-out pressures and relieving rates. Select 152 at www.HydrocarbonProcessing.com/RS

5

6

7

8

Relieving rate, lb/hr

40,000

400 Pressure, psig

romance.hoerbiger.com

...R...O...MA...NCE

Fluid Flow and Rotating Equipment

Fluid Flow and Rotating Equipment For a new HDS unit, the peak temperatures can be incorporated into the selection of the design temperatures for both the shell and tube sides of the feed/ effluent exchangers and the air cooler. A conservative approach is to set the design temperature of the effluent side of the feed/effluent exchangers to the same temperature as the HDS reactor. With the dynamic simulation, the peak temperatures can be used to reduce the design temperature, which may result in cost savings for the feed/effluent exchangers. Also, the dynamic simulation is useful in the design of the reactor effluent air cooler in terms of selecting fin tube type based on the design temperature. For the examples shown in TABLE 1, if the peak temperature exceeds the existing equipment design conditions, it may be possible to rerate the impacted exchangers or consider reusing some of the existing exchangers in a different sequence. If rerating or reusing the impacted exchangers is not a viable option, then the exchangers will need to be replaced. Calculating the relieving rate. Dynamic simulation can be used to calculate the required relieving rate of a PSV following the loss of recycle gas if the settleout pressure is greater than the setpoint of the PSV. For a revamp case, the reactor throughput may be increased, resulting in a higher operating pressure in parts of the reactor loop. The reactor loop settle-out pressure may increase to a value higher than the PSV setpoint at the separator. For the given example, the reactor loop PSV is located on the separator. The set pressure is 290 psig, and the normal operating pressure is 210 psig (FIG. 2). When the recycle compressor stops, the PSV opens and the relieving rate peak flow is approximately 56,000 lb/hr. The pressure continues to increase after the PSV initially opens, but the dynamic model does not predict the PSV to be fully open since the separator pressure does not reach 10% overpressure (319 psig). The pressure peaks at around 302 psig. The results indicate that the PSV is adequately sized for this settleout scenario. Takeaway. Dynamic simulation is the best method to predict the transient responses of the temperatures and pressures in an HDS reactor loop from the

loss of feed or recycle gas. For loss of feed, the transient temperature response predicts the magnitude of the heatwave needed to evaluate the design temperatures of the effluent side of the feed/effluent exchangers. Dynamic simulation allows the calculation of a realistic relieving rate from the products separator during a settle-out pressure caused by the loss of recycle gas. OSCAR GARCIA is a process consultant in downstream process engineering in the process plants and industrial business of Wood Group Mustang, where he has worked on multiple refining and chemical projects. Mr. Garcia has 17 years of industry experience. He holds a BS degree in chemical engineering from The University of Texas at Austin and an MS degree from Texas A&M University-Kingsville. RAY SHIPMAN is a process manager in downstream process engineering within the process plants and industrial business of Wood Group Mustang. He has more than 35 years of industry experience, including 20 years at Wood Group Mustang, where his responsibilities include feasibility studies, front-end design and detailed engineering of refinery and petrochemical processing units. Mr. Shipman is a registered professional engineer in Texas and holds a BS degree in chemical engineering from Rice University. CHUNG TONG has more than 35 years of industry experience, including 20 years at Wood Group Mustang, where he is the process engineering manager and the discipline technical authority in the downstream process engineering department within the process plants and industrial business. His responsibilities include proposal development, feasibility studies, front-end design and detailed engineering of refinery and petrochemical processing units. Mr. Tong is a registered professional engineer in Texas and holds a BS degree from National Taiwan University and an MS degree from the University of Houston, both in chemical engineering. R. E. (ED) PALMER is the manager of downstream process engineering within the process plants and industrial business of Wood Group Mustang. He has more than 40 years of industry experience, including 20 years with Wood Group Mustang. He is responsible for directing all process design activities for downstream projects and studies, including selecting the lead process engineer and supporting team for each engagement. Mr. Palmer is a registered professional engineer in Texas and a member of the American Institute of Chemical Engineers. He holds a BS degree in chemical engineering from the Missouri University of Science and Technology and has authored numerous technical articles and presentations for industry publications and meetings.

DEFINING THE LIMIT AS STANDARD Electric actuators for the oil and gas industry

Safe, explosion-proof, tough. AUMA offer a large portfolio of actuator and gearbox type ranges. ■ Automating all types of industrial valves ■ High corrosion protection ■ Integration into all commonly used control systems ■ Global certifications and approvals

Find out more on our automation solutions

www.auma-usa.com

Select 153 at www.HydrocarbonProcessing.com/RS anzeige_drittel_seite_hydrocarbon_processing.indd04.02.2016 1 14:26:21

VERSATILE. Always a leading innovator, ROSEN not only supplies pipeline customers with the latest diagnostic and system integrity technologies but also offers flexible solutions and all-round support for plants & terminals. www.rosen-group.com

Select 61 at www.HydrocarbonProcessing.com/RS

Special Report

Fluid Flow and Rotating Equipment J. M. HANCOCK and S. RODRIGUES, Nalco Champion, Oegstgeest, The Netherlands

Extend ethylene plant run length with compressor chemical treatment An aging ethylene plant historically suffered from plant run length limiting compressor fouling that resulted in a loss of polytropic efficiency and intercooler backpressure buildup. The target run length of the ethane-fed plant was five years, and the need to reduce throughput as the end of run approached was not uncommon. Due to the past performance of the plant—i.e., the compressor’s vulnerability to fouling—run length limiting compressor fouling was expected. This plant assumption was challenged in 2007, when a revamp expansion project was completed to include a second compressor train (CT-2), which also experienced significant fouling. This led to a series of events that resulted in better mitigation of compressor fouling, which, in turn, redefined the limitations of both plants. The new mitigation strategy resulted in two milestone successes. The first was a two-year extension of the target run length from five to seven years. The 2012 scheduled plant shutdown was rescheduled for 2014, and this record-setting plant run length was achieved without any load reduction. The second success was the ability to improve plant margins by operating at a higher cracking severity. Previously, high coil outlet temperatures resulted in increased fouling in the compressor. The improved onsite compressor fouling mitigation strategy keeps this impact under control, and high cracking severity is now an economic advantage. The historical compressor fouling and changes to the treatment philosophy that afforded an extended run length are discussed here. Lessons learned from this success have resulted in a new approach to operations that will be applied for subsequent runs. Challenge: Compressor fouling. Although mechanisms of

compressor fouling have been discussed within the industry,1–4 incidences of unplanned plant shutdowns and lost production attributed to fouling provide evidence that gaps remain. A delicate balance exists between the cost of a mitigation strategy and the economic return from the investment. Further complicating the issue, no singular solution or “easy fix” is available. Every plant has a different tendency to foul and requires a bespoke mitigation strategy to achieve site-specific goals. Typically, these include various combinations of chemical approaches: wash oil, attemperation water injection, chemical inhibitors and/or chemical dispersants.1–4 To demonstrate the varied approaches, a survey of 36 plants presented at the 2013 European Ethylene Producers Committee

(EEPC) Conference concluded that 46% of plants reported use of a chemical antifoulant program, 35% reported use of water injection, and 19% reported use of wash oil injection.5 Despite these mitigation strategies in place within the industry, surveys conducted by the EEPC found that fouling is the second leading cause for compressor failure in an ethylene plant.6 In terms of monetary losses, blade fouling can cause energy efficiency losses as great as 1%. For turbomachinery, that can use up to 70 MW of power and translate into an annual loss of $300,000.7 The ethylene plant added the second compressor train (CT-2) in 2007. The five-stage compressor maintains inlet temperatures of 40°C–45°C and outlet temperatures of 80°C–92°C. Intermittent washing was performed weekly into each compressor section to wet the internal surfaces and minimize the accumulation of polymers. Attemperation water in the form of boiler feedwater was injected continuously and directly into each compressor to lower the compressor operating temperature. After three years of operation, the cracked gas compressor (CGC) train exhibited fouling of such severity that the medium-pressure (MP) rotor had to be replaced. In 2011, the MP rotor was replaced once more after a mechanical failure, to which fouling was a main contributor. FIGS. 1 and 2 show the extent of fouling in the intercoolers and rotors of the machine. A collaboration between the plant and a chemistry service provider was initiated in August 2011, after the second rotor replacement, to develop and implement a complete treatment regime to improve reliability. At startup, the compressor train was untreated, followed by a brief and unsuccessful treatment program. Following a compressor audit and deposit analysis, a mitigation strategy was implemented: • Application of a proprietary antifoulant to compressor Stages 1–4 • Application of a proprietary antifoulant to compressor intercoolers Stages 2–4 • Intermittent application of wash oil to Stages 1–4, with a frequency dependent on fouling tendency (1–3 times per week) • Continuous boiler feedwater application to each stage • Proactive performance monitoring coupled with a well-defined response protocol for adjusting antifoulant injection rates. The main contribution to compressor fouling is unwanted polymerization of reactive monomers present in the cracked Hydrocarbon Processing | AUGUST 2016 29

Fluid Flow and Rotating Equipment gas stream via Diels-Alder reactions and free-radical polymerization mechanisms. The chemical structures of the main culprits are shown in FIG. 3. These compounds undergo unwanted polymerization, particularly given the high discharge temperatures of > 90°C.2–4 The chemical composition of foulant differs progressively throughout the compressor train due to the changing composition of cracked gas. A major contributor to fouling in the early stages is the unwanted reactions of styrene, indene and cyclopentadiene. In later stages, radical polymerization of butadiene and isoprene becomes more prevalent. Although each stage of compression can suffer from fouling, some stages of a plant are more vulnerable than others. The proprietary antifoulant includes multifunctional products that mitigate fouling by several approaches: • Chemically reacting with fouling precursors to prevent unwanted polymerization • Changing the physical interaction between the foulant and the equipment surface to reduce deposition • Designing dispersants to have favorable intermolecular interactions with both foulant polymers and solvent to prevent agglomeration and deposition of foulant. The antifoulant is either dosed with wash oil when injected continuously (preferred), with flash water, or into CGC suction lines. The dosage is dependent on fouling severity and discharge temperatures. Given the impact of temperature on fouling, a higher antifoulant dose rate is recommended during

FIG. 1. CT-2 fouled interstage cooler before mitigation strategy.

FIG. 2. CT-2 fouled rotor before mitigation strategy.

Styrene

1,3-Butadiene

Divinylbenzene

Isoprene

Indene

Piperylene

Cyclopentadiene

Vinylacetylene

FIG. 3. Chemical structures of reactive monomers that can polymerize and contribute to compressor fouling.

30 AUGUST 2016 | HydrocarbonProcessing.com

the summer months in warm climates due to higher compressor discharge temperatures. The application of the new treatment program had an immediate impact on compressor performance. Six months after the mitigation strategy was implemented, improvements in performance were seen: • Stage 1: 15% recovery of polytropic efficiency • Stage 2: Increase of polytropic efficiency from 55% to 90% • Stage 4: Stabilization of polytropic efficiency from 92% to 94% • No ∆P increase observed in Stages 2–4 intercoolers • Rotor speed dropped by 250 rpm • Steam turbine condensate flow reduced by 3 tph. Three years after the mitigation strategy was implemented, the compressor was opened as part of a general turnaround overhaul. It was found to be remarkably clean, as shown in FIG. 4. Additionally, during these years the cracked gas load on the compressor was increased by 5%, which translates into a throughput increase of the complete production unit. The improvement in reliability and performance is illustrated in FIG. 5, where the MP rotor casing’s axial displacement is stabilized following the antifoulant treatment program. The chemical treatment program was begun on a clean machine following an unplanned stop due to MP rotor failure. Once applied, severe fouling was not observed. The expansion project led to the reevaluation of the existing compressor train (CT-1) treatment philosophy. The three-casing, five-stage compressor has inlet temperatures of 40°C–45°C and outlet temperatures of 85°C–105°C. At start of run, the scheduled plant run length target was five years; however, due to the expansion project, the forecast turnaround for the plant was extended by two years. A team was formed to mitigate fouling of CT-1 to meet this deadline without reduction of throughput. The existing program was adapted to reflect the following fouling mitigation strategy: • Application of antifoulant to compressor Stages 1–5 • Application of antifoulant to Stage 2 compressor intercooler • Intermittent application of wash oil to Stages 1–5, with frequency dependent on fouling tendency (1–2 times per week) • Transition between two different wash oils to take advantage of the higher final boiling point. The wash oil would remain in the liquid phase at a higher temperature • Continuous attemperation water application to each stage. A discharge temperature of < 90°C is preferred, but specific constraints interfered with achieving this objective; < 100°C was attained • Close daily performance monitoring, weekly reporting and proactive response adjusting antifoulant injection rates. In 2013, diligent monitoring of the compressor indicated a decline in polytropic efficiency, mainly on Stages 2 and 5. Consequently, another audit was conducted to identify remedial actions that would allow attainment of the revised seven-year run length target. A few observations that influenced the performance of the program are shared here as lessons learned. Discrepancies between theoretical (intended) and actual wash oil injections during intermittent flushes due to deviation from the desired pump rate were found. The intermittent wash oil injection rates and frequencies were modified for additional protection of vulnerable stages.

Benchmark.

We’ve raised the bar with best-in-class solutions for oil and gas applications. Welcome to a new standard of innovation and technology.

As you have come to expect, Dresser-Rand provides safe, reliable and efficient rotating equipment for nearly every application in the oil and gas market. But there’s more. The new Dresser-Rand business now has expanded resources and more experience as a member of the global Siemens family. The Dresser-Rand business combines one of the industry’s most extensive portfolios of rotating

equipment with a universe of intelligent solutions and one of the world’s largest technical support and service center networks. We offer more choices—where you need us— all from a single supplier. So what can you expect from us? High-quality products and services that turn longer run times for your equipment into an everyday experience. This, combined with local around-the-clock support, is the new standard.

dresser-rand.com Select 57 at www.HydrocarbonProcessing.com/RS

Fluid Flow and Rotating Equipment selectivity—i.e., ethylene product yield is reduced due to an increase in secondary reactions.8 Furnace operating pressure is a function of CGC suction pressure. Efficiency deterioration on CGC at the same load results in a suction pressure increase and a backpressure to radiant coils. Lessons learned. A high-efficiency furnace and high conversion

C-2 MP casing axial displacement

FIG. 4. CT-2 during scheduled turnaround after antifoulant treatment program was implemented. 0.8 0.7 0.6 0.5 0.4 0.3 0.2 0.1 0 -0.1 -0.2 Nov.-08

Antifoulant treatment started

July-09 April-10

Dec.-10

Aug.-11 April-12 Date

Dec.-12

Sept.-13 May-14

FIG. 5. CT-2 MP casing axial displacement trend over time.

Antifoulant dosage rates were modified to protect vulnerable stages, as detected by monitoring stage vibrations and intercooler ∆P. The investment and focus on compressor monitoring and treatment returned a significant value to the plant; the extended run-length goal of seven years was achieved successfully without throughput reduction. The compressor performance achieved was unprecedented for the plant, with a record-setting plant run length. Moving forward, the protocol relating to the development and implementation of chemical treatment programs has been redefined, and a new performance record has been established, altering the previous “mindset” regarding the limitations of the CT-1 compressor. Another success was the ability to improve plant margins by operating at a higher cracking severity. Before applying the improvements to mitigate compressor fouling, trial periods of high coil outlet temperatures were showing an unacceptable impact on fluctuations of intercooler ∆P and rotor axial displacement— an indication of fouling in the compressor train. Therefore, the cracking severity adjustment was limited by the delicate balance between yields and compressor reliability. The new compressor treatment program contributed to the plant’s ability to maintain a higher coil outlet temperature than was previously possible, which resulted in accessing the potential of recently installed new cracking furnaces, boosting the ethylene yield by 3 vol%–4 vol%. By maintaining the same CGC throughput in the CT-1 and CT-2, the plant generated a significant production gain of 5%. The implemented compressor treatment program has decoupled the connection between yields and compressor reliability, and improved ethylene selectivity is now an economic advantage for the plant. The chemical reaction in the cracking furnace is sensitive to pressure. Higher operating pressure has a negative impact on 32 AUGUST 2016 | HydrocarbonProcessing.com

are key parameters for the operation of a competitive plant. Although the seven-year run length was successful due to improvements made in controlling compressor fouling, the backpressure to the furnace was higher than the design suction pressure because the comprehensive mitigation program was not started on a clean system. The design suction pressure is 0.6 bar, but the suction pressure gradually increased by a factor of almost 2. Proactive monitoring helped identify this issue at an early stage, and immediate corrective actions were put in place by increasing the antifoulant dosage to arrest the fouling tendency. The mitigation strategy applied to the system resulted in stabilization of the backpressure to achieve the plant run-length goal. Stabilizing the backpressure yielded significant savings for the plant. A lesson learned is to not wait until deterioration of compressor efficiency is observed to develop a comprehensive mitigation strategy. Although the backpressure was stabilized, a reduction to the design suction pressure was not possible. The key to achieving the maximum value of an antifoulant program is to apply the defined strategy at the beginning of run on a clean machine. Inspection during the scheduled general shutdown showed that the CT-1 compressor was not severely fouled, despite the seven-year run. The CT-2 compressor casing and wheels were found to be very clean. The success at the plant demonstrates the value of remaining diligent in applying updated fouling mitigation strategies. This approach challenged the existing mindset of plant limitations, and the plant has a new run-length target of seven years. Sharing the experience with the industry is a powerful example of how improved profitability can be achieved without CAPEX investment in aging plants, and improved fouling control can lead to implementation of operating strategies that were previously deemed out of reach.

ACKNOWLEDGMENTS The authors would like to acknowledge Walter Militello for his technical expertise and advice during the application of the compressor treatment modifications. They would also like to thank Dr. Theodore Arnst and Dr. Debby Rossana for their technical reviews and contributions during the preparation of this article. LITERATURE CITED Complete literature cited available online at HydrocarbonProcessing.com. JESSICA M. HANCOCK joined Nalco Champion in 2009, and is currently a global marketing manager. She provides technical support for petrochemical plant process treatment programs and stewards the development and launch of innovative treatment programs into the petrochemical industry. She obtained a PhD in chemistry from the University of Washington. SEBASTIAN RODRIGUES holds a BS degree in chemistry. He has 15 years of experience with Nalco Champion, and his responsibilities include providing technical support to ethylene producers within the Gulf region.

Special Report

Fluid Flow and Rotating Equipment M. RAJASEKHAR, V. D. THAKARE, G. SRIVARDHAN, V. K. JAYANTI, D. K. R. NAMBIAR, S. R. SINGH and V. SHUKLA, Engineers India Ltd., New Delhi, India

Boost capacity of SRUs with mixing devices for oxygen enrichment The requirement for additional sulfur recovery unit (SRU) capacity in a refinery may stem from hydroprocessing operations; a refinery expansion project; a shift from lighter to heavier, sourer crudes to improve margins; or the need to remove extra sulfur content to meet tighter environmental regulations. To meet this demand, the typical practice has been to enrich the incoming process air with oxygen (O2 ), thereby reducing the flow of nitrogen (N2 ) through the SRU and allowing an increase in the acid gas load that can be handled. For better utilization of O2 -enriched air, it is important to ensure complete mixing between the O2 stream and the process air. Incomplete mixing may lead to flame instability in the main combustion chamber. The extent of mixing of the two streams in seven different mixing configurations using computational fluid dynamics (CFD) is analyzed here. The best of these seven configurations has been implemented at some of the refineries and is performing to expectations. SRU operation and capacity augmentation. In refineries, the SRU is a vital unit that removes gaseous hydrogen sulfide (H2S) from acid gas streams by converting it to elemental sulfur through a modified Claus process.1 In the modified Claus process, the overall reaction is divided into two steps—a combustion reaction followed by the Claus reaction, as shown in Eqs. 1–3:

Combustion reaction: H2S + 1.O2 1 SO2 + H2O

(1)

Claus reaction: 2H2S + SO2 1 3S + 2H2O

(2)

Overall reaction: 3H2S(g) + 1.5O2(g) 1 3S(l)+ 3H2O(g)

(3)

Benefits of O2 enrichment. In the event of a refinery expan-

sion or a shift to heavier and sourer crudes to improve margins, augmenting SRU capacity by adding more trains to meet the increased load is not always a cost-effective option. Instead, enriching the process air entering the SRU with O2 proves to be a successful approach. O2 enrichment is a low-cost process that increases flame stability and thermal efficiency, with a negligible space requirement compared to a new SRU. Besides being a highly flexible method for expanding SRU capacity, the approach also reduces emissions by decreas-

ing N2 flow into the tail gas treating unit (TGTU). This, in turn, results in a higher partial pressure of H2S in the amine absorber, leading to better absorption and lower sulfur emissions. Specific instances can be found where O2 enrichment has been used solely to increase the furnace temperature.2 SRU capacity enhancement through O2 enrichment. As per the stoichiometry of the overall reaction (i.e., Eq. 3), it is evident that 1 kmol/hr of H2S needs 0.5 kmol/hr of O2 . If all of the O2 is supplied through air, then approximately 1.88 kmol/hr of N2 accompanies the 0.5 kmol/hr of O2 (on the basis of 21% O2 in air). N2 accounts for more than 55% of the feed volumetric flow, contributing in no small measure to the large pressure drop through the SRU. By increasing the O2 concentration, a drastic increase of the acid gas flow into the reactor can be realized at the expense of N2 , as shown in TABLE 1. Therefore, O2 enrichment is a simple yet effective option to overcome SRU capacity limitations while staying within the same pressure drop limits. The O2 content in the combustion air can range from 21%– 100% through use of appropriate technologies and depending on the level of O2 enrichment. For low-level enrichment (21%–28% O2 ), the O2 feed can be “hot-tapped” into the combustion air pipeline while the sulfur plant is in operation, or through a diffuser, which provides good mixing and completely avoids downtime and associated production losses. However, this method is not advised for O2 enrichment levels above 28%, as the safe handling of O2 at the resulting high temperatures requires special burners featuring separate ports for the entry of air and O2 .3 Care should be taken during O2 enrichment to ensure complete mixing between the O2 and combustion air, or it may TABLE 1. SRU capacity enhancement vs. level of O2 enrichment O2 enrichment, % 21

28

50

75

100

Acid gas, H2S

100

121

169

203

225

O2

50

61

85

102

113

N2

188

156

84

33

0

Total flow

338

338

338

338

338

Flow to reaction furnace, kmol/hr

Hydrocarbon Processing | AUGUST 2016 33

Fluid Flow and Rotating Equipment lead to flame instability in the main combustion chamber. Any unreacted O2 can result in the formed SO2 being further oxidized to SO3 , which can then break through to the downstream catalytic convertors and poison the catalyst bed. A portion of SO3 may also combine with H2O to form H2SO4 , which is a highly corrosive agent capable of damaging downstream equipment.2 Many refineries practice direct injection of O2 into the air line for low-level O2 enrichment without the necessary precautions to prevent improper mixing, and these plants end up with the aforementioned problems. Mixing devices like static mixers, oxynators, simple tee mixers and impingement jet mixers4 are used for continuous mixing of gas/gas streams to achieve desired levels of mixing, dispersion and heat transfer. The choice of a suitable mixing device is usually guided by pressure drop considerations, especially in revamp situations. Space constraints may also limit the piping length downstream of the mixing device before the air/O2 mixture enters the main combustion chamber. In such an event, discrimination between various mixing device designs is facilitated by a tool such as CFD. The use of CFD simulations to make an informed choice of a mixing device is illustrated here. Seven different static mixing configurations were simulated; the best were selected after ranking them on various parameters, such as the extent of mixing at various cross-sections downstream of the point of injection of O2 , pressure drop and ease of installation. CFD modeling: Assumptions. The flow is assumed to be

steady and incompressible for the CFD simulations. In general, compressibility effects become significant only at Mach numbers exceeding 0.3. For the flow velocities encountered in practice, incompressibility can be safely assumed.

FIG. 1. Phase 1 geometric configurations (1, 2 and 3).

FIG. 2. Phase 2 geometric configurations (4 and 5).

34 AUGUST 2016 | HydrocarbonProcessing.com

Geometrical details. A total of seven different geometrical configurations were simulated in this study, which was carried out in two phases. By way of a preliminary study, three geometrical arrangements were first explored in Phase 1, and the best of these was then chosen as the base model to develop four more variants in Phase 2 studies. In all cases, a pipe length 15 times the diameter of the air pipe and spanning the point of introduction of O2 into the pipe was modeled in the commercial CFD software. The dimensions for the main pipe and the O2 inlet pipe, along with the process operating parameters considered in the simulations, are as follows: 1. Main pipe for air flow (14-in. nominal bore): 9,500 kg/hr at 0.91 kg/cm2g 2. Branch pipe for O2 introduction (3-in. nominal bore): 500 kg/hr at 1 kg/cm2g. Phase 1 geometrical configurations. The first arrangement (FIG. 1) considered was a simple T joint in which O2 is introduced into the main pipe through a branch pipe joined at an angle of 90°. In the second and third configurations (FIG. 1), O2 enters the main pipe through a 90° bend and is released radially through eight and four rectangular slots of different dimensions, respectively, provided on the 3-in. pipe. Phase 2 geometrical configurations. In Configuration 4, (FIG. 2) as in Configuration 3 (FIG. 1), O2 enters through the branch pipe and goes through a 1.5-in. diameter, 90° bend. Further downstream, its velocity is increased by means of a converging section wherein the pipe diameter is halved; from that point, it exits tangentially through five slots. Configuration 5 (FIG. 2) differs from Configuration 4 only at the O2 release point, as it contains guide vanes that provide additional swirl, which promotes better mixing of O2 . In Configurations 6 and 7 (FIG. 3), O2 is introduced into the main pipe through four circumferentially placed rectangular and circular nozzles, respectively. For simplicity, it has been assumed that the O2 flow distributes uniformly among the four nozzles—in practice, this can be ensured by providing a suitably sized ring header. Meshing and boundary conditions. A fine mesh was used for all of the configurations to ensure that all mesh quality parameters were within specified bounds. The first three configurations were simulated with mass flowrates of 9,500 kg/hr of process air and 500 kg/ hr of O2 applied at the respective inlets. The mass fraction of O2 in the process air was set at 0.21. The target mass fraction of O2 for complete mixing was 0.2495, based on an O2 mass balance. The best-performing of the three, along with the four new configurations of Phase 2, were next simulated at actual operating conditions from one of the refineries—i.e., a mass flowrate of process air of 8,390 kg/hr (design value) with

Fluid Flow and Rotating Equipment the mass fraction of O2 in the incoming process air at 0.2236 (equivalent to 19.67 vol% O2 ). Mass flowrate of O2 was retained at 500 kg/hr for these simulations. The target mass fraction of O2 indicating complete mixing works out to be 0.2673 in this case, based on O2 mass balance. Simulation results and discussion. CFD simulations were

carried out for seven different geometries. Except for the hypothetical case of perfect mixing, at any cross-section downstream of the O2 inlet, the mass fraction of O2 will exhibit a distribution about the target value computed from an O2 mass balance. The standard deviation of this distribution provides a measure of the extent of mixing of the two streams at that particular location. It is reasonable to assume satisfactory mixing if the O2 mass fraction is within ±3% of its target value over most of the crosssection. The contour plots of the O2 mass fraction on planes at respective distances of 1D, 3D, 5D and 7D (D = diameter of main pipe), from the point of injection of O2 , give a concise snapshot of the effectiveness of each mixing arrangement. For Configuration 1 (FIG. 1), the contours of the mass fraction of O2 at these four locations are depicted in FIG. 4. The range has been limited to 0.2–0.29 for better clarity. The contours of mass fraction show an accumulation of O2 toward the top of the pipe, resulting in the stratification of O2 concentration. This accumulation also results in insufficient mixing along the length of the pipe.

Even after 15D, complete mixing was found to be unachievable in this configuration. The standard deviation of the spread about the mean value of 0.06 and 0.05 at the 5D and 7D planes, respectively, confirms the incomplete mixing. Configuration 2 follows the similar trend of Configuration 1. For Configuration 3, all of the planes show considerably improved distribution of O2 , as compared to the previous configurations. A more uniform concentration of O2 was obtained at the 7D plane. At 7D, the standard deviation is only 0.008, which is nearly 3% of the mean mass fraction on the plane and meets the defined criteria for complete mixing. In Phase 2, Configuration 3 was compared with the four previously discussed configurations. For this purpose,

FIG. 3. Phase 2 geometric configurations (6 and 7).

Enhance Steam Turbine Performance

Increase reliability and efficiency on your process steam turbines with the Sentinel® FBS, only from Inpro/Seal®. Carbon rings are weak, brittle and wear quickly, making them an ineffective sealing solution by decreasing equipment reliability and efficiency. The Inpro/Seal Sentinel® FBS (Floating Brush Seal) is a drop-in replacement for existing carbon rings that effectively seals against steam loss, reducing maintenance and downtime, while also protecting downstream carbon rings from contamination and high pressure. For more information, visit www.inpro-seal.com or contact +1 309-787-4971.

EXPER T EN G IN EER IN G . PR OV EN R ESU L TS.

Select 154 at www.HydrocarbonProcessing.com/RS

Hydrocarbon Processing | AUGUST 2016 35

Fluid Flow and Rotating Equipment Between Configuration 6 and Configuration 7, the latter is more attractive on parameters like pressure drop in the O2 line and ease of fabrication. As a result of this evaluation, Configuration 7 has been implemented at some refineries and has been performing to expectations. LITERATURE CITED Bishtawi, R. E. and N. Haimour, “Claus recycles with double combustion process,” Fuel Processing, Vol. 86, No. 3, 2004. 2 Asadi, S., M. T. Hamed Mosavian and A. Ahmadpour, “Effect of O2 concentration on the reaction furnace temperature and sulfur recovery using a TSWEET process simulator,” Journal of Chemical Engineering & Process Technology, Vol. 4, No. 2, 2011. 3 Nasato, E., U. Parekh and P. Morris, “Oxygen enrichment of sulfur recovery units to boost capacity, conserve capital and improve environmental performance,” Sour Oil and Gas Advanced Technology Conference, Abu Dhabi, UAE, 2008. 4 Paul, E. L., Ed., Handbook of Industrial Mixing Science and Practice, John Wiley and Sons, Hoboken, New Jersey, 2004. 1

FIG. 4. Contours of mass fraction of O2 on four planes until x = 2.5 m.

M. RAJASEKHAR is a post-graduate student in chemical engineering from IIT Roorkee, with more than four years of experience in CFD and energy efficiency improvement studies for refineries and petrochemical plants. He works as a senior engineer at EIL.

FIG. 5. Contours of O2 mass fraction on downstream planes for Configuration 7.

Configuration 3 simulations were repeated at conditions similar to those considered for the four other geometrical models—namely, a mass flowrate of 500 kg/hr for O2 and 8,390 kg/hr for process air, with the mass fraction of O2 in the process airstream set to 0.2236. As before, a quantitative assessment of the effectiveness of each of the mixing configurations is facilitated by computing the standard deviation of the spread of O2 concentration around its mean value at the downstream planes. The deviation on mass fraction of O2 on the 7D plane was observed to be the lowest for Configuration 4. It is well within the acceptable criteria of ±3%, but at the expense of higher pressure drop (i.e., 729 mbar). Although Configurations 3, 4, 5 and 6 offer acceptable deviations at the 7D downstream cross-section, with pressure drops of 113 mbar, 729 mbar, 735 mbar and 79.13 mbar, respectively, Configuration 7 is much simpler and more flexible in terms of operation and fabrication, and causes minimal pressure drop of 72.9 mbar. Also, the contours of mass fraction show complete mixing by 7D itself (FIG. 5). O2 is introduced through a ring header sized to ensure uniform distribution of velocity and mass flow to each of the four nozzles at design, as well as at turndown conditions. Implementation of new configuration. While both Con-

figuration 3 and Configuration 4 appear to provide a comparable degree of mixing, the latter incurs a higher pressure drop and is also more difficult to implement in the field. Among Configurations 3, 6 and 7, the first incurs a higher pressure drop and higher deviation from the mean compared to the other two.

36 AUGUST 2016 | HydrocarbonProcessing.com

V. D. THAKARE is a post-graduate student in chemical engineering from the National Institute of Technology at Surat. He works as an engineer at EIL. He is actively involved with CFD simulations, both for research on new technologies and troubleshooting operational problems in refinery equipment.

G. SRIVARDHAN is a post-graduate student in chemical engineering from IIT Kanpur. He works as a senior engineer at EIL. He has more than six years of experience in CFD projects and energy efficiency improvement studies for refineries and petrochemical plants.

V. KAMESH JAYANTI is deputy manager at the R&D division of EIL. He joined EIL in 2006 after completing his MTech degree in chemical engineering from the Indian Institute of Technology (IIT) at Kharagpur. He has more than seven years of experience in design, adequacy checking for revamp projects, commissioning and troubleshooting in SRU, TGTU, ARU and O2 enrichment technology for capacity enhancement of the SRU. D. K. R. NAMBIAR is assistant general manager in the R&D division of EIL. Dr. Nambiar is a graduate of the Indian Institute of Technology at Bombay, with a PhD from the Indian Institute of Science in Bangalore. At EIL, he has been associated with the establishment of a CFD facility at the R&D center in Gurgaon. The facility is used to execute projects for both in-house and external clients. SHEO RAJ SINGH is a graduate in chemical engineering from IIT Kanpur. He joined EIL in 1981 and holds the position of head of R&D. During his long association with EIL, he has been involved in developing and implementing several process technologies and computational tools to improve the design and operation of industrial processes. VARTIKA SHUKLA is an executive director with Engineers India Ltd. (EIL), leading its research and development (R&D) division. She joined EIL after graduating in chemical engineering from IIT Kanpur in 1988, and was the head of EIL’s process division before her current assignment. She has more than 26 years of consulting experience in the process design of refinery, gas processing and petrochemical units.

SUPERIOR SPRAY. SERIOUS RESULTS. OPTIMIZE

VALIDATE

BUILD

PERFORMANCE

DESIGN

TO ORDER

Let us assist with nozzle selection, determination of injector placement in the vessel and spray direction to ensure optimal performance based on your exact operating conditions.

Prior to fabrication, we can use Computational Fluid Dynamics (CFD) to validate injector spray performance. We can also use failure analysis modeling to verify the injector design will withstand process conditions such as thermal stresses, heat transfer, vortex shedding and more.

Whether you need to cool gas, dissolve salts in overhead lines or inject chemicals or torch oil into a gas stream, we can meet your needs. Injector types include hydraulic, gas atomizing and steam versions equipped with a wide range of nozzles.

TALK WITH YOUR LOCAL SPRAY EXPERT TO FIND OUT HOW WE CAN HELP Companies like Technip, Mustang Engineering, Bechtel, Shell and many others rely on us to manufacture B31.1, B31.3 and U-Stamp code-compliant injectors and conduct radiographic, hydrostatic, ferrite tests and more. Let us do the same for you. Our local experts and Spray Technology Centers are right where you need them – in your area.

For unmatched service and support, visit spray.com/localexpert or call 1.800.95.SPRAY. Select 67 at www.HydrocarbonProcessing.com/RS

| Bonus Report LNG Designs for onshore and offshore liquefied natural gas (LNG) projects are advancing with the rise of floating LNG (FLNG) projects and small-scale liquefaction technologies, as well as the continued construction and startup of traditional onshore LNG terminals. In this bonus report, a cylindrical hull FLNG concept is explored. This concept is a further development of the proven cylindrical floating production, storage and offloading (FPSO) design. It provides a more stable and economical platform for the offshore gas pretreatment and liquefaction processes. It also introduces a better option for producing, storing and offloading LNG in challenging operating areas. Presently, PETRONAS’ SATU FLNG is moored at the Kanowit gas field, 180 km offshore Bintulu, Sarawak, Malaysia. With a production capacity of 1.2 MMtpy, the FLNG facility will change the landscape of the LNG industry and support PETRONAS’ global LNG portfolio, as well as enhance its reputation as a preferred and reliable LNG supplier.

Bonus Report

LNG L. ODESKAUG, Sevan Marine ASA, Norway; and S. MOKHATAB, Consultant, Dartmouth, Nova Scotia, Canada

Cylindrical hull concept improves design for offshore FLNG production As demand for natural gas remains high, the development of offshore stranded gas fields via floating liquefied natural gas (FLNG) production technology will aid sustained growth. Offshore FLNG production offers a potential solution for many projects—in particular, small-scale to mid-scale capacity (0.5 MMtpy to 3 MMtpy). However, it is only in the last 10 years that a few FLNG projects have progressed to the detailed design and construction stage. In fact, some special challenges still exist in design, construction and operation of an offshore FLNG facility (mainly due to vessel motion, limited space and weight, facility operations and safety requirements) that require capital-intensive mitigating designs. The cylindrical hull FLNG concept, which is a further development of the proven cylindrical floating production, storage and offloading (FPSO) design, provides a more stable and economical platform for offshore gas pretreatment and liquefaction processes. The cylindrical hull has significantly less pitch and roll motions than a conventional ship-shaped hull, and eliminates the need for expensive turret and swivel solutions. Therefore, this design concept introduces a better option for producing, storing and offloading LNG in challenging operating areas. The development of the cylindrical hull concept is described with regard to its applicability to offshore FLNG applications. FLNG background. Floating above an offshore natural gas field, an FLNG facility will produce LNG at sea. However, some hurdles must be overcome to make this concept economically viable for offshore FLNG production. The key technical challenges that a floating gas liquefaction facility must combat can be summarized below: • Space and weight requirements: Floating systems are space limited, requiring more compact and lighter equipment to fit the deck space. These systems have high equipment density to overcome space and weight constraints. High equipment density substantially increases the potential for explosions in the event of an ignited gas release, which would have higher impact severities, perhaps escalating to total facility loss. • Facility operations: Bad weather that adversely impacts operational efficiency should be expected offshore, and extreme environmental conditions may require sudden shutdown of the facility. • Feed gas flexibility: The FLNG facility may be required to handle various feed gas compositions, including

carbon dioxide (CO2 ) and nitrogen (N) contents of several percent, either during initial operation or at a future location. This requires increasing the flexibility and availability of the gas pretreatment processes to ensure proper operation of the facility without too many shutdowns or retrofits. • Safety: Design and operation of FLNG facilities present a set of safety challenges, mainly due to the potential inventory of hazardous, flammable process and refrigerant fluids, as well as the consequence of any loss of containment (resulting from external impacts and escalation from topside events). • Vessel motion: Vessel motion is a key limiting factor in deploying FLNG facilities in harsh environments. Once an FLNG facility is in operation, moving decks mainly affect the operation and performance of process equipment having liquid distribution under the effect of gravity (most notably the separation columns and heat exchangers) by creating significant liquid maldistribution that results in a loss of mass or heat transfer efficiency, process upsets, etc. The sea’s wave motions may also cause sloshing in partly filled membrane tanks, leading to high impact pressures on the thermal insulation. In addition, safe offloading of LNG product to visiting LNG carriers under harsh environmental conditions requires more robust mooring and loading-arm technologies than those developed for sheltered, land-based ports. • Operation and maintenance: Operating and maintaining an FLNG facility of any design would require a substantial workforce. Accommodating, supplying and providing adequate emergency evacuation protection for such a large offshore population is a challenge and would be a major operating cost component for such a project. In light of the abovementioned challenges, an essential need exists to move innovative technical solutions for stranded offshore gas off the drawing board. The key aspect in developing a successful offshore FLNG project is the proper design of the hull necessary to provide (1) a seaworthy and stable platform for production and product offloading, as well as safe accommodation of the crew in a remote, potentially hostile environment; and (2) enough deck area to accommodate the topside process/utility units, required product storage and offloading systems, and support facilities. Hydrocarbon Processing | AUGUST 2016 39

LNG The geostationary cylindrical hull concept, as shown in provides a more stable and economical hull configuration for offshore FLNG production, storage and offloading than the traditional ship-shaped vessels. It will provide several advantages compared with conventional ship-shaped hulls, enhancing offshore FLNG project development and operations: • No expensive turret mooring and swivel system. • No need to rotate, even in the harshest environmental conditions, making it tolerant for weather spreading. • Large LNG storage capacity of approximately 240,000 m³. • Large deck area with high load capacity. • Insignificant bending stresses due to global loads on the hull that eliminate typical wave-induced fatigue loads and minimize hull deflections, resulting in simplified topside design. • Favorable motion characteristics (lower pitch and roll motions), resulting in a stable platform for the processing units. • Excellent area segregation, resulting in reduced probability and consequences of major accident hazards. In cases FIG. 1,

FIG. 1. Cylindrical FLNG production unit.

FIG. 2. Schematic of cylindrical hull arrangement.

40 AUGUST 2016 | HydrocarbonProcessing.com

where segregation by physical distance is not sufficient, physical barriers (e.g., fire and blast walls) are provided. • Simple arrangement for lowering seawater intake, resulting in reduction in size of the cooling water system. The cylindrical FLNG unit, which has secured approval from the American Bureau of Shipping (ABS), as well as from DNV-GL, will be capable of operating in water depths of between 30 m and 3,000 m, and in harsh environments such as extreme cyclonic situations. The unit design is based on environmental load calculations for a 100-yr return storm in the Barents Sea. The cylindrical hull design concept has been proven within the drilling and offshore oil and gas production sectors and is being considered by several major oil companies for field development. During 3Q 2015, Sevan Marine was awarded a feasibility study with an oil major company to explore the use of cylindrical hull design for a specific FLNG development. Cylindrical hull design and arrangement. The circular

hull is made up of a double bottom and a double side, varying in width at the vertical part of the hull. The double bottom has a slightly larger diameter than the main hull, which gives favorable damping effects (FIG. 2). The process deck is a fully plated deck and will ensure proper segregation between the processing areas and the containment system below. The main philosophy is to locate the part of the process that contains hydrocarbons aft of the deck, and the non-hydrocarbon systems at the forward part. In the meantime, the layout of main equipment will follow a homogeneous weight distribution to decrease oscillations and improve stability. A utility deck (including main utilities) is arranged below the process deck. The central control room will be protected inside the living quarters. The flare boom will be located on the starboard side of the process deck. The FLNG unit will, if possible, be moored so that the flare is downwind of the dominant wind direction. The cargo tanks are arranged in the hold space below the double process deck (surrounded by a double bottom and double sides), and include six LNG, one central LPG and six condensate storage tanks (FIG. 3). The six LNG tanks (with storage capacity of between 180,000 m3 and 220,000 m3, and a self-supporting, prismatic Type B (SPB) or membrane cargo containment system, respectively) will be arranged around the circumference of the hold space. One tank is arranged in the center for use as an LPG tank in case of LPG production (with storage capacity of 20,000 m3); it is otherwise used as an LNG tank. The six condensate tanks (with combined storage capacity of 40,000 m3) are located between the six LNG tanks. The condensate tanks extend from the top of the double bottom to the process deck. It should be noted that liquid motions in tanks are mainly due to roll/pitch and sway/surge motions. In the case of the cylindrical FLNG unit, the unit is known to have very reduced motions and, as a consequence, sloshing loads would be minimized. The remaining heave motion will introduce some hydrostatic loads, especially due to the height of the tanks. However, such loads are investigated, and as the cargo containment system is designed to withstand high sloshing loads, dynamic cargo pressure would not be a critical item.

LNG Gas processing. The topside processing plant (FIG. 1) comprises the main units illustrated in FIG. 4. All processing units have been configured as single trains except the liquefaction unit, which is split into two separate trains. 8P W.B.

P

.9 W.B

W.B .4P

Cond. 1P

LNG2P

LNG3P

2P W.B.

W.B.1 1P

W.B.5 P

3P

W.B .10P

Cond. 2P

W.B.6P

. W.B

LNG1P

Cofferdam

W.B.1P

W.B.12P

W.B.7P

LPG Cond. AFT

W.B.1S

W.B.12S

LNG2S

0S

Cond. S

W.B .9

S

W.B. 8S

W.B.7S

W.B.6S

W.B .3

.1 W.B

S

W.B.2 S

LNG1S

LNG3S

11S W.B.

The LNG offloading area is located aft of the FLNG unit. Three alternative LNG offloading systems can be used for the cylindrical FLNG unit: side-by-side (with conventional LNG carriers), tandem (with dedicated LNG carriers), and the arc loading system (with conventional LNG carriers). These would allow loading operations in sea states up to 1.5-m, 5.5-m and 4.5-m significant wave heights, respectively. The arc loading system is a novel development using an L-shaped dynamic positioning vessel that attaches to the LNG carrier and guides it during approach and departure, and keeps the LNG carrier at a safe distance of 150 m–200 m from the FLNG facility during the offloading operation. The hold space is separated from the process area by the double, fully plated process deck, acting as a cofferdam between the process deck and the cargo tanks. The hold space will have a controlled atmosphere to reduce/eliminate the risk of fire and explosion, and also to avoid condensation and a humid atmosphere. The living quarters (LQ) and helideck will be located at the fore end of the FLNG unit, as far away as possible from the processing areas. The FLNG unit shall, if possible, be oriented so that the dominant wind directions will minimize the likelihood of gas release or smoke from a fire drifting toward the LQ and primary evacuation points from the unit. Due consideration will also be given to the requirements for helicopter approach, supply boat and export LNG carrier operations, cold venting and radiation from flaring.

Cond. 1S 5S W.B.

S

.4 W.B

FIG. 3. Cargo tank configuration.3

Drier Steam Means Higher Profits Steam drum design is critical to maintain steam dryness and water quality for optimum performance of your boiler. If water is allowed to carryover, then damage can occur and energy is lost. Carryover is your boiler’s enemy. Dyna-Therm’s high performance steam drums have been protecting downstream equipment including superheater tubes and turbines for decades. We offer proven designs for the following: • High pressure • Intermediate pressure • Low pressure • Retrofitting of existing drum internals No steam production rates are too high and no carryover problems are too difficult for us to solve—steam qualities of 99.995% with .001 PPM/TDS are possible!

Let us design the steam drum that best fits your system.

Performance is what we guarantee!

High efficiency steam drums and separators!

281-987-0726 www.dyna-therm.com SEPARATION AND STEAM DRUM SOLUTIONS SINCE 1961 Select 155 at www.HydrocarbonProcessing.com/RS

Hydrocarbon Processing | AUGUST 2016 41

LNG and another degree of complexity and safety considerations to the FLNG unit design. Dehydration. Molecular sieves are used to dry the gas leaving the AGRU to below 0.1 ppmv to avoid hydrate formation in the cryogenic systems. They can also be used for removal of mercaptans and other sulfur compounds (if present in the feed gas) to meet the product specification of 10 ppm. The molecular sieve unit utilizes high-quality molecular sieves in a three-bed adsorber configuration, where two adsorbers are in adsorption mode while the third is undergoing thermal regeneration via a hot, dry gas. Note that various techniques can be used to reduce the size and improve the performance of the molecular sieve units in FLNG projects. For example, using split-bed configurations of dense molecular sieves can reduce bed voidage and reduce vessel volume. Using high-quality molecular sieves with superior properties and improved regeneration methods can extend their lifetime and improve reliability while providing cost savings. Heavy hydrocarbon removal. Removal of the heavy hydrocarbons (HHCs) from the gas to be liquefied is necessary to avoid waxing and plugging in the main cryogenic heat exchanger. The simplest solution is to use a scrub column operating at liquefaction pressure and thermally integrated with the main cryogenic heat exchanger in the liquefaction unit, in which the separated HHC fraction is recycled to the condensate stabilization unit. The scrub column is easy to operate and can produce a highpressure residue gas to reduce liquefaction unit horsepower; however, component separation is not very sharp, and LPG recovery is limited due to the high operating pressure and relatively high reflux temperature. Increased split, which also raises the efficiency of the LNG production, can be achieved by introducing a fractionation column. This configuration is selected as base case. A front-end natural gas liquid extraction unit utilizing conventional turboexpander technology can handle a wide variety of feed gas compositions and effectively remove HHCs, but it contains rotating equipment that impacts the capital investment and reliability of the FLNG facility. Natural gas liquefaction. Two types of refrigeration cycles (mixed refrigerant and turboexpander) can be proposed for offshore FLNG production using the cylindrical floater. However, considering the main criteria that influence the commercial acceptance of small- to mid-scale FLNG projects (weight, footprint, vessel motion, cost, equipment count, startup time and safety), the dual-nitrogen (N2) expander cycle has been selected as the base case. A major benefit of using nonflammable N2 as the cycle fluid is that it is inherently safe and eliminates Nitrogen-rich the need for refrigerant makeup and storoffgas to fuel age of hazardous hydrocarbon refrigerants. This provides the lowest practiLNG to Liquefaction Feed gas storage cal risk of fire/explosion and allows for Inlet Mercury Acid gas Dehydration End-flash unit (including separation removal removal more compact equipment spacing on an scrub column) FLNG vessel. N2 is also maintained in Heavy hydrocarbons LPG (if present the gaseous phase at all points during the in feed gas) refrigeration cycle, so distribution in the Condensate to storage Condensate To LPG heat exchangers is not a concern, unlike stabilization fractionation with other refrigeration cycles. As a result, liquefaction process perFIG. 4. General scheme of gas processing route in cylindrical FLNG unit. formance is less sensitive to vessel moveOffgas

Field production upon arrival at the FLNG production facility will be processed in a slug catcher, which separates the gas, hydrocarbon liquids (condensate) and aqueous phase. The flash gas is further separated in a downstream high-pressure separator to remove any liquid entrainment prior to entering the gas pretreatment section. The condensate is processed in the condensate stabilization unit to reduce the vapor pressure and allow storage in atmospheric storage tanks. The following sections discuss proposed gas treatment technologies in treating the sour feed gas to meet LNG feed gas specifications, as well as the appropriate natural gas liquefaction process for offshore FLNG applications. Mercury removal. Removal of mercury is required to avoid the risks of mercury attack on the brazed aluminum heat exchangers and equipment in the cryogenic section. The mercury removal unit can be positioned upstream or downstream of the acid gas removal unit. Installing vessel(s) of non-regenerative sorbents before the acid gas removal unit removes all the mercury and ensures no mercury contamination through the rest of the FLNG production system. The mercury-contaminated wastes should be sent onshore for proper disposal at a hazardous waste facility. Acid gas removal. The acid gas removal is based on using a promoted amine solvent process (typically activated methyldiethanolamine), where the amine solvent is continuously regenerated with heat input. In the case that mercaptans (R-SH) and other organic sulfur components are present in the feed gas, a mixture of chemical (amine) and physical solvents can be used to allow for complete CO2 removal, while achieving hydrogen sulfide (H2S) removal comparable to alkanolamines (amine solvent). Generally, this option will result in an expensive design with a hydrocarbon coabsorption that may be too large to be acceptable.1 The optimum solution in many cases is the distribution of the mercaptans removal capabilities over the optimized mixed chemical-physical solvent in the acid gas removal unit (AGRU), as well as the molecular sieve unit (MSU). In the case that the feed gas CO2 content is high, membrane separation is a suitable method for bulk CO2 removal, where further treatment with amine is required to meet required H2S and CO2 specifications. The discharged acid gas stream can be routed to the flare stack to ensure its safe disposal (in the case of low H2S content), or reinjected to a suitable reservoir to minimize environmental impact (if the concentrations and flowrate of acid gas components are too high). However, acid gas injection will require an additional dehydration/compression system that adds costs

42 AUGUST 2016 | HydrocarbonProcessing.com

Identify potential motor failures before they happen? Absolutely.

Motors, like any other machines, are subject to wear and aging. Fortunately, condition monitoring solutions from ABB can identify possible weaknesses and defects at an early stage, while there is still time to take action. – Life Expectancy Analysis Program (LEAP) assesses the condition of stator winding insulation and evaluates the insulation’s remaining life. – MACHsense-R monitors the rotor and bearings of the motor from a remote location, measuring vibration and temperature across multiple points. ABB conditioning monitoring services – minimizing unplanned downtime, improving reliability and enhancing performance. So you can identify possible failures before they happen. For more information on condition monitoring services from ABB, visit new.abb.com/motors-generators/service.

ABB Motors and Generators Service Team Tel 1-800-Help365 Email [email protected] Select 63 at www.HydrocarbonProcessing.com/RS

LNG ment in offshore FLNG projects. The N2 expander design is simple and flexible to changes in feed gas composition, requires minimal operator intervention, and has good turndown capabilities. An important attribute is the ability to quickly start up and shut down in a safe and controlled manner.2 Each liquefaction train is driven by a refrigerant cycle compressor powered by an air-cooled aeroderivative gas turbine in which air cooling is provided by mechanical refrigeration in standard water chilling packages. The pressure of the liquefied product is lowered in a flash drum close to the LNG storage pressure, where N2 (being the lighter component) is flashed off and removed. The flashed liquid is then pumped into the LNG storage tanks. Fuel gas for the liquefaction system drivers and electrical power generation is generated as a mixture of lean-end flash gas (EFG), natural boiloff gas (BOG) from the LNG storage tanks and a supplemental fuel gas from the raw feed gas. Excess BOG/EFG, as well as vapor return during LNG transfer, is recompressed for reliquefaction. Note: The end-flash N2 removal process is well suited to feed gases with N2 content of up to 2 mol%, where, for a feed gas with more N2 content, there is a justification to remove the N2 content either before or during liquefaction. FLNG production capacity and performance. The nameplate LNG production capacity for the cylindrical FLNG unit is 2.45 MMtpy by using the largest-available gas turbines as

drivers in the liquefaction trains and a patented dual-N2 expander cyclea with integrated HHC removal. This is the rundown rate to LNG cargo tanks on a 365-stream-day basis. With corrections for boiloff loss, transfer loss and availability loss, the expected offloaded volume will be 2.25 MMtpy. The plant’s thermal efficiency, considering only the fuel gas consumption of the liquefaction system’s gas turbines, is 94%. Total facility efficiency, considering all power consumption in process systems, marine systems and domestic systems, is 92% (i.e., 8% of the feed is consumed as fuel, and 92% is converted to LNG and condensate product). Takeaway. The cylindrical FLNG production concept offers a safe, reliable and cost-efficient solution to the emerging FLNG market. It offers a more stable platform for the gas processing plant than the traditional ship-shaped vessels. For all areas where a turret and swivel arrangement for a ship-shaped vessel would be required, the geostationary cylindrical hull eliminates the need for a costly turret mooring and high-pressure swivel system. The cargo tanks arrangement has been optimized to use a maximum number of existing standard components for both the LNG and LPG storage tanks. The SPB cargo containment system provides better resistance to sloshing loads, while membrane technology leads to an optimum use of the space available. A large storage capacity is in line with production requirements, operational good practices and a reduced cargo tank weight. NOTE a This dual-N2 expander cycle is used as base case for the topside design and is based on the proven standard dual-expansion cycle, with optimization that improves efficiency, availability and operation/startup on large-scale N2 cycles. ACKNOWLEDGEMENT Thanks are due to Inga Bettina Waldmann of KANFA Aragon AS (Norway) for reviewing this manuscript and providing useful comments. LITERATURE CITED Mokhatab, S., W. A. Poe and J. Y. Mak, Handbook of Natural Gas Transmission and Processing, 3rd Ed., Gulf Professional Publishing, Burlington, Massachusetts, 2015. 2 Mokhatab, S., J. Y. Mak, J. Y. Valappil and D. A. Wood, Handbook of Liquefied Natural Gas, Gulf Professional Publishing, Burlington, Massachusetts, 2014. 3 Odeskaug, L., “The cylindrical hull concept for FLNG application,” OTC-25703, Paper presented at the 2015 Offshore Technology Conference, Houston, Texas, May 4–7, 2015. 1

TRI-CON

TRI-CHECK

TRI-BLOCK

LARS ODESKAUG is the deputy chief operating officer of Sevan Marine ASA in Norway, with more than 25 years of international experience in the oil and gas industry at the senior corporate and project management level. He was the managing director of Hitec Marine (Norway) from 1994–2002. From 2002–2005, he was the president of Remora Technology AS (Texas, US), and from 2005–2011 he was the CEO of TORP LNG AS (Texas, US). Mr. Odeskaug has given numerous presentations at international energy conferences and has published several articles related to innovation in the oil and gas industry.

WWW.ZWICK-VALVES.COM

THE NEXT GENERATION

TRI-SHARK

44

TRI-CONTROL

TRI-JACK

Select 156 at www.HydrocarbonProcessing.com/RS

SAEID MOKHATAB is an internationally recognized gas processing consultant who has been actively involved in several large-scale gas field development projects, concentrating on design, precommissioning and startup of processing plants. He has presented on gas processing technologies worldwide and has authored or co-authored nearly 250 technical publications, including two well-known Elsevier handbooks. He has held technical advisory positions in leading professional journals, societies and conferences in the field of gas processing, and has received a number of international awards and medals in recognition of his outstanding work in the natural gas industry.

Process Engineering and Optimization A. A. JAIN and A. GUPTA, Reliance Industries Ltd., Jamnagar, India

Choose the most appropriate modeling approach for reactors The efficient operation and design of any reactor depends on the know-how regarding the chemical reaction and the hydrodynamics inside the reactor. Mathematical models are developed to understand these complicated phenomena. Model analysis is carried out to understand the influence of various input parameters on the composition of products and the hydrodynamics inside the reactor. As shown in FIG. 1, such analysis, along with experimental investigations, plays an important role in the design and optimization of a commercial reactor. Various approaches to modeling reactors have been explored. These models can be divided into three general types: datadriven models, equilibrium/thermodynamic models and ratebased models. The approach to modeling a reactor depends on the understanding of the hydrodynamics inside the reactor, the information available with respect to the required kinetics or equilibrium, and the availability of the required input information. It also depends on the objectives set out before modeling the reactor through optimization studies, reactor design, scaleup studies, hydrodynamics studies, troubleshooting, etc. Each modeling approach has its own set of advantages and disadvantages that are explained here through the examination of a case study of a fluidized bed gasifier (FBG). This article is intended to be useful in developing an appropriate strategy to model reactors. Data-driven models. The results of data-driven models are based on data rather than on intuition or on the understanding of the phenomena. One of the advantages of these models is that no prior knowledge of the complex phenomena occurring inside the reactor is required for its formulation.1 Regression analysis and artificial neural networks (ANNs) are the more popular types of data-driven models. Methods based on regression analysis. Here, a data function is assumed. Given two input parameters, x1 and x2 , with an output of “y,” a few of the data-fitting functions generally assumed are: • Multiple linear form, y = a1x1 + b1x2 + f1 • Power form, y = a1x1b1x2c1 • Exponential form, y = a1 + b1x1c1 + d1x2e1 • Quadratic form, y = a1 + b1x1 + c1x2 + d1x1x2 • Mixed form, y = a1 + b1 × exp(c1 + d1x1 + e1x2).

In these models, the regression coefficients (a1 , b1 , c1 , d1 and e1 ) are extracted using various optimization routines, such as genetic algorithm, pattern search methods, simulated annealing, etc., by minimization of objective function (Eq. 1—yreal,i is ith experimental data, ysimulated is ith simulated results). The quality of fit is measured by root mean square error (RMSE) (Eq. 2). Objective function =∑

n i=1

( yreal ,i − ysimulated ,i )

2

(1)

n

Root mean square error (RMSE) =

∑i=1( yreal ,i − ysimulated,i )

2

n

(2)

Artificial neural network models. Artificial intelligence

(AI) technology systems are widely accepted and offer an alternative method to tackle ill-defined problems. They have been used in diverse applications in control, robotics, pattern recognition, forecasting, medicine, power systems, manufacturing, optimization, signal processing and social/psychological sciences. AI systems comprise areas like expert systems, ANNs, genetic algorithms, fuzzy logic and various hybrid systems that combine two or more techniques. The detailed overview of the algorithm scheme of ANN can be found in literature.2 Geometry • Reactor design • Internals design Hydrodynamics • Mass/heat/momentum transfer, etc.

Closure • Mass and energy balance equations

Kinetic studies • Heterogeneous reactions • Homogeneous reactions • Catalyst activation/ deactivation

Reactor model • Model analysis Design and optimization

Design Model verification Pilot plant

Operating experience

Commercial reactor

FIG. 1. A schematic representation of approach to reactor development. Hydrocarbon Processing | AUGUST 2016 45

Process Engineering and Optimization In an ANN, a certain percentage of the data (typically 70%– 75%) is used for training. From the remaining data, a certain percentage of the data (typically 15% each) is issued for validation and testing of the network. Sigmoid (input layer → hidden layer) and linear (hidden layer → output layer) transfer functions (Eq. 3) are used within the network. On this trained, validated and tested network, further simulations are performed to predict outputs for various cases. 1 S(t)= (3) 1+e (−t ) Equilibrium-based models. The overall mass and energy balance equations are solved using these models, and the main outcomes are the amount, composition and heating values of the product gases. These models neither take into account the reactor design nor the process and hydrodynamics inside the reactor. The model calculates the chemical equilibrium composition (most stable composition) by minimization of Gibbs free energy. One of the key advantages of this model is that no prior knowledge of complex reactions and hydrodynamics is required. It is computationally inexpensive and can provide a guideline for process design and evaluation. An equilibrium model can be used to study the influence of various operating parameters, as it provides qualitative trends. The two general approaches for equilibrium models are stoichiometric and nonstoichiometric. While the two approaches are essentially equivalent, the fundamental difference is that stoichiometric models assume a clearly defined reaction mechanism that incorporates all the chemical reaction and the major species involved, while the non-stoichiometric equilibrium model is based on the minimization of Gibbs free energy, where no specific assumptions on the chemical reactions are made. The elemental composition of the fuel, which can be obtained from the ultimate analysis and the operating conditions, Product gases “out”

Particle level

Cyclone

Solid-gas reactions, pyrolysis, combustion, gasification

Freeboard

Particles/fly ash

Bubbles Fluidized bed Solids feed Bottom ash Inlet feed gas (air/O2/steam)

FIG. 2. A schematic representation of a bubbling fluidized-bed gasifier.

46 AUGUST 2016 | HydrocarbonProcessing.com

is only required as input to these models. Results of equilibrium models have shown good agreement with experimental data of certain reactors because the general assumptions of the models are in agreement with the actual conditions in the reactor. The general assumptions for the equilibrium model are that reaction rates are fast, residence time is long enough to reach equilibrium state, and the reactor is zero dimensional. Rate-based models. These models are more realistic compared to the data-driven and thermodynamic models. They take into account the fluid dynamics, transport processes and chemical reactions inside the reactor. As shown in FIG. 1, depending on the system, modeling of the chemical reactions involves consideration of heterogeneous and homogenous reactions. Two types of rate-based models exist: chemical reaction engineering models (CRE) and computational fluid dynamic (CFD) models. The mass and energy balances are solved in both models. In the CRE model, the momentum equations are not solved; instead, semi-empirical correlations are used to describe the fluid dynamics (bubble diameter, bubble velocity, bubble voidage, velocity of gas in emulsion phase, etc.). In a CFD model, the momentum equations are explicitly solved. Modeling chemical reactions (source terms) are common to both models. Case study. Various prominent coal gasification technologies3 have been developed and are used worldwide, including moving-bed, fluidized-bed (bubbling and circulating fluidizedbed) and entrained-bed gasifiers. The moving bed and fluidized bed, among others, are considered more apt for handling highash coal.4 A fluidized bed has certain advantages over a moving bed in terms of scaling up and environmental issues. Movingbed gasifiers generate tarry products, whereas a fluidized-bed gasifier yields only gaseous product as the volatiles are cracked, facilitating more environmentally friendly products and easier plant operation.5 The advantages of a fluidized-bed gasifier are well documented.6 Good gas-solid contact, excellent heat transfer characteristics, better temperature control, large heat storage capacity, a good degree of turbulence and high volumetric capacity are a few prominent advantages. A schematic representation of a bubbling fluidized-bed continuous gasifier (BFBG) is shown in FIG. 2. Typically, in a BFB continuous gasification of coal, the coal particles (0 mm–5 mm) are continuously fed into the BFB reactor at a point above the gas distributor. These particles react with the gasifying, and fluidizing, agents—steam, air/oxygen, carbon dioxide (CO2 )— to produce gases that are primarily composed of syngas—i.e., carbon monoxide (CO) and hydrogen (H2 )—at temperatures above 800°C. During this operation, multiple reactions (gas-solid heterogeneous and gas-gas homogenous) occur simultaneously in the reactor. The heterogeneous reactions consist of gas-solid reactions, such as pyrolysis, char combustion and char gasification, while the homogenous reactions consist of the water-gas shift reaction, H2 combustion, etc. The product gases coming out of the bed are passed through a cyclone separator for separation of elutriated solids. The product gases are further treated using a gas cleaning system. The superficial gas velocity varies between three to nine times the minimum fluidization velocity. The key challenges in the modeling of a BFBG are complex reaction ki-

Process Engineering and Optimization netics, gas-solid fluid dynamics and particle behavior. The two main output parameters that gauge the performance of a gasifier are: Inlet carbon – Outlet carbon Inlet carbon Chemical energy Cold gas efficiency (CGE) = Coal energy Carbon conversion (X) =

Chemical energy (MJ/kg) = y H2 141.80 + y CO10.08 + y CH4 55.58 Coal (thermal) energy (MJ/kg) = 33.855°C + 144.9H + 10.5S

Here, C, H and S = weight % content, dry basis; and y is the mass fraction of gases. A set of 25 experimental data obtained for a high-ash Indian coal in a pilot-scale setup were simulated using the various modeling approaches. Apart from the operating conditions of the gasifier, the main input parameters for a gasifier model are the ultimate and the proximate analysis of the coal (C, H, N, S and O, volatile matter, fixed carbon, ash and moisture). Regression analysis. The experimental data could not fit into

any of the regression models, and the relationships between the input and output parameters were too scattered to fit any of the models. Similarly, regression-based models (multiple linear regression, power regression analysis) with six inputs from a dataset of 106 experiments (fixed carbon, volatile matter, mineral matter, air feedrate/kg of coal, steam feed/kg of coal and tem-

perature and output in these cases were the rates and heating values of the product gases) were developed.7 The model results did not yield a good match with the experimental data. Artificial neural network. An ANN model (FIG. 3) that indicated a good match with the experimental data was developed. The model was later used for sensitivity analysis (effect of carbon content in coal, effect of temperature, etc.) that did not show the expected results, even qualitatively. The lack of anticipated results is due to the lack of a large set of experimental data for training the ANN model. Similarly, a neural network was created8 with eight inputs with a single hidden layer, and output of CO, CO2, H2 and methane (CH4). The network was trained and validated with 18 experimental data obtained from literature for a BFB wood gasifier. It indicated that the developed network showed a good match with the experimental data, although no results of sensitivity analysis using the developed model were discussed in the work. To train an ANN model to be used for optimization studies, an exhaustive set of data is required. ANNs cannot be used to study effects of parameters if they are out of the range of the training data and, therefore, are not useful for scale-up studies. Equilibrium model. Simulation results with a non-stoichiometric approach for calculating the equilibrium compositions were completed using a commercial process simulation software. The results showed an error margin of between 20% and 40% (FIG. 4) with respect to experimental CGE, while the carbon

Select 157 at www.HydrocarbonProcessing.com/RS

Hydrocarbon Processing | AUGUST 2016 47

Process Engineering and Optimization conversion calculated at equilibrium was always 100%. These results did not match with the actual experimental results, while the carbon conversion ranged between 60% and 90%. Although the model did not match the actual experimental data, it gave a fair indication of the operational limits and the qualitative change in the outlet gas composition and generation rate with changes in various operating parameters. Due to the high temperature at which entrained-flow gasifiers operate, they have shown gas composition nearly equal to the equilibrium compositions.10 For a BFBG, the modified equilibrium models may prove to be useful, as shown in the results of previous studies.11 Input layer (i)

C H

Hidden layer

Output layer

Weights

N

Wj,i

S O

Wk,j

VM

X

j=1 k=1

FC M

CGE

j=2 k=2

T O St

FIG. 3. ANN model structure to predict carbon conversion (x) and cold gas efficiency (CGE) from coal gasification in a BFB gasifier. 1.0 CGE (CRE) ± 20 error CGE (EM) ± 40 error

0.9

Simulated results, CGE

0.8

0.7

0.6

0.5

0.4 0.3 0.3

0.4

0.5

0.6 0.7 Experimental results, CGE

0.8

0.9

FIG. 4. A schematic parity plot comparing the errors obtained for the simulation of 25 experiments with a CRE model and the equilibrium model12 adapted for CGE.

48 AUGUST 2016 | HydrocarbonProcessing.com

1.0

Rate-based models. A CRE model based on the two-phase

theory12 showed a good match with the experimental data (FIG. 4), with < 20 % error. The two-phase theory assumes that the fluidized bed is divided into the emulsion and bubble phases. The emulsion phase consists of solids and the volume of gases required to keep the solids at minimum fluidization velocity (Umf ), while the bubble phase consists of extra gases flowing through the bed. Other theories exist to model a BFBG. They estimate the fluid dynamics (bubble diameter, bubble velocity, bubble voidage, velocity of gas in emulsion phase, etc.) using different correlations. Nine chemical reactions (four heterogenous and five homogenous) have been considered in this model. The kinetics for these chemical reactions have been taken from data reported in literature.13 Experimental studies using TGA, autoclave, a fixed-bed reactor, etc., can be conducted to estimate the kinetics for a particular coal. The temperature profiles and the hydrodynamic profile inside the reactor showed a good match with the experimental results. This model was then used for sensitivity analysis (including pressure effects) and scale-up studies. In the past decade, several CFD models for various reactors have been developed. In the case of a BFBG, based on how the solids are treated, two major CFD models exist: if the solid has been assumed as a continuum, then the Eulerian framework is applied to describe the motion of the solids; and if the solid particles are individually tracked, then the equation of motions are solved to track the motion of particles. The gas phase is described by the Eulerian framework and modeled similarly to single-phase flow, wherein interaction with the solid phase is accounted for by an additional term. The two approaches are termed as Eulerian-Eulerian (EE) and Eulerian-Lagrangian (EL) models.14 The source terms (kinetics) are the same in both the CRE and CFD models, but the main distinction is the handling of the fluid mechanics. In the EL model, the solid phase exchanges mass, momentum and energy with the gas phase, where each individual particle is solved in a Lagrangian frame of reference. The discrete element/particle method inspired by molecular dynamics is commonly applied. In a DEM/ DPM model, the collision between particles may either be based on a hard-sphere or soft-sphere approach.15 Closure relations in these models are not simple; several equations with semi-empirical parameters must be solved simultaneously. A comparative analysis of CFD models for a BFBG can be found in literature.14 A CFD model was not formulated due to two main reasons: it was too computationally intensive, and it involved an extended simulation time (a BFBG involved complex gas solid hydrodynamics coupled with several chemical reactions). A CRE model proved a robust computational tool in attaining the objectives for design/scale-up/optimization studies of a BFBG. In literature, researchers have simulated a bench-scale BFBG using CFD and have shown a good match with the experimental data. The simulation time for CRE models is significantly less when compared to CFD models. The coupling of CRE models with CFD models—wherein the insights (flow and mixing knowledge) obtained from a CFD model are utilized in quantifying flow and mixing in a CRE model—have been reported.16 Similarly, the use of ANNs to estimate the kinetics in a rate-based model has also been reported. A summary of the advantages and disadvantages of the various approaches to model the reactor are listed in TABLE 1.

ADVERTORIAL

Shell Global Solutions

WHERE COULD YOU BE LOSING MONEY?

In recent years, while working with customers around the globe, we have identified several areas in which businesses like yours could find further opportunities to drive out costs and capture additional margin. 1. COST BASE “The pressure to reduce costs is relentless. However, we lack the tools to monitor and control energy, utility and waste management costs fully, and I do not believe that our staff are fully aware of the effects of their actions on plant profitability. Consequently, we regularly exceed our operating budget.” Realistic targets are vital. It may be useful to evaluate how your operation performs compared with other similar assets; this can provide a road map for future improvement projects. Furthermore, there are tremendous insights to gain from seeking out best practices and key lessons from other refineries. 2. MAINTENANCE AND AVAILABILITY “Any attempt to enhance the reliability of our assets severely affects our cost reduction drives. My mandate is to cut costs and simultaneously improve reliability, but these seem to be conflicting objectives.” The industry’s high performers in this area are using various tools and techniques to strike a balance between good, safe maintenance practices and higher margins. These include introducing risk-based rather than time-based processes that focus on the most critical areas so that you can plan your preventive maintenance to achieve the desired plant availability for optimal effort. 3. UNIT UTILISATION “In our part of the world, the market for middle distillates is booming and the margins are high, but throughput constraints on our hydrocracker mean that we cannot take maximum advantage of this, and we do not have the capital required to build a new unit.”

reactor internals, can often unlock yield improvements for relatively minor capital expenditure. 4. CRUDE FLEXIBILITY “Of course there are cheaper crudes out there, but they are heavier, sourer and more contaminated. I have concerns about their impact on our assets, so I will only process the conventional crudes that we have tried and tested.” Because crude feedstock is such a large component of a refinery’s costs, there are substantial benefits from replacing a small part of the crude slate with alternative, cheaper crudes. Before changing the crude diet, it is essential for a refiner to take advice on key issues such as hardware constraints and crude sourcing, but this can be an effective way to cut costs for minor capital expenditure if the risks are mitigated.

In today’s volatile refining sector, cost reduction and margin improvement have become business imperatives for many refiners. By applying tried-and-tested techniques, and successful practices from other refineries around the world, you may be able to identify opportunities to maintain or enhance your competitiveness. To learn more, please visit:

In recent years, there has been a strong trend in the industry towards revamps rather than grassroots capital projects, chiefly because they offer an enhanced return on investment. One of the simplest types of revamps, installing higher-activity catalysts and latest-generation Select 92 at www.HydrocarbonProcessing.com/RS

www.shell.com/rapid

Process Engineering and Optimization TABLE 1. Approaches to model a reactor (Case study: BFBG, adapted)17 Approach

Essential aspects

Results

Advantages

Disadvantages

Types

Data-driven models

No governing equations

Outlet gas composition, temperature and carbon conversion

No understanding of any complex process in the reactor is required

1.  No insights into the complex relationships between the input and output parameters 2. Quality and quantity of data is required for building a robust model

1.  Regression analysis 2. Artificial neural networks (ANN) 3. Fuzzy rule-based systems (FRBS), etc.

Thermodynamic models

1.  Based on Gibb’s minimization theory 2. Calculates equilibrium gas composition and temperature

Outlet gas composition, temperature and carbon conversion

No understanding of any complex process in the reactor is required

No insight into the hydrodynamics of the reactor

1.  Equilibrium models (EM) 2. Modified-equilibrium models (MEM), etc.

Chemical reaction engineering (CRE) models

1.  Based on first principles, solves the mass and energy balance equations 2. Instead of solving the momentum equations, semi-empirical correlations are used 3. Ideal reactors/ combination of ideal reactor assumptions for phases (CSTR, PFR, compartments)

Profile of gas/ solid species composition, temperature and hydrodynamics across the reactor

1.  Computationally less intensive than CFD models 2. Can be used to study the influence of various input parameters on the performance of the reactor 3. Gives sufficient details for engineering applications

Flow structure and the range of applicability depends on the correlations used

1.  Davidson-Harrison model (DHM) 2. Kunii-Levenspiel model (KLM), etc.

Computational fluid dynamic (CFD) models

1.  Mass, energy and momentum balance equations are solved 2. Constitutive relations and closure laws are adopted

Profile of gas/ solid species composition, temperature and hydrodynamics across the reactor

Useful for exploring in-depth hardware details

1.  Computationally very expensive and timeconsuming solution 2. Uncertainty of various parameters 3. Availability of in-depth experimental data to validate the model results

1.  Eulerian-Eulerian model (EEM) 2. Eulerian-Lagrangian model (ELM), etc.

Conclusions and summary. The relative advantages and

disadvantages of various approaches to model a reactor were critically analyzed, and suggestions for potential improvements in these models have been presented. An ANN is useful when a large amount of quality data is available to train and test the neural network. It is also useful when sufficient details of the hydrodynamics and the kinetics involved in the reactor are not known. It is not useful for scale-up studies of a reactor. For reactors where high residence time and rapid chemical reactions are seen, an equilibrium model may be sufficient to model the reactor. In the case of a BFBG, an equilibrium model did not show a good match with experimental data, but it did provide a fair indication on the operational limits. It is also useful in understanding the qualitative change with alterations in input parameters. Results with rate-based models, where the hydrodynamics and the chemical reactions in the reactor are taken into account, best described the experimental data of a BFBG. These models provide sufficient knowledge for the scale-up and optimization of the reactor. CFD models are computationally intensive, especially for the simulation of commercial-scale reactors. However, with the availability of cheap computational power, use of this tool to simulate reactors is increasing. Models where flow and mixing knowledge obtained from a CFD model are utilized in quantifying flow and mixing in a CRE model have been reported in literature. 50 AUGUST 2016 | HydrocarbonProcessing.com

LITERATURE CITED Sözen, A., E. Arcaklioğlu and M. Özkaymak, “Turkey’s net energy consumption,” Applied Energy, Vol. 81, No. 2, 2005. 2 Kalogirou, S. A., “Artificial intelligence for the modeling and control of combustion processes: A review,” Progress in Energy and Combustion Science, Vol. 29, No. 6, 2003. 3 Gräbner, M., Industrial coal gasification technologies covering baseline and high-ash coal, Wiley-VCH Verlag GmbH & Co., Weinhein, Germany, 2015. 4 Collot, A. G., “Matching gasification technologies to coal properties,” International Journal of Coal Geology, Vol. 65, 2006. 1

Complete literature cited available at HydrocarbonProcessing.com. ANKIT A. JAIN is a research scientist in the refining research and development department of Reliance Industries Ltd. His current area of research is the modeling of multiphase flows and reaction engineering, and he also has experience working as a process design engineer. Dr. Jain earned his BS degree and PhD in chemical engineering from the National Institute of Technology (NIT) at Surat and the Indian Institute of Technology (IIT) at Bombay, respectively. AJAY GUPTA is the assistant vice president and head of the fixed-bed process development group of refining research and development at Reliance Industries Ltd. in Jamnagar, Gujarat, India. He earned his BS, MS and PhD degrees in chemical engineering from the Indian Institute of Technology (IIT) at Delhi. His experience includes the modeling and simulation of chemical processes; implementation of advanced process control strategies in fluid catalytic cracking (FCC) and crude distillation units in several refineries; development of reactor models for various chemical processes, including effect of scale; and the application of computational fluid dynamics for solving problems in the petroleum refining and petrochemical industries.

GasProcessingConference.com

AMERICAS

September 13–14, 2016 Norris Conference Centers – CityCentre Houston, Texas

Explore the latest gas processing developments, from field gathering to plant design to LNG technology, at the 2016 GasPro Americas Conference Register today for the must-attend event for the gas processing industry Join us at the GasPro Americas Conference, scheduled September 13–14, 2016 at Norris Conference Centers - CityCentre in Houston, Texas, USA. The 2016 event has 18 sessions organized into three tracks that will benefit CEOs, COOs, CTOs, directors, country managers, regional directors, project managers, technical directors, and heads of engineering for technology and operating companies. This conference will focus on gas supply, procurement, purchasing, transportation, trading, distribution, operations, safety, the environment, regulatory affairs, technology development, business analysis, LNG, and more. All segments of the gas processing industry—the upstream, midstream and downstream sectors—will be discussed. Session topics include: • The State of Natural Gas in the Americas • Water Treatment • Gas Treating • Separation/Dehydration • Cryogenics: Rejection, Ethane, Methane and Nitrogen • Syngas Production and Utilization

USB Sponsor:

• IOT and the Future of Big Data for the Natural Gas Industry • NGL Recovery • Fractionation • LNG/FLNG • Liquefaction/Regasification • Metering/Custody

Host Sponsors:

Save 15%

on Registration

when you use code: HP16

Operators eligible for a free pass! Contact Melissa Smith, Events Director, at [email protected] or + 1 (713) 520-4475 for details.

To attend, sponsor or exhibit, visit GasProcessingConference.com

Exhibitors:

Supported by:

Industry Insights Knowledge Sharing Access to Experts

Tuesday, September 13 Agenda 7:30–8:30 a.m.

Registration and Continental Breakfast

8:30–8:35 a.m.

Opening Remarks: Adrienne Blume, Executive Editor, Hydrocarbon Processing and Editor, Gas Processing

Session 1: The State of Natural Gas in the Americas 8:35–10 a.m.

Speakers Include: Lee Nichols, Editor-Associate Publisher, Hydrocarbon Processing and Construction Boxscore Database Anne Keller, Manager, NGL Research, Wood Mackenzie

10–10:30 a.m.

10:30– 11:45 a.m.

11:45 a.m.–12:45 p.m.

12:45–2 p.m.

Coffee and Networking Break

Session 2: Condensate Removal

Session 3: Water Treatment / HSE

Speakers Include:

Speakers Include:

Haiming Jin, Senior Process Engineer, SNC Lavalin

Chuck Miller, Director-Global Natural Gas Processing Sales & Marketing, Emerson Process Management

Guof Chen, Senior Process Engineer, Joule Processing

2:30–4 p.m.

Dewitt Dees, Chief Executive Officer, RWL Water North America

Session 5: Gas Treating

Session 6: Cryogenics: Rejection - Ethane, Methane, Nitrogen

Session 7: Flaring / Emissions

Speakers Include:

Speakers Include:

Speakers Include:

David B. Engel, Managing Director, Nexo Solutions

Scott Schroeder, PE, Senior Technical Consultant – Gas Processing, Advisian

Martin Dean Layfield, Global Segment Leader, Gas Value Chain, DNV GL

Stephen Morgan, Sales Manager, Chart Industries

Donald Kendrick, Senior Vice President of Technology, ClearSign Combustion Corporation

Jennifer Dyment, Marketing Manager, AspenPlus®, AspenTech

Rafael Aguilar, Senior APC Engineer, Honeywell Process Solutions

Coffee and Networking Break

Session 8: Separation / Dehydration

Session 9: Syngas Production and Utilization

Session 10: IOT and the Future of Big Data for the Natural Gas Industry

Speakers Include:

Speakers Include:

Speakers Include:

Shabbir Husain, Senior Process Engineer, Chevron Energy Technology Company

Esben Lauge Sorensen, Syngas Technology Specialist, Haldor Topsoe, Inc

Hong Qin, GHG Data Compliance Lead, Wood Group Mustang

Laura Aiken, Project Engineer, Bechtel Corporation, USA

Robert Schuetzle, Chief Executive Officer, Greyrock Energy

Ray Ozdemir, Framergy, Inc 4 p.m.

Steve Arendt, Vice President, Global Oil, Gas & Chemicals Sector, ABS Group

Lunch

Scott Alvis, Vice President, Business Development, Optimized Gas Treating, Inc 2–2:30 p.m.

Session 4: Best Practices / Operations / Project Management

Christopher Molnar, Vice President, Industrial Heaters & Systems, Chromalox Marcus Wagner, Chief Executive Officer, AcctTwo

Closing Remarks: John Royall, President and Chief Executive Officer, Gulf Publishing Company

Register at GasProcessingConference.com

Trends & Developments Networking Best Practices Past Attendees Include: • • • • •

Chevron WorleyParsons Group Linde Process Plants Deloitte Stabilis Energy

• • • • • • • •

Chart Energy & Chemicals Bechtel Corporation Optimized Gas Treating Emerson Framergy Enbridge Atlas Copco SNC Lavalin

• • • • • • • •

Advisian Sterne Agee DNV GL Honeywell UOP ClearSign Combustion Nexo Solutions Chromalox AspenTech

• Greyrock Energy • Alta Mesa Services • Williams • GasTech Engineering LLC • DOW Chemical • SGC Energia • BP ...to name a few.

Wednesday, September 14 Agenda 7:30–8:30 a.m.

Registration and Continental Breakfast

8:30–8:35 a.m.

Opening Remarks: John Royall, President and Chief Executive Officer, Gulf Publishing Company

Session 11: The Future of LNG in America 8:35–10 a.m. 10–10:30 a.m.

10:30 a.m.– 12:20 p.m.

Speakers Include: Andrew Slaughter, Executive Director, Deloitte Center for Energy Solutions, Deloitte Jason Feer, Global Manager, Poten & Partners Coffee and Networking Break

Session 12: LNG / FLNG

Session 13: Fractionation

Session 14: NGL Recovery

Speakers Include:

Speakers Include:

Speakers Include:

Steve Stump, President, Stabilis Energy

Dr. Rameshwar Hiwale, Principal Process Engineer, Linde Process Plants, Inc

Joseph Lillard, Engineering and Product Manager, Atlas Copco

Jason M. Manning, Principal Process Engineer, Black & Veatch Maurice “Zickie” Allgrove, Vice President-Transport, America, Advisian

Sachin Joshi, Technology Process Manager, GTC Technology US, LLC

Ben Adamson, Engineering Director, Refrigeration Engineering International Pty Ltd Kevin Currence, Technology Manager – Gas Processing, Black & Veatch Danielle Goldbeck, Process Engineer, Linde Engineering North America, Inc

12:20–1:20 p.m.

1:20–2:35 p.m.

Lunch

Session 15: Liquefaction / Regasification

Session 16: Metering / Custody

Session 17: Alternative Applications / GTL

Speakers Include:

Speakers Include:

Speakers Include:

Thomas Quine, President, Northstar Industries LLC

Daniel Hackett, Business Development Director-Ultrasonics, Emerson Process Management

Robert Walsh, Senior Vice President, Intrexon Corporation

George Arnett, Systems Sales Manager, Chart Energy & Chemicals Dr. Delmar “Trey” Morrison, Principal Engineer, Exponent, Inc

2:35–3:05 p.m.

Seshasai (Sai) V. Yesantharao, Solutions Architect, Honeywell Advanced Solutions

Coffee and Networking Break

Session 18: Policy: The Importance of Legislative and Regulatory Compliance 3:05–4:30 p.m. 4:30 p.m.

Speakers to be announced Closing Remarks: John Royall, President and Chief Executive Officer, Gulf Publishing Company

Register at GasProcessingConference.com

Registration to the two-day conference includes: • Access to keynote addresses and technology presentations organized in three tracks • Breakfasts, lunches and refreshment breaks • Access to sponsors and exhibitors throughout conference • Conference proceedings (post-event)

Save 15%

on Registration when you use code:

Admission Rates for Single Attendee

Early Bird (by August 16)

Regular Admission

$500

$600

Prices listed above are in $USD. Group discounts available. Contact Melissa Smith for details.

HP16

Want a free pass to GasPro? Employees of operating companies, E&P companies, or refiners may be eligible to attend this event for free. To see if you qualify for a free pass, contact Melissa Smith, Events Director, at [email protected] or + 1 (713) 520-4475.

Sponsorship and Exhibit Opportunities The 2016 GasPro Conference provides an excellent forum to connect face-to-face with top operators and technology leaders in the gas processing industry. Benefits of sponsoring/exhibiting include: • Increase brand awareness • Network face-to-face with decision makers • Attend the conference sessions for free • Generate new sales leads Contact Melissa Smith, Events Director, at [email protected] or + 1 (713) 520-4475.

Register at GasProcessingConference.com

Process Engineering and Optimization J. C. GENTRY and M. BHARGAVA, GTC Technology US LLC, Houston, Texas; and M. J. BINKLEY, GTC Process Equipment Technology, Euless, Texas

Distillation—Then and now Distillation is the most often used means to separate two or more components, exploiting the physical properties of different boiling points. Separation by distillation is completed in stages with chemical equilibrium at each point in the process to build a composition profile. The products are taken out at the peak of the composition, usually at the top and bottom of a tower, with intermediate levels of composition between the extremes. Separating components to increase value. The main reason systems are designed to separate components is that separated components have more value than the mixture of components. The raw material of hydrocarbons is a mixture, as in crude oil or natural gas condensate. Chemical intermediates and reaction products will have a mix of byproducts or unreacted components. Separating these byproducts or components permits efficient recycle or use of products as purified substances. Another reason for separating components is to prepare a feed for further processing. An example is removing lights and heavies from an isomerization unit feed. While separations are used in virtually all chemical process systems, distillation is the most prominent. It requires a large consumption of energy and capital expenditure, but it remains the most widely used separation technique because it is well-understood and proven in a wide spectrum of applications.1 Many academics have developed heuristics—a hands-on interactive approach to learning that enables a person to learn something for themselves—for designing separation systems, and others have proposed algorithms for finding the perfect method to design a separation system. These methods are useful for training purposes, but often get lost or overlooked within the urgent schedule of producing a process design.2,3 Most designs are created with a series of practical constraints, such as: • Project schedule • Available plot space • Flexibility to handle feedstock variations • Easy understanding and control of operations. With this in mind, some simplified ideas are proposed here that build upon previous work, but which may be more practical and useful to today’s designers.

Maintain component separation. It seems obvious to fol-

low the basic principle, “Once components are separated, do not remix them again.” However, numerous commercial operations do, in fact, have remixing embedded in the process. One classic example was a C8 aromatics processing plant that used super-fractionation to separate the ethylbenzene and pxylene (α EB:pX = 1.06). The feed to this unit originated from three types of processes: • Hydrogenated pyrolysis gasoline comprising approximately 60% EB • Catalytic reformate comprising approximately 18% EB • Aromatics transalkylation comprising 1%–2% EB. The source material also had benzene/toluene (B/T) fractions, combined in upstream units and operating in parallel to remove the BT cut. The material then proceeded to a series of two major EB fractionators with a total number of 600 trays to remove the ethylbenzene. The first system of three towers in series concentrated the stream to approximately 50% EB purity. The final two towers in series produced an EB product of 99.3% purity (FIG. 1). Upon examination of the fundamental separations taking place in the unit, the stream value was downgraded by the upstream mixing, as the most difficult components to separate were equalized by combining the streams containing different Pygas, 60% EB

50% purity

Reformate, 18% EB

300 trays #1

300 trays #2

Xylenes

Xylenes

99.3% EB

TDP, 1.5% EB

FIG. 1. Key components are remixed in a series of two major EB fractionators, with a total number of 600 trays to remove the ethylbenzene. Hydrocarbon Processing | AUGUST 2016 51

Process Engineering and Optimization EB content. These streams were separated again at substantial cost. The remedy was to simply segregate the streams from an upstream fractionation train, permitting complete shutdown of one of the major super-fractionation towers (FIG. 2). In older plants, process designs are often created out of expediency and morphed into an inefficient configuration after the original construction. Evaluating complex systems helps eliminate instances of separation that lead to remixing. Complete easy separations first. One of the classic heuristics advocated by most scholars, this advice follows common sense.4 It is natural to follow the path of least resistance. In a scientific view of multi-component distillation, the “easy first” approach is described by a pre-fractionation system followed by a main fractionation system. The pre-fractionation system removes the “clutter,” placing the extraneous components into proper zones better suited for broad separations; Pygas, 60% EB

Reformate, 18% EB

300 trays #1

300 trays #2

Xylenes

Xylenes

99.3% EB

TDP, 1.5% EB

FIG. 2. Segregating the streams from an upstream fractionation train avoids remixing.

A A,B

B A,B,C

B,C

C Prefractionator arrangement FIG. 3. A prefactionator arrangement places the extraneous components into proper zones more suited for broad separations.

52 AUGUST 2016 | HydrocarbonProcessing.com

meanwhile, the main fractionation system will separate the more difficult components with the closest relative volatility (FIG. 3). The conventional approach to the three-component separation is to carry out the separations in a series operation using two fractionation columns. The obscure downside to this approach is the “back-mixing” of components that takes place in the bottom of the first column (FIG. 4). The middle boiling product is concentrated at an intermediate point within the column, only to be downgraded to a lower concentration at the bottom of the column. Therefore, components that were previously separated have been recombined. A convenient way to circumvent this thermodynamic inefficiency and follow the concept of prefractionation is to use a dividing wall within the column to separate the vessel into different functional zones (FIG. 5).a Dividing wall column (DWC) distillation is not new to the field of chemical engineering; it traces its roots to 1935. However, DWC technology has failed to achieve popularity due to a lack of understanding about what is happening inside the column, difficulty in simulating the designs and general inertia in adopting new technologies. With better understanding, faster computer simulators and robust mechanical designs, this hesitancy is gradually changing. Consider functional separation. Many applications can benefit from DWC technology and alternative types of thermal coupling. The principle of a DW to create separation zones presents other options for operations inside a column. Consider the case of liquefied petroleum gas (LPG) recovery from refinery fuel gas. The two conventional approaches to remove C3 (or C3/C4) from a mixture of C1 to C5+ components are distillation with refrigeration, and absorption into a heavy oil followed by stripping of the gases from the absorbing material. Refrigerated distillation has an additional cost for the refrigeration duty, high-pressure operation and expensive equipment. Absorption followed by stripping requires two columns to make the separation. The absorption section also builds a concentration peak of the intermediate component, only to downgrade the concentration that was achieved by mixing with the heaviest components (in violation of “no remixing”). This process has introduced a new method for making the absorption and stripping in the same column by using a dividing wall. In this case, the wall separates the top of the column into two zones: absorption and stripping. The column operates at moderate pressure, with cooling water as the cold utility. The absorption oil captures the C3+ components and allows the C2 components to escape as a vapor without condensing. The C3 (or C3/C4) components are brought to the bottom of the wall near a peak in composition, where they are passed to the stripping section of the column on the other side of the wall for removal into the overhead as a condensed liquid product. Ideally, sufficient heavy components in the feed stream that can be recycled to the top of the absorption zone are available so that no external feed material is used. As shown in FIG. 6, this method is applied to LPG recovery.b The general concept of absorption and stripping in the same column with a TDW can be applied where a non-condensable stream is mixed with a heavy oil product, such as a hydroprocessing stabilizer.c Different variants of the concept can be used in revamp situations of existing two-column systems.

Process Engineering and Optimization

Tray number

Column 2

Column 1

Functional separation extended. Thermodynamic inefficiency as component B is concentrated, then Continuing with the concept of keeping diluted within the first column and separated again in a second column 0 components separated, certain applicaA tions exist where two streams are com5 bined into a common fractionation sys10 Feed tem out of convenience to separate some 15 (A,B,C) B light and heavy components. These 20 grouped fractions may not have the same 25 composition and would be more valuB 30 able if kept segregated. Rather than have (B,C) 35 two separate fractionation systems, a top Remixing 40 dividing wall column (TDWC) can be 45 used to segregate the streams in the same 0 0.2 0.4 0.6 0.8 1.0 column into two overhead products with Composition: Mole fraction C a common bottom product. Remixing leads to thermodynamic inefficiency One example is xylene fractionation within a paraxylene complex. If the FIG. 4. Back-mixing in the bottom of the first column is one of the downsides of separations paraxylene recovery section is based on in a series operation. selective adsorption, then a strict requirement exists to keep the C9+ components out of the C8 A fraction. Multiple feeds enter into a common xylene tower to accomplish the C9+ removal. However, the C8 fractions will have different levels of EB or pX comingled together inside the tower, only to be separated again in the adsorption section of the plant at a higher cost. B If the xylene column was designed with a TDW, then the streams with low EB (or high pX, for example) could be fed to AB a different entry point in the adsorption system to reduce the ABC B separation energy and debottleneck that part of the plant. Still another application of functional utility within a distillation column is to exploit the potential for heat recovery at BC a higher temperature in multi-component distillation. Typically, a three-cut distillation tower using a conventional sidedraw or DW with side-draw will have the lightest component B at the top of the column. The latent heat from condensing duty will be made at the equilibrium temperature of this component at column pressure. Often, this pressure is too low for recovering useful heat. The side-draw will have only sensible C heat to transfer at the intermediate temperature. FIG. 5. A DW within the column is used to separate the vessel into One strategy using a TDWC is to intentionally skew the different functional zones.a temperature profile, such that the intermediate cut can also be taken as a distilled product, with latent heat available at a higher temperature (FIG. 7). • Manufacturing—Despite the most clever process design, manufacturing flaws or installation errors can completely Here, the feed side will distill the top product at low negate the benefits of the design. Trusting the party overhead temperature, while the intermediate cut will be controlling the entirety of the process design, hydraulic taken out as overhead from the opposite side of the wall at a design, manufacturing and installation is recommended. higher temperature. The enthalpy from the mid-fraction can • Tray spacing and valve type—Some of the earliest types be used to preheat the main column feed, or to reboil a sepaof contacting devices were bubble cap trays, which were rate distillation system to save overall energy. prominent in the 1940s and 1950s. In the 1960s, trays with moveable valves came into service, affording higher Back to the basics. Stepping back again into history, the bacapacity and fouling resistance. sic principles of distillation column design must always be fol• Installation—Improved modern tray designs have lowed, even with advanced schemes. These principles include: many different features, such as optimized active area • Optimization of feed location—This avoids pinch in the and downcomer area, flow promotion, elimination of stripping vs. rectification zones. liquid gradient and bubble promotion. The trays may • Balancing the NTS vs R:D ratio—Rules of thumb, such be installed at closer spacing, and there may also be as the recommended R/D being 1.2 × minimum R/D, DWs. High-performance trays require more care depend on capital costs vs. long-term energy costs and during inspection and installation. could change over time. Hydrocarbon Processing | AUGUST 2016 53

Process Engineering and Optimization

Top liquid product Feed

Absorption zone

Bottoms product

Bottom product is the absorbing medium

FIG. 6. C3 (or C3 /C4) components are passed to the stripping part of the column on the other side of the wall for removal into the overhead as a condensed liquid product. This method is applied to LPG recovery.b

Lowest temperature

Feed (A,B,C)

A

B

Intermediate temperature for heat integration

 %Flood VLoad + [GPM× (FPL/13,000)] =  (1) 100 AA ×CAF Winn combined the classic entrainment flooding “C-factor” of Souders-Brown with tray spacing and liquid load variables. Details of the Souders-Brown entrainment model are thoroughly discussed by Henry Z. Kister.6 Equation 13 is the most reliable flood correlation among the classics. Recently, Resetarits and Ogundeji tweaked Equation 13 for a slight improvement.7 Their conclusion completely supports the fundamentals of the original F. W. Glitsch model. Findings. Fractional distillation is a common-sense approach to separate chemical components. Much can be learned from the experience of industry predecessors. Many of the original design principles apply, as the basic mass transfer operations do not change. New techniques include: • The concept of an “outside-in” approach to purifying an intermediate cut in multi-component distillation, removing the clutter of light-most and heavy-most components first, to reach high-purity intermediate products by pre-fractionation. • DWCs to accomplish the pre-fractionation within the same column, and advanced thermal coupling systems to move closer to ideal efficiency. • Extensions of segregated columns to retain the value of compositional profiles and recover usable heat from within the distillation system. These new techniques are based on sound engineering principles and will undoubtedly become a core part of the technology base. NOTES GT-DWC and GT-TDWC, GTC Technology US LLC (patent pending). b GT-LPG MAX, GTC Technology US LLC (patent pending). c GT-Advanced Stabilizers, GTC Technology US LLC (patent pending). a

LITERATURE CITED Complete literature cited available at HydrocarbonProcessing.com.

C FIG. 7. Using a TDWC, the intermediate cut can also be taken as a distilled product, with latent heat available at a higher temperature.

• Rating methods—Given the preponderance of computer rating methods for trays and packings, it is prudent to utilize such information. One of the classic rating methods, and perhaps the best overall tool for tray design, is an equation (known as Equation 13) from the Ballast Tray Design Manual Bulletin 4900.5 This model was developed by Francis W. Winn, who spent years in technical management at Fractionation Research Inc. before he worked for Fritz W. Glitsch & Sons in the 1960s (Eq. 1). 54 AUGUST 2016 | HydrocarbonProcessing.com

JOSEPH C. GENTRY, is vice president of technology, R&D and engineering, for GTC Technology US LLC. He previously worked for ARCO Chemical Co. and Lyondell Petrochemical Co. in the olefins and aromatics areas. Mr. Gentry earned a BS degree in chemical engineering from Auburn University and an MBA from the University of Houston. He is the inventor of several patented separations technologies and has specialized in their applications for the petrochemical industry. MICHAEL J. BINKLEY is a consultant for the GTC Process Equipment Technology (PET) group. He is a registered professional engineer in Texas. Mr. Binkley has focused his 45 years of experience in mass transfer/separations equipment development and applications with Glitsch Inc. (1969–2001) and GTC (2002–2016). His first seven years of process engineering were with the PPG Chemicals Division in Lake Charles, Louisiana. Mr. Binkley is an inventor of several separations equipment advancement-related patents, as well as numerous product trademarks. With several GTC co-inventors, he has seven patent applications pending review by the US Patent and Trademark Office. Mr. Binkley earned his BS degree in chemical engineering from Texas Tech University. MANISH BHARGAVA is licensing manager for advanced distillation systems for GTC Technology US LLC. Mr. Bhargava has over 15 years of industry experience in the process industry. Prior to joining GTC, he worked at KBR as the principal technical professional, and with DCM Shriram Consolidated Ltd. as a process engineer. He graduated with an MS degree in chemical engineering from the Illinois Institute of Technology.

Maintenance and Reliability G. MURTI, The Augustus Group, Montgomery, Texas

Design operations-and-maintenance-friendly pressure vessels—Part 1 Many scholarly articles have been published on the design, selection and fabrication of pressure vessels.1–3 The articles, books and training materials published to date focus on the requirements of vessel designers and the manufacturers. None of the published literature appears to address the needs and concerns of the ultimate beneficiaries—the end users. For vessel designers, fabricators and quality control inspectors, this is a onetime responsibility. They probably never look back once a vessel is out of their territory. The reason is simple: almost all pressure vessels are designed per specific requirements of the asset owners. A few proprietary designs are exceptional cases, such as reactors, desalters and coalescer vessels, etc. However, only the internals would be proprietary, and the base vessel would still be required to comply with the owner’s general specifications. In such cases, it is important that the end users do not feel overwhelmed by the proprietary designs and adhere to their company’s basic vessel specifications. Note: This article is the first in a series of design articles intended to increase awareness of operational and maintenancerelated concerns and what design engineers can do to provide user-friendly and fit-for-purpose equipment for the hydrocarbon processing industry (HPI). This series is based on the actual implementation of what is narrated and the satisfactory experiences drawn by the end users. Readers are advised to make their own engineering judgment on the validity of the design improvements suggested herein and to develop their own needs. If engineering or other professional services and judgments are required, then

the assistance of a competent professional authority should be sought. Throughout this article, the terminology “vessel” is used to represent pressure vessels, drums, columns, towers, heat exchanger shells and any equipment designed using pressure vessel codes such as ASME-VIII, EN 13445, PD 5500, etc. The terms “codes, standards, specifications, regulations and recommended practices” are used to broadly define the overall design requirements, recommendations and practices prevailing in the industry. The terms “vessel fabricator” or “manufacturer” have same meaning, as do the terms “owner” and “end user.” Ensuring functional safety and longterm service. Provided here are useful

ideas to ensure that a vessel meets functional safety requirements and provides operation-and-maintenance-friendly service to end users over the long term. The tips provided are simple to implement, do not interfere with proprietary designs and do not violate any of the code requirements—rather, they exceed them. The tips are also generic and do not require any code-specific calculations. Therefore, they are applicable to pressure vessels built to any code. This article essentially covers what the codes and design books would not reveal. The tips would lower ownership cost if implemented at design stage and assist the end users to meet their health, safety and environmental regulations, as well as reduce or eliminate field modifications during service life of the vessel. Carrying out field modifications, which invariably requires hot work, is one of the most painstaking exercises in operating plants.

There are many reasons why pressure vessels should be ergonomically designed. Pressure vessels are probably the longestserving equipment in the HPI. Their life often exceeds the working life of plant personnel, and the vessel can be passed on to the next generation. Even though a facility may cease oil and gas production, a well-maintained vessel would survive. Good pressure vessel design in the initial stages is also important, as there is practically no involvement of vessel designers and fabricators in subsequent field modifications, if any are ever undertaken. For pressure vessels, there are no performance tests to be conducted prior to dispatch. This is true even for proprietary designs. During plant commissioning, vessel manufacturers’ representatives are generally not needed—unlike rotating machinery, where the designs are proprietary and machinery manufacturers are usually involved in site performance tests, troubleshooting, modifications works, etc. Many recommendations exist to enhance vessel design. Design vessels to match outer diameter to piping specs. With the

advancement of computer-added design and drafting, the subjective visualization of the actual size of vessels has diminished. Computer printouts are exchanged, and the design is prepared with minimal manual intervention. A typical process software would carry out the inside diameter calculation of a vessel by taking various process parameters into consideration. The software is not programmed to standardize the vessel’s outside diameter. The mechanical design software would determine the inside diameter, calculate Hydrocarbon Processing | AUGUST 2016 55

Maintenance and Reliability made to standardize a vessel’s outside diameter. A typical plate bending machine (FIG. 1A) can roll a plate into a cylindrical

373 mm (FIG. 1B). The fabricator used SA516, Gr. 60 plates and rolled the exchanger shell. This exchanger failed in service, and replacement exchangers were ordered. Due to NACE compliance, the plates needed to underA vessel must meet functional safety requirements go an HIC (hydrogen-induced and provide operation-and-maintenance-friendly cracking) test. HIC tests have a service. These measures in the design stage are simple 28-day soaking period, and fabrication activities came to a standto implement, do not interfere with proprietary designs still until the results were known. and often exceed code requirements. The problem was referred to the plant engineering team. The fabricated shell was replaced The selected plate thickness is added shell with any inner diameter, as long as with 16-in. (nominal pipe size) NPS seam(twice) to the vessel’s inner diameter, and it is higher than the roller diameter. Plate less pipe, 21.4-mm WT (Sch. 80). Saddle and nozzle projections were adjusted to the vessel’s outer diameter is established thickness is the only limitation. and passed on to a vessel fabricator. The ASME-VIII and TEMA do not en- ensure no changes to plant piping. Using thickness formulae in ASME-VIII, Div. 1, courage standardization of vessels based seamless pipe of appropriate specification section UG-27, “Thickness of shells un- on the outer diameter. This is in contrast avoided HIC test and saved fabrication der internal pressure” is based on vessel to the piping codes to which a vessel is costs, and units were ready within two inside dimensions. Tubular Exchanger invariably attached. For example, ASME- weeks. Similarly, another exchanger with Manufacturers Association (TEMA) B31.3, section 304.1.2 and ASME-B31.1, an original shell OD of 440 mm, 22-mm standard, section N-1.1 1, defines the section 104.1.2 perform calculations thickness was redone using 20-in.-NPS seamless pipe, 508-mm OD and 26-mm nominal diameter as the inside diameter based on the outer diameter. of the exchanger shell. It is recommended that efforts be made WT (Sch. 80). These examples show the Supplemented by the American So- to round off the vessel’s calculated outer importance of engineers reviewing and ciety of Mechanical Engineers (ASME) diameter to match with the corresponding standardizing the shell dimensions proand TEMA, and assured by the fabrica- pipe outer diameter, e.g., up to 80-in. out- duced by the computers, where possible. The advantages of vessel OD standardtors that they can build a vessel to match er diameter (OD). Beyond 80 in., it may any inside or outside diameter, no effort is be rounded off in multiples of 6 in. ASME ization are: • Vessel fabricators can acquire piping standards B36.10 and B36.19 cover seamless and welded pipes pipes up to 80 in. NPS. Seamless pipes are for vessel fabrication, wherever available up to 24 in. Submerged arc-weldavailable. Short lengths of pipes ed pipes (SAW) are available from 16 in. of approved specifications to 48 in. in North America and up to 64 are usually readily available at in. in Asia. Use of pipes in lieu of plates is fabricators’ works and with recommended wherever available. operating companies. Such pipes with traceable documentation Practical examples. A heat exchanger can be approved for fabrication shell was designed with an inner diameter of vessels and exchangers. (ID) of 337 mm, a wall thickness (WT) • Less time and effort, and lower cost of 18 mm and an OD = 337 + 18 + 18 = FIG. 1A. A typical plate bending machine. at the vessel manufacturer’s shop. Standard templates can be used to verify the OD of rolled plates, and there is no need to fabricate custom-made templates for each ordered vessel. • Saddle design is standardized, as the outer curvature of vessel is standardized. • Synchronized pressure/ temperature rating of the connected piping and the piping components welded to the vessel, such as nozzleFIG. 1C. The use of pipes in lieu of plates can save cost and improve integrity. By using singlenecks and welding-fittings. FIG. 1B. Part of an exchanger GA drawing piece 24-in. and 18-in. NPS pipes, longitudinal • Extra-long vessels can be fabricated (non-standard shell diameter). and circumferential welds can be avoided. with one piece of pipe. If plates pressure wall thickness (including allowances) and pick up the next commercially available plate thickness.

56 AUGUST 2016 | HydrocarbonProcessing.com

Maintenance and Reliability are used, then a limitation of plate width exists. FIG. 1C shows a typical pig launcher barrel fabricated using plates with many longitudinal and circumferential welds. • If using pipes, a host of timeconsuming and expensive quality tests can be avoided. Such tests have already been carried out as part of pipe quality testing procedures. Use standard materials where possible. In a new grassroots project, it is

easy to adhere to exotic specifications, even for small items. The delivery periods for such materials get overlapped by big-ticket items. However, when it comes to maintenance replacements, procurement of items with exotic specifications in small quantities is a bottleneck. In the example given earlier, forged tubesheet to SA-266 was used. Ordering new sheets per SA-266 would take up to four months for delivery. Recovering old tubesheet was not an option as the diameter changed. The tubesheet was replaced with SA-516 plate, which is widely available, easy to fabricate, permitted by the code and posed no quality compromise. SA-516 should have been used in the first instance.

• Lack of a stagnant dead area, which is unnecessary from a design point of view and detrimental due to the potential for crevice corrosion • Better liner accuracy for volume control • Lower costs, as less welding is involved and a davit is not required, as shown in FIG. 2B • Time savings: a fabrication shop would produce dished-end in 24 hr to 48 hr. In the case of tall and slim columns, as shown in FIG. 2C, a split in the middle is very maintenance friendly. FIG. 2D shows the original vessel design (simulated image) with a side manway. The change was incorporated during a mechanical review of drawings by the maintenance team.

It is recommended to ensure saddle projection exceeds the bottom nozzles and boot projection. The advantages/disadvantage of tall steel saddles (FIG. 3B) are:

Bottom nozzles/boot should stay within saddle height. A majority of

company specifications require a 6-in. high saddle for smaller vessels and a 12in. high saddle for large vessels, as measured at the vertical axis of the vessel. This design does not consider the nozzles and boots projecting beyond the saddle. FIG. 3A shows such a vessel. It is an unnecessary cost-saving exercise and creates subsequent problems onsite.

FIG. 2C. Manway on the top of a smalldiameter vertical vessel.

Do not provide side manways on small vessels. Many company specifi-

cations require at least one manway on a vessel, regardless of the vessel dimensions. Such manways do not serve the purpose and have many disadvantages: • Technicians cannot easily get inside • Fabrication issues: excessive welding heat-input on the self-reinforcing pad tends to distort the vessel • Extra fittings, such as weldolet and debit arrangement, may be needed • Projection of manway adds extra dead volume to vessel and may be a surprise to unsuspecting process engineers • Crevice corrosion in the deadvolume stagnant area, in FIGS. 2A and 2B (left). One recommendation is for small vessels, perhaps up to 42-in. NPS, to be provided with either a flanged dishedend or a blind flange, depending on the availability. FIG. 2B (Option 1 and Option 2) show the suggested configurations. Advantages include:

FIG. 2A. Manway provided on a smalldiameter vessel.

FIG. 2B. Various options for manways on small vessels.

FIG. 2D. Original design of the vessel manway on the side (simulated image). Hydrocarbon Processing | AUGUST 2016 57

Maintenance and Reliability • Ease of transportation: In the upstream industry, a majority of vessels are generic gas and oil separators and free-water knockout drums. Vessels are moved from one field to another to meet fluctuating production patterns. In such scenarios, vessels with tall saddles are easy to transport and do not require wooden pallets, in case nozzles and boots project beyond the saddles. • If the saddle projection is short, it must be compensated with an increased civil foundation height. While this may look like a better option from a cost perspective, it has disadvantages. The piping stresses are transferred to the ground via a concrete pedestal. Concrete is not appropriate for handling piping stresses, as cracks that are difficult to repair can

FIG. 3A. Vessel with a short saddle.

FIG. 3B. Vessel with a tall saddle.

FIG. 4. Avoid inside projection on manways.

58 AUGUST 2016 | HydrocarbonProcessing.com

appear. Steel saddles distribute loads much more effectively than concrete. Field preference is for steel saddles that overlap all bottom appurtenances by at least 6 in. (150 mm). • The only disadvantage for tall steel saddles is that fireproofing costs are marginally higher and may require fireproofing per relevant specification. Tall concrete pedestals would not require fireproofing, as concrete is treated as inherently fire-resistant.

vacuum design. The vessel was designed for a high-altitude location where atmospheric pressure was 93.3 kPa. The design engineer used the default software value for an atmospheric pressure of 101.325 kPa. When reduced to 93.3 kPa, the inside projection vanished. It is imperative that the software default values are not used without knowing their implications. What might appear as mundane data at the vessel design stage may be the root cause of an unpleasant experience for operation and maintenance personnel.

Avoid inside projection on manways.

an existing vessel is an expensive proposition. It is worth adding a spare nozzle that can be used at a later date. The most useful spare nozzle is at the top of the vessel, preferably of the same size as the main inlet nozzle.

Vessels are usually designed for full vacuum or half vacuum for steam-out conditions. Under certain conditions of full vacuum, the design may dictate the manway to be strengthened in addition to an external reinforcing pad. The most popular method to provide the required extra strength is to project the neck inside and weld circumferentially, as shown in FIG. 4 (left). This design makes exiting the vessel a very unpleasant, if not impossible, exercise. Another disadvantage of inside projection is that it does not allow complete vessel drainage. The small hole in the projected nozzle does not help effective drainage. An inside reinforcing pad could be a solution, but it would not drain the vessel completely. Inside projection can be easily avoided by slightly reducing the vacuum required for steam-out conditions, thereby eliminating the projection. The vessel design engineer must adjust the vacuum numbers and communicate back to the process engineer for endorsement. A vessel need not be designed for full vacuum, as it will theoretically not achieve it. In one case, the vessel design software dictated inside projection due to full

Spare nozzles. Fitting new nozzles on

Cost implications. The initial vessel cost should not be evaluated in terms of CAPEX alone. If maintenance costs and possible field modifications (OPEX) are combined, all suggested measures eventually reduce the ownership cost of the vessel. Traditionally, the OPEX for static equipment like pressure vessels has been considered to be very low, as compared to CAPEX. That is not true for poorly designed vessels. Producing the correct products in the first instance is a win-win situation for all. LITERATURE CITED Heinze, A. J., “Pressure vessel design for process engineers,” Hydrocarbon Processing, May 1979. 2 Smolen, A. M. and J. R. Mase, “ASME pressurevessel code: Which division to choose?” Chemical Engineering, January 1982. 3 Pullarcot, S. K., Practical guide to pressure vessel manufacturing (Mechanical engineering), Ed. 1, FACT engineering and design organization, Mercel Dekker Inc., Basel, New York, 2005. 1

D. GOPALKRISHNA MURTI, P.Eng., is a senior consultant affiliated with the Augustus Group. He has more than 40 years of experience in design and project engineering; resolving complex design and code incompatibility issues; field troubleshooting; and plant safety, integrity and reliability management. He works within the onshore and offshore, LNG/NGL processing, refining, petrochemical and power industries. Mr. Murti has to his credit over 30 revisions/additions/new standards to API, ASTM, ASME, NFPA, BSI, etc., all based on field experience, and has authored several articles on design issues. He obtained his BS degree in mechanical engineering from Jiwaji University in India. He is a registered engineer in Canada and India, with licensing in Texas (TBPE) in process.

Process Control and Instrumentation N. KASIRI and P. JOUYBANPOUR, Iran University of Science and Technology, Tehran, Iran; and M. REZA EHSANI, Isfahan University of Technology, Isfahan, Iran

Utilize genetic programming to develop new point efficiency correlation Distillation is one of the most used separation operations in the chemical and petroleum industries. It is a process in which a mixture of some components, with different boiling points, can be separated by heating the mixture to a temperature between the boiling points of the components.1 In distillation columns, one of the most commonly used simplifying assumptions in the mathematical modeling is the presumption of the existence of ideal trays on which the equilibrium of phases takes place. Point efficiency is the single-most important factor in distillation column simulation. The commonly used tray efficiency evaluation schemes possess little accuracy and are not recommended. The purpose of this study was to develop a new correlation for calculating point efficiency in distillation columns using genetic programming (GP). Previously published literature was reviewed and parameters reportedly affecting point efficiency were selected. The developed equation using GP, which is based on 52 experimental data sets, differs from published results in description and solution precision. Comparison with experimental data demonstrates that the new model is capable of generating point efficiency estimates with less than 1.85% relative error. Considerable attention has been given to the understanding and improvement of the performance of distillation trays. The most important parameter in the design and analysis of traytype columns is tray efficiency.2,3 This is only realistically applicable if a precise method for the estimation of tray efficiency can be devised. The commonly used tray efficiency evaluation schemes, providing a single efficiency for each tray, possess little accuracy and are, therefore, not recommended. Since efficiency varies from one section to another, it is best to apply efficiency separately for each section. In practice, efficiency data and prediction methods are often too crude to give a good breakdown between the efficiencies of different sections, and overall column efficiency is applied over the entire column.4 This is due to the fact that each point location on each of the trays provides an opportunity for mass transfer to take place, from which different components take different opportunities. So, point efficiency has been studied extensively.2,3,5,6,7 One of the first correlations for point efficiency was published by AIChE4 and is shown in Eq. 1.

EOG = 1- exp (-NOG)

(1)

Standart et al. questioned the validity of these assumptions. In one case, they observed just one point efficiency greater than unity, which is inconsistent with the assumptions and Eq. 1.4 Most point efficiency models were developed from the experimental data of systems (Eq. 2).6 −0.0029 ⎡ ⎤ × ⎢ ⎥ DG (1−φe ) ⎢1+m ρ MV ⎥ ρ ML DL (A H / A B ) ⎥ EOG =1− exp⎢ ⎢ −0.3195 ⎥ 0.4136 0.6074 ⎛ A H ⎞ ⎛ hl ⎞ ⎢⎛ ρV VH h fe ⎞ ⎥ ⎜ ⎟ ⎟ ⎜ ⎟ ⎢⎜ µ ⎥ ⎠ ⎝ DH ⎠ ⎝ A B ⎠ ⎣⎝ ⎦ V

(2)

Its average error is 6.95%, which is too high. The objective in the present work is to find a mathematical model for point efficiency that adequately fits and describes the available experimental data. The application of genetic programming. Evolutionary optimization methods have been applied extensively to problems in computer science. They were proposed by Holland,8 while Goldberg9 contributed most to their practical use in many areas. GP is the most recent and very promising evolutionary optimization method. Koza10 demonstrated the applications of GP to robotics, games, control and symbolic regression.11 These applications are based on the principles of the biological evolution and Darwinist principles, such as genetic combining, natural selection and survival of the fittest. GP is one of the many computer algorithms that has been shown to provide reliable solutions to complex optimization problems.10–13 It is also considered a powerful evolutionary optimization method applicable to many process-based problems. It is used here to find an appropriate correlation for point efficiency in distillation columns using a set of available experimental data. Genetic programming application. In this work, the GP algorithm10 was coded in the MATLAB environment. Initially, Hydrocarbon Processing | AUGUST 2016 59

Process Control and Instrumentation the group of genes expected to reflect unknown mathematical relations, hidden in the experimental data consisting of two parts, are chosen. The first part contains terminal genes consisting of dimensionless groups. These genes are, in fact, independent variable symbols, numerical and logical constants. The second part contains the function genes (i.e., *, –, +, /, ^, log, exp, etc.), which suffice for the efficient connection of the terminal genes. Function genes can take one or more arguments. Random solutions of various forms and lengths are generated by means of selected genes. Of course, the maximum size of each solution must be limited to a reasonable value. Then, each solution is evaluated on the basis of experimental data for independent variables through several fitness cases. The obtained results are compared with the corresponding experimental data for the dependent variable. The criterion for fitness of the individual solution is the extent of differences (error). Usually, better individual solutions are those with errors closer to zero. The verification of solutions during several iterations, known as generations, is then performed by means of genetic operations. It turns out that for the progress of the population, reproduction, crossover and mutation are used. The reproduction operation ensures the survival of the fittest solutions of population and their advance in unchanged form into next generation. The crossover is an operation that ensures the exchange of genetic material between solutions. The crossover point is randomly chosen, either on function genes or terminal genes. FIG. 1 shows the crossover of two solutions on function genes. Two new child solutions result from two parent solutions. They are, in fact, mathematical expressions usually written as: (x + y) / z + xz and x(1 – yz). They consist of the genes x, y and z, which are variable symbols of the solution. In defining randomly generated crossover points, two parent crossover fragments are obtained. Child 1 is produced by deleting everything below the crossover point of Parent 1, and then inserting the crossover fragment of Parent 2 at the crossover point of Parent 1. Child 2 is produced in a similar manner. Therefore, two child organisms are created: (1 – yz) / z + xz and x(x + y). It can be seen that both child organisms include the genetic mateParent 2

Parent 1

/

x Z

x

x

y

z

– 1 x

Crossover

Child 1

z Child 2

/

x

–

z

y

z

x

z

x

1

FIG. 1. The crossover of two solutions ensures the exchange of genetic material.

60 AUGUST 2016 | HydrocarbonProcessing.com

y

rial from their parents. It is very important to preserve syntactic structure of the solutions during the crossover operation. The mutation is an operation that creates a new individual by randomly altering a single function with a number of the population. That is, it consists of randomly changing a functional input, or constant, in one of the mathematical expressions making up the present population. This creates one new child solution for the new population by randomly mutating a randomly chosen part of one selected solution. FIG. 2 shows the traditional mutation, which consists of randomly selecting a mutation point in a tree and substituting the sub-tree rooted there with a randomly generated sub-tree. After the completion of reproduction, crossover and mutation, a new generation is obtained that is also evaluated and compared with the experimental data. The process is repeated until the termination criterion of the process is fulfilled. This can be a prescribed number of generations or a sufficient quality of solutions (criterion of success). Of course, because of the probabilistic nature of GP, the satisfactory solution cannot be obtained in every run of genetic programming. Several independent runs are needed to reach the given criterion of success.14 Background on point efficiency. A comprehensive composite database for distillation sieve-tray efficiency is used to develop point efficiency based on a model that considers the fluid on the distillation tray to be contained in a liquid-continuous region. This model allows estimates of the portion of the mass transfer that occurs in each region, and the mass-transfer resistance that occurs on the liquid and vapor sides of the interface. For most cases, most of the mass transfer occurs within the liquid-continuous region. The liquid-side resistance is often significant. Point efficiency touches many parameters, such as physical properties of mixture, hydraulic characters and the geometry of the column. Some of these parameters, and their influence on efficiency according to reviewed publications, consist of: • The slope of the equilibrium line, which is a direct function of relative volatility. Efficiency decreases when the slope of the equilibrium line and relative volatility increase. Lower volatility reduces the significance of the liquid phase resistance, therefore raising efficiency.4 • An increase in liquid viscosity is also usually associated with a decrease in liquid phase diffusivity, and the tray efficiency will decrease accordingly.4 • Lower liquid viscosity usually implies higher liquid diffusivity and lower resistance to mass transfer in the liquid phase. Therefore, efficiency increases as liquid viscosity diminishes.4 • With trays operating in froth regime, an increase in weir height will directly increase the efficiency. Weir height is especially important in liquid-limited systems, or systems where a slow chemical reaction is taking place.4 • The hole diameter has an important effect on the efficiency. Trays with large holes have a different hydrodynamic behavior and are less efficient than trays with small holes. In the spray regime, entrainment is very strongly dependent upon the hole diameter.15 • Tray spacing has no important effect on tray efficiency in small columns. Therefore, it is not taken into consideration in this work.15

Process Control and Instrumentation • The reflux ratio was stated to have a small effect on tray efficiency. It was not considered in this work.4 • At high vapor and low liquid rates, entrainment becomes significant. The liquid entrainment and Reynolds Number rise as the vapor velocity rises, and efficiency increases accordingly.7 • Vapor and liquid diffusivity have a direct influence on efficiency. Increased diffusivities will increase efficiency. Liquid diffusivities generally increase with pressure. However, in most distillation systems, the influence of the vapor diffusivity will dominate the influence of the liquid diffusivity.4 Generally, diffusivity, density and viscosity are affected by composition variations and can affect efficiency. In this research, experimental data used is dependent on most of these parameters, so as to obtain the most descriptive correlation to be employed in simulations. Composite database. The correlation has been based on a

generalized model of mass transfer in a five-sieve tray distillation column that was developed to account for experimental measurements on a 0.6-m diameter sieve tray indicating the presence of vapor entrainment into the downcomer for three systems. The databases16 used here consist of the results of three sets of experiments, all under total reflux conditions. Two of these demonstrate wide-boiling-point-range systems of methanol/water and isopropanol/water, and the remaining

set belongs with a close-boiling-point-range system of methylcyclohexane/toluene. The data was all for binary separations. The entire database consisted of 52 data sets. On the basis of measured and calculated data, an attempt was made to determine an equation for the point efficiency of distillation columns using genetic programming. In this work, 47 out of the total 52 data sets were randomly used to develop the desired correlation, and the rest were used for error evaluations. Method to calculate the point efficiency. The modeling

of point efficiency by GP involves finding a mathematical model, in symbolic form, that provides a good fit between independent variables and the associated dependent variable. Mutation point

Mutation point

3 x

y

Randomly generated sub-tree

x

y

/

y x

2

/

y x

2

FIG. 2. A mutation of one solution.

VIEW ON DEMAND WEBCAST Operational Excellence by Using Dynamic Simulation for Steam Shedding Studies Presenter: Dhaval Dave

Senior Consultant Schneider Electric Software

The goal of any steam network is to ensure that steam can be provided to all consumer facility safely, efficiently and without interruption. However, after a steadystate simulation is used to design a steam network, dynamic simulation can be used to evaluate the “as designed” process and control strategy to maximize likelihood that it can provide stable and uninterrupted operation. In fact, dynamic simulation can be used for: • Design validation

• Sensitivity on design boiler rating

• Relief valve size & safety validation

• Determination of amount of steam to shed and which unit

• Control system validation • Testing/Tuning of feed forward controllers

Moderator: Lee Nichols

• Determination of the duration of steam shedding.

View on Demand at: HydrocarbonProcessing.com/Webcasts

Editor/Assistant Publisher Hydrocarbon Processing

Hydrocarbon Processing | AUGUST 2016 61

Process Control and Instrumentation TABLE 1. The best genetically evolved model for point efficiency +, –, *, /, ^, log, exp, m, DV /DL, ρMV/ρML, hL/DH, Re, (1-φe), EOG, constants

Chosen genes Genetically evolved model

EOG

Average model error as compared with the rest of the experimental data

1.85%

Average model error as compared with all experimental data

1.75%

For each data set within the column, the point efficiency (EOG ) is considered to be a function of the slope of the equilibrium curve (m), mass density (ρ), mole density (ρM), molecular diffusivity (D), molecular viscosity (µ) in both of the phases, height of the outlet weir (h w ), liquid volumetric flowrate (Q L), sieve tray perforation diameter (DH), and velocity of the vapor through a perforation hole (VH) once the velocity of the vapor over the bubbling surface of the tray has been assumed.5–7 These factors were organized into dimensionless groups consisting of the slope of the equilibrium curve (m), Reynolds number (Re), the proportion of each mole density and molecular diffusivity in both phases, the ratio of the liquid inventory (hL) to the sieve tray, perforation diameter (DH) and the effective froth density (φe). The Peng-Robinson equation of state was employed to evaluate the densities of gases and liquids for methanol/water and isopropanol/water systems, and a non-random two-liquid (NRTL) model was used for the methylcyclohexane/toluene system. The viscosity was evaluated by the Brokaw model for gases and the Hayduk-Laudie model for liquids.17 To determine these dimensionless groups, the following steps should be taken:5,6 1. The effective froth density includes many factors. The correlation (Eq. 3) used to calculate the effective froth density is: φe = exp(−12.55K s0.91 )

(3)

where KS is the density-corrected vapor velocity (Eq. 4) over the bubbling surface of the tray, expressed in m/s; ⎛ ρV ⎞0.5 K s = ⎜ ⎟ VS ⎝ ρ L −ρV ⎠

(4)

2. The Reynolds number (Eq. 5) includes the direct proportion of vapor mass density (ρV), velocity of the vapor through a perforation hole (VH ), effective froth height (hfe ) and the inverse of vapor molecular viscosity (µV ). Re =

ρV VH h fe

(5)

µV

The effective froth height (Eq. 6), expressed in meters, is given by: 2

⎛ Q ⎞ h fe = hW +C ⎜ L ⎟ ⎝ φe ⎠

3

62 AUGUST 2016 | HydrocarbonProcessing.com

(6)

⎡ ⎤ ⎛ ρ ⎞ ⎢ ⎥ ⎜ MV ⎟+Re ρ ⎝ ML ⎠ ⎥ = 1− exp⎢⎢−0.867789 − ⎞ ⎥ ⎛ D V ⎢ (1− φe ) (exp(m) + 10.6125) ⎜2 D − (1− φe )⎟ ⎥⎥ ⎢⎣ ⎠ ⎦ ⎝ L

where Q L is the volumetric flowrate of liquid per length of weir that is expressed in m3/s/m (Eq. 7), and C is a constant: C = 0.501+0.439exp(−137.8hW )

(7)

The weir height, hW , is expressed in meters. 3. To calculate the ratio of the liquid inventory to the sieve tray perforation diameter (hL/DH), the liquid inventory height (Eq. 8) should be expressed in meters and evaluated from: h L = h fe ×φe (8) 4. The slope of the equilibrium curve (m) is an experimental data. 5. The ratio of vapor mole density to liquid mole density should be calculated. 6. The ratio of vapor molecular diffusivity to liquid molecular diffusivity should be calculated. Therefore, the slope of the equilibrium curve (m), the Reynolds number (Re = ρVVHhfe/µV), the ratio of vapor mole density to liquid mole density (ρMV/ρML), the ratio of vapor molecular diffusivity to liquid molecular diffusivity (DV/DL), the ratio of the liquid inventory to the sieve tray perforation diameter (hL/DH), and 1-φe are chosen as independent variables, while point efficiency (EOG) is the dependent variable. The best genetically evolved model. The best-fitted correlation of the present work is summarized in TABLE 1. Row 1 shows the chosen genes. Function genes and terminal genes are separated by commas. Row 2 shows the genetically evolved models on the basis of chosen genes. In Row 3, the model error is compared to the rest of experimental data (the five-member set), and in Row 4, a comparison with all experimental data is provided, showing the less-relative error compared with other research.6 FIG. 3 shows the differences between this model and the experimental data. TABLE 2 illustrates the composite coverage of the database and the range of dimensionless group values. Results and efficiency changes. The influences of the dimensionless factors on point efficiency can now be studied using the GP-based correlation within its validity range, while the others are kept constant. Some of these factors—consisting of the effective froth density, the ratio of the proportion vapor molecular diffusivity to liquid molecular diffusivity, and the slope of the equilibrium curve (m)—have an inverse influence on ef-

REGISTER

45TH TURBOMACHINERY & 32ND PUMP SYMPOSIA

The premier conference for Turbomachinery and Pump professionals.

DEVELOPED FOR THE INDUSTRY, BY THE INDUSTRY.

ONLINE REGISTRATION OPEN NOW TPS.TAMU.EDU

Process Control and Instrumentation rest of the experimental data (five data sets) demonstrated that the new model is capable of generating accurate point efficiency estimates with an average of 1.85% relative error (TABLE 1). The results highlighted the fact that new correlation should be generally applicable for the prediction of sieve tray point efficiency in distillation within its validity range of the initial data (TABLE 2). This successful fitting supports the assertion that GP is an effective method for developing various process correlations, especially for the same entity of point efficiency with more independent variables, such as material property, and with more comprehensive data sets.

1.00 Calculated point efficiency by new correlation Experimental data

0.95 0.90 Point efficiency

0.85 0.80 0.75 0.70 0.65 0.60 0.55 0.50

0

10

20

30 Run number

40

50

60

ACKNOWLEDGEMENT The authors would like to express their gratitude to Dr. J. Ivakpour for his generous assistance.

FIG. 3. The comparison between the model prediction and the experimental data.

C D DH EOG hfe hL hW KS m QL Re VH VS

TABLE 2. The composite coverage of the database and the range of dimensionless group values Parameters

Minimum

Maximum

Max/min

m

0.224

2.8

12.5

ρMV/ρML

0.0007

0.0047

6.714

DV /DL

1.9407 × 1.0e + 04

5.1902 × 1.0e + 04

2.674

hL/DH

3.6305

6.0872

1.677

1-φe

0.4533

0.6977

1.539

Re = ρVVHhfe/µV

5.3755 × 1.0e + 04

2.2822 × 1.0e + 05

4.245

ficiency, while the others, consisting of the ratio of the proportion of vapor mole density to liquid mole density and Reynolds number, have direct influence on efficiency. The factors create exponential changes in efficiency. When the slope of the equilibrium curve is increased, the mass transfer resistance increases but the molar rate of diffusion decreases, thereby decreasing efficiency. By increasing the ratio of the densities, the mass transfer resistance is decreased and the molar rate of diffusion is increased, so efficiency is increased. The ratio of the diffusivities has an inverse influence on efficiency. These systems are liquid phase-rate controlled where the mass transfer is more accomplished, so the liquid diffusivity has more influence. Therefore, by increasing the ratio of the diffusivities, the mass transfer resistance is increased and efficiency is decreased. As 1-φe increases, efficiency is decreased. When the effective froth density is decreased, it causes the mass transfer to decrease, therefore making distillation more difficult. In the developed equation, the Reynolds number has the largest influence on efficiency. When it is changed within its validity range, the most significant change occurs on efficiency. Increasing the Reynolds number causes a decrease in the mass transfer resistance, so efficiency is increased. The new model. A new correlation to predict the point efficiency of sieve trays was developed by using GP and based on 47 data sets. The genetic process automatically eliminated the superfluous independent variable (hL/DH) from the model. The developed correlation is highly accurate, as it is based on data collected from extensive experimentation. Comparison with the 64 AUGUST 2016 | HydrocarbonProcessing.com

NOMENCLATURE Constant Molecular diffusivity, m2/s Sieve tray perforation diameter, m Point efficiency Effective froth height, m Liquid inventory on tray, expressed in height of clear liquid, m Height of weir, m Density-corrected vapor velocity over the bubbling surface, m/s Slope of equilibrium curve Volumetric liquid flowrate, m3/min Reynolds number Vapor velocity through perforation, m/s Vapor velocity over bubbling surface of the tray, m/s

Greek letters: ρ Density, kg/m3 µ Viscosity, N.s/m2 φ Froth density Subscripts: e Effective f Froth V Vapor L Liquid M Molecular LITERATURE CITED Bjorn, I. N., U. Gren and F. Svensson, “Simulation and experimental study of intermediate heat exchange in a sieve tray distillation column,” Computers and Chemical Engineering, Vol. 26, 2002. 2 Chen, G. X. and K. T. Chuang, “Determining the number of gas-phase and liquidphase transfer units from point efficiencies in distillation,” Industrial & Engineering Chemistry Research, Vol. 33, 1994. 1

Complete literature cited available online at HydrocarbonProcessing.com. DR. NOROLLAH KASIRI graduated with his BSc degree from Glamorgan University before pursuing an MSc degree and PhD at Swansea University, Wales, UK. Dr. Kasiri joined the School of Chemical Engineering at Iran University of Science and Technology (IUST) as an assistant professor, where he established the CAPE center. Over the past 20 years of CAPE activity, he has managed professional chemical, process and reservoir engineers, resulting in the presentation and publication of over 200 papers, the conclusion of 70 research projects and the development of 14 software packages. He is currently with IUST as an associate professor. PARVIN JOUYBANPOUR began her career in chemical engineering at the Science & Research Campus of Islamic Azad University, where she graduated with a BSc degree, followed by post-graduate studies at the CAPE center at Iran University of Science and Technology (IUST), where she earned her MSc degree. She earned a second MS degree in industrial management in Sweden. DR. MOHAMMAD REZA EHSANI serves as a professor in the chemical engineering department at Isfahan University of Technology (IUT). He began his career in chemical engineering at Sharif University of Technology, where he graduated with a BSc degree. Dr. Ehsani earned his MSc degree and PhD at UMIST University, Manchester, UK.

Water Management C. MCKNIGHT and B. RUMBALL, Syncrude Canada Ltd., Edmonton, Canada

Experience with naphtha in sour water emulsions generated in a fractionator overhead accumulator

Bitumen processing plant operations. Syncrude Canada

operates an integrated oil sands plant near Fort McMurray, Alberta, Canada. The oil sands are mined, and the hydrocarbon is then extracted as Athabasca bitumen. The bitumen is topped in atmospheric and vacuum distillation units, and the residue is converted into lower-boiling materials by either carbon rejection (coking) or hydrogen addition (hydroprocessing). The resulting distillates are further hydrotreated and then blended to form a sweet premium synthetic crude oil blend. Fluid coking technology is the engine for pitch conversion at Syncrude. The three fluid cokers (Labeled 8-1, 8-2 and 8-3) process 200 Mbpd of resid (with a material boiling temperature above 524°C, also called pitch) when operating at design conditions. Resid is converted to lighter products, coke and gases by contacting with hot coke solids in a fluidized bed reactor. Any unconverted resid, or “recycle,” is condensed in the scrubber and returned to the reactor for further reaction. Steam and reaction products ranging from light hydrocarbon gases to heavy gasoil are sent from the scrubber to the fractionator for separation. In the fractionator, heavy and light gasoil are fractionated and condensed into liquid products. The overhead from the fractionator contains steam, light hydrocarbon gases and naphtha. This gaseous stream passes through a bank of heat exchangers that condenses the water and naphtha. The resulting gas/liquid/liquid stream is sent to the fractionator overhead accumulator. In this horizontal drum, the large gas flow is separated from the two liquids, water and naphtha, which are also separated from one another. The gaseous prod-

ucts are sent to a knockout drum and, eventually, to an offgas compressor. The water is sent to the sour water processing facilities. A portion of the naphtha is recycled to the fractionator, and the balance is sent downstream for further processing. Historically, Syncrude Canada has suffered from poor naphtha/water separation in the overhead accumulators of all three cokers. Generally, poor separation manifests as high naphtha content in the sour water stream. Typically, Cokers 8-1 and 8-2 run approximately 1 vol%–2 vol% naphtha in the sour water. However, if the condensing heat exchangers leak recycled cooling water (RCW—i.e., water that is recycled from the oil sands extraction process and contains bitumen and clay fines) into the stream, then stable emulsions can form. This is thought to be due to RCW contaminants, most likely the fine solids. These emulsions then cause much higher naphtha losses to the water stream. Fortunately, this sour water is directed to a holding tank, where naphtha can be skimmed and sent to the light slop system for reprocessing in the cokers and eventual inclusion into Syncrude Canada’s Syncrude sweet premium hydrocarbon product. During the first three runs of Coker 8-3, even higher levels of naphtha were found in the water stream, near 5 vol%. Due to this poor separation and the apparent potential for improvement, efforts were undertaken to redesign the gas/liquid/ liquid separator internals through the process licensor. These new internals were installed in 2012 and have been operating since early July of that year. 8

Demulsifier injected 96/11/15 7:00

7 Naphtha in 8-1, C4 sour water, %

A number of conversion processes, both hydrogen-addition-based and carbon-rejection-based, crack heavy oil into lighter hydrocarbon products. The lightest products from these processes include incondensable gases, water from steam, and light hydrocarbons or naphtha. These products are generally discharged from the main unit fractionator into a condensing train and an overhead accumulator drum. Syncrude Canada Ltd. recently attempted to modify an overhead accumulator drum to improve naphtha/sour water separation. Unfortunately, performance declined due to the formation of naphtha in sour water emulsion. The rapid response to alleviate the problem through the use of a demulsifier is described here. Also included are potential mechanical solutions developed through the application of computational fluid dynamics and cold flow modeling.

8-1 naphtha in sour water 16 naphtha in sour water

6 5

Demulsifier injected

4

96/11/12 7:00

Pumps off

3 2 1

0 96/9/22 0:00

96/10/2 0:00

96/10/12 0:00

96/10/22 0:00

96/11/1 0:00

96/11/11 0:00

96/11/21 0:00

96/12/1 0:00

FIG. 1. Historical experience with chemical demulsifier. Hydrocarbon Processing | AUGUST 2016 65

Water Management After installation of the new internals in 2012, very high levels of naphtha in the water were observed, approximately 15 vol% to 20 vol%. Samples of the water were taken after the separator water drawoff pump. The samples showed a milky appearance, suggesting a strong oil-in-water emulsion. Unfortunately, the naphtha recovery and light slops reprocessing systems were being overwhelmed due to these large volumes of naphtha. The work undertaken to find rapid solutions to the problem and determine how the new internals might have impacted the emulsion formation is described below. Rapid response. Initial work focused on options to break the

naphtha/sour water emulsion. The most promising options included the use of packed-bed coalescers and the addition of a chemical demulsifier. Fortunately, Syncrude Canada attempted the use of a chemical demulsifier in the mid-1990s, with some success, as shown in FIG. 1. Typical naphtha in sour water ranges from 2 vol% to 5 vol% before the demulsifier addition, and less than 1 vol% after addition. Field experimentation was immediately conducted by sampling the emulsion directly into varying amounts of demulsifier. As shown in FIG. 2, the result was remarkable. At the 1-ppm level, the emulsion broke very quickly. As a result, efforts were turned toward field implementation. FIG. 3 shows the timeline from the startup of the unit in July, through the realization of the problem in August, and to the successful demulsifier injection in September. The graph shows the light slops generation from the poor naphtha/sour water separation, as well as the building tank level as the plant was not able to process the amount generated.

Initially, the demuslifier was injected directly into the accumulator drum liquid. The effect was immediate and allowed the plant to reduce light slops tank levels through reprocessing. After a brief outage of the demulsifier, the problem resumed, as shown by the spike in light slops generated. After this point, the demulsifier was injected upstream of the accumulator, and the best effect was observed. The demulsifier is still used to manage the emulsion problem. Problem analysis. In parallel to the demulsifier work, effort was directed at understanding what caused the emulsion formation. Extensive process reviews were undertaken and confirmed that the unit was operating in the same manner as before the unit turnaround. This work included radioactive tracer studies conducted pre- and post-turnaround. As shown in FIG. 4, bitumen and steam are fed to the fluidized-bed reactor section with thermally cracked products and steam flowing overhead from the scrubber to the fractionator. Steam, incondensable gases and naphtha then proceed to the overhead accumulator. Typical volume ratios are shown, with 99.5 vol% of the flow being gas and only 0.5 vol% comprising the condensed sour water and naphtha. FIG. 4 also shows that possible RCW leakage into the process from the condensing exchangers was discounted, as this unit uses regular clean cooling water and not RCW. Having confirmed no operating differences, the accumulator modifications were examined. FIG. 5 shows the simple design provided with the original plant design. The two inlets impinge on a shelf with the liquid falling to the pool below, and the vapors turning past the vessel head and back toward the demister

FIG. 2. Samples with demulsifier added and control. FIG. 4. Schematic of reactor, fractionator and overhead accumulator. 250

5

Tank almost full

4

Gpm

Meters of oil in tank

200

2

100

0

Demulsifier injected July

August

FIG. 3. Injection of demulsifier in 2012.

66 AUGUST 2016 | HydrocarbonProcessing.com

September

Shelf inlet Demister

0

FIG. 5. Original accumulator design.

Water Management pad, and then proceeding to the single gas exit. Water and naphtha move through the demister pad, the water is collected in the boot and withdrawn, and the naphtha is taken off from above. FIG. 6 shows the modified accumulator internals. Instead of the inlet shelf, a “slotted T” distributor was used to distribute the gas and liquids. Furthermore, calming baffles and a packed bed were provided to improve liquid/liquid separation, and a less dense demister pad was retrofitted. Of the changes, the slotted T appeared to be the greatest concern area for generating emulsion. Here, energy from the incondensable gases could be imparted into the naphtha and sour water. Bench-scale surrogate experiments. To understand the

emulsification process, an engineering consulting firm was commissioned to develop rapid bench-scale experiments. To model the process, it used compressed air to represent the incondensable gases, with canola oil and tap water representing the naphtha and sour water, respectively. [Using actual materials (sour water and sour naphtha) poses severe safety issues due to the amount of hydrogen sulfide dissolved in the real commercial streams.] Within one month, from commissioning to first operation, data was generated. Initially, the shelf inlet was modeled, as shown in FIG. 7. One surprising result was the identification of emulsion entering from the white inlet tube. It appeared that some emulsion had formed when canola oil and water were mixed upstream of the shelf. Clearly, once the oil and water

condense, even turbulence in the piping upstream of an accumulator can cause emulsion formation. For contrast, FIG. 8 shows the results of passing the flow through a single-slot simulation of the slotted T. The white emulsion formed is apparent, although the flowrates being used were extreme. FIG. 9 shows settling curves for emulsion formed under varying inlet geometries and operating conditions. Overall, the slotted T geometry appeared to have the potential to form emulsions, but only under extreme flow conditions. Inlet gas phase flow structure. As shown in FIG. 4, gas is the majority of the volume flow in the accumulator. To obtain a better understanding of the flow structure, a computational fluid dynamics (CFD) simulation was commissioned. Results Slotted T inlet Demister

Calming baffles/packed bed/gap

FIG. 6. Modified accumulator design (slotted T picture inset).

InstruCalc CONTROL VALVES • FLOW ELEMENTS • RELIEF DEVICES • PROCESS DATA InstruCalc 9.0 calculates the size of control valves, flow elements and relief devices and calculates fluid properties, pipe pressure loss and liquid waterhammer flow. Easy to use and accurate, it is the only sizing program you need, enabling you to: Size more than 50 different instruments; Calculate process data at flow conditions for 54 fluids in either mixtures or single components and 66 gases, and; Calculate the orifice size, flowrate or differential range, which enables the user to select the flow rate with optimum accuracy.

NEW VERSION

Order Direct from the Publisher. GulfPub.com/InstruCalc or call +1 (713) 520-4426.

Updates and What’s New in InstruCalc Version 9.0 ENGINEERING STANDARD UPGRADES Control Valve Revisions: • Updated to ANSII/ISA 75.011.01-2012 • Calculation accuracy changed for critical flows • Viscosity correction factor changed • Pressure drop calculation revised to agree with Crane Technical paper No 410. • Option of Cv Units (English) or Kv units (Metric) added. • Option of either aerodynamic noise calculation by ISA 75.17 method or InstruCalc method

• Calculation accuracy added (input data within acceptable limits) Relief Devices: • Pressure Relief Devices Program follows API 520 Pt 1, 9th edition dated 7/14 OPERATIONAL IMPROVEMENTS The ability to have more than one calculation open at a time has been added. Each instance of the program is framed in a different color. The user can have multiple “what if” scenarios displayed for making engineering decisions. Hydrocarbon Processing | AUGUST 2016 67

Water Management

FIG. 7. Simplified shelf design operating with air, canola oil and water.

FIG. 10. CFD gas-phase-only simulation of the slotted T distributor.

FIG. 8. Single slot operating with air, canola oil and water.

FIG. 11. Drawings of cold-flow Plexiglas mockups: Original shelf design (left) and modified internals with slotted T distributor (right).

average. Combining these results with the bench-scale work suggested that the inlet distributor was capable of forming the naphtha-in-water emulsion.

FIG. 9. Settling curves for emulsion formed for various conditions.

from this “gas-only simulation” are shown in FIG. 10 for the new slotted T inlet. Remarkably, the gas flow was not uniform but showed peak velocities (red areas) almost four times the 68 AUGUST 2016 | HydrocarbonProcessing.com

Cold-flow experiments on emulsion formation. The initial work described previously led to a high-level understanding of the potential problems with the new internals. However, a more detailed look at the geometry was required. As a result, a scaled cold-flow model of the drum was commissioned and constructed. FIG. 11 shows profile and end-on views of the original shelf and slotted T designs that were replicated in Plexiglass. Tap water and Exxsol D95 were used to simulate the sour water and naphtha, respectively. Cold-flow observations immediately suggested that mixing in the slotted T was indeed much higher than the original shelf configuration. The gas and liquid entered the vessel and

Starting oil/water interface

100 90 80 Exxsol D95 oil, vol %

70 60

Froude number match

Oil-free surface

Water Management

Oil draw port

50 40 30 Shelf Tee Mod. Tee, no pool baffles Mod. Tee, with pool baffles

20

FIG. 12. Open-face slotted T design.

impinged on the closed center portion of the slotted T. This action caused intensive mixing of the gas, with the water/oil mixture forming an emulsion. A simple solution, requiring only the removal of the slotted T face, was suggested during the experiment (FIG. 12). Sampling of the cold flow unit under various operating scenarios was also completed. Results shown in FIG. 13 suggest that the slotted T leads to higher oil-in-water content than the original shelf design. Furthermore, removing the face of the slotted T can return the separation to the level of the original shelf. The inset picture shows various samples of water and oil recovered from the cold-flow experiments. (Note: In the picture, water is dyed dark blue for contrast from the clear Exxsol D95).

10 0 0.0 0.5

1.0

1.5

2.0

2.5

3.0 3.5 4.0 4.5 5.0 Sample port height, in.

5.5

6.0 6.5

7.0

7.5

FIG. 13. Cold-flow emulsification results for inlet configurations at different sampling heights.

Energy dissipation, m2/sec3

Original shelf T (50)

Slotted T (97)

Energy dissipation: The key to emulsion formation. Fi-

nally, gas-phase CFD was also used to understand the three inlet configurations, the original shelf design, the slotted T design and the open-face slotted T. Under nominal gas flow conditions, energy dissipation over the inlets was estimated (FIG. 14). The slotted T design had the greatest energy dissipation, followed by the original shelf and open-face slotted T. These results are entirely consistent with the cold-flow work described in the preceding section. Recommendations. Great care must be taken when modifying the internals of gas/liquid/liquid separators. Relatively small changes to internals can lead to increased energy dissipation into the liquids, causing emulsion formation and poor separator performance. Fortunately, the demulsifiers have allowed Coker 8-3 to continue operation. They have been used continuously over the recent three-year run. Once the accumulator vessel is available for mechanical work, the face of the slotted T can be removed to reduce emulsion formation and to minimize the need for a demulsifier. Since naphtha in sour water separation has been a longterm issue at Syncrude Canada for all three cokers, this work suggests that all aspects of separation should be examined. These aspects include the nature of the naphtha (i.e., higher density than typical due to fractionator ammonium chloride deposition issues), energy dissipation in the upstream piping, and the design of the gas/liquid/liquid separator. Improving these components, as well as the continued use of a demulsifier, may be required to solve Syncrude Canada’s naphtha-in-sour-water separation problem.

Open face T (45)

0

50

100

FIG. 14. Energy dissipation for shelf, slotted T and open-face T. ACKNOWLEDGMENTS The authors greatly appreciate the work provided by R. Carruthers (GE Water and Process Technologies), A. Gulamhusein (BC Research Inc.) and A. Mezo (Coanda Research and Development) in support of this project. CRAIG MCKNIGHT has worked at Syncrude Canada for 29 years. During that time, he has been responsible for research projects in all areas of Syncrude’s upgrading operations. In 1998, he was seconded to Syncrude’s upgrading expansion project and worked on flowsheeting options and economics. Mr. McKnight returned to research activities in 1999 when he was seconded to ExxonMobil. There, he was named “Innovator of the Year” for his work on sulfur removal from gasoline. He is working to improve fluid coker, LC-Finer and naphtha recovery unit operations. Mr. McKnight graduated from Queen’s University in Ontario, Canada in 1987 with an MS degree in applied science. BOYD RUMBALL has worked at Syncrude Canada for 36 years. At present, he is a senior associate in operations support for extraction and upgrade. In this role, he works as the fluid coking technical advisor in the conversion area. Previously, he served as the Fluid Coker 8-3 ready-for-operations technical leader for the Upgrader Expansion (UE-1) project. He was responsible for the commissioning and initial startup of Coker 8-3. He has also had work assignments in site engineering services, operations and advanced control. Boyd holds a BS degree in chemical engineering from the University of Waterloo in Ontario, Canada. Hydrocarbon Processing | AUGUST 2016 69

Environment and Safety M. YANG, NEL, Glasgow, UK

Understand the sources of oil pollution in water Oil in water is one of the most common types of water pollution. It has been previously reported that, in the UK alone, over one-fourth of all serious water pollution incidents involved oil. Oil is also a highly visible form of pollution that can easily spread. Even small quantities can potentially cause harm to any aquatic environment. For example, five liters of oil can form a film on the surface of a small lake, reducing the level of oxygen in the water and making it difficult for fish to breathe. Oil is also a risk to sewage treatment works, where accidental discharges can be difficult and costly to clean up. A small quantity of oil can have a disproportionate impact: it can taint drinking water even at extremely low concentrations. Many potential sources can lead to water contamination by oil. It is important that these potential sources are known so that the prevention, treatment and quantification of oil in water contamination and pollution can be understood. Sources of oil contamination in water. According to the Scottish Environmental Protection Agency (SEPA), the most common types of oil pollutants are diesel, central heating oil, waste oil and, to a lesser extent, petrol. The most frequent causes of oil pollution are spills during delivery, leaks from poorly maintained or damaged tanks, lack of oil separators, waste oil disposal into drainage systems, waste oil dumped onto or into the ground, and, in some cases, waste oil burned in the open. In the majority of these incidents, no emergency oil spill response plans and/or equipment are available. Other sources of pollution include corroding pipelines below and above ground, and illegal oil discharges at sea. For large power plant operations, a significant amount of cooling water is used to cool equipment such as pump motors, compressors and transformers. Oil leaks sometimes occur, resulting in oil pollution of the cooling water. Oily wastewater can also come from boiler feed, leaks from lubrication systems and from drip pans, such as those positioned below transformers. For oil refineries, a huge amount of water is used for operations that include desalting and cooling. Conservation of Clear Air and Water in Europe (CONCAWE) reported that the total aqueous effluent from EU refineries in 2010 reached 1,583 MMt. In particular, process water that results from desalting can contain a significant amount of crude oil. Legal aspects of oil in water. In many countries, regulations related to aqueous discharges in surface waters set maxi70 AUGUST 2016 | HydrocarbonProcessing.com

mum values to the quantity of a limited number of contaminants that can be released in the surface waters. The discharge of water contaminated with oil and hydrocarbons is strictly controlled and regulated. In the EU, mineral oil and hydrocarbons are List-1 substances under the original Dangerous Substances Directive 76/464/EEC. The directive accounted for discharges to inland surface waters, territorial waters, inland coastal waters and ground water. The 76/464/EEC directive was integrated into the Water Framework Directive (2000/60/EC) that was adopted in September 2000. In 2013, 76/464/EEC was subsequently repealed and has since been codified as 2006/11/EC. The 2000/60/EC directive is less prescriptive and integrates the various components more effectively. However, it is very complex, as it takes into account the characteristics of the water in which the discharge takes place, as well as the environment, the local species, human activities and their interactions. A priority list of 33 substances, the discharge of which must be reduced, has been established (EU Decision 2455/2001/CE). Appearing at the top of the list are 11 substances and products that include heavy metals, pesticides and polycyclic aromatic hydrocarbons (PAH). Discharge of these 11 substances and products will be banned by 2021. Trade effluent, the liquid waste “produced in the course of any trade or industry” that is discharged to the wastewater system, requires a consent from representative authorities. Typical consent parameters and quoted example figures are: • pH limits typically fall in the range of 6 pH–10 pH, and 5pH–11pH • Suspended solids are typically limited to 1,000 mg/l • Chemical oxygen demand (COD) limit is typically 2,000 mg/l • Biochemical oxygen demand (BOD) is typically limited to 1,000 mg/l • Oil must not exceed 100 mg/l. For surface water discharges—or mineral oil and hydrocarbons contained in the discharge water—that are discharged to groundwater, the sea, rivers or lakes, the limits depend on the potential local environmental impact. Two types of consent exist: numeric and descriptive. In numeric terms, wastewater discharge of up to 5 mg/l or 10 mg/l of oil and hydrocarbons, respectively, into the control water may be permitted by governmental agencies, depending upon the quantity of the discharge and the receiving environment. The descriptive term “no visible oil” has also been used.

Environment and Safety Effluent from industrial processes is normally discharged to a sewer, subject to the approval of water treatment companies. Concentration of oil in these discharged waters can vary significantly, from mg/l levels to hundreds of mg/l. For refineries in Europe, an annual average maximum of 5 mg/l in the effluent was stipulated in PARCOM Recommendation 89/5. Discharge of mineral oil and hydrocarbons to the control water is strictly regulated. Control water covers virtually all fresh and saline natural waters up to the offshore territorial limit, including rivers, streams, lakes, estuaries, coastal waters and ground water. Oil-in-water treatment technologies. Oil in water can be present in both dispersed and dissolved forms. Dispersed oils in water are found in the form of small droplets that may range from sub-microns to hundreds of microns. Dissolved oils are those present in a soluble form. Examples include benzene, toluene, ethylbenzene and xylene (BTEX). Many treatment technologies can be used to treat oil-contaminated water. They range from mechanical and physical/ chemical to biological. Treatments include: • Gravity and enhanced gravity • Gas flotation • Filtration (sand, walnut shell, membrane, etc.) • Absorption and adsorption • Chemical/oxidation • Biological. For dispersed oils, gravity, gas flotation and filtration-based technologies are commonly used. Separation efficiency depends on numerous factors (oil droplet size, density of the oil, viscosity and density of the water phase, bubble size, flowrate, temperature, etc.). For dissolved oils, gravity or filtrationbased methods are ineffective. Absorption, adsorption, chemical/oxidation and biological-based technologies are required. European standards (BS EN 858-1:2002 and BS EN 8582:2003) for the design and use of prefabricated oil separators are available. These standards refer to two classes. Class 1 separators, which are designed to achieve a concentration of less than 5 mg/l of oil under standard test conditions, should be used when removal of very small droplets is required. Class 2 separators are designed to achieve a concentration of less than 100 mg/l under standard test conditions, and are suitable for handling discharges where a lower quality requirement applies (e.g., where the effluent passes to a foul sewer). Until recently, these standards were widely used by UK regulators. For oil refineries, a typical wastewater treatment plant consists of primary and secondary oil/water separation to remove most of the oil, followed by biological treatment and tertiary treatment (if necessary) to remove the remaining oil and other contaminants. The primary and secondary oil/water separations are usually achieved with an API gravity separator followed by a dissolved air flotation (DAF) or induced gas flotation (IAF) unit. Water discharge from the flotation units is then routed to an aeration tank/clarifier that constitutes the biological system. A tertiary treatment may be added prior to the discharge of the treated water. According to CONCAWE, the concentration of “oil” in the discharged effluents from refineries in Europe has consistently been less than 2 mg/l since 2000.

Oil-in-water analysis methods. Oil in water is a method-

defined parameter, and oil in produced water can be present in different forms: free oil, dispersed oil and dissolved oil. Free oil usually refers to oil floating on the surface of water or very large oil droplets that settle to the surface very quickly. Dispersed oil is oil in produced water in the form of small droplets that range from sub-microns to hundreds of microns. Dispersed oil may contain aliphatic, aromatic hydrocarbons and other organics—e.g., acids and phenols. Dissolved oil refers to oil in produced water in a soluble form. Aliphatic hydrocarbons generally have very low solubility in water. It is the aromatic hydrocarbons—in particular, the single-ring BTEXs; two-ring naphthalene, phenanthrene and dibenzothiophene (NPDs); and those organic acids (e.g., fatty acids and naphthenic acids) and phenols—that form the bulk of dissolved oil (FIG. 1). The amount of dissolved and dispersed oil in water can increase or decrease depending on the processing conditions being utilized (temperature, pressure, flowrate and treatment technologies). Some of these constituents may be present but might not contribute to the measured oil in water. The relative contribution that these components make to the oil in water content depends on the method used for analysis.

MEASUREMENT METHODS Three main types of reference oil in water measurement methods exist: infrared absorption, gravimetric, and gas chromatography and flame ionization detection. Infrared absorption. In a typical infrared absorption-based method, an oily water sample is first acidified and then extracted, typically by a chlorofluorocarbon (CFC) solvent. Once the extract is separated from the water sample, it is dried and purified by the removal of polar compounds. A portion of the extract is placed into an infrared instrument, where the absorbance is measured. By comparing the absorbance obtained from a sample extract to those that are prepared with known oil concentrations, the oil concentration in the original sample can be calculated. A well-known example of an infrared-based reference method is IP 426/98. Gravimetric. Gravimetric-based methods measure anything

extractable by a solvent that is not removed during a solvent Total oil

Dissolved

Aromatics

Acids

Dispersed

Phenols

Aromatics

Acids

Aliphatics

BTEX

Fatty acids

BTEX

Fatty acids

PAHs

Naphthenic

PAHs

Naphthenic

FIG. 1. The amount of dissolved and dispersed oil in water can increase or decrease, depending on the processing conditions being utilized. Hydrocarbon Processing | AUGUST 2016 71

Environment and Safety evaporation process and is capable of being weighed. In a typical gravimetric-based method, an oily water sample is acidified and then extracted by a solvent. After separating the solvent (now containing oil) from the water sample, it is placed into a flask that has been weighed. The flask is placed into a temperaturecontrolled water bath, and the solvent is evaporated at a specific temperature, condensed and collected. After the solvent is evaporated, the flask containing the residual oil is dried and weighed. If the weight of the empty flask is known, then the amount of residual oil can be calculated. A good example of gravimetricbased reference methods is the USA EPA Method 1664. Gas chromatography and flame ionization detection. Unlike infrared and gravimetric methods, the use of GC-FID offers the potential for obtaining details of the different types of hydrocarbons in the oil fraction. Like the other reference methods, an oily water sample is acidified and extracted by a solvent. The extract is then dried and purified before a small amount is injected into a GC instrument. With the help of a carrier gas and the chromatographic column, different groups of hydrocarbons are detected as they leave the column at different times. Carrier gas is used to move the components through the column, while the column acts to separate the different groups of hydrocarbons so they leave the column and are detected at different times. An example of a GC-FID-based reference method is the ISO 9377-2. Choosing the best reference method. It should be emphasized that different methods will produce different results. Therefore, it is not possible to compare results obtained from the different reference methods. Also, separate methods require different instruments and procedures, affecting costs (CAPEX and OPEX), training, health and safety. Until recently, infrared-based methods were commonly used with available portable, fixed-wavelength instruments. Due to the use of CFC and the lack of compositional details of the measurement method, they are becoming less popular. Gravimetric methods are simple and relatively cheap, but again they do not provide details of the composition. Due to the evaporation procedures involved in gravimetric methods, some loss of volatile components also occurs. GC-FID methods do not require the use of CFCs, have no issues with the loss of volatiles, and have the potential to provide detailed information on composition. However, they necessitate sophisticated instruments that require skilled operators. Sampling and sample handling. While measurement meth-

ods are crucial for obtaining good results, it is important to understand that measurement results are only as good as the provided samples. If the samples used for the analysis are not representative of the flow stream, then the results obtained will be of little use. To obtain a representative sample, a number of aspects must be considered: • Location of withdrawn sample • Selection of the appropriate sampling devices • Isokinetic sampling, if samples are taken from a pipe (isokinetic samples are taken such that the velocity of fluid in the sampling pipe is the same as that in the main flow pipe)

72 AUGUST 2016 | HydrocarbonProcessing.com

• Sample bottles, which must be scrupulously clean. For regulatory compliance monitoring, samples may have to be taken at specified locations. For process control and optimization, this is not an issue. Once a representative sample is obtained, the sample must be properly handled. This will depend on when, where and how the sample is to be analyzed and whether the samples are for regulatory compliance or process optimization. Sample handling may include acidification, transportation and storage. Acidification serves two main purposes: to preserve the samples by killing bacteria that can degrade oil; and to dissolve precipitates, such as iron oxide and calcium carbonate, which can stabilize an emulsion and prevent a complete separation between solvent extract and water after the extraction process. Generally, the pH value needs to be lower than 2. This is often achieved by adding a small volume of diluted HCl solution. If a water sample is to be transported in addition to acidification, then the sample should be stored and transported in a suitable, sealed container to prevent ingress of light. Exposure to light may degrade hydrocarbons in the water sample and change the oil concentration. Similarly, if oil in water samples are to be stored for whatever reason, according to the ISO 5667-3 standard, they should ideally be stored in a refrigerator with a temperature of between 1°C and 5°C. The ISO standard also states that the maximum recommended preservation time before analysis for an oily water is one month. Future trends/needs. With increasing environmental aware-

ness, a tighter regulation with a reduced oil discharge limit (5 mg/l is widely accepted in the EU) may be introduced. A similar change has already taken place for the offshore oil and gas industry in the North Sea: the oil in produced water performance standard of 40 mg/l was reduced to 30 mg/l from January 2007. A risk-based approach for the management of produced water in the North Sea has been implemented since 2013. Treatment technology continues to evolve and improve, particularly membrane technologies that enable oil-contaminated water to be treated and increasingly made available for reuse for agriculture, livestock feeding, refill of aquifers and even drinking water. With a rapidly growing world population, water reuse will become increasingly important. Portable oil-in-water analysis methods and sophisticated online oil-in-water monitors have also been developed, bolstering the monitoring and optimization of treatment processes.

DR. MING YANG is the environmental consultancy services manager at NEL, a provider of technical consultancy, research, testing and program management services. Since joining NEL in 1998, he has been responsible for over 30 international conferences related to produced water, oil-in-water measurement and multiphase separation. Dr. Yang has also initiated and led several joint industry projects, and has presented and chaired many produced water-related events. In addition to publishing a book chapter on oil in produced water measurement, he has established a one-day training course that has been conducted numerous times globally. He was one of two authors who originally drafted the UK guidance notes on sampling and analysis of produced water and other hydrocarbon discharges. He joined NEL after working at Heriot-Watt University, where he was involved in research projects related to produced water characterization and re-injection. He also conducted research projects related to production chemicals and multiphase separation at the University of Manchester.

Environment and Safety S. V. BAPAT, Petrokon Utama Sdn Bhd, Brunei Darussalam

Consider post-design changes to confine a hazardous area Hazardous area classification (HAC) is carried out while performing basic design engineering, and HAC drawings are produced as one of the deliverables. The objective of the area classification is to avoid simultaneous occurrence of a flammable mixture and an ignition source. HAC drawings are used to select the electrical equipment, to plan the maintenance activities, to control the traffic movement within internal roads and to make the decisions on facility layout (FIG. 1). Note: HAC classification does not consider the toxicity of the handled fluids. The hazardous area and the zone sizes can be appropriately confined by following the steps outlined in this work. The hazardous area represents the volume of the plant, which contains significant quantities of flammable mixtures, as it exists during normal operating conditions, startup or shutdown. Flammable mixtures can form when the surrounding air mixes with the hydrocarbon vapors released from potential leak sources, such as flanged joints, pump seals, sample points, sumps and vents. Each source will generate a hazardous area around it. Its dimension—i.e., the hazard radius—will depend on the fluid category (i.e., the nature of the hydrocarbon fluid as liquefied gas, gas, boiling liquid or stabilized liquid), operating pressure and release hole size. Determining whether the hydrocarbon vapor is lighter or heavier than air is also an important criterion. Each plant consists of many such sources; therefore, several hazardous areas exist that may overlap each other or remain independent of one another. Generally, the sources at the periphery govern the boundary of a hazardous area. Depending on the release duration and type of ventilation, the area will be designated as Zone 0, 1 or 2 (FIG. 2). Any ignition source that can provide either ignition energy (such as the hot surfaces of motors, electric heaters and junction boxes) or generate sparks (such as during maintenance activity or by incoming traffic vehicles) may cause ignition of a flammable mixture. Therefore, the area is classified so that the likelihood of a flammable mixture spreading over the ignition sources is avoided or minimized. The key steps in area classification are: 1. Identify the release source and establish the size 2. Identify the fluid category that could be released through each source, along with its operating temperature and pressure

3. Estimate the hazard radius from the standard,1 or by performing dispersion calculations 4. Establish the duration of release and nature of ventilation, and determine the zone type 5. Identify the cloud limits and build the hazardous area boundaries 6. Perform analysis 7. Recommend the gas group (allowable energy) and the temperature class (allowable maximum surface temperature) for electrical equipment. Analysis of HAC classification. Similar to piping and in-

strumentation diagrams (P&IDs), HAC drawings remain live during the “define-and-execute” phase, and then throughout the plant life. Changes in hazardous area may occur at several different phases, as outlined in the following sections. 1. During the engineering phase, or while performing the detailed design (DD). During the DD, the location of each flange and equipment is known, Uncertified electrical items

Internal and external roads

Control room

Suction intake of diesel generator

Suction intake of air compressor

Helideck and boat landing

Hazardous area based on various release sources– Zones 0, 1, and 2 Fla m Sep

a ra ti

m a ble m ixtur

e

o n dist a n c e fr o m z

Unrestricted vehicle movement

s one

Admin. building

Unrestricted public movement

FIG. 1. Example of a hazardous area classification. Hydrocarbon Processing | AUGUST 2016 73

Environment and Safety and the detail of each source of release is available. Sometimes new equipment is added, or a few flanges are intentionally introduced to ease the installation. Sometimes in DD, the flare or vent loads are increased. The increased number of sources (i.e., new flanges) changes the previously prescribed hazard boundary, while the increased quantity of the released gas increases the zone area. 2. During a capacity revamp, or rejuvenation phase (brownfield projects). Capacity revamps can be performed either by changing the operating parameters of the existing facility or by installing new equipment (piping, vent stack, etc.) in an existing facility. When the operating parameters are changed, one of the governing parameters for the HAC will be an increase in operating pressure. Usually, an increase in operating pressure will be followed by the recalibration of trip setpoints and the installation of new relief valves. Due to the increase in operating pressure, the release rate through the existing sources will increase, thereby also increasing the hazard radii. When revamping an existing facility, new sources of release will be added. Depending on the operating pressure, fluid category and the size of source and its location, the new hazard radius may stretch the existing boundary to a new limit. Therefore, the existing HAC zones extend beyond their current limits and overlap their surroundings. The zones may even encroach on existing noncertified electrical equipment, utility systems (e.g., fire water hydrants safety showers), air compressor suction ducts, internal and external roads, and platforms on the helideck. 3. During the temporary operation phase that may determine the HAC. Examples of these temporary operations are: • Change in service of equipment—For example, a dedicated diesel tank and its associated system are used for storing and transporting kerosine • Performing intelligent pipeline pigging— Different fluid is used as a pig-driving medium at higher operating pressures

Zone 1 Flammable atmosphere is likely to occur in normal operation Zone 2 Flammable atmosphere is not likely to occur in normal operation; or, if it occurs, will exist for short time

Atmospheric roof tank Zone O Flammable atmosphere is continuously present or present for long periods Stored liquid

Example of hazardous area classification

FIG. 2. Hazardous area Zones 0, 1 and 2, based on various release sources. The sketch shows the objective of area classification to separate the sources of release from the sources of ignition.

74 AUGUST 2016 | HydrocarbonProcessing.com

Dyke

• Storing and transporting the fuel for temporary power generation—For example, at the time of commissioning. During the operations described, a new fluid category will be introduced. If it belongs to the “determining category,” then the hazard radii and the zone boundary will also increase. 4. During the reassessment phase, or during the “legacy-as-building” phase. In this case, the latest standards are applied to an existing facility after the “as-building” of P&IDs and plot plans, and then the extent of the HAC zones are reassessed. The new standards2 may define the zone limit to lower the flammability limit instead of being previously applied at the time of plant design. To reduce the flammability limit, additional air volumes will be needed. As a result, a larger volume and hazard radius will be required. 5. During the drawing preparation phase, due to human error. Human error is common to all previously described points. Sometimes, the criteria are applied without completely understanding the requirements. However, the wrong selection of fluid category, operating pressure or release hole size will affect the recommended radii and extend the zone size. A few examples can be considered: • A new facility was designed to gather production from onshore wells. It consisted of flowlines, production and test manifolds, a manual depressurization facility and a vent. The size of the Zone 2 cloud generated by the released gas was so large that it not only went beyond the plant fence, but also approached a nearby public facility. To limit the cloud size within the fence, the release rate and quantity were lowered by reducing the initial depressurization pressure and installing an in-line restriction. • An offshore platform was designed to produce fluids at high pressure (25 barg). It had a transformer (which was not certified) and was located outside Zone 2. The new wells to be drilled on that platform were intended to produce fluids at elevated pressure (75 barg). Each new flowline had a sample point. It was observed that the hazardous area generated from the sample point of the new flowlines encroached on the transformer. As a result, it was recommended to not install the sample, to not produce in the elevated-pressure regime and to not extend the deck to relocate the transformer. • In an existing onshore gas distribution station, the manual isolation valve was installed in the underground valve pit. When the HAC drawing was created, it was observed that Zone 2 overlapped the facilities around it, as well as the adjacent public road. The criterion of the drain sump, which requires a larger radius, was applied to the underground pit. • An HAC study was performed during the basic engineering of an offshore drilling platform.

Environment and Safety Based on its outcome, a non-rated distributed control system (DCS) and instrumented protective system (IPS) were ordered. During the DD, the DCS and IPS locations were classified as Zone 2. It was not possible to relocate the sources of release. As a result, the procured equipment was declared unfit. To avoid the procurement of zone-rated items, a separation distance from Zone 2 was provided by the deck extension. • In a crude oil export terminal, new crude transfer pumps were installed between the existing internal road and the tank dyke. It was observed during the HAC study that the edge of Zone 2, generated from the pump flange, encroached on the road. This meant that the flammable cloud would remain on the road. It was not possible to relocate the road or the flanges. To mitigate this issue, finger gates were installed to control the traffic movement, and the vehicles entering the facility were recommended to be equipped with a muffler. • An onshore hydrocarbon fluid receiving facility included pig launchers and receivers, as well as drain sumps. The sumps were designed to receive the drains after launching and receiving the pigs. The facility HAC drawings were “as-built” and issued to the latest standard. As the size of Zones 1 and 2 overlapped the facility, they encroached on the road to the facility. However, since pigging is an irregular operation, it was suggested to block the roads from traffic when pigging was scheduled to be performed. Other alternatives included sump modification to make vapor tight, sump relocation away from the road and the installation of a vapor barrier wall. However, these alternatives were not considered due to practical limitations and the temporariness of the operation. Recommendations. The hazardous area and the zone sizes can be confined under several considerations: 1. First, give weight to HAC requirements while preparing the basis for design (Bf D) and basic engineering. Involve competent personnel while preparing the Bf D and during the engineering phases. The initial study should address HAC requirements to avoid major rework. 2. Plan the piping routing study early. Consider grouping flanges and locating the release sources away from uncertified ignition sources. Proposals to locate the new sumps in the existing plant should be critically reviewed. If a pressurization unit is being considered, then it will be important to consider the effect of power loss, the reliability of the power supply and the possibility of supplying air from the hydrocarbon-free area. 3. If the controlling hazardous area is generated by sources that revive only during temporary operations, sampling, etc., then judge whether operating procedures can be used to counter this generation.

4. It is also important to select the right fluid category. The hazard radius depends on the fluid category; do not apply a worst-case approach. If fluid category C is selected over fluid category B, then the hazardous area will be larger than needed. If dispersion calculations are performed to estimate the radius, then the composition considered must represent the intended fluid category. 5. Select the operating conditions that are representative of normal conditions. Sources operating at high operating pressure require a large radius. If a pressure trip setpoint or the relief valve set pressure is selected, then verify that it is representative and not excessively high compared to the maximum operating pressure range. Explore the possibility of reducing the initial depressurization pressure considered for the vent rate calculations. 6. The size of the release source should be precise. Sources of releases are represented by the flanges, pump seals, sample points, restriction orifices from vent lines, and other items. Review each of them and ensure that their sizes are considered in estimating the hazard radius and vent rate calculations. Also, ensure that the dispersion modeling is representative. 7. Consider performing dispersion calculations, as these will help when the hazard radius becomes excessively high—particularly when the fluid category is C and the release size is large. 8. Consider the operation of the pressurization unit. If an uninterrupted electric supply is available, then one of the options could be to consider the air pressurization unit to provide the sufficient number of air changes. However, this unit has other limitations, such as the requirement of hydrocarbon-free air, the location of the suction duct away from the HAC boundary limit and the requirement of airtight doors. 9. Finally, it is important to isolate all ignition sources. Ignition sources can include incoming traffic, maintenance activities, hot surfaces or static electricity generation. ACKNOWLEDGMENT Gratitude to my family, management and colleagues. LITERATURE CITED SPE International, “Hazardous area classification for electrical systems,” Petrowiki, June 1, 2015, online: http://petrowiki.org/Hazardous_area_classification_for_ electrical_systems 2 Energy Institute, Area Classification Working Group, “Hazardous area classification,” 2016, online: https://www.energyinst.org/technical/safety/ei-15-hazardous-area-classification 1

SANJAY V. BAPAT is a process engineering manager at Petrokon Utama Sdn Bhd in Brunei. He has worked for 25 years in the process engineering of chemical, petrochemical, fertilizer, and oil and gas projects. He is a Fellow of IChemE, a chartered engineer and a Brunei Shell Petroleum-approved Level 2 technical authority. He is responsible for delivering front-end, conceptual design and detailed design deliverables, along with assurances of compliance with process safety requirements. Mr. Bapat develops concepts and methodologies and strategizes design condition selection, hazardous area classification, flare and blowdown studies, consequence modeling and surge analysis studies. He has presented papers at IChemE’s Third Hazards AP symposium and the Brunei International Conference on Engineering and Technology. He holds a postgraduate degree in chemical engineering. Hydrocarbon Processing | AUGUST 2016 75

REFOCUSED REIMAGINED REENERGIZED

AFPM Q&A AND TECHNOLOGY FORUM

Industry Experts. Peer to Peer Discussions. Relevant Knowledge Exchanges. Innovative Formats.

Q&A and Technology Forum Marriott Waterfront Baltimore September 26 – 28, 2016

2016

AFPM

Official Conference Show Publication

afpm.org/Conferences

Heat Transfer C. WRIGHT, Global Heat Transfer, Leeds, UK

Sample heat transfer fluids to offset carbon effects on thermal plant efficiency The long-term viability of a plant depends on maintaining continuous output and reducing production costs. Carbon accumulation in a heat transfer plant can lead to a reduction in plant efficiency. A number of interventions are available to counteract the buildup of carbon, including the adoption of a program of routine sampling of the heat transfer fluid (HTF), which potentially reduces the energy consumed and avoids unwanted interruptions to production that may occur through failure of component parts. An HTF is central to manufacturing and is used as a heat carrier in the processing of food, chemicals and energy.1 The main types of HTF media include air and other gases, water and steam, mineral-based HTFs, synthetic-based HTFs, molten salts and liquid metals. The cost efficiency of any manufacturing plant is critical to its long-term survival and includes the cost of the HTF, the storage of the energy produced, the cost of the heater, and the cost of land and property for the operation.2 The efficiency of a plant also includes the HTF’s physiochemical properties, such as mechanical and chemical stability, and the compatibility between the HTF and the heater and/ or storage material. However, efficiency can be measured more simply as production output vs. the energy consumed to generate the output. Therefore, revenue is equal to the production output divided by the energy consumed to produce the output. Energy consumption can be increased by technical inefficiencies, meaning that while output is maintained, the cost of energy consumption increases. The net effect is an increase in the cost of operation and a reduction in revenue. Maintaining continuous output and stable energy consumption are critical to a plant’s long-term cost efficiency. The condition of the HTF can negatively impact the energy consumed by a plant, as discussed in this article. Heat transfer fluid condition. Past research has shown that the condition of an HTF is improved by regular sampling with the optimal sampling frequency reported to be once every three months for a mineral-based HTF. Once sampled, an HTF is chemically analyzed to assess and ascertain its state of thermal cracking, the degree of oxidation, the system’s safety, the extent of HTF contamination and the degree of system wear. Prior research used the tests reported in TABLE 1 and ranked the occurrence of events. Results showed that total acid number (TAN) and closed flashpoint temperature ranked as the most frequently occurring events (i.e., ≥ 1 in 3 systems). Interestingly, when HTF systems were sampled less frequently (i.e.,

every 2–3 years), the buildup of carbon was the second-mostcommon event (i.e., ≥ 1 in 3 systems). This dropped to ≤1 in 20 cases when sampled at least once per year. Carbon occurs as a byproduct of thermal cracking and oxidation of an HTF. All HTFs will degrade over time, and this is why carbon needs to be routinely monitored. When carbon is formed, it becomes suspended in the HTF and acts as a sticky substance that will adhere to the internal surfaces of the HTF system, including the heater. FIG. 1 depicts a system that uses a mineral-based HTF. Also shown are the potential sources of carbon formation, which include: • Adherence of the carbon to the internal pipework, including the heater, which acts as an insulator, increasing the heat needed to heat the HTF • Adherence of carbon to internal pipework and a reduction in the diameter of the pipework • The formation of carbon sludge that accumulates in the expansion tank and circulates around the HTF system has the potential to bake onto internal pipework. Changes in pump efficiency. In normal function, the speed of the pump in a thermal plant drives the flow of the HTF. The efficiency of the pump is affected by the buildup of carbon as Atmosphere Expansion tank: Carbon sludge formation 1S1

Expansion tank

Pipework: Carbon sludge circulated through system Pipework: Carbon buildup and blockages

Oil/gas separator

TE2

Heater: Carbon baked onto internal surface

Stop valve Return Supply

Stop valve

Heater PI Pressure gauge TIC TEI

Gate valve Pump

Drain Strainer Stop valve

FIG. 1. Typical system that uses a mineral-based HTF. Note: The sources of carbon formation are highlighted. Hydrocarbon Processing | AUGUST 2016 77

Heat Transfer it increases the kinematic viscosity of the HTF. Kinematic viscosity of a non-Newtonian fluid, such as an HTF, is calculated by dividing the absolute viscosity (also known as dynamic viscosity) by its mass density. The viscosity of an HTF is temperature dependent, so the reference temperature must be standardized for the results to have any meaning. In the laboratory, kinematic viscosity is normally presented at 40°C and 100°C. Increased carbon in the HTF raises kinematic viscosity. This means the pump must consume more energy, assuming a constant temperature to pump the HTF and to overcome the increased resistance to flow presented by the higher viscosity of the HTF, as shown in the bottom left of FIG. 2. Likewise, the buildup of carbon on internal pipework reduces the inNormal carbon levels Normal kinematic viscosity

Pump

Pump

This is the case for a virgin HTF and reflects normal pump speed and resistance to flow Increased kinematic viscosity

Increased carbon levels

Pump speed needs to increase to sustain the same flow

Pump

Pump

Pump speed needs to increase to overcome the increase in kinematic viscosity, which increases the resistance to flow

Pump speed needs to increase to overcome the increase in kinematic viscosity and increased resistance to flow

FIG. 2. The response of changes in pump speed to changes in HTF kinematic viscosity and resistance to flow. Reduce the operating temperature to reduce the rate of thermal cracking “temperature”

Replace the existing HTF with a synthetic-based HTF “recharge”

Drain some of the HTF and top up with a virgin HTF “dilution”

To manage carbon forming in the HTF

Install a nitrogen blanket to reduce the exposure of the HTF to oxygen, which can lead to carbon sludge formation “nitrogen blanket”

Install a filter to extract carbon from the circulating HTF “filtration”

Incorporate a program of routine sampling and chemical analysis “sampling”

Add antioxidant packs to reduce oxygenation of the HTF, which can lead to carbon sludge formation “antioxidants”

FIG. 3. Options for managing the buildup of carbon in an HTF system.

ternal diameter of pipes and increases the resistance to flow. For turbulent flow, there is increased resistance, and much higher pressures are required to drive HTF flow. The point at which turbulent flow occurs is referred to as the critical velocity and calculated as: (viscosity × Reynold’s number) / (2 × density x radius). Therefore, radius and viscosity contribute to the overall resistance experienced under turbulent flow conditions. For changes in internal diameter, the pump will consume more energy to achieve sufficient pressure to drive flow (FIG. 2, top right). In real life, viscosity changes and carbon buildup will occur during the process of thermal degradation, and changes in resistance to flow will occur as a result of the increased viscosity and reduced conductivity of the pipework. When combined, energy consumption and the demand on the pump are increased (FIG. 2, bottom right). The buildup of carbon on the heater’s internal surfaces cannot be ignored. While this does not form part of the model presented in FIG. 2, carbon buildup would lead to carbon lining the internal surfaces of pipework. As carbon is a good insulator, more energy would be needed to achieve the same operational temperature. The net effect is an increase in the energy consumed to heat the HTF. Practical methods to reduce carbon. Monitoring carbon is important for efficient operations and lower OPEX. A number of options exist to help reduce carbon buildup. These options are outlined in FIG. 3 and include: • Sampling—Prior research has shown the effectiveness of increased sampling, which has been associated with improved HTF condition. This is based on the association between fluid cleanliness and component life. • Temperature—HTF manufacturers generally recommend that an HTF be sampled at least once per year when operating near its upper operating temperature. Some manufacturers recommend that this sampling be conducted twice yearly if the operating temperature is 20°C below its upper operating temperature. This scenario relates to Arrehenius’ Law, which shows that a correlation exists between the rate of a reaction and temperature, meaning the rate of a fluid’s degradation doubles for every 10°C rise in temperature. • Dilution—The HTF in the system is partly drained and then filled with virgin HTF to dilute the existing fluid. This process effectively removes some of the carbon and other degradation byproducts.

TABLE 1. Routinely conducted tests when chemically analyzing an HTF Test

Unit

What this test relates to

Carbon residue

% weight

The extent of thermal cracking and/or oxidation

Total acid number

mg KOH per g HTF

The extent of oxidation

Open and closed flashpoint temperature

°C

The extent of thermal cracking and/or system safety

Kinematic viscosity

mm /s

The extent of thermal cracking and/or oxidation

Water content

ppm

The degree of contamination

Ferrous wear debris (e.g., insoluble debris)

ppm

The degree of system wear

Elements (e.g., silicon)

ppm

The degree of contamination

78 AUGUST 2016 | HydrocarbonProcessing.com

2

Heat Transfer • Filtration—Filtration effectively removes containments from the system. If contaminants are left in the HTF, then they can catalyze the degradation of the fluid. In existing systems, filters with finer pores than the strainer (FIG. 1) can be used as a temporary or permanent addition to an HTF system and enable the continuous filtration of particles. The effectiveness of this approach can be demonstrated by incorporating an assessment of fluid cleanliness (i.e., ISO 4406:1999) to quantify the number and distribution of particulates suspended in the HTF. • Recharge—An option to drain and refill an HTF system with a virgin HTF always exists. The client can choose to use either a mineral- or synthetic-based HTF. Syntheticbased HTFs can be used at much higher temperatures and are more resistant to thermal degradation. • Nitrogen blanket—The effect of oxygen (FIG. 1) is detrimental to an HTF at temperatures exceeding 60°C and leads to the formation of corrosive acids, carbon sludge and carbon fouling. The use of a nitrogen blanket to prevent the HTF from coming into contact with air is an appropriate countermeasure. • Antioxidants—Oxygen can significantly increase the degradation of an HTF and damage HTF system components. Antioxidant packs or repellents are used to deplete the oxygen in the HTF. Recommendations. The revenue from an efficient operation can be understood in terms of the ratio of the revenue gained from the production output relative to the cost of energy to produce the output, as depicted in FIG. 4. In Scenario 1, this model assumes that an HTF is not routinely managed and accepts that the accumulation of carbon will eventually lead to an increase in energy consumption. In the longer term, this could also lead to component failures and interrupt operation output, as seen in FIG. 4. In Scenario 2, the effect of carbon is still an influencing factor, but HTF sampling has been incorporated into the model. This

=

Production output Energy consumed to produce the output (and replacement parts)

+

1 Cost of routine heat transfer fluid maintenance

Increas e

Decrease

Kinematic viscosity of the HTF Diameter of internal pipework Heater efficiency

Leads to

Increas

e

Revenue

Expansion tank: Carbon sludge formation Pipework: Carbon sludge circulated through system Pipework: Carbon buildup and blockages Heater: Carbon baked onto internal surface

FIG. 4. The effect of carbon formation on plant efficiency and revenue. In Scenario 1, carbon accumulation leads to increased kinematic viscosity and reduced pipework diameter and heater efficiency. If left unmanaged, carbon accumulation will eventually lead to reduced production output, increased energy consumption and an overall decrease in revenue. In Scenario 2, the sources of carbon are the same; however, routine sampling is incorporated. This process can be used to maintain kinematic viscosity, internal pipework diameter and heater efficiency.

scenario represents an additional cost and loss of revenue, but it is a proactive approach to avoid the longer-term detrimental effects of carbon accumulation. This scenario works to maintain constant energy consumption. The net effect is that revenue will remain relatively constant as output and energy consumption are consistently maintained. Another advantage is that the cost of replacing component parts is potentially avoided as routine maintenance is used to correct increases in the levels of carbon and sustain the HTF and the HTF system. Options available for the management of carbon include the management of operating temperature, dilution of the HTF, installation of a temporary or permanent filtration unit, the complete replacement of the HTF, and strategies to manage oxidation of the HTF. ACKNOWLEDGEMENTS The author would like to acknowledge the writing support provided by Red Pharm Communications, a part of the Red Pharm Co. LITERATURE CITED Wagner, W., “Heat transfer technique with organic media,” Heat Transfer Media, 2nd Ed., begellhouse, Graefelfing, Germany, 1997. 2 Tian, T. and C. Y. Zhao, “A review of solar collectors and thermal energy storage in solar thermal applications,” Applied Energy, Iss. 104, 2013. 1

CHRIS WRIGHT is a research scientist and holds a BSc degree and a PhD from the University of Leeds in the UK. His research focuses on the use and maintenance of heat transfer fluids in manufacturing and processing, including specialist chemicals, food, pharmaceutical and solar sectors.

Introducing the

US GAS PROCESSING PLANT DIRECTORY

US GAS PROCESSING

PLANT DIRECTORY 500+ Plants with Information about Name, Capacity, Plant Scope and Detail, and Owner/ Operator Information. Gain the market knowledge to grow your business and inform your decisions in the booming United States gas market. $1,195 per edition. Searchable, digital format. Group rates and site licenses are available.

Order at GulfPub.com/GPPD or call + 1 (713) 520-4426.

Hydrocarbon Processing | AUGUST 2016 79

www.ConstructionBoxscore.com

Logon for a

FREE ONLINE DEMO!

, G IN IN F E R L A B O L G E H T R O F E MARKET INTELLIGENC S IE R T S U D IN G N /L G IN S S E C O R P S A G D N A L A IC M E H C O R T E P Hydrocarbon Processing’s Construction Boxscore Database,

the most reliable source to track active construction projects in the refining, petrochemical, gas processing, LNG and solids industries throughout the world, now reaches further and is more powerful than ever before!

Welcome to the NEW ase Construction Boxscore Datab

• Project details on thousands of active projects and global construction contracts, including contact information for key personnel • Advanced search that filters the listings by project type, scope, region, investment and more • Daily updates for new and newly updated projects • The weekly Boxscore Update e-newsletter with new listings and trends analysis

For more information, contact: Norman Adams at [email protected] or +1 (713) 525-4626 J’Nette Davis-Nichols at [email protected] or +1 (713) 520-4426

BOB ANDREW, TECHNICAL EDITOR [email protected]

Innovations Breakthrough in activation of methane Turning methane into a high-value chemical was achieved by German chemical company Grillo after years of intensive research. The new process leads to highpurity methanesulfonic acid (MSA) by direct sulfonation of methane with sulfur trioxide (FIG. 1). Initial large-scale production is planned for 2019. Methane is the main component of natural gas and is, thus far, primarily being burned for heat and energy. Industry and science have been searching for a material use for methane. Besides direct sulfonation, which has now been achieved by Grillo, research focuses on direct oxidation of methane to methanol and oxidative coupling to ethylene. Grillo’s chemicals division has solved the challenge of methane’s limited reactivity by utilizing a tailored reaction environment and specific activators. The process has been continuously optimized and now achieves almost full conversion at mild reaction conditions. The process, Grillo-Methane-Sulfonation, is cost-competitive and based on natural gas and sulfur trioxide (SO3) as feedstocks. It is free of environmentally problematic intermediate and byproducts. MSA is a modern, green product. The very strong acid is not oxidative and is readily biodegradable and toxicologically unproblematic. MSA is being used in the electroplating, electronics, industrial cleaning and pharmaceutical industries. It is experiencing strong overall growth and, at a significant scale, potential applications are many.

Techniques (HFT)’s QuickPurge system (FIG. 2) reduces waiting times, producing zero color welds and no loss of corrosion resistance caused by oxidation. A 36-in. diameter stainless steel pipe, for example, can be purged down to 100 ppm of oxygen in under 32 minutes. This can be compared with purging using foam dams that will outgas heavily into the weld zone and take hours instead of minutes. High-speed pipe purging system QuickPurge is one of the leaders in the field of purging tube, pipe and pipeline joints over a 6-in. diameter, where oxygen levels are required to be 10 ppm or less. Combining IntaCal with the integrated PurgeGate device makes it possible to safely inflate the dams with argon gas and purge the space between the dams where the weld joint is located. It is almost impossible for the inflatable dams to burst as a result of undue pressure or accidental flow increase. All systems are manufactured as standard with a hose for connecting to Weld Purge Monitor, which can read oxygen levels down to 10 ppm. Materials chosen for QuickPurge are resistant to the higher weld heat present, while exhibiting lower outgassing rates to prevent weld contamination. For heattreated chrome steel pipe joints, HFT manufactures the HotPurge range for the higher and longer temperature exposure requirements. The sleeve between the

dams on the QuickPurge systems reduces the volume to be purged by two thirds. Sleeve lengths for each size have been carefully calculated so that the QuickPurge systems can be pulled around 90° elbows for purging connecting joints.

FIG. 1. Molecular structure of methanesulfonic acid.

FIG. 3. Pressure-and-vacuum relief for storage tanks.

Select 2 at www.HydrocarbonProcessing.com/RS

Reduced storage tank emissions Pentair Valves & Controls’ Anderson Greenwood 4000 Series of pressure and vacuum relief valves (FIG. 3) are compliant with the latest API 2000 standard. Developed accordingly, the valves are the first high-capacity, full lift valves verified to meet the requirements of the 7th Ed. of API 2000, which covers normal and emergency vapor venting requirements for bulk liquid storage tanks.

FIG. 2. QuickPurge system for welding large pipelines.

Select 1 at www.HydrocarbonProcessing.com/RS

Big pipes welded faster, less contamination Time is money, and waiting for pipe joints to be purged and ready for welding can take up to four hours, depending on the diameter of the pipe and the chosen purging method. Huntingdon Fusion

Hydrocarbon Processing | AUGUST 2016 81

Innovations The series has been engineered to provide increased flow capacities and will fully open at 10% overpressure, helping protect tanks from physical damage caused by internal pressure fluctuations. As a result, the valves can be set more closely to a storage tank’s maximum allowable working pressure (MAWP) or maximum allowable working vacuum (MAWV), enabling customers to fill and empty tanks more quickly and operate them at higher pressures. The 4000 Series valves remain closed longer and enhance flow capacity, resulting in increased productivity and reduced evaporation. Select 3 at www.HydrocarbonProcessing.com/RS

New passive harmonic filters Common electrical components like motors, pumps, fans, automation equipment, DC fast chargers, equipment with front-end, six-pulse rectifiers, among others, introduce harmonics into the electrical system. High harmonic distortion can cause failures or malfunctions of electrical devices. They also cause a temperature rise in the electrical network and equipment, resulting in losses and shorter service life. Passive harmonic filters, such as the ECOsine Passive by Schaffner EMC,

tuned to the offending harmonic order can correct the sine wave before the harmonics can cause damage. Schaffner EMC introduces the FN3415, the newest addition to the ECOsine family of passive harmonic filters for 60 Hz, 480 or 600 VAC applications (FIG. 4). Available in various NEMA enclosures, they are easily installed and commissioned, providing immediate benefit to the electrical system by limiting the amplitude of the existing harmonics, reducing losses and allowing equipment to operate more efficiently with a longer service life. Proven in industrial settings, Schaffner passive harmonic filters are ideal for use with virtually any kind of power electronics with front-end, six-pulse rectifiers, including non-linear loads such as motor drives/VFDs, factory automation equipment, fans, pumps, chillers and other HVAC equipment, DC fast chargers, mission critical systems/datacenters, water/wastewater equipment, oil and gas drilling, and more. Designed to operate at 99% efficiency, these proven products reduce harmonics to meet IEEE-519 (2014) with TDD and THDI performance at full, light and noload conditions below 5%, while THDv ratings are below 3%. Schaffner also offers more basic designs for installations with less critical harmonic mitigation needs. NEMA 1 and NEMA 3R enclosures, wallmounted or free-standing, are standard. Select 4 at www.HydrocarbonProcessing.com/RS

Noise mapping software upgrade FIG. 4. Harmonic filters reduce electrical distortion.

FIG. 5. Acoustic contours from sound planning tool.

82 AUGUST 2016 | HydrocarbonProcessing.com

Following a successful trial in Germany, noise-mapping software developer SoundPLAN has released version 4.0 of

FIG. 6. Monitoring vibration in gears and bearings.

its SoundPLANessential software, used to calculate noise emissions from roads, railways and industry sources (FIG. 5). The software is designed for users with an occasional need for a noise map, helping them achieve reliable tabular and graphical results for a standard noise calculation with minimal time investment. SoundPLANessential 4.0 introduced a new rail standard for noise predictions (Schall 03-2012). Improvements in 4.0 include better positioning of receiver points, enhanced manipulation of objects within the maps, better use of colors for clarity and a new display option of geometry bitmaps in a 3D view. Select 5 at www.HydrocarbonProcessing.com/RS

New vibration technology boosts maintenance profitability Technologies for monitoring machine vibration levels have been in use for decades. The further development of such technologies, however, has been slow. With the rapid advancement of digital component technologies providing increasing opportunities, time has caught up with conventional vibration monitoring methods. HD ENV by SPM is a new vibration enveloping technology using sophisticated digital algorithms to provide earlier warning of machine faults than conventional vibration monitoring technologies. HD ENV is extremely sensitive to changes in vibration levels, making it possible to identify gear and bearing faults (FIG. 6) at a very early stage and closely monitor their continued development. This provides the opportunity to optimize corrective and predictive maintenance scheduling, potentially driving maintenance profitability to new levels. The early fault detection capability of HD ENV is attributed to digital signal processing algorithms that, combined with low-noise hardware and ICP-compatible accelerometers, extract and enhance relevant gear and bearing information with clarity from noisy industrial environments. HD ENV presents disturbance-free vibration data in high-definition quality. Clear spectrums and time signals, where the source of the signal is easily identified, provide a snapshot of machine condition, alerting maintenance of potential problems. Select 6 at www.HydrocarbonProcessing.com/RS

MARKETPLACE / [email protected]  /  +1 (972) 816-3534

Why Should You Filter Your Water?

SALE ▪ RENT ▪ LEASE Superheat & Saturated Boilers to 250,000pph 750psig Custom Design & Manufacture Too!

Scale  formation  reduces  the  heat  transfer  rate  and increases  the  water  pressure  drop  through  the  heat exchanger  and  pipes.  In  fact,  one  study  has  shownthat .002" fouling will increase pumping needs by 20%.    

▪In Stock Assorted Sizes ▪ Ultra Low Nox Available ▪ SCRs Available

The Best Engineered Water Filteration Solution Always Costs Less 2672 S La Cienega Blvd, Los Angeles, CA 90034 USA (800) 226-1942  (310) 839-2828   Fax (310) 839-6878 www.tekleen.com    [email protected] Select 201 at www.HydrocarbonProcessing.com/RS

SURPLUS GAS PROCESSING/REFINING EQUIPMENT 25 MMCFD x 1100 PSIG PROPAK REFRIGERATION PLANT 28 TPD SELECTOX SULFUR RECOVERY UNIT 1100 BPD LPG CONTACTOR x 7.5 GPM CAUSTIC REGEN NGL/LPG PLANTS: 10–600 MMCFD AMINE PLANTS: 60–3300 GPM SULFUR PLANTS: 10–180 TPD FRACTIONATION: 1000–25,000 BPD HELIUM RECOVERY: 75 & 80 MMCFD NITROGEN REJECTION: 25–100 MMCFD MANY OTHER REFINING/GAS PROCESSING UNITS We offer engineered surplus equipment solutions.

Select 202 at www.HydrocarbonProcessing.com/RS

pull out all the stops

4.25" wide x 4" high

ARTICLES FOR DISTRIBUTIONCMYK color Repurpose editorial content for distribution • Electronic Reprints • High-Quality Glossy Handouts • Personalized Direct Mail Products • Cross Media Marketing • Plaques and Framed Prints

For additional information, please contact Foster Printing Service, the official reprint provider for Hydrocarbon Processing.

Call 866.879.9144 or [email protected]

Bexar Energy Holdings, Inc.

Phone 210-342-7106 • Fax 210-223-0018 www.bexarenergy.com • Email: [email protected]

Select 203 at www.HydrocarbonProcessing.com/RS

Chemical Resistant Epoxy Room Temperature Curing EP41S • Withstands alcohols and fuels • Fast curing • Abrasion resistant • For high strength coatings, liners and sealants

Gulf Publishing Company is seeking an ADVERTISING REPRESENTATIVE to cover India for World Oil, Hydrocarbon Processing and use this logo on white or lighter backgrounds Gas Processing. This individual will be responsible for print, digital and event sales. They must live in India and have a background use this logo on black or darker inTheadvertising sales. Experience backgrounds. Printing Service is in white. in the oil and gas market a plus. Please forward all resumes to [email protected] C: 56 M: 0 Y: 100 K: 27 PMS 370 C

Hackensack, NJ 07601 USA +1.201.343.8983 • [email protected]

www.masterbond.com Select 204 at www.HydrocarbonProcessing.com/RS

Detailed and up-to-date information for active construction projects in the refining, gas processing, and petrochemical industries across the globe ConstructionBoxscore.com

Hydrocarbon Processing | AUGUST 2016 83

ADVERTISER INDEX  /  HydrocarbonProcessing.com The first number after the company name is the page on which an advertisement appears. The second number is the Reader Service Number. There are two ways readers can obtain product and service information: go to www.HydrocarbonProcessing.com/RS, follow the instructions on the screen, and your request will be forwarded for immediate action, or go online to the advertiser's website listed below.

Company

Page

RS#

Website

ABB Motors and Generators Service Team..... 43

(63)

(73) (56)

Auma Riester GmbH & Co Kg ....................... 27 (153) www.info.hotims.com/61390-153

www.info.hotims.com/61390-155

Gulf Publishing Company Construction Boxscore Database...............80 HP Webcast .............................................61 HP Marketplace ....................................... 83

(51)

www.info.hotims.com/61390-51

Software ................................................ 67 US Gas Processing Plant Directory............. 79

(63)

www.info.hotims.com/61390-63

Bluebeam Software Inc.................................8

(57)

Events—GasPro..................................... 50A

www.info.hotims.com/61390-56

Baldor Electric Company ............................. 43

Dresser-Rand..............................................31

Heat Transfer Research Inc. ......................... 23 (151) www.info.hotims.com/61390-151

(69)

Hoerbiger .................................................. 26 (152)

Clock Spring Company ................................ 47 (157)

Inpro / Seal Company ................................. 35 (154)

www.info.hotims.com/61390-69 www.info.hotims.com/61390-157

www.info.hotims.com/61390-152 www.info.hotims.com/61390-154

Company

Page

RS#

Linde Engineering North America ................ 87

(66)

Website

Dyna-Therm ...............................................41 (155)

www.info.hotims.com/61390-73

Axens ........................................................ 88

RS#

www.info.hotims.com/61390-57

AFPM......................................................... 76

Ariel Corporation.........................................16

Page

Website

www.info.hotims.com/61390-63

Ametek Process Instruments ....................... 22

Company

www.info.hotims.com/61390-66

Merichem Company......................................6

(84)

www.info.hotims.com/61390-84

Rosen Swiss AG .......................................... 28

(61)

www.info.hotims.com/61390-61

Saudi Aramco .............................................. 5

(82)

www.info.hotims.com/61390-82

Shell Research Ltd ......................................49

(92)

www.info.hotims.com/61390-92

Siemens AG ................................................. 2

(75)

www.info.hotims.com/61390-75

Spraying Systems Co ...................................37

(67)

www.info.hotims.com/61390-67

Turbomachinery Laboratory........................ 63 Zwick Valves ..............................................44 (156) www.info.hotims.com/61390-156

ZymeFlow Decon Technology .......................18

(93)

www.info.hotims.com/61390-93

This Index and procedure for securing additional information is provided as a service to Hydrocarbon Processing advertisers and a convenience to our readers. Gulf Publishing Company is not responsible for omissions or errors.

Catherine Watkins, Publisher Phone/Fax: +1 (713) 520-4421 E-mail: [email protected] www.HydrocarbonProcessing.com SALES OFFICES—NORTH AMERICA IL, LA, MO, OK, TX Josh Mayer Phone: +1 (972) 816-6745, Fax: +1 (972) 767-4442 E-mail: [email protected] AK, AL, AR, AZ, CA, CO, FL, GA, HI, IA, ID, IN, KS, KY, MI, MN, MS, MT, ND, NE, NM, NV, OR, SD, TN, TX, UT, WA, WI, WY, WESTERN CANADA Ryan Akbar Phone/Fax: +1 (713) 520-4449 Mobile: +1 (832) 691-6053 E-mail: [email protected] CT, DC, DE, MA, MD, ME, NC, NH, NJ, NY, OH, PA, RI, SC, VA, VT, WV, EASTERN CANADA Merrie Lynch Phone: +1 (617) 357-8190, Fax: +1 (617) 357-8194 Mobile: +1 (617) 594-4943 E-mail: [email protected] CLASSIFIED SALES Gerry Mayer Phone: +1 (972) 816-3534, Fax: +1 (972) 767-4442 E-mail: [email protected] DATA PRODUCTS J’Nette Davis-Nichols Phone/Fax: +1 (713) 520-4426 E-mail: [email protected]

84 AUGUST 2016 | HydrocarbonProcessing.com

SALES OFFICES—EUROPE

SALES OFFICES—OTHER AREAS

FRANCE, GREECE, NORTH AFRICA, MIDDLE EAST, SPAIN, PORTUGAL, SOUTHERN BELGIUM, LUXEMBOURG, SWITZERLAND, GERMANY, AUSTRIA, TURKEY Jim Watkins Phone: +33 (0)6 76 35 11 52 E-mail: [email protected]

CHINA—Hong Kong Iris Yuen Phone: +86 13802701367 (China) Phone: +852 69185500 (Hong Kong) E-mail: [email protected]

ITALY, EASTERN EUROPE Fabio Potestá Mediapoint & Communications SRL Phone: +39 (010) 570-4948 Fax: +39 (010) 553-0088 E-mail: [email protected] RUSSIA/FSU Lilia Fedotova Anik International & Co. Ltd. Phone: +7 (495) 628-10-333 E-mail: [email protected] UNITED KINGDOM/SCANDINAVIA, NORTHERN BELGIUM, THE NETHERLANDS Michael Brown Phone: +44 161 440 0854 Mobile: +44 79866 34646 E-mail: [email protected] REPRINTS Rhonda Brown, Foster Printing Service Phone: +1 (866) 879-9144 ext. 194 E-mail: [email protected]

BRAZIL—Rio de Janeiro Marco Antonio Monteiro Mobile: +55 21 99616-4347 Fax: +55 21 2240-5077 E-mail: [email protected] INDIA Catherine Watkins Phone/Fax: +1 (713) 520-4421 E-mail: [email protected] INDONESIA, MALAYSIA, SINGAPORE, THAILAND, AUSTRALIA—Perth Peggy Thay Publicitas Singapore Pte Ltd Phone: +65 6836-2272, Fax: +65 6634-5231 E-mail: [email protected] JAPAN—Tokyo Yoshinori Ikeda Pacific Business Inc. Phone: +81 (3) 3661-6138, Fax: +81 (3) 3661-6139 E-mail: [email protected] KOREA Young-Seoh Chinn JES Media, Inc. Phone: +82 (2) 481-3411/3, Fax: +82 (2) 481-3414 E-mail: [email protected]

ALISSA LEETON, CONTRIBUTING EDITOR [email protected]

Events AUGUST GTL Technology Forum, Gulf Publishing Company Events, Aug. 2–3, Norris Conference Centers—CityCentre, Houston, Texas GTLTechForum.com (See box for contact information) NAPE, Aug. 10–11, George R. Brown Convention Center, Houston, Texas (See box for contact information) 10th Annual Maintenance and Reliability Symposium, The American Chemistry Council (ACC) Annual Meeting, Aug. 18–19, Moody Gardens, Galveston, Texas P: 281-538-9996 [email protected] www.smrphouston.org AFPM Cat Cracker Seminar, Aug. 23–24, Royal Sonesta Hotel, Houston, Texas (See box for contact information) ONS 2016, Aug. 29–Sept. 1, Stavanger, Norway P: +47 51-84-90-40 [email protected] www.ons.no/2016

SEPTEMBER World Heavy Oil Congress, Sept. 6–9, Stampede Park, Calgary, Alberta, Canada P: +1 403-209-3555 or +1 888-799-2545 [email protected] www.worldheavyoilcongress.com 48th Annual ECC Conference, Sept. 7–10, JW Marriott Desert Springs, Palm Desert, California [email protected] www.ecc-conference.org

Oil Sands Trade Show, Sept. 13–14, Suncor Community Leisure Centre, Fort McMurray, Alberta, Canada P: +1 403-209-3555 or +1 888-799-2545 [email protected] www.oilsandstradeshow.com Gas Pro Americas, Gulf Publishing Company Events, Sept. 13–14, Norris Conference Centers—CityCentre, Houston, Texas GasProcessingConference.com (See box for contact information) AIChE 61st Annual Safety in Ammonia Plants and Related Facilities Symposium, Sept. 18–22, Grand Hyatt Denver, Denver, Colorado P: +1 800-242-4363 www.aiche.org/conferences Process Safety Summit, Sept. 25–27, Abu Dhabi UAE, Sofitel, Abu Dhabi P: +971 4-364-2975 [email protected] www.oilandgasprocesssafety.com National Association of Corrosion Engineers (NACE) Corrosion Technology Week, Sept. 25–29, Royal Sonesta Houston Galleria, Houston, Texas P: 281-228-6413 [email protected] www.ctw.nace.org AFPM Q&A and Technology Forum, Sept. 26–28, Baltimore Marriott Waterfront, Baltimore, Maryland (See box for contact information)

EUROCORR 2016—European Corrosion Congress, Sept. 11–15, Le Corum Congress Center, Montpellier, France www.eurocorr2016.org

Operational Excellence in Refining and Petrochemicals, Sept. 26–28, Norris Conference Centre—Houston CityCentre, Houston, Texas P: +44 0-20-7036-1300 [email protected] www.opexinrefiningand petrochem.com

Turbomachinery & Pump Symposia, Sept. 12–15, George R. Brown Convention Center, Houston, Texas P: +1 979-845-7417 [email protected] www.pumpturbo.tamu.edu

Polyurethanes Technical Conference, Sept. 26–28, Hilton Baltimore, Baltimore, Maryland P: 202-249-6121 [email protected] www.polyurethane. americanchemistry.com

OCTOBER Yokogawa Users Conference and Exhibition, Oct. 3–6, Renaissance Orlando at SeaWorld Hotel, Orlando, Florida www.yokogawausers conference.com NAPE, Oct. 12–13, Colorado Convention Center, Denver, Colorado (See box for contact information) National Safety Council (NSC), Oct. 15–21, Anaheim Convention Center, Anaheim, California P: +1 630-285-1121 [email protected] www.congress.nsc.org ACC Annual Meeting, Oct. 16–19, Moscone Center, San Francisco, California P: +1 202-293-4103 www.acc.com

The Abu Dhabi International Petroleum Exhibition & Conference (ADIPEC), Nov. 7–10, Abu Dhabi National Exhibition Centre P: +971 0-2-6970-500 [email protected] www.adipec.com API 11th Annual Cybersecurity Conference for the Oil & Natural Gas Industry, Nov. 9–10, Westin Houston Memorial City, Houston, Texas (See box for contact information) API Fall Refining and Equipment Standards Meeting, Nov. 14–17, Hyatt Regency New Orleans, New Orleans, Louisiana (See box for contact information) European Autumn Gas Conference, Nov. 15–17, The Hague, Netherlands P: +44 0-20-3772-6080 [email protected] www.theeagc.com

49th GOMA Symposium, Oct. 19–21, Jure Hotel, Šibenik, Croatia P: +385 1-48-73-549 [email protected] www.fuels.goma.hr

FEBRUARY 2017 Egypt Petroleum Show (EGYPS), Feb. 14–16, CICEC, Cairo Egypt P: +971 0-4445-3726 [email protected] www.egyptpetroleumshow.com

RIO Oil & Gas 2016 Expo and Conference, Oct. 24–27, Rio de Janeiro, Brazil P: +55 21-2112-9080 [email protected] www.riooilgas.com.br/en Emerson Global Users Exchange, Oct. 24–28, Austin Convention Center, Austin, Texas [email protected] www.emersonexchange.org/ americas/ LARTC 5th Annual Meeting, Oct. 25–27, Mexico City, Mexico P: +44 0-20-7384-8022 [email protected] www.lartc.events.gtforum.com/

NOVEMBER Women’s Global Leadership Conference, Gulf Publishing Company Events, Nov. 1–2, Hyatt Regency Houston, Houston, Texas WGLconference.com (See box for contact information)

Hydrocarbon Processing/ Gulf Publishing Company Events P: +1 713-520-4475 [email protected] [email protected] American Fuel & Petrochemical Manufacturers (AFPM) P: +1 202-457-0480 [email protected] www.afpm.org/Conferences American Petroleum Institute (API) P: +1 202-682-8195 [email protected] www.api.org NAPE P: +1 817-847-7700 [email protected] www.napeexpo.com

Hydrocarbon Processing | AUGUST 2016 85

MIKE RHODES, MANAGING EDITOR [email protected]

People

Courtney McShane has joined Willbros Group Inc. as director of business development for the Marcellus Shale region and Eastern Seaboard. Prior to joining Willbros, she represented several pipeline construction business units for PLH Group. Ms. McShane will be responsible for the continued expansion of Willbros’ pipeline construction and maintenance services. Total has named Philippe Sauquet, the former head of the refining and chemicals division, to lead its newly created gas, renewables and power unit. He is also appointed executive VP of strategy and innovation and will remain a member of Total’s executive committee. He will assume his new role on September 1. In 2014, Mr. Sauquet was named president of refining and chemicals and joined Total’s executive committee. Total has made other executive appointments. Bernard Pinatel is joining Total as president of refining and chemicals and a member of Total’s executive committee. Namita Shah will serve as executive VP, people and social responsibility, and as as a member of Total’s executive committee. Ms. Shah will lead the HR division and oversee the newly created Total Global Services. JeanJacques Guilbaud will be senior advisor to the chairman and CEO.

Enerfab Inc. has made two executive promotions: Scott Anderson will serve as chief operating officer (COO) and president of Enerfab Process Solutions and Fabricated Products; and Robert Sylvester will serve as COO of Enerfab Power & Industrial. Enerfab Process Solutions and Fabricated Products serves a variety of industries, including chemical, petrochemical, oil and refining, natural gas and water treatment, with large-scale fabrication of ASME pressure vessels, API storage tanks, process columns and other industrial equipment. In his role as COO, Mr. Anderson will oversee two subsidiaries, Brighton Tru-Edge and Enerpipe. He joined Enerfab in 1990 and was named president of the division in 2012. Enerfab Power & Industrial serves the power generation and utility industries, including onsite construction, fabrication and maintenance. Mr. Sylvester previously served as president of NGS Investments and president of Enerfab Electric. PSG, a Dover company and a manufacturer of pumps, systems and related flow-control solutions, has appointed Heather Graham as VP of HR, where she will report directly to PSG President Karl Buscher. Prior to joining PSG, Ms. Graham was the global HR director for PetroTechnical Services at Schlumberger.

86 AUGUST 2016 | HydrocarbonProcessing.com

Evonik’s Oil Additives business line has appointed Mukund Bhure as its Global Industrial Lubricants marketing manager. Mr. Bhure joined Evonik’s Oil Additives business line in April 2010 from Chevron Lubricants. He has more than 24 years of sales and marketing experience with lubricants and specialty chemicals. The US Department of Commerce has confirmed four appointees to the newly established US-Mexico Energy Business Council. The four appointees are members of the National Electrical Manufacturers Association (NEMA): Julian Alzate, director of international relations at Schweitzer Engineering Laboratories Inc.; R. Craig Breese, president of Honeywell Mexico, Honeywell Intl. Inc.; Vernon Murray, VP and general manager for Mexico and Northern Latin America at Emerson Process Management; and Darryl Wilson, VP and chief commercial officer (CCO) at GE Energy Connections. All appointees will serve twoyear terms on the council. In addition to the committee of privatesector appointees, the US-Mexico Energy Business Council will include representatives from the US Department of Commerce, the Department of Energy, the Mexican Ministry of Energy and the Ministry of Economy.

Emerald Polymer Additives, a business group of Emerald Performance Materials, has named John Zuppo as president of its Polymer Additives and Nitriles (PANIT) business group. Mr. Zuppo has been serving the Emerald’s PANIT group for the last five years, most recently as commercial vice president. Previously, he served as VP of procurement for all of Emerald’s business groups. Prior to joining Emerald, Mr. Zuppo served as director of direct raw material procurement for Ferro, and fulfilled a variety of roles at The Goodyear Tire and Rubber Co., including procurement, business development, sales and marketing, and research and development. He earned a BS degree in chemical engineering from The University of Akron and an MBA from Case Western Reserve Weatherhead School of Management. Hoover Container Solutions has announced three new changes to its senior management team: Joseph Levy has been appointed senior VP and chief financial officer (CFO); Johan Wramsby has been named senior VP and chief operating officer (COO); and Arash Hassanian will assume the role of senior VP of global sales and marketing.

ABB has appointed Chris Shigas as its US director of media relations. He joins ABB from agency French/ West/Vaughan, where he was a senior VP working in media relations, corporate communication and crisis communication. His campaigns have earned dozens of national awards. Prior to his agency experience, he was an Emmy award-winning television news producer. Mr. Shigas has over 20 years of experience in public relations and broadcast journalism. Alan Kelly, president of ExxonMobil Fuels, Lubricants and Specialties Marketing Co., has announced his intention to retire effective August 1, after more than 34 years of service. As of this issue’s publication, it is anticipated that the board of directors of ExxonMobil Corp. will appoint Bryan Milton as Kelly’s successor and VP of the corporation. Mr. Milton presently serves as president of ExxonMobil Global Services Co. He joined Exxon Chemical in 1986, where he worked in various plant and developmental engineering roles, and held various leadership positions within ExxonMobil Chemical Co. He was appointed manager of the Baton Rouge chemical plant in 2006, and in 2008 was named executive assistant to the chairman and CEO of ExxonMobil Corp.

Tighter emissions standards costing you? Meet or exceed regulations and save with our patented flameless thermal oxidizer. Selas Linde North America‘s Thermatrix® flameless thermal oxidizer is a field-proven, cost effective solution for destruction of a wide range of VOCs and HAPs. From skid-mounted units to treat lower flows to large scale, plant-integrated systems, we have a solution for emission treatment.

→ → → → → → →

99.99% or greater destruction efficiency Low NOx: < 5 ppmv @3% oxygen Undetectable CO emissions Flameless - can be installed near the emission source Treats batch or continuous streams Energy efficient operation Low cost of ownership

Learn more!

Selas Linde North America A division of Linde Engineering North America Inc. Five Sentry Parkway East, Suite 300, Blue Bell, PA 19422 USA Phone +1 610.834-0300 [email protected] www.leamericas.com

Select 66 at www.HydrocarbonProcessing.com/RS

Select 51 at www.HydrocarbonProcessing.com/RS

Technology and Business Information for the Global Gas Processing Industry

GasProcessingNews.com | JULY/AUGUST 2016

Special Supplement to

SMALL-SCALE

PROCESSING

SOLUTIONS Opportunities and challenges for commercialization and deployment of small-scale LNG and GTL

FLNG DEVELOPMENTS

Nearshore solutions can cut costs, reduce technical complexity of floating LNG

GAS TREATING

Design H2S removal and dehydration units for operation under ultra-high pressure

CONTENTS

EDITORIAL COMMENT Small-scale gas processing operations are ramping up around the world, particularly in the LNG sector. New LNG import terminals are opening in Europe and Asia as cost-effective, efficient processing technologies are taking hold in the market. For example, Spain is using LNG produced on a small scale for bunkering, while Finland recently opened ADRIENNE BLUME, the first of several planned small-scale Editor LNG import terminals. According to a report released earlier this year by Visiongain, the small-scale LNG market could see capital expenditures of $2.5 B in 2016. CAPEX for small-scale gas processing is forecast to increase over the following years in both mature and emerging markets. The growing popularity of flexible, less-expensive, smallscale LNG and GTL processing technologies is being driven by the need for more localized supplies to meet specialized demand, as well as the desire to capitalize on small and stranded gas reserves. In most cases, these technologies also offer shortened construction timelines. New projects will meet the need for distributed power and fuel for road and marine transport, filling the gaps left by gas pipelines and tanker deliveries. A number of small-scale processing technologies are available from specialized startup companies, as well as from major oil and gas technology firms. Many of these companies have seen major progress on commercial installations in recent years. These developments, which offer promising supply alternatives in today's difficult market and into the future, are detailed in this issue's Special Report, Executive Viewpoint and Regional Perspectives articles. GP

www.GasProcessingNews.com

PUBLISHER EDITORIAL

Editor Technical Editor Editor/Associate Publisher, Hydrocarbon Processing

MAGAZINE PRODUCTION

Vice President, Production Manager, Editorial Production Artist/Illustrator Senior Graphic Designer Manager, Advertising Production

P. O. Box 2608 Houston, Texas 77252-2608, USA Phone: +1 (713) 529-4301 Fax: +1 (713) 520-4433 [email protected] Catherine Watkins [email protected] Adrienne Blume Bob Andrew Lee Nichols

GasProcessingNews.com | JULY/AUGUST 2016

11 SPECIAL REPORT: SMALL-SCALE PROCESSING SOLUTIONS 17

Opportunities and challenges for small-scale LNG commercialization R. S. Bhullar

21

Market development is key to success for small-scale LNG S. Bonini and A. Chandra

LNG TECHNOLOGY 25

Develop successful nearshore FLNG solutions— Part 1: Gas pretreatment strategies S. Mokhatab

GAS TREATING 29

35

Design for ultra-high-pressure H2S removal from natural gas P. Roberts Manage activated carbon effects on MDEA solution foaming D. Engel, S. Williams and A. Heinen

PLANT DESIGN 41

Prevent hydrate formation with high-pressure deethanizer design C. C. Chen and Y.-S. Liu

DEPARTMENTS Sheryl Stone Angela Bathe Dietrich David Weeks Amanda McLendon-Bass Cheryl Willis

Gas Processing News .......................................................... 6 US Industry Metrics ............................................................. 8 New in Gas Processing Technology ................................ 46

COLUMNS

ADVERTISING SALES

See Sales Offices, page 45.

Copyright © 2016 by Gulf Publishing Company. All rights reserved.

Publisher's Letter ................................................................. 4 The transition into a new era

Regional Perspectives.........................................................11 Natural gas monetization options for Iran: LNG or GTL?

President/CEO CFO Vice President Vice President, Production

John Royall Pamela Harvey Ron Higgins Sheryl Stone

Other Gulf Publishing Company titles include: Hydrocarbon Processing, World Oil and Petroleum Economist.

Executive Q&A Viewpoint ................................................. 13 Too much of a good thing: Methanol as a solution to gas oversupply in the Marcellus

Cover Image: The small-scale LNG plant in Kwinana, Australia is an example of Linde’s StarLNG standardized LNG concept. Photo courtesy of The Linde Group.

P. O. Box 2608, Houston, Texas 77252-2608, USA | Phone: +1 (713) 529-4301 | GasProcessingNews.com

Dear Reader,

The transition into a new era

These are unprecedented times in the oil and gas industry. New technologies have produced gluts of both oil and natural gas, which have provided cheap feedstocks for the downstream processing industries. The cyclical nature of the oil and gas business has had vastly different effects on each region. To oil-exporting nations, reduced oil prices equate to low government revenues. In turn, little money is available to fund social, industrial or infrastructure projects. To other nations, low oil and natural gas prices have been a boon to the processing industries. They have seen a boost in the construction of additional downstream processing capacity, as well as cheap fuel prices for consumers, which spur consumption. As with all cyclical industries, sometimes change is a necessity to ensure the strength and viability of an organization. Change allows a business to build on its strengths and evolve into an even more exceptional enterprise. Gas Processing is undertaking such a transition. As of August 1, I have assumed the role of Publisher for two of Gulf Publishing Company’s exceptional brands, Hydrocarbon Processing and Gas Processing. Most of my life has been spent in and around the oil and gas industry. I was born in Chicago, but was largely raised in the oil patches of Iran, the UAE and Texas. As a result of my upbringing, I was familiar with drill collars and Christmas trees (assemblies on surface and subsea wells) before I had heard of Barbie dolls. For the past 20 years, I have represented Gulf Publishing Company’s Hydrocarbon Processing, World Oil and Gas Processing publications in France, Spain, Germany, Switzerland, Austria, Belgium and the Middle East/North Africa regions. My contributions included display and digital advertising sales, organizing industry forecast presentations and event participation, and conducting marketing seminars for clients. During this time, I have found no other publications as highly regarded in the downstream industry as Hydrocarbon Processing and Gas Processing. This is not arrogance or by accident; it is based on audited circulation numbers and years of industry professionals’ testimonials. Both publications reached this position by applying the highest levels of editorial integrity. It is proof that, although consumer and industry publications are largely giving up on editorial standards, our industry still values and needs a trusted source of information. As we look into the future of our industry and publications, technical content is king. This will never change. We will continue to bring our readers the highest-quality technical and operating articles in the industry. It is what we have been doing since 1922, and will continue to do now and into the future. What has, is and will continue to change is how that information is disseminated. Gas Processing is ready to engage and listen to the industry to provide editorial and data content in the most useful medium and format possible. I want to thank you for your devotion to Gas Processing. It is gratifying to hear how our publication, website and newsletters provide interesting and, more importantly, sound technical content to make your work and our industry a better place. I also want to thank all of the advertisers who support this publication. Please share your comments, ideas and news with us; we highly value your feedback. It is a dynamic time in the downstream industry, and I can guarantee that Gas Processing will continue to be at the forefront of technology, trends and data intelligence.

Catherine Watkins Publisher, Hydrocarbon Processing and Gas Processing

4 JULY/AUGUST 2016 | GasProcessingNews.com

Modular, scalable, cleaner energy. Fueling the future of natural gas. A skid-mounted, plug-n-play natural gas liquefaction plant that provides a cleaner more abundant LNG fuel source for remote locations. GE’s small scale LNG plants mean faster commissioning times and reduced installation costs.

Imagination at work.

www.geoilandgas.com/smallscalelng

GAS PROCESSING NEWS

BOB ANDREW, Technical Editor

GTI celebrates 75 years Gas Technology Institute (GTI) recently celebrated an important milestone—its 75th anniversary. As a leading research and development (R&D) and training organization addressing global energy and environmental challenges, GTI spent the last seven-plus decades developing high-impact technologies and providing technical insight to unlock the potential of natural gas and other energy resources. GTI’s aim is to make these technologies economically and environmentally sustainable while simultaneously reducing energy costs for consumers. GTI also has a solid reputation for developing gas distribution tools and technologies that reduce energy delivery costs. Experts have created a wealth of solutions to enhance the safety and integrity of the nation’s vast pipeline infrastructure and target critical global methane emissions issues. The organization has achieved many successes over its history, focusing on initiatives aligned with important national priorities. GTI has provided solutions to critical challenges along the entire gas value chain and improved the ways of producing, transporting, and using energy to benefit the general public. R&D projects through the decades have expanded the supply and reduced the environmental footprint of unconventional gas production. Cleaner ways to use abundant coal resources have been generated and brought to market through GTI-patented processes. Groundbreaking efforts to convert wood into renewable “drop-in” gasoline in GTI’s integrated biorefinery have resulted in a sustainable biofuel with 74% lower greenhouse gas emissions than petroleum-based fuel. Promoting the clean and efficient use of energy resources across all end-use markets— residential/commercial, industrial, power generation, and alternative compressed natural gas and hydrogen transportation—and contributing to the progress of US green building practices are among GTI’s contributions. GTI is developing a portfolio of affordable equipment and appliances with reduced energy consumption and emissions, which will contribute to improving air quality for decades to come.

Gas jet ejector extends field life GEA has designed, manufactured and delivered a gas jet compressor system for an offshore platform in the North Sea consisting of a jet pump (ejector) with a total length of 4.5 m, and silencers to reduce the sound pressure level. At a newly drilled well, natural gas is available with a pressure of up to 600 bar. The gas pressure decreases in the course of exploitation of the gas field. At a defined gas pressure, the gas delivery rate is so low that the mechanical compressor is no longer economical. Gas jet compressors are designed to increase the natural gas pressure and to allow extension of the use of the installed mechanical compressors. In this way, the gas field life is prolonged. The system compresses a low-pressure natural gas (8 barg) by using a high-pressure natural gas (90 barg) to an intermediate gas pressure (12 barg). GEA managed to increase its portfolio by designing the ejector body for 137 barg, and on the motive side for 213 barg. The material of construction is corrosion-resistant and corrosion-tested duplex stainless steel, which is coated with sprayed aluminum. Ejectors are reliable in operation at low operating costs. They are nearly maintenance-free because they are simply constructed, having no moving parts. Almost any vacuum duty can be accommodated, including large suction flows.

Engineer chosen for Elba LNG project

HDR Inc. has been selected as the owner’s engineer for Kinder Morgan’s LNG export project at Elba Island near Savannah, Georgia. The project is in the engineering, procurement and construction phase. The Elba liquefaction project will include installation of 10 liquefaction units for an output capacity of approximately 2.5 MMtpy of LNG, and modifications to the existing Elba Terminal to allow for export. The project will use Shell’s small-scale liquefaction units, which will be integrated into the existing Elba terminal and enable rapid construction compared to traditional large-scale plants. Kinder Morgan is also working with ABSG Consulting Inc., a subsidiary of ABS Group of Companies Inc., for the development of a maintenance and integrity program for the terminal. ABS Group will develop the program for the base plant, balance of plant and the terminal upgrade.

6 JULY/AUGUST 2016 | GasProcessingNews.com

3D-printed polymer turns methane to methanol Lawrence Livermore National Laboratory (LLNL) scientists have combined biology and 3D printing to create the first reactor that can continuously produce methanol from methane at room temperature and pressure. The team removed enzymes from methanotrophs—bacteria that eat methane—and mixed them with polymers that were then printed or molded into innovative reactors. LLNL found that isolated enzymes offer the promise of highly controlled reactions at ambient conditions with higher conversion efficiency and greater flexibility. Advances in oil and gas extraction techniques have made vast new stores of natural gas, composed primarily of methane, available. However, a large volume of methane is leaked, vented or flared during these operations, partly because the gas is difficult to store and transport compared to more valuable liquid fuels. Methane emissions also contribute approximately one-third of net global warming potential, primarily from these and other distributed sources, such as agriculture and landfills. Existing industrial technologies (such as steam reformation) to convert methane into more valuable products operate at high temperature and pressure, require a large number of unit operations and yield a range of products. As a result, these technologies have a low efficiency of methane conversion to final products, and can operate economically only at very large scales. A technology to efficiently convert methane to other hydrocarbons is needed as a profitable way to convert stranded sources of methane and natural gas to liquids for further processing. The only known catalyst (industrial or biological) to convert methane to methanol under ambient conditions with high efficiency is the enzyme methane monooxygenase (MMO), which converts methane to methanol. The reaction can be carried out by methanotrophs that contain the enzyme, but this approach inevitably requires energy for upkeep and metabolism of the organisms. Instead, the LLNL team separated the enzymes from the organism and used the enzymes directly.

The Green Solution to Sulfur Recovery

The LO-CAT process, available exclusively from Merichem, is a patented liquid redox system that uses a proprietary chelated iron solution to convert H2S to innocuous, elemental sulfur. The catalyst is continuously regenerated in the process.

LO-CAT Total Package

Flue Gas

The LO-CAT technology is applicable to all types of gas streams including air, natural gas, CO2, amine acid gas, 2S concentrations. With over 35 years of continuous improvement, LO-CAT units are very reliable and require minimal

Treated Gas Chemical Addition

From engineering and fabrication, to installation supervision, training, and startup, through process warranties and onsite service, Merichem provides a total sulfur recovery solution. Each system is

Proprietary Sulfur Filter System

Raw Gas

aggressive schedules can be accommodated. Full equipment packages are provided for stick-built or

ABSORBER VESSEL

Solution Circulation Pump

Sulfur Slurry

OXIDIZER VESSEL Air

Direct-Treat

Air Blower

LO-CAT Direct Treatment Scheme Chemical Addition

Flue Gas

When treated gas cannot be combined with air, a direct-treat design is employed. This is achieved by use of two separate vessels, an absorber and an oxidizer. The absorber treats the sour gas, producing sweet gas in a single pass. The oxidizer serves two purposes: The regeneration of spent catalyst and the concentration of

Raw Gas

Proprietary Sulfur Filter System

system takes the sulfur-rich slurry, washes it and produces an elemental sulfur cake. LO-CAT AutoCirc Scheme When treating a gas that can be mixed with air, the operating and capital expenses. By combining the absorber and oxidizer in one vessel, the solution circulation pump is eliminated resulting in reduced electrical consumption. The single vessel approach also minimizes footprint.

Sulfur Slurry

Air Air Blower

AutoCirc

Merichem Company

Merichem Process Technologies 5450 Old Spanish Trail Houston, Texas 77023

Tel: +1 713.428.5000 www.merichem.com

US INDUSTRY METRICS

A. BLUME, Managing Editor

US natural gas spot prices at Henry Hub and NGL spot prices at Mont Belvieu, $/MMBtu

Oct. 2015

Jan. 2016

April 2016

3 Monthlyprice price(Henry (HenryHub) Hub) Monthly 12-monthprice priceavg. avg. 12-month Production Production

0

J J A S O N D J F M A M J J A S O N D J F M A M J 2014 2015 2016

2 1 0

Gas prices, $/Mcf

4

40 20

US gas plant field production, Mbpd

$/MMBtu

5

July 2015

5

60

120

10

0

6

80

US natural gas plant field production of NGL, LPG, ethane/ethylene and propane/propylene, Mbpd

Natural gasoline Isobutane Butane NGPL composite Propane Ethane Natural gas spot prices (Henry Hub)

15

7

Production equals US marketed production, wet gas. Source: EIA.

25 20

US gas production (Bcfd) and prices ($/Mcf) 100 Production, Bcfd

In the US, Henry Hub natural gas spot prices began climbing in late June, rising to just under $3/MMBtu by early July. Meanwhile, prices for NGL bounced up briefly in June, but then slid again by early July. Domestic production of NGL, LPG, ethane/ethylene and propane/propylene slipped in April (the latest month of data available as of the time of publication) from highs in March. In contrast, natural gas production dropped sharply in March from a high seen in February. The portion of ethane recovered as a percent of total NGPL production has been increasing this year due to higher ethane use by USGC petrochemical producers. GP

July 2016

100 80 NGL LPG Ethane/ethylene Propane/propylene

60 40 20

April May June July Aug. Sept. Oct. Nov. Dec. Jan. Feb. Mar. April 2015 2015 2015 2015 2015 2015 2015 2015 2015 2016 2016 2016 2016 Source: US EIA

Source: US EIA

InstruCalc CONTROL VALVES • FLOW ELEMENTS • RELIEF DEVICES • PROCESS DATA InstruCalc 9.0 calculates the size of control valves, flow elements and relief devices and calculates fluid properties, pipe pressure loss and liquid waterhammer flow. Easy to use and accurate, it is the only sizing program you need, enabling you to: Size more than 50 different instruments; Calculate process data at flow conditions for 54 fluids in either mixtures or single components and 66 gases, and; Calculate the orifice size, flowrate or differential range, which enables the user to select the flow rate with optimum accuracy.

NEW VERSION

Order Direct from the Publisher. GulfPub.com/InstruCalc or call +1 (713) 520-4426.

Updates and What’s New in InstruCalc Version 9.0 ENGINEERING STANDARD UPGRADES Control Valve Revisions: • Updated to ANSII/ISA 75.011.01-2012 • Calculation accuracy changed for critical flows • Viscosity correction factor changed • Pressure drop calculation revised to agree with Crane Technical paper No 410. • Option of Cv Units (English) or Kv units (Metric) added. • Option of either aerodynamic noise calculation by ISA 75.17 method or InstruCalc method

8 JULY/AUGUST 2016 | GasProcessingNews.com

• Calculation accuracy added (input data within acceptable limits) Relief Devices: • Pressure Relief Devices Program follows API 520 Pt 1, 9th edition dated 7/14 OPERATIONAL IMPROVEMENTS The ability to have more than one calculation open at a time has been added. Each instance of the program is framed in a different color. The user can have multiple “what if” scenarios displayed for making engineering decisions.

NEWSLETTER ISSUE 1: July 2016 / www.GasProcessingNews.com

Technology and Business Information for the Global Gas Processing Industry

Expanded Panama Canal to boost gas tanker traffic to Asia Adrienne Blume, Editor, Gas Processing and Executive Editor, Hydrocarbon Processing Panama government officials estimate that 20 MMtpy of LNG will pass through its newlyexpanded waterway (FIG. 1), the equivalent of nearly one tanker per day. Most of the supplies will travel to Asia, the world’s largest LNG-importing region. The economies of South Korea and Taiwan stand to benefit greatly from the $5.3-B expansion. Japan and China are also expected to import LNG through the enlarged canal. The expansion project, which broke ground in 2007, was completed in late June. Its debut is serendipitous for US gas producers, as the shale boom has sent domestic supplies surging, and drillers are looking to send their fuel to markets abroad. Previously, the Panama Canal could accommodate ships carrying 5,000 containers, but the expansion will allow vessels with 14,000 containers to pass. The new locks can able to accommodate vessels up to 160 feet wide and 1,200 feet long, which means that ships carrying LNG will finally gain access. The larger waterway will help US gas producers avoid longer trips around South America, cutting transits to Asian markets and increasing profits. The volume projected by the Panama Canal Authority represents about 8% of global LNG trade, equivalent to nearly 300 ships a year. Two shipping alliances, CKHYE and G6, have already announced plans to upgrade the size of the ships they use for the US-to-Asia route via the Panama Canal. The Canal expects its total revenues to increase 17% in 2017 to $2.8 B as a result of the expansion.

FIG. 1. The third set of locks for the expanded Panama Canal (shown at right), and the existing locks (shown on left) in May 2016. Photo credit: Panama Canal Authority.

In Partnership with

NEWSLETTER ISSUE 1: July 2016 / www.GasProcessingNews.com

In Partnership with Technology and Business Information for the Global Gas Processing Industry

Japan: An unlikely new champion for coal? Gavin Sutcliffe, Head of Conference and Governing Body, Gastech Japan is traditionally respected as one of the world’s most efficient and technologically advanced societies, embracing an inclusive and responsible approach to economy, society and environment. However, Japan’s latest Basic Energy Plan is undergoing major revision by the Ministry of Economy, Trade and Industry (METI), and is now expected to promote an increased use of coal-fired power generation as public resistance to nuclear power remains as robust as ever since 2011. As recently as last year, METI was committed to a balanced energy roadmap for Japan, with a fuel mix that included around 20% from nuclear, 26% from coal, 27% from LNG and 22%–24% from renewable energy sources. If nuclear power is to remain offline, other fuels will have to fill the huge shortfall left if the energy-intensive society is to remain functional, and METI’s revised plans are likely to be made public in early 2017. Ironically, this begrudging acknowledgement by the government that the public will not stomach perceived environmentally “risky” nuclear fuel has paved the way for coal’s resurgence with plans for many new coal-fired power plants to be built. The likely result is that Japan will miss its carbon obligations agreements signed during the COP21 in Paris last year. Approximately 40 new stations powered by coal are planned to open over the next decade or so, putting Japan’s emissions targets in serious jeopardy. METI plans to make the increased long-term investment in coal and “clean-coal” technology, to fill the large gap left by nuclear. Japan certainly appears isolated among its G7 peers as being the sole country still building unabated coal-fired power plants. In May of this year, a report compiled by Oxford University’s Smith School of Enterprise and Management called Japan’s plans to invest billions of dollars in coal-fired generation “flawed,” and the report’s main author, Ben Caldecott, remarked, “Does Japan seriously think that there will still be coal-fired power stations in the system in the 2070s? Because that is what they are committing themselves to with the plans they have laid out.” The report asserts that the plans to take Japan’s energy policy in the opposite direction of other countries will impact the country negatively over the coming decades. Coal and clean-coal technology industries are powerful, influential businesses in Japan, where lobbying keeps coal at the forefront of government policy, even as capital expenditure costs for renewable energies have fallen between 35% and 45% in the past decade. With even the cleanest new-generation coal technology producing twice the CO2 emissions as gas-fired generation, the heat is now on the Japanese government, both internally and externally, to reexamine its commitments to coal and fundamentally review its position as a leading member of the world’s most industrialized nations.

Coal will continue to challenge the position of natural gas—in particular, LNG—as Japan’s most flexible and economical fuel. This is one of many key geo-strategic energy debates that will take place at the world’s largest natural gas and LNG event of 2017, the Gastech Conference and Exhibition, which arrives in Chiba, Tokyo, April 4–7. The Japanese government will be present, and the event is being hosted by 10 of Japan’s most important energy stakeholders and investors, including: Tokyo Gas, JERA, INPEX, Mitsubishi Corporation, Mitsui & Co., JAPEX, Sumitomo Corporation, JX Nippon and Itochu. With the Gastech conference now recognized for five decades as the thought-leadership platform for commercial and technical leaders alike, the issues of how gas and LNG as primary fuels of a low-carbon energy mix can continue to stave off challenges from coal and other fossil fuels will be fiercely debated. Gastech is calling for global professionals working across technical and commercial disciplines to provide outstanding and innovative original submissions to speak at the conference next year, and encourages those interested to connect with the organizing team as soon as possible. With more than 40 major technical sessions running over four days and three days of strategic-level commercial content, the opportunities for leading gas and LNG professionals to engage with the world’s largest customer community has never been more exciting or accessible. Gastech’s governing body of 50 leading international professionals will be voting on the best papers soon. Do not miss out on the chance for your organization to make an impression to more than 2,500 executive delegates and many thousands more visitors and exhibitors alike! Visit www.gastechevent.com/call-for-papers to review themes and make your submission.

REGIONAL PERSPECTIVES: IRAN

Natural gas monetization options for Iran: LNG or GTL? E. SALEHI, APED Engineering Consultants, Calgary, Alberta, Canada

Iran has the world’s largest proven natural gas reserves at 1,201 Tcf, according to BP’s Statistical Review of World Energy,1 and it is the fourth-largest natural gas producer in the world. However, it exports only small amounts of natural gas to Azerbaijan, Armenia and Turkey—approximately 9 Bcmy. Potential exists for Iran to export approximately 30 Bcmy to the EU over the long term with the development of LNG facilities. Iran has so far prioritized its gas allocation to satisfy domestic demand for heat, power and industry use, and for reinjection to aging oil fields to maintain production. With new phases of the giant South Pars gas field coming onstream, however, any surplus gas will need to be monetized through pipeline export, petrochemical use, gas-topower, LNG and/or GTL (FIG. 1). Iran has demonstrated impressive development in the petrochemical and gasto-power industries. However, no significant movement has been seen on LNG or GTL. The question remains of which option will be the best pathway for the monetization of Iran’s gas reserves.

Iran needs to invest in both LNG and GTL technologies; however, better opportunities exist to develop GTL over LNG. The country could become a global GTL hub if it is willing to embrace emerging and smaller-scale technologies. Saturated LNG market. The global

LNG market is oversupplied, and is expected to remain so until 2025 due to the large number of LNG plants anticipated to come onstream in Australia and the US. Slowdowns in LNG demand from China, Japan and South Korea are magnifying this supply glut. As a result, project developers are hesitant to commit over the near term to capital-intensive liquefaction projects. Financing may become more of a challenge if oil prices remain low. According to BP statistics,1 global liquefaction capacity is more than 300 MMtpy, with another 130 MMtpy of capacity under construction. More than 60 LNG projects with a combined liquefaction capacity of 650 MMtpy are expected to compete for market share between 2020 and 2025. The majority of these

FIG. 1. Iran’s existing natural gas infrastructure. Source: US Energy Information Administration, IHS EDIN.

projects are located in the US, Canada and Australia. By contrast, worldwide LNG demand has hovered at 240 MMtpy over the last few years. For Iran to compete with these large, emerging LNG players, it would need to include LNG buyers in the ownership structure of its liquefaction projects. However, LNG is not necessarily the most economic or practical option for Iran to monetize its natural gas in the present market situation. The LNG market will continue to be oversaturated in the foreseeable future, and it will be a challenge for Iran to finance LNG projects. Potential in small-scale GTL. The

country has a much better chance of becoming a hub for GTL. Only five commercial GTL plants are in operation around the world. All are licensed by either Sasol or Shell and are located in South Africa, Nigeria, Malaysia and Qatar, with a combined production capacity of 260 Mbpd. Sasol and Shell are the pioneers of the GTL industry, although other GTL

FIG. 2. Small-scale GTL projects, like the ENVIA Energy GTL plant under construction in Oklahoma City, Oklahoma, US, offer opportunities for alternative GTL industry development across the world. Photo courtesy of Velocys. Gas Processing | JULY/AUGUST 2016 11

REGIONAL PERSPECTIVES: IRAN technology licensors in the market include Air Liquide/Lurgi and BP/Davy Process Technology. However, neither partnership operates a GTL plant at commercial scale. The next tier of GTL technology licensors are emerging technology companies that provide mini- and micro-scale GTL technologies. These technologies are more efficient compared to conventional GTL technologies, and they are

AdvaSulf

cost-effective for smaller-scale applications like associated gas flaring. In line with the shale boom in North America, tremendous advances have been made in small-scale GTL technology over the past decade (FIG. 2). These smaller-scale technologies convert less-valuable natural gas and associated gases into higher-value liquid products. At present, large oil and gas companies are evaluating options for

TM

HySWEET® COSWEET

TM

SweetSulf

TM

Sprex® AdvAmine

TM

A UNIQUE TASTE OF SWEET FOR YOUR GAS 50 YEARS OF EXPERIENCE IN GAS SWEETENING AND SULPHUR RECOVERY PROCESSES With its unique and complete proprietary technologies portfolio, PROSERNAT offers optimized solutions to bring on specs any type of gas contaminated with CO2, H2S, COS and organic sulphur species, while producing sulphur with the most stringent emissions standards. www.prosernat.com

12JULY/AUGUST 2016 | GasProcessingNews.com

investment in these technologies. Iran could potentially invest in, or acquire, some of these game-changing emerging technology companies. Actions for Iran. Although larger-scale GTL projects are economically challenging in the present price environment of below-$50/bbl oil (as of the time of publication), Iran needs to develop a long-term plan for GTL in its natural gas downstream sector. One step toward creating a GTL hub in Iran would be to establish a team of experts inside the oil ministry. This team would comprise a mix of market and business analysts, as well as engineers with specialties in natural gas monetization techniques. It would define Iran’s natural gas monetization roadmap and consult the oil minister on development options. The team would also be responsible for evaluating the most feasible alternatives for Iran’s natural gas monetization, including residential, industrial, enhanced oil recovery, pipeline export, petrochemical, gas-to-power, GTL, LNG and other options. It is recommended that the team members seek expertise from world-scale consulting firms. Lastly, Iran should carefully watch how Qatar is changing its strategy for gas export development amid the downturn in the oil and gas market. Qatar’s actions could serve as a regional model for Iran’s development of its new gas reserves going forward. GP 1

LITERATURE CITED BP, BP Statistical Review of World Energy, June 2015, available online: https://www.bp.com/content/dam/bp/pdf/energy-economics/statisticalreview-2015/bp-statistical-review-of-world-energy2015-full-report.pdf

EBRAHIM SALEHI is a business development and project director at APED Engineering Consultants. He is a chemical engineer with over 10 years of oil and gas industry experience within operating and EPC companies. Dr. Salehi’s specific areas of expertise include flare gas recovery, natural gas monetization, conventional and unconventional oil and gas field development, and biofuels. His strong research and development background, combined with a big-picture understanding of the oil and gas market, has also led to publication of several articles and patents. He holds a PhD in chemical engineering from the University of Calgary.

EXECUTIVE VIEWPOINT

Too much of a good thing: Methanol as a solution to gas oversupply in the Marcellus SAM GOLAN, CEO, Primus Green Energy

SAM GOLAN is CEO of Primus Green Energy. He is a seasoned general manager with experience as an executive leading multinational engineering, project management and manufacturing software companies from the entrepreneurial stage to an established market presence, mergers and acquisitions, and initial public offerings. In the past, he served as the general manager at Cimatron Technologies in North America. At Cimatron, Mr. Golan led the development of integrated “lean design and manufacturing” solutions for the automotive, process and aerospace industries; provided clients with comprehensive, cost-effective solutions to streamline design and manufacturing cycles; enabled collaboration with outside vendors; and shortened product delivery times. He also cofounded and managed Smart Team, a product lifecycle management company acquired by Dassault Systems SA. Mr. Golan holds a bachelor’s degree in economics and business administration.

The shale revolution has caused a natural gas boom in the Marcellus shale play that has been a victim of its own success. The surge in production, along with reduced demand as a result of an unusually mild winter in the Northeast, has caused prices to drop by comparison with those in other regions of the country. Earlier this year, the price of natural gas from the Dominion Transmissions North Point/Leidy Hub in Pennsylvania was approximately $0.95/MMBtu below the price at Henry Hub in Louisiana, the main trading point for US natural gas. Stock prices for many leading producers in the Marcellus are also falling. With a lack of pipelines to transport natural gas to consumer-rich markets in the Northeast, Marcellus producers have been forced to cut production, slash expenditures and merge with companies with deep enough pockets to weather the downturn. Analysts see no relief in sight. Although the spread between Marcellus and Henry Hub prices has narrowed in recent months as new pipelines have drawn down supply, a pipeline buildout extensive enough to handle the excess is years away. The same is true of export terminals to ship LNG to foreign markets. Methanol production as a solution. What is the answer to this conundrum? In a word: methanol. Methanol (CH3OH) is a chemical molecule that can be used in many ways—from serving as the basic chemical building block for paints, plastics and solvents to innovative applications in energy, transportation fuels and fuel cells. It is one of the top five global chemical commodities in the world. At present, the US lower 48 states import methanol from large methanol plants on the Gulf Coast or abroad. However, methanol demand can also be satisfied on a regional basis by producing it from low-cost natural gas through

Primus Green Energy’s STG+ gas-toliquids (GTL) technology. Methanol produced regionally through the STG+ technology is competitive with methanol imported from the Gulf Coast and overseas due to lower-cost natural gas feedstock and reduced transportation costs. Another advantage is the ability to react quickly; local producers can meet regional demand in days, rather than weeks. The Primus STG+ technology is flexible and produces high-quality gasoline, diluent and chemicals, including methanol, from natural gas. Due to the cost savings and advantages afforded by local methanol production, Primus is developing a series of North American methanol plants. These plants range in size from 160 metric tpd to 640 metric tpd and can be expanded by adding trains. As a result of its efficient, integrated design, the STG+ methanol technology is cost effective at scales as small as 2 MMcfd (50 Mcmd) of natural gas, although it can be expanded into trains eight times larger. Unlike traditional, large, stickbuilt methanol plants, the STG+ systems are simple to deploy. The modular units are fabricated at a central location and then transported to the plant site for final assembly and commissioning. The methanol system is modular and can be integrated within existing chemical and gas processing plants or as a stand-alone system, depending on requirements. Primus’ methanol system is inexpensive to operate, converting 1 MMBtu of natural gas into 10 gal of methanol at a production cost comparable to that of larger Gulf Coast methanol plants. The technology can produce methanol from a variety of feed gas options, including natural gas, associated gas, ethane and NGL. Primus estimates demand for methanol in the Marcellus region to be at least 500 Gas Processing | JULY/AUGUST 2016 13

SMART SOLUTIONS IN GAS PROCESSING For more than 40 years, CB&I has excelled at producing innovative answers for the gas processing industry. We’ve been awarded more than 300 gas processing projects, including 200 turbo-expander plants. Our installed processing capacity exceeds 26 BSCFD. We provide smart solutions for fast-track modular plants or complete turnkey installations. Our portfolio of Randall Gas processing technologies includes very high ethane and propane recovery, and high propane recovery while in rejection mode. Contact CB&I to learn more about our modular gas plant approach and superior gas processing technologies. Let us help you improve the economics on your next gas processing project. NGL – ETHANE RECOVERY TO 99% LPG – PROPANE RECOVERY TO 99.7% FRACTIONATION GAS TREATING LIQUID PRODUCT TREATING NITROGEN REJECTION MODULAR GAS PROCESSING PLANTS

A World of Solutions Visit www.CBI.com

19M052016H

EXECUTIVE VIEWPOINT metric Mtpy. Through its first announced project in the Marcellus, Primus plans to meet about 10% of that demand (about 55 metric Mtpy). When the project comes online in 4Q 2017, it will be the first GTL plant in the region. Due to the scalability of the pre-fabricated systems, additional trains can be added incrementally, according to need and availability of capital. Plans call for the addition of three more trains to Primus’ Marcellus methanol project, bringing the total capacity to 640 metric tpd in the coming years. Primus is also discussing the development of purpose-built plants for larger industrial customers and planning four more methanol plants in North America. These plants will target areas with cost-advantaged natural gas that are underserved in terms of methanol production. Advantages of the STG+ process. STG+ has low capital and operating costs, high liquid product quality, zero wastewater, process simplicity and a favorable conversion yield. The three-step STG+ process takes place in a continuous gas-phase closed loop, with no intermediate condensation steps. The process starts with steam methane reforming, in which natural gas or other hydrocarbon gases react with steam at high temperature and pressure to produce synthesis gas, or syngas (a mixture of H2 and CO). As mentioned, STG+ can accommodate a range of natural gas feedstock types, including pipeline gas, wellhead gas with no limits on C2+, gas containing up to 25% CO2 and high-ethane residue gas from processing plants. In the second step, the syngas is converted to methanol in a fixed-bed catalytic reactor. In the third step, the water/methanol mixture is separated from other gases and fed to a distillation system that is designed to meet the operator’s methanol purity requirements. The on-spec methanol is collected from the distillation system and sent to storage. Any unconverted gas is recycled, with a portion being used to fuel the reformer. The process water is recycled as steam for the reformer. The STG+ technology can also be used to produce gasoline by replacing the distillation unit with the back end of the STG+ natural gas-to-gasoline system, in which methanol-rich gas is converted to dimethyl ether (DME), which is converted to raw gasoline. In the fourth and final

step of the natural gas-to-gasoline process, durenes (undesirable components of gasoline) are removed, producing a highquality gasoline end product with low benzene and zero sulfur. The ability to convert to gasoline production adds another layer of flexibility to the technology. Meeting local methanol demand with low-cost gas. The combination of

an abundant supply of low-cost natural

gas and a strong demand for methanol is creating an opportunity in the Marcellus for STG+ to utilize locally produced gas to meet regional methanol needs. The STG+ technology bridges mismatches between abundant natural gas supply and restricted takeaway capacity. It also delivers profitability, even in the face of market downturns, by turning low-cost natural gas into liquid end products. GP

BETTER ECONOMICS THROUGH INTEGRATED TURNKEY CAPABILITIES Greater long-term returns from a single-source provider – concept, engineering, design, fabrication, manufacturing and complete construction services. We have the experience, expertise and complete turn-key service capabilities needed to optimize efficiency and improve your project economics, whether for new or modified facilities; gathering systems; or compressor stations. The right solution from the BCCK family of companies: BCCK Engineering, NG Resources and NG Field Construction will minimize your risks and maximize your project margins.

Don’t settle for less, log onto www.bcck/family or contact a representative at 432-685-6095.

Gas Processing | JULY/AUGUST 2016 15

Big LNG expertise. Also available in small LNG plants. Air Products has contributed to the success of more LNG operations than any other company. And we bring our full capabilities to LNG projects of any scale, from peak-shaving plants producing less than 0.1 MMTPA to the largest base-load facilities, on land or off-shore. Our LNG team can help you get a plant up and running at the highest efficiency—on time, on budget, and in any climate. To learn more, call 1-800-654-4567 (US), 1-610-481-4861 (worldwide) or visit us online.

tell me more airproducts.com/lngplant DOWNLOAD OUR FREE GAS WEIGHT & VOLUME CONVERSION APP. ©2015 Air Products and Chemicals, Inc.

SPECIAL REPORT: SMALL-SCALE PROCESSING SOLUTIONS

Opportunities and challenges for small-scale LNG commercialization R. S. BHULLAR, Fluor Corp., Aliso Viejo, California

The rapid development of the unconventional gas supply chain in North America could have a game-changing impact on large-scale and mega-gas processing/liquefaction facilities. However, there are relatively fewer risks and, therefore, greater opportunities associated with the development of small-scale gas processing and LNG facilities and infrastructure. Issues related to large global hydrocarbon development are different from those related to small-scale gas processing facilities. Gas is anticipated to be the leading fuel of choice in the future and will play a key development role in Asian and Latin American economies. New development drivers and regulations provide great opportunities for small-scale and mid-scale gas and LNG projects, without regard to supply and demand dynamics. A comprehensive review of the developing trends and opportunities related to small-scale gas processing and LNG is offered here. New opportunities are discussed, including traditional power and fuel replacement, marine bunkering, environmental drivers and regulations, feedstock and petrochemicals, and infrastructure opportunities, along with challenges that small-scale project developers may face. Understanding the issues, drivers, economics and challenges is vital to the economic leverage and future success of these projects. Large-scale projects are capital-intensive and extremely risky, and they use complex technologies and equipment. Additionally, they face a constantly changing regulatory environment. Presented here are simpler technologies and infrastructures, quick deployment options, smaller sizes for projects and more targeted approaches for reducing risk and bringing these projects online with relatively smaller capital requirements. Evolving LNG supply/demand picture. Natural gas will be a leading fuel of choice going forward, and LNG will be a significant component of this gas supply chain. However, the industry faces significant challenges. Shale gas supply development and the US transition from a net importer to a net exporter of energy have created a permanent, game-changing market shift. Mega-projects were planned and completed by overambitious producing nations like Australia to meet the gas needs of the US, which never materialized. At present, an excess of LNG supply exists on the order of 100 metric MMtpy. This excess has created unutilized capacities and a large imbalance in the supply chain that will grow as more projects come online. This excess gas supply is shaping the emerging picture of global LNG supply and trade, as shown in FIG. 1. In the US

and Canada, several LNG receiving and regasification terminals were built and never utilized or brought onstream. Some of these terminals are being converted to liquefaction facilities for LNG export. Much of the North American gas is being exported to Japan, Korea, Latin America and Europe. This growth is being further facilitated by the opening of a parallel Panama Canal branch, which will lower shipping costs to Asia. China, on the other hand, is playing a relatively small role in the global LNG picture. The country is developing pipeline infrastructure to import Russian gas, while making significant strides in small-scale LNG and gas supply infrastructure. North American gas supplies would have a relatively small impact on the European market, which is well supplied with gas by pipelines and other infrastructure, with supply arrangements already locked in by long-term sales and purchase agreements (SPAs) from customers in the Middle East and North Africa. The biggest challenge to the gas processing/LNG industry going forward is North American gas supply—specifically, lowcost shale gas production. At the time of publication, the cost of shale gas at Henry Hub was in the range of $2/MMBtu to $2.50/MMBtu, and is expected to hover in that price range for the foreseeable future. If the cost of liquefaction is added to this price, along with the costs of pipeline and marine transportation, then the gas can be supplied to Asian-Pacific countries for less than $6/MMBtu. These low costs could damage the economics of LNG projects in Australia that are completed or waiting to be brought onstream. For many Australian and other higher-priced LNG projects, the breakeven cost is $11/MMBtu–$14/MMBtu.

FIG. 1. Global trade flows for LNG, 2015. Gas Processing | JULY/AUGUST 2016 17

SPECIAL REPORT: SMALL-SCALE PROCESSING SOLUTIONS

In the US, small-scale industries for LNG distribution from micro-LNG plants are developing. In these projects, a small-scale or micro-LNG producer draws feed gas from a gas pipeline grid, liquefies the gas in a small liquefaction plant, and then trucks the produced LNG fuel to local harbors. shows the breakeven costs for typical projects from a study by Deutsch Bank. These costs can make participating in the LNG/gas processing industry very challenging for the players involved and for future players waiting to join the club. For future projects, the same approaches as those taken in the past will not work. New strategies must be developed and followed. The merits of small-scale to mid-scale LNG projects are discussed here, although some of these concepts can be applied to larger projects when they again become economical.

FIG. 2

Completely integrated LNG/gas supply chain. Mega-proj-

ects are out of favor at present, and it is inconceivable that new projects would be financed and executed in a similar manner as the traditional SPA-based projects. A lack of financing due to the uncertainty of the long-term SPAs greatly complicates these project developments. The structures of the new projects would need to be completely different, from a cost-and-supply point of view. Standalone mega-liquefaction projects with breakeven costs upwards of $12/MMBtu–$14/MMBtu are not feasible at present. The new bar for future projects is much lower—around $4/MMBtu– $6/MMBtu. The structures of the new projects must include the total integration of the downstream supply side into the upstream supBrass LNG

FOB cost, breakeven Shipping cost

LNG Canada Kitimat LNG Sabine Pass* Mozambique Tanzania Pluto Gordon Ichthys Prelude FLNG Wheatstone OC LNG Browse Shtokman 0

2

4

6

8 $US/MMBtu *Assumes $4/MMBtu Henry Hub spot gas price

10

12

14

16

FIG. 2. Breakeven costs for typical LNG projects. Source: Deutsche Bank.

18 JULY/AUGUST 2016 | GasProcessingNews.com

ply side. Cost distribution, financial risks and rewards must be shared with all of the stakeholders and beneficiaries in the supply and distribution chain for projects to move forward. Reducing capital costs of projects. For new projects to be

realized, the overall capital cost of the entire project must be addressed, first and foremost. Several considerations and factors should be kept in mind: • Formation of alliances and partnerships • Risk-sharing from “cradle to grave” • Addition of modular capacity as needed or as justified • Next-generation modularization to reduce cost and schedule • Standardization of design • Minimization of cost on high-cost items • Minimization of time to market • Use of prefabricated equipment like tanks and compressors, rather than custom-designed, fieldfabricated equipment. Project financing will need to be much broader and more complex, with larger alliances and partnerships to share the risks and rewards of these projects going forward. The time has come to take a serious look at the rapidly escalating construction costs of large-scale projects in the industry. One potential idea is to execute project construction differently to increase cost control availability. Significant advances have been made to reduce construction costs with the next generation of modularization. The new techniques being developed lower construction costs by 30% and reduce schedule by 25%, resulting in quick payout and cost savings. Everything must be challenged, starting from codes and standards to the way things are built—for example, the use of expensive pipe racks; the optimum use of proper materials; and even sparing philosophy, reliability and availability. Project developers have more flexibility in small-scale and mid-scale LNG facilities due to the smaller sizes of these projects. The highest-cost items for most LNG projects are: • Storage tanks • Jetties • Marine facilities • Boiloff gas (BOG) handling units • LNG vaporization • Infrastructure, including pipe racks. For small-scale LNG facilities, tremendous opportunities exist to reduce costs in these critical areas. Unlike the custom, stickbuilt, large-capacity storage tanks for mega-size LNG projects, the small-scale facilities can use prefabricated tanks. They can also utilize pressurized storage, whereas large, conventional LNG terminals must use atmospheric storage. New technologies, such as pressurized storage tanks, are available off-the-shelf from multiple suppliers in increments of 1,000 m3, which are suitable for small-scale LNG. One major advantage is BOG handling. The application of this technology in small-scale LNG has been demonstrated and proven. For small-scale LNG facilities, the generated BOG is put directly into the gas pipeline or into power generation to supplement utilities. For most small-scale LNG plants, the BOG handling cost can be significantly reduced by allowing the pressure to increase in the storage tanks. Then, the high-

OPTIMAL FILTRATION FOR MAXIMUM PERFORMANCE Optimize Your Filtration Performance • Glycol Dehydration • Rich Amine Treatment • Injection Water • Compressor Protection • Turbine Protection • Hydrocarbon Condensate • Molecular Sieve Drier Protection

Optimizing the reliability of your ltration systems is essential in maximizing your operational performance. At Jonell, our emphasis is to provide the most advanced technologies to meet your increasing production and processing needs. Jonell is a leading designer and manufacturer of ltration/separation products and services for Gas Processing, Transmission, and LNG Industries. Regardless of your application, Jonell’s ltration/separation products and services can maintain and enhance your operational performance.

Phone: 254-559-7591 E-mail: [email protected]

www.jonellinc.com

SPECIAL REPORT: SMALL-SCALE PROCESSING SOLUTIONS pressure gas can be put into the pipeline or burned for auxiliary power generation. Unlike conventional facilities, LNG vaporization for smallscale LNG can be simply executed with an atmospheric vaporizer using ambient air. This technology has been available in nitrogen facilities for many years. Drivers for small-scale LNG. The drivers for small-scale LNG

are many and depend largely on geographic location. For Asia and most of the undeveloped parts of Latin America, the biggest need is for power to fuel economic growth. Unlike North America or Europe, which have mature and developed gas markets, these developing regions do not need natural gas in its gaseous form; they need electric power to sustain their developing standards of living. Some communities are using small, diesel-based generation, with the obvious negative impact of environmental pollution (unless they use expensive low-sulfur diesel). Most developing regions need less than 50 MW of power generation, with occasional required capacities of between 100 MW and 150 MW. Small-scale LNG is perfectly suited for such an application. The author has designed modularized LNG facilities for small-scale power generation in a cost-competitive way that can be brought online relatively quickly, unlike conventional facilities. The overall cost and complexity must be balanced with the costs and time needed for turnaround. One of the biggest drivers in Europe and North America is diesel replacement, for which natural gas is well suited. The EU

Introducing the

US GAS PROCESSING PLANT DIRECTORY

US GAS PROCESSING

PLANT DIRECTORY 500+ Plants with Information about Name, Capacity, Plant Scope and Detail, and Owner/ Operator Information. Gain the market knowledge to grow your business and inform your decisions in the booming United States gas market. $1,195 per edition. Searchable, digital format. Group rates and site licenses are available.

Order at GulfPub.com/GPPD or call + 1 (713) 520-4426.

20 JULY/AUGUST 2016 | GasProcessingNews.com

has the major initiative of sulfur emission control areas (SECAs), and the US has a parallel program. Both initiatives include a large push for small-scale LNG facilities. Although diesel prices are somewhat low at present due to the glut of oil being processed, the natural gas option will be cheaper and better in the long run. Bunker oil has traditionally been used as marine transport fuel. The EU and US emissions initiatives mandate the replacement of bunker fuel with natural gas to reduce pollution caused by marine vessels. This replacement is a crucial milestone for the industry and has sparked the development of LNG engines for marine, truck, bus and rail transport across the world. Engines are already available for commercial use. A large market also exists to convert diesel generator sets into dual-fuel sets using LNG as an alternative fuel. Furthermore, new crude supply tankers, cruise ships and passenger ferries are being built to run on LNG. As these initiatives are realized, opportunities for infrastructure development are emerging. These opportunities include small LNG terminals that can be supplied by smaller marine vessels, trucks or trains. For domestic gas use, small-capacity ISO containers are already available that can be trucked to remote locations, as seen in Northern Europe. In the US, small-scale industries for LNG distribution from micro-LNG plants are developing. In these projects, a small-scale or micro-LNG producer draws feed gas from a gas pipeline grid, liquefies the gas in a small liquefaction plant, and then trucks the produced LNG fuel to local harbors, such as Los Angeles or Long Beach in California, where new environmental initiatives are being enforced. These small-scale LNG plants can even supply fuel for small trucking or bus fleets. Takeaway. Although the future for standalone mega-projects is on shaky ground, tremendous opportunities for growth exist in small-scale and mid-scale LNG. These opportunities may come from monetizing small pockets of stranded gas, growing power and transportation fuel needs, tightening marine fuel regulations, the replacement of diesel with gas, and a host of other infrastructure and supply chain opportunities. Additionally, the opening of the second lane of the Panama Canal will further reduce the timeline of LNG delivery to customers and shipping costs to smaller Asian countries. For serious investors, the opportunities for small-to-mid-scale LNG are wide-ranging and varied, and limited only by the imagination. GP ROMEL S. BHULLAR is a senior technical director and a senior fellow with Fluor Corp.’s office in Aliso Viejo, California. As part of the energy and chemicals group, he has over 35 years of experience in conceptual development, feedstock, engineering and construction, commissioning and startup, gas processing, liquefaction and regasification, pipelines and infrastructure for both onshore and offshore gas projects. His main focus is solving complex issues related to technical processes, process control, safety, process integration, process automation systems and subsystems, control rooms and other enterprise integration functions. Mr. Bhullar has executed multiple large- and mega-sized upstream and downstream projects for national and multinational energy companies. At present, he is working as a technical advisor on Fluor’s team for Asia’s newest LNG trading terminal in Singapore, operated by SLNG. Mr. Bhullar has authored several dozen articles in the areas of advanced process control, process safety, process automation and integration in major national and international professional magazines. He has also presented at major technical conferences, including Hydrocarbon Processing’s International Refining and Petrochemical Conference (IRPC), and conducted training and technical workshops around the world.

SPECIAL REPORT: SMALL-SCALE PROCESSING SOLUTIONS

Market development is key to success for small-scale LNG

S. BONINI, Muse, Stancil & Co., London, UK; and A. CHANDRA, Muse, Stancil & Co., Houston, Texas

The small-scale LNG (SSLNG) industry continues to grow, driven largely by rising environmental factors. When combined with the low cost of natural gas vs. competing fuels, LNG makes a good societal and economic choice. The growth of interest in LNG as a primary energy source has led to technological developments in end-user markets, driven foremost by China, with the US and Europe considerably behind in terms of installed capacity. This article examines the factors behind the SSLNG renaissance and concludes that it is a sustainable trend. Although the recent collapse in liquid fuel prices has slowed the conversion rates from diesel and gasoline, the environmental imperative to improve air quality in cities around the world continues. This environmental issue will determine the SSLNG sector’s future. The collapse in natural gas prices in the US, and for LNG globally, makes natural gas an attractive primary energy source with clear environmental benefits. Forecasters anticipate that the spread between oil and gas is likely to open up again, given that oil prices have more upside potential than natural gas prices in most markets. SSLNG market progress. The SSLNG

industry generally encompasses projects producing less than 1 metric MMtpy of LNG, which is equivalent to a natural gas flow of 140 MMscfd, or 1.72 MMgpd of LNG. The average SSLNG plant is only one-tenth of this production; it requires approximately 14 MMscfd of feed gas. The industry is structurally very different from the large global LNG business, even though the underlying gas processing needs are the same and are

achieved in similar ways. The primary challenge for SSLNG lies in developing the markets and infrastructure downstream of production units. In terms of capacity, according to International Gas Union statistics, SSLNG provides approximately 5% of total LNG production capacity at 20 metric MMtpy of installed capacity. China has 75% of total SSLNG capacity, with 15 metric MMtpy of total production across 125 plants, suggesting an average production of only 0.12 metric MMtpy. These Chinese plants are liquefying gas primarily for use in trucks. The large-scale LNG industry is driven by utility companies and follows a well-established and readily understood model. A large-scale LNG project is driven by the presence of substantial natural gas reserves that are cheap to produce. These supplies are purchased by a few buyers, which are typically large, sophisticated industrial utilities and power companies purchasing under long-term contracts for at least a decade. The LNG is imported to terminals, where is it is regasified and sent out into a high-pressure transmission system. In this scenario, both the sellers and the buyers are involved in the energy industry. The small-scale industry is far more fragmented, with a wide variety of players and drivers. As a result, it is less understood and far less developed. SSLNG is typically sold to much smaller, fragmented buyers whose motivation could be environmental and/or economic. In the small-scale world, LNG will often be distributed to displace diesel usage. The buyers are often not energy companies at all; they are simply buying a fuel for their own use. Typical drivers for SSLNG production include:

• Air quality issues around particulate emissions from diesel engines and, in the marine world, the introduction of Emission Control Areas (ECAs) • The drive to reduce natural gas flaring, particularly in US shale areas • The displacement of liquid fuels with cheaper LNG (this is the most prevalent driver for the US). Marine regulation—A key driver. Due

to the sharp increase in world trade, a rise in air pollution from shipping has been seen. Unlike the land transport sector, shipping has historically been much more lightly regulated and has had minimal air emissions regulations placed upon it. The International Maritime Organization (IMO) began work on this problem in 1997 and developed international standards, policies and agreed zones in which nations could implement ECAs. In these coastal areas, all marine traffic is required to control emissions of some or all of the following: SOx , NOx and ozone-depleting substances, volatile organic compounds (VOCs) and particulates. FIG. 1 shows the designated areas for control. These zones have been agreed upon internationally, and individual states are now implementing regulations within their jurisdictions. The US began drawing up its regulations in 2009 and is now beginning enforcement. Europe has implemented its ECAs in the Baltic and North Seas. Meanwhile, China announced at the end of 2015 that it will be establishing ECA measures across a large area of the Pearl River Delta. The regulations will be tightened over time and differ between zones, with a trend of stricter controls and cleaner operations. Gas Processing | JULY/AUGUST 2016 21

SPECIAL REPORT: SMALL-SCALE PROCESSING SOLUTIONS The introduction of ECAs has meant that ships must either fit scrubbers to remove SOx , or switch to expensive ultralow-sulfur-content fuels while in the ECA. An alternative is to move to fuels such as natural gas. The LNG carrier fleet has, for some years, been developing dual-fuel (diesel/natural gas) propulsion systems. A tremendous effort is being made globally to develop LNG as a marine transport fuel, with a strong emphasis on smaller, nearshore and inland vessels. The LNG volumes required for the marine industry are significant when studied collectively, but they are small on an individual basis. The LNG fuel tank for the world’s first LNG-fueled container vessel is 900 m3, or approximately 400 metric t. Existing large-scale LNG export facilities supply cargoes of 170,000 m3 and larger, and load a ship every three days. A whole new set of infrastructure and production units are required to serve the transportation market. It is an entirely different business, with different commercial models. China: A success story. China represents 75% of installed global capacity and has experienced strong, rapid growth, despite its comparatively low gas reserves.

The country has achieved success in SSLNG with strong central policy directives aimed at increasing the use of natural gas as a vehicle fuel, displacing diesel from larger trucks as part of the effort to improve air quality in cities. China has an LNG-fueled heavy truck fleet of approximately 200,000. Energy prices are regulated in China, with the government setting key prices to achieve particular policy objectives. The country keeps natural gas prices low and sets locally produced LNG prices to be at least 20% cheaper than diesel, making it more attractive for end users to switch. When oil prices were high, the SSLNG producers had a good business model with low input prices and a significantly higher LNG price. Recent drops in diesel prices have caused Chinese regulators to lower LNG prices to protect the trucking industry at the expense of SSLNG producers, and the growth has slowed dramatically. The SSLNG plants are built predominantly alongside pipelines providing pipeline-quality gas. A significant number of plants take feed from coalbed methane or stranded inland gas reserves located away from gas infrastructure. SSLNG provides a monetization route for these reserves.

This non-free market approach may appear strange to US readers. The Chinese have kickstarted the use of LNG as a transport fuel, rapidly increasing SSLNG production capacity, the truck fleet and associated infrastructure. The effort has been so successful that the Chinese government is looking to repeat the program for marine transport for inland river traffic. As previously mentioned, China has announced its first ECA area, as well as its first LNG marine bunkering facility. The expectation is that these moves will result in the same sort of growth in SSLNG serving the Chinese marine sector. This approach is currently inconceivable in the US, although it is worth remembering that, until the 1980s, US wellhead prices were determined by the federal government. It does show that the technologies are viable and can be rapidly deployed, if a coordinated approach is made with clear policy directives. US experience and outlook. The US experience has been built on a free-market approach and is helped by a key marine regulation. The country’s SSLNG industry has received a tremendous boost in recent years as a result of the natural gas bonanza. The US also has ac-

New ECA?

ECA ECA

New ECA?

ECA

New ECA? New ECA? New ECA? New ECA? Existing ECA Possible future ECA

FIG. 1. Existing and possible future ECAs. Source: DNV.

22 JULY/AUGUST 2016 | GasProcessingNews.com

SPECIAL REPORT: SMALL-SCALE PROCESSING SOLUTIONS cess to large, cheap natural gas reserves for the foreseeable future. When oil prices were high, the price differential between natural gas and gasoline/diesel made switching between the fuels attractive. Considerable investment followed in natural gas engine technology and distribution infrastructure, most notably by oil tycoon T. Boone Pickens’ Clean Energy initiative. However, the subsequent crash in oil prices has reduced the price differential and, therefore, the motivation to switch away from gasoline and diesel. Nonetheless, progress continues to be driven by the US gas reserve base and the move toward “greener” transportation. The most notable recent developments in SSLNG production in the US are taking place in Florida around interconnected markets in port and rail infrastructure. These developments are driven by both economic and environmental factors. New Fortress Energy, a wholly owned subsidiary of New York-based global investment management firm Fortress Investment Group (FIG), also owns Florida East Coast Railway. The railway has ordered LNG-fueled locomotives from GE and is actively converting its fleet. FIG also purchased Raven Transport, a trucking company with one of the largest LNG-fueled fleets in the US. Raven owns nearly 200 LNG-fueled trucks and plans to reach approximately 500 units by 2019. The “chicken-and-egg” problem that faces many small-scale developers that try to develop downstream markets with third parties can be alleviated by partnering or cooperating with a group of affiliated companies. FIG, which is developing rapidly in the SSLNG space, has taken the latter approach. In other developments, the JAX LNG plant at Jacksonville will shortly be supplying LNG to the port as a bunkering fuel. This facility was developed and is owned by Crowley Marine and Pivotal LNG, with the primary focus on the marine market. It will be the supply source for the world’s first LNG-powered container ships based there—Isla Bella and Perla Del Caribe, planned for use within the Caribbean. Crowley is already shipping 10Mgal ISO tanks to Puerto Rico for use in the Coca-Cola bottling facility, providing a cleaner and cheaper alternative to diesel. European SSLNG efforts. The SSLNG sector has seen the most activity in north-

Although the recent collapse in liquid fuel prices has slowed the conversion rates from diesel and gasoline, the environmental imperative to improve air quality in cities around the world continues. This environmental issue will determine the SSLNG sector’s future. ern Europe, around the Baltic and North Seas, where the first ECAs have been implemented. The Norwegian government has established a NOx fund that charges all NOx emitters in the zone while offering subsidies to companies that reduce emissions using new technologies. In recent years, this has led to payments of $58 MM to LNG-fueled vessel owners. Two offshore platform supply vessels, three passenger ferries and one gas carrier are now operating in the region, with 17 more LNG-powered vessels under construction and planned. The LNG-powered fleet is growing rapidly, with 80 or so forecast to be in operation by the close of 2016. Of this total, 56 are forecast to be operating in the Baltic and North Seas. SSLNG production has grown around Norway to support the marine market. Norway also uses SSLNG to supply remote coastal communities, and it serves these markets with a number of SSLNG carriers. SSLNG for international trade. This review has looked at the principle growth areas for SSLNG as a local clean fuel supply, and where the LNG is produced and consumed within the same country, or even the same province. This scenario is quite different from the large-scale industry, which exists to transport large, remote and stranded reserves to populous markets. SSLNG is now carving out niche markets. As mentioned above, Crowley Marine is supplying LNG in ISO tanks to Puerto Rico. The company recently announced the purchase of 19 additional ISO tanks to meet increasing demand. Meanwhile, FIG recently received approval from the US Department of Energy to export LNG to Jamaica in ISO tanks. Also this year, Hawaii Gas announced its intention to import LNG from Clean Energy in ISO tanks. In Thailand, Thai-based LNG Plus International Co. Ltd. recently complet-

ed a small-scale gas-fired power plant in Myanmar, supplied with LNG by road tanker from Thailand. In June, Japan’s Ministry of Land, Infrastructure, Transport and Tourism announced that it will be seeking to trial LNG bunkering in the port of Yokohama. All of these developments are significant for SSLNG production and supplychain technologies. It is particularly notable that the US is leading the charge with new entrants and not with traditional LNG players. SSLNG is truly a separate industry with a shared technology. GP SIMON BONINI has over 30 years of experience in the LNG and international energy industries, and has managed all aspects along the LNG chain, including developing and implementing new strategies. He spent 17 years at BG Group developing several businesses, including the Trinidad export project, BG’s shipping business, BG’s position in Lake Charles, and the global LNG trading business. Mr. Bonini also worked at Woodside, establishing its US LNG strategy, and at Centrica, founding its LNG import business. Most recently, he has been active as COO and board member of 4Gas and Dragon LNG. He was also one of the founding partners of Parallax Energy and the CEO of Louisiana LNG and Live Oak LNG. Mr. Bonini holds a first class degree in chemical engineering from Imperial College in London, and an MBA degree from INSEAD in France. He is also a fellow of the Institute of Chemical Engineers. AJEY CHANDRA is a director at Muse, Stancil & Co., and the managing partner of the Houston office, where he also leads the midstream practice area for the firm. He joined Muse in 2014 after 28 years of experience in various facets of the midstream industry, including operations, engineering, business development, management and consulting. During his career, Mr. Chandra has had a wide variety of assignments covering all aspects of the energy industry, and he has had several long-term expatriate assignments overseas, including Europe and Southeast Asia. His operating, consulting and management experience includes working at Amoco, Purvin & Gertz, Hess and NextEra Energy Resources prior to joining Muse. Mr. Chandra holds a BS degree in chemical engineering from Texas A&M University and an MBA degree from the University of Houston. He has also attended executive education classes at Harvard Business School and is a registered professional engineer in Texas. Gas Processing | JULY/AUGUST 2016 23

Same great people. Same great service. Same great company.

You know what we do. We can do more together.

Valerus is becoming SNC-Lavalin’s Production & Processing Solutions Valerus is now a part of the larger SNC-Lavalin family. Together, we provide unmatched modular equipment facilities on a turnkey basis.

To learn more, visit snclavalin.com/en/oil-gas/processing-treating © 2016 SNC-Lavalin.

LNG TECHNOLOGY

Develop successful nearshore FLNG solutions— Part 1: Gas pretreatment strategies S. MOKHATAB, Consultant, Dartmouth, Nova Scotia, Canada

Gas pretreatment strategies. In a typical nearshore FLNG facility (FIG. 1), feed gas is transported via pipeline and flexible riser to the FLNG facility, where impurities are removed (in the gas pretreatment section) and the gas is liquefied before being stored onboard the facility. Pursuant to LNG, other liquid products (LPG and condensate) will be stored and subsequently offloaded to marine carriers for delivery to market. While each facet of the nearshore FLNG production facility is important, the gas pretreatment section of the facility plays a critical role in treating the raw feed gas to meet final sulfur specifications and purity levels required by the natural gas liquefaction unit. The specifications to be met include hydrogen sulfide (H2S) removal to under 4 ppmv, carbon dioxide (CO2 ) to below 50 ppmv, total sulfur to under 10 ppmv as S, water to less than 0.1 ppmv, and mercury (Hg) to the level of 0.01 µg/ Nm3. Also, heavy hydrocarbons (HHCs) shall be removed to below freezing limits in cryogenic heat exchangers (typically in the range of 0.1 mol% for C6+ and a few ppm for aromatics). The general nitrogen (N2 ) specification for rundown LNG is 1 mol% maximum. TABLE 1 shows feed gas impurities in various locations. Different locations require different gas treatment schemes. The following section briefly discusses the different technologies in treating sour feed gas to the natural gas liquefaction unit, and the technology in the removal of contaminants to meet environ-

mental and emissions regulations and LNG feed gas specifications. Acid gas removal. The acid gas removal unit (AGRU) mainly

removes the acidic components, such as hydrogen sulfide (H2S) and carbon dioxide (CO2), from the feed gas stream. This process helps meet the sales gas H2S specification and avoid CO2 freezing (and subsequent blockages) in the cryogenic exchanger, respectively. It also removes some amount of carbonyl sulfide (COS), mercaptans (R-SH) and other organic sulfur species that contribute to sulfur emissions. Three solvent absorption processes (chemical absorption, physical absorption and the mixed solvents) are commonly used for acid gas removal. These processes can also be used in FLNG production facilities. Most commonly, H2S and CO2 are removed from the natural gas feed stream in a chemical solvent unit utilizing an aqueous amine solvent. With the exception of methyldiethanolamine (MDEA), amines are generally not selective and will remove both CO2 and H2S from the gas. When used in treating sour gases to meet the tight CO2 specification for LNG production, the activity of CO2 absorption is too slow with pure MDEA, which must be enhanced with a promoter (i.e., piperazine). In contrast, a feed gas with 10 mol% CO2–12 mol% CO2 can be handled by a promoted MDEA process. The advantage of the amine technology is that the solubility of aromatics and heavy hydrocarbons in the aqueous solvent is low, resulting in lower hydrocarbon losses. However, the disadvantage is the high energy consumption for the regeneration of aqueous amine. Physical solvents, which can be applied advantageously when the partial pressure of the acid gas components in the feed Flare

Liquefaction

Gas pretreatment

Inlet facilities

Helideck

Process utilities

Power generation

Turret– swivel

Movement has been seen in offshore floating liquefied natural gas (FLNG) developments. However, nearshore projects that utilize a barge-based floater located in a nearshore environment, and that also take advantage of onshore support, may be more secure than offshore alternatives. One reason for this scenario is that the offshore projects could face greater technical challenges and higher costs. As such, great interest exists in developing robust, reliable and innovative natural gas pretreatment and liquefaction solutions for nearshore, lower-cost FLNG projects. At the nearshore FLNG facility, the need exists for a compact, flexible and energy-efficient pretreatment package to remove contaminants. This package must also deliver feed gas with the required specifications by the natural gas liquefaction unit to maintain continuous uptime of LNG production. Part 1 of this article describes the appropriate processing technologies for designing a robust pretreatment section and shows how the integration of treatment technologies and expert know-how make a difference.

LNG storage tanks

LPG storage tanks

Condensate storage tanks Feed gas

Electrical rooms Hull utilities Buoy

Riser

Mooring line

FIG. 1. Typical layout of a nearshore FLNG facility (modified after Festen et al., 2009).1 Gas Processing | JULY/AUGUST 2016 25

LNG TECHNOLOGY gas is high (typically greater than 60 psi), will lower the energy consumption but will coabsorb more hydrocarbons. For this reason, physical solvents are generally suitable to treat lean gases. When the need exists to treat rich gases, the NGLs must first be removed by chilling. As an alternative, the use of hybrid (mixed) chemical and physical solvents is beneficial, where they can be formulated to allow for complete CO2 removal, while achieving H2S removal comparable to alkanolamines. In hybrid systems, mercaptans (R-SH) and other organic sulfur components, if present in the feed gas, can also be removed by the physical solvent portion. Generally, this option will result in an expensive design with a hydrocarbon coabsorption that is too large to be acceptable.3 In many cases, the optimum solution is the distribution of the mercaptans removal capabilities over the optimized mixed chemical-physical solvent in the AGRU and the molecular sieve unit (MSU). In this option, the regeneration of the MSU gas can be integrated with the AGRU using a shared regeneration system. The treated regeneration gas can then be recycled either to the inlet of the MSU or the inlet of the AGRU absorber.3 A number of processes (membranes, cryogenic fractionation and adsorption) are also available to remove H2S and CO2 from natural gas. Membrane separation, which offers several advantages compared to an amine unit (i.e., greater turndown capability, and reduced installation costs and plot space) is only suitable for bulk CO2 removal, where further treating with amine is required to meet H2S and CO2 specifications. The membranes require a suitable pretreatment system to remove particulates and to avoid liquid formation in the membranes. Improper pretreatment generally leads to performance decline rather than to complete nonperformance. The main limitation of the membrane system is linked to the significant loss of hydrocarbons in the CO2 discharge. This constriction is partly due to the relatively large membrane surface area TABLE 1. Feed gas impurities in some areas2 South America

Southeast Asia

H2S, ppmv

5–1,000

5–200

2–50

1,000–2,900

Total sulfur, ppmv

5–1,000

5–250

2–60

0–400 (R-SH) 1–40 (COS)

Australia

Middle East

CO2, %

2–55

9–50

2–30

2–7

Hg, µg/Nm3

0–100

200–2,000

50–200

0–50

H2O, ppmv

Saturated

Saturated

Saturated

Saturated

TABLE 2. Molecular sieve vessel size comparison of a typical molecular sieve unit4 Single-layer molecular sieve, 1⁄8 in. Molecular sieve quantity/vessel, kg

Split configuration Split (combining configuration 1 ⁄8-in. and 1⁄ 16-in. using dense molecular sieves) particles

27,000

21,250

21,200

Vessel internal diameter, m

3.3

3.3

3.15

Vessel height, m

4.47

3.65

3.4

Vessel volume, m3

38.2

31.2

26.5

26 JULY/AUGUST 2016 | GasProcessingNews.com

that would be required to reach a 50-ppm CO2 spec. Membrane systems perform well at reduced feed flowrates, but their performance drops when design flowrates are exceeded. Additional modules must, therefore, be added in parallel to accept higher flowrates. As a result, the membrane separation process does not realize the economies of scale as the flowrate is increased. Cryogenic fractionation appears to offer a good prospect for removing CO2 and H2S from natural gas. However, this technology requires substantial energy to provide necessary refrigeration. It also requires pretreatment of feed gas to remove components with a freezing point above the operating temperature to avoid freezing of lines and blockages of process equipment. Adsorbents for acid gas removal are generally limited to small gas streams operating at moderate pressures. For example, molecular sieve technology may be a cost-effective method to remove low CO2 contents up to 2 mol% in small-scale facilities (with a feed gas flowrate of about 40 MMscfd), where an amine absorption unit is not considered suitable from a capital expenditures (CAPEX) point of view. Note: The discharged acid gas stream can be routed to the flare stack to ensure its safe disposal, in the case of low H2S content, or reinjected to a suitable reservoir to minimize environmental impact if the concentrations of acid gas components are too high. However, acid gas injection will require an additional system for dehydration, unless water is knocked out at 800 psi–900 psi. This injection will prevent corrosion and hydrate formation, as well as compression, all of which add costs, complexity and safety considerations to the nearshore FLNG facility design. For high H2S contents, the discharged acid gas stream can also be routed to an onshore plant for sulfur recovery. However, this method poses additional export and handling issues. Water removal. Molecular sieves are used to dry the gas leaving the AGRU to below 0.1 ppmv to avoid hydrate formation in the NGL recovery unit. They can also be used for the removal of mercaptans and other sulfur compounds to meet the product specification of 10 ppm. Molecular sieve units, if properly designed, can economically handle only feed gases containing a maximum of 1,500 ppmv RSH.3 While moisture removal is traditionally done with the smaller-pore-sized 3A and 4A molecular sieves, mercaptans/sulfur removal is accomplished with the larger-pore-sized 5A and 13X types. The 5A type molecular sieve is used for trace removal of H2S and the removal of light mercaptans (C1/C2–SH), while the 13X molecular sieve is used for the adsorption of heavy/ branched mercaptans. However, coadsorption of benzene, toluene, ethylbenzene and xylene (BTEX) components with concentrations higher than 30 ppmv on 13X molecular sieves will result in increased length of the molecular sieve bed. These components can also cause transients in the concentrations of these components in the regeneration gas. Transients can cause separation problems in the physical absorption process used to recover the mercaptan species from the spent regeneration gas. The practical solution for such a purpose is to use a 5A molecular sieve for removing light mercaptans in the gas phase, as this product has no BTEX capacity. The heavier mercaptans are then removed with the LPG and condensate (C5+) cuts, which may be further treated downstream using a caustic scrubber

LNG TECHNOLOGY process, followed by a molecular sieve unit to dry the treated liquids to meet the required product specifications.3 The key role of the molecular sieve units in gas pretreatment increases the need to understand the design principles and operation of such units to optimize the size and improve the performance of the molecular sieve units in FLNG projects. In recent years, various techniques have been proposed to reduce the unit size. For example, using split-bed configurations of dense molecular sieves can reduce bed voidage and vessel volume (TABLE 2). Using high-quality molecular sieves with superior properties and improved regeneration methods can extend bed lifetime and improve reliability while providing cost savings. Mercury removal. Removal of mercury using nonregenerative metal-sulfide sorbents or regenerative silver-impregnated molecular sieves is required to avoid the risks of mercury attack on the brazed aluminum heat exchangers and equipment in the cryogenic section. The mercury removal unit can be positioned upstream or downstream of the AGRU. Installing vessel(s) of non-regenerative sorbents before the amine unit removes all mercury and prevents contamination through the remainder of the FLNG production facility. Although this method appears to be costly, it is actually very simple, since no regeneration equipment is required. It is also a very safe and conservative approach to handling mercury in the feed gas. Installing a nonregenerative mercury removal sorbent downstream of the amine unit, just before the molecular sieve unit, reduces the size of the molecular sieve beds to some extent, but it also poses the risk of mercury contamination of the solvent system. Adding a silver-impregnated mercury sieve section to the molecular sieve beds to simultaneously remove water, mercaptans and mercury provides another option with a potentially longer service life. However, this option requires a separate vessel of nonregenerative metal-sulfide adsorbent for treating a relatively high mercury content in the regeneration water that would result in additional costs.5 Note: The mercury-contaminated wastes should be sent onshore for proper disposal at a hazardous waste facility. HHC removal and NGL recovery. Removal of HHCs (C6+ and aromatics) from the gas to be liquefied is necessary to avoid waxing and plugging in the main cryogenic heat exchanger (MCHE). The usual solution is to use a scrub column ahead of the liquefaction unit operating at liquefaction pressure and thermally integrated with the MCHE (FIG. 2). Although this method has been widely used, it has limitations in terms of inlet feed gas operating pressure and composition. In fact, a significant reduction in the scrub column pressure (to below the critical point) may be necessary, resulting in reduced liquefaction efficiency and increased power consumption. In addition, when the gas becomes lean in C2 or C3+, it is difficult for the column to operate stably and efficiently due to insufficient liquid reflux in the column. An alternative to using a scrub column is to use an NGL extraction unit to recover the C2+ or C3+ hydrocarbons from the treated/dried gas. Conventional turboexpander technology can be used to produce a lean gas for liquefaction to comply with LNG product specifications. Although propane and butane pose no freezing problem, they are removed together with the heavy hydrocarbons and

can be separated and sold as liquid products. In addition, the extracted ethane is returned to the natural gas stream and used as refrigerant makeup or to supplement the fuel gas. Although a front-end NGL extraction unit utilizing conventional turboexpander technology can handle a wide variety of feed gas compositions and effectively remove HHCs, it contains rotating equipment that impacts the capital investment and reliability of the FLNG facility. Today’s proprietary NGL recovery processes may reduce capital costs through the use of high-efficiency expanders/compressors and compact heat exchangers, but they may prove difficult and complex to operate. This complexity makes these processes less desirable for most FLNG facilities that prefer operational simplicity and minimum maintenance designs. Note: In case the need exists for the removal of small quantities of HHCs from pipeline-quality gas to meet the more stringent specification for LNG, applying a silica-gel-based adsorption process to adsorb heavy hydrocarbons at high pressure (without removing lighter ones) is an economical option over other existing processes. In this case, the desorbed heavier components from the adsorption unit may be preferentially fed to the fuel gas system, which avoids the need for LPG removal and storage. Nitrogen removal. The presence of more than approximately 1 mol% of nitrogen (N2 ) in LNG may lead to auto-stratification and rollover in storage tanks, presenting a significant safety concern. A higher percentage of N2 content in the feed gas also impacts the liquefaction process itself by reducing liquefaction efficiency (additional refrigeration requirements per unit of LNG produced, due to the need to condense N2 in the feed gas). In addition, high-N2 -content feed gas may require treatment of (or the spiking of higher-Btu gas into) the BOG so that it may be used as the fuel gas for the gas turbine(s) on the FLNG facility. Therefore, the need exists for an efficient technique for the removal of N2 from LNG, even for relatively low N2 levels. For feed gas containing N2 levels of approximately 1 mol% to 2 mol%, N2 can be removed in the end-flash section within an FLNG production facility. When N2 is present in high concentrations (greater than 5 mol%), it should be removed in the front section of the liquefaction unit to minimize liquefaction energy requirements. Several N2 rejection methods exist, including cryogenic separation, membranes and molecular sieve technology. NG

MCHE

NG LNG

NGL

Reflux Overhead

Precooling

Dry NG

Propane refrigerant

Scrub column

Lean gas

Liquefaction MCHE

Mixed refrigerant

NGL

FIG. 2. Integrated LNG unit, scrub column.6 Gas Processing | JULY/AUGUST 2016 27

LNG TECHNOLOGY H2S, CO2, RSH Regen gas absorber

Common regenerator CO2 Feed gas from HP separator

Mercury removal (nonregenerative sorbents)

Bulk CO2 removal (membrane)

H2O Absorber (optimized mixed physical/ chemical solvent)

H2O and R-SH removal (molecular sieve technology) Hot regen gas

Mercury

To N2 removal/ HHCs removal/NGL recovery liquefaction unit (conventional turboexpander technology)

NGL fractionation

C2 C3 C4 C5+

Treatment/ drying

FIG. 3. Typical integrated pretreatment scheme for nearshore FLNG facility receiving raw, high-CO2-content feed gas.

Feed gas from HP separator

Mercury removal (non-regenerative sorbents) Mercury

CO2

H2O

CO2 removal (promoted MDEA solvent)

HHCs H2O removal (silica gel technology)

H2O Residual H2O To N2 removal/ removal liquefaction unit (molecular sieve technology)

HHCs

FIG. 4. Typical integrated pretreatment scheme for nearshore FLNG facility receiving lean, pipeline-quality gas.

However, the only existing, viable, large-scale rejection technology is the use of cryogenic separation. In fact, the applications of membranes and molecular sieve technologies are generally limited to small scales. Membrane systems typically produce a waste N2 stream with fairly high hydrocarbon content (revenue loss), and the stream cannot be vented directly to the atmosphere. Therefore, waste N2 must be reinjected for sequestration or disposed of by other means. Molecular sieve technology is uneconomical when used to remove high levels of N2 .5 Although economic justification exists to remove N2 from the feed gas before liquefaction, it is possible to remove N2 within the liquefaction process. When N2 removal is performed in the liquefaction section, it avoids the N2 rejection unit (NRU) product compression system, with refrigeration provided by a liquefaction unit refrigeration system. It also avoids losses associated with reheating and cooling feed gas for N2 rejection. However, in this scheme, a high level of heat integration with the liquefaction system adds to process complexity and operational risk, as neither the NRU nor the liquefaction system is conventional.7 When an NRU is to be installed in conjunction with the AGRU and NGL recovery units, the opportunity exists to integrate both facilities by eliminating repeat heat exchange equipment and recompression. For example, the selection of the NGL recovery unit outlet pressure can be set to match the best 28 JULY/AUGUST 2016 | GasProcessingNews.com

efficiency point of the NRU columns, and the rejected N2 from the NRU can be used to strip the AGRU solvent. Such a simple integration concept can be incorporated into the design to achieve higher energy efficiency and reduce equipment counts while maintaining the operability of the overall process design. Integrated pretreatment scheme. Commercial process technologies, like those discussed above, can be integrated and configured into various FLNG pretreatment schemes, each offering unique benefits. FIGS. 3 and 4 show typical pretreatment schemes for nearshore FLNG facilities based on two different types of supplied feed gas. In the case of supplying pipeline-quality gas to the facility, a quick-cycle silica gel adsorption unit allows the single-step removal of both heavy hydrocarbons and water from natural gas; however, it does not generally achieve the water dewpoint of 0.1 ppm typically required in LNG production facilities. Although trim layers of molecular sieves could be added to the bottom of this unit to obtain the required water specification, the large number of repeated cycles would impose a challenge for the molecular sieve by reducing its performance and lifetime. A more practical approach is to add a small molecular sieve dehydration unit downstream of the quick-cycle unit to remove the residual ppm levels of water from the gas. In these integrated treating schemes, the main objective is to have an optimized, compact solution that can provide great process flexibility, safety and systems reliability while providing significant energy and capital cost savings. Note: The optimum solution will vary from project to project, as each feed gas is different. Takeaway. A key step in the development of an attractive nearshore FLNG solution is the selection of an appropriate gas pretreatment system that best meets the project objectives. Several technology options can be integrated into the design of the gas pretreatment section in nearshore FLNG projects. When determining the optimal integrated pretreatment scheme, safety, weight, costs (CAPEX and OPEX), reliability and operational flexibility must be considered. Next issue. Part 2 of this article will appear in the September/ October issue of Gas Processing. GP ACKNOWLEDGMENT Thanks are due to Scott Northrop for reviewing this manuscript and providing useful comments. LITERATURE CITED Complete literature cited available at GasProcessingNews.com. SAEID MOKHATAB is an internationally recognized gas processing consultant who has been actively involved in several large-scale gas field development projects, concentrating on design, precommissioning and startup of processing plants. He has presented many invited talks on gas processing technologies worldwide and has authored or co-authored nearly 250 technical publications, including two Elsevier handbooks referenced by practitioners in the field. He has held technical advisory positions for leading professional journals, societies and conferences in the field of gas processing, and has received a number of international awards and medals in recognition of his outstanding work in the natural gas industry.

GAS TREATING

Design for ultra-high-pressure H2S removal from natural gas P. ROBERTS, formerly with Advisian, WorleyParsons, Twickenham, UK

What are the pressure design limits for natural gas H2S removal and associated dehydration facilities? This question was asked in the search for facility designs to remove H2S from sour gas at a pressure of 160 bara. The inlet gas to the facility is at 160 bara, and it requires export with minimum pressure loss for reinjection. The existing facilities consist of mixed metal oxide beds to remove low levels of H2S and non-regenerable molecular sieve beds to remove the water formed in the reaction to remove H2S. The level of H2S increased over time, and was close to exceeding the capacity of the beds. The decision was made to install an acid gas removal unit (AGRU) and an associated dehydration system to address the increased levels of H2S in the feed. Two main design options were presented: 1. High-pressure (HP) design: Pressure letdown to 80 bara AGRU and 80 bara triethylene glycol (TEG)/ molecular sieve dehydration plus compression (standard technology) 2. Ultra-high-pressure (UHP) design: 160 bara AGRU and 160 bara TEG/molecular sieve dehydration.

A substantial financial incentive was identified to avoid the requirement of recompression and operate at UHP design. Here, the work carried out to validate the design of the UHP facilities is examined. It also discusses the work undertaken to verify the design and operation of an AGRU and an associated dehydration unit at UHP, as well as the challenges—both expected and unexpected—that were encountered. At the start of the design, it was anticipated that the main difficulty in designing the UHP system would be the AGRU, due to the large increase in pressure from any previous design for natural gas. Therefore, it came as a surprise that the project could not move forward—not from a lack of confidence for an AGRU to operate at UHP, but due to difficulties in finding a suitable means of dehydration at this pressure. Amine unit background. Amine units have been used for

years to treat natural gas as the primary method of acid gas removal, although very few plants operate at pressures above 80 bara. Two examples exist, both in the North Sea: • 100 bara: Statoil’s Sleipner platform • 112 bara: Gaz de France’s K-12B platform.

TABLE 1. UHP amine design challenges Point number

Challenge

Mitigation actions

1

Increased amine losses (carryover) and poor performance of absorber internals due to a reduced liquid-vapor density difference

The density difference of approximately 50 kg/m3 is greater than some well-established processes, and not significantly less than a typical amine absorber Water wash possible to minimize amine losses Add a design margin to the absorber diameter If the absorber is trayed, increase the tray spacing and size of downcomers

2

Uncertainty of physical properties

Physical properties of the feed gas are well known at 160 bara A proprietary Peng-Robinson package can be used in excess of 200 bara Sour gas injection facilities designed up to 800 bara

3

Increased hydrocarbon absorption into the amine, resulting in additional flash gas and potential for foaming

Published data exists up to 140 bara, showing that methane absorption is almost linear with pressure and appears suitable for extrapolation Additional laboratory testing Accommodate for additional flash gas in the design

4

Reduced efficiency of H2S absorption and failure to meet product specification due to hydrocarbon coabsorption and non-ideal behavior

Published data for CO2 up to 200 bara shows that UHP operation affects CO2 loadings at fixed partial pressure

5

Mechanical considerations for 160-bara operation

Use existing refinery plant experience (hydrotreater) Gas Processing | JULY/AUGUST 2016 29

GAS TREATING Many refining applications exist up to 200 bara, but these process hydrogen-rich gas rather than natural gas. Potential challenges of designing and operating an amine unit in hydrocarbon service at UHP were identified, as shown in TABLE 1. The challenges that require more mitigation are Points 3, 4 and 5. Points 3 and 4 concern hydrocarbon absorption into amine at HP, and Point 5 concerns mechanical considerations. Therefore, it was decided to obtain further data from the following sources: • A UHP amine literature survey • Additional laboratory testing to meet areas not covered by published data • Refinery data that may to be used to assess mechanical and operational issues. UHP amine literature survey. Literature is available for the ab-

sorption of UHP natural gas and CO2 in amine; however, there is little information for H2S. Two main sources of data exist: 1. The University of Alberta Gas Liquids Engineering Ltd. has published data in conjunction with the university at pressures up to 140 bara1 60

HC absorption, scf/100 gal

50 40

Methane

30

Ethane

20 10

Propane

0 0

10

20

30

40

50

60 70 80 90 100 110 Partial pressure, bara

120 130 140 150

FIG. 1. Solubility of C1–C3 in aqueous solutions of MDEA.1

100

0.0045

No methane 50 bara 100 bara 150 bara 200 bara

Exp. this work Carrol et al., 1998 Culberson and McKetta, 1951 Sol. Data Ser., 27/28

0.0040

Partial pressure CO2, bara

0.0035 xch4, mol/mol

2. Statoil, in conjunction with the University of Trondheim, has published data for pressures up to 200 bara.2,3 University of Alberta. One article1 tabulates the solubilities of methane, ethane and propane in 35 wt% MDEA for pressures up to 130 bara. FIG. 1 was developed from the data in the article. The required partial pressure of methane (for the facility at 160 bara) is 148 bara, which is only slightly in excess of the experimental data. Several conclusions can be drawn: • Methane absorption is approximately linear and, in fact, flattens out slightly at higher pressures • The absorption of ethane and propane is minor compared to methane at typical natural gas concentrations. Statoil research center’s laboratory in Trondheim. The chart in FIG. 2 2 shows that the relationship of methane absorption to pressure is approximately linear from 70 bara, with some flattening out as the pressure approaches 200 bara. This confirms the data in the previous paper from the University of Alberta (FIG. 1). The chart in FIG. 3 3 shows how the equilibrium of CO2 in amine is affected by high partial pressure and consequent high absorption of methane. The CO2 equilibrium is shifted to the left (i.e., reducing the CO2 loading in MDEA) for high pressures of methane at the same partial pressure of CO2. The methane effectively displaces the CO2 in solution. This research was used to stabilize operation on the Sleipner CO2 removal unit. The reduced loading capability at HP was overcome by increasing the amine circulation and the reboiler duty.4 The results are relevant to the facility, as H2S loading may also be affected by the high partial pressure of hydrocarbons. The literature survey described above is summarized in TABLE 2. Experimental results. In addition to the published data, one company a has provided its own experimental data, summarized in TABLE 3. For the facility design case, the following addition data was commissioned from new experimental measurements: • Confirmation and extension of data of CH4 + CO2 + MDEA (pressure 120 bara–200 bara) • Extension of solubility data for CH4 + H2S + MDEA for lower temperatures and extension of pressure range (30°C, pressure 120 bara–200 bara). The new experimental results are shown in FIG. 4.

0.0030 0.0025

10

1

0.0020 0.0015 0.0010 50

70

90

110

170 130 150 Partial pressure, bara

FIG. 2. Solubility of methane in MDEA.2

30 JULY/AUGUST 2016 | GasProcessingNews.com

190

210

230

250

0.0 0.0

0.2

0.4

0.6 0.8 Loading, molCO2/molMDEA

1.0

FIG. 3. Effect of hydrocarbon absorption on CO2 equilibrium curve.2

1.2

GasProcessingConference.com

AMERICAS

September 13–14, 2016 Norris Conference Centers – CityCentre Houston, Texas

Explore the latest gas processing developments, from field gathering to plant design to LNG technology, at the 2016 GasPro Americas Conference Register today for the must-attend event for the gas processing industry Join us at the GasPro Americas Conference, scheduled September 13–14, 2016 at Norris Conference Centers - CityCentre in Houston, Texas, USA. The 2016 event has 18 sessions organized into three tracks that will benefit CEOs, COOs, CTOs, directors, country managers, regional directors, project managers, technical directors, and heads of engineering for technology and operating companies. This conference will focus on gas supply, procurement, purchasing, transportation, trading, distribution, operations, safety, the environment, regulatory affairs, technology development, business analysis, LNG, and more. All segments of the gas processing industry—the upstream, midstream and downstream sectors—will be discussed. Session topics include: • The State of Natural Gas in the Americas • Water Treatment • Gas Treating • Separation/Dehydration • Cryogenics: Rejection, Ethane, Methane and Nitrogen • Syngas Production and Utilization

USB Sponsor:

Exhibitors:

• IOT and the Future of Big Data for the Natural Gas Industry • NGL Recovery • Fractionation • LNG/FLNG • Liquefaction/Regasification • Metering/Custody

Host Sponsors:

Save 15%

on Registration

when you use code: GP16

Operators eligible for a free pass! Contact Melissa Smith, Events Director, at [email protected] or + 1 (713) 520-4475 for details.

To attend, sponsor or exhibit, visit GasProcessingConference.com

Supported by:

GAS TREATING The existing experimental data, up to 120 bara, shows the absorbed methane increasing in an approximately linear man­ ner with increasing pressure. The new experimental data showed that the rate of increase declines with increasing pres­ sure at UHP. This result reflects the trend shown in FIGS. 1 and 2 for methane absorption at UHP. A similar equilibrium shift was observed with H2S as CO2 , as seen in FIG. 3.3 Two main conclusions from the literature survey and the experimental results were drawn: • Acid gas loading in the amine will be lower than expected due to the high absorption of methane. The equilibrium shift reduces the acid gas loading in MDEA. This reduction will be offset by the higher partial pressure of acid gases at UHP. A simple mitigation option exists in increasing the circulation rate by approximately 10%. • Increased hydrocarbon adsorption will result in a higher flash gas flowrate, which can be accommodated in design. Reference plants. As previously stated, the highest known operating pressure of an AGRU treating natural gas is 112 bara; however, it is common for AGRUs to operate at higher pres­ sures in refineries. The example refinery absorber, situated in the residue hydrotreating unit (RHU), cannot be used for assessing hydro­ carbon absorption since the hydrocarbon content is low. How­ ever, it is a good reference for safety/operational issues in a UHP absorber.

CH4 solubility

Extrapolation of existing data

Existing data

New experimental data from company

The biggest safety question for any HP amine unit is the HP/low pressure (LP) interface, which becomes a larger prob­ lem at UHP. The blow­by case can usually be designed with adequate low­level protection in the absorber and flash drum relief for gas breakthrough. TABLE 4 summarizes the UHP re­ finery reference and the facility requirements. A site visit was made to the refinery to gather more informa­ tion about the design and operation of the amine unit. • The RHU was commissioned in the 1980s. Apart from refinery shutdowns, it has operated without interruption since that time. • The RHU contains an amine unit in a reactor recycle loop. • The amine unit was originally designed with MEA but switched to MDEA (40 wt%) and MEA (5 wt%). • The amine unit has had high availability and operates without the problems commonly associated with amine plants, such as foaming and column flooding. • The main findings from discussions with personnel at the site were mechanical and operational. Mechanical. Notes on the mechanical condition of the amine unit included: • The mechanical integrity of the amine absorber column after 30 years of operation is very good. The inspection report illustrates no major repair work to the vessel shell or head, only minor corrosion and pitting. The bottom of the vessel is lined with stainless steel and in good condition. • The high head amine booster centrifugal barrel pumps (nine stages) are original to the facility and require a major overhaul only once every 10 years. They have proven extremely reliable. • The existing high­differential­pressure­rich amine letdown valves are a Masoneilan­Lincoln­Log type (designed for HP liquid letdown/cavitation service). TABLE 3. Internal data for company experimental data Solute

0

50

100 150 Total pressure, bar

200

250

FIG. 4. CH4 absorption data against pressure for CH4 + H2S + MDEA.

TABLE 2. Summary of UHP amine literature survey Reference

Pressure range up to Solute Results

1

130 bara

CH4

The methane absorption is approximately linear and flattens out slightly at higher pressures The absorption of ethane and propane will be minor compared to methane

2

200 bara

CH4

CO2

Maximum pressure

Temperature

150 bara

40°C–100°C

CO2 + CH4

130 bara

40°C–60°C

H2S + CH4

160 bara

≥ 60°C

CH4

TABLE 4. Refinery AGRU reference Pressure, bara Gas flow, MMscfd Amine Amine circulation, m3/hr

Refinery

Facility requirements

180

160

137

260

MDEA

MDEA

55

80

–

0.2

Composition, %

The relationship of CH4 absorption to pressure is approximately linear from 70 bara up to pressures in excess of 200 bara

CO2 H2S

1

0.1

Methane

19

92.7

The CO2 equilibrium line moves to the left at increased pressure due the effect of coabsorption of hydrocarbons and reduced fugacity

Ethane

2

4

Propane plus

–

3

H2

78

–

32 JULY/AUGUST 2016 | GasProcessingNews.com

GAS TREATING

Amine conclusions. A firm basis exists in both the experimental and operational data to proceed with the UHP amine unit design. The published data, supplemented by further experimental work by one company,a covered the range of operation required. With this data, the UHP AGRU can be designed with allowances for increased circulation rate and flash gas flow. The observations at the refinery showed that the same basic design principles apply to UHP as at HP. At UHP, those problems outside of the normal operating envelope—for instance, in pump design—can be overcome. Dehydration. Since amines are in aqueous solution, the sweetened gas leaving the amine column is water saturated and, therefore, must be dehydrated. The process technology options considered were molecular sieve and TEG dehydration. Molecular sieves have been used for adsorption at UHP; however, no references combine online regeneration. Reference plants exist for TEG up to 160 bara. Molecular sieve. A choice of designs exists for molecular sieves: • Combine gas dehydration with natural gas sweetening, provided that the H2S loading is not too high. • Locate the sieve downstream of the amine AGRU for dehydration. The required inlet H2S concentration (0.1 mol%) is well within the limit for gas sweetening; therefore, molecular sieves could be used for combined dehydration and acid gas removal. This would eliminate the need for the upstream AGR; however, acid gas removal would be required from the molecular sieve regeneration gas, which may be at UHP or HP. COS formation during the adsorption cycle effectively blocked the use of molecular sieve H2S removal technology, which directly impacted the total sulfur specification of the product gas. COS formation is not an issue particularly related to UHP, but the quantity is a concern for meeting the tight total sulfur product gas specification. For the design case with 1,000-ppmv H2S in feedgas up to 80 ppmv, COS would be formed even for low-COS molecular sieves, which would exceed the treated gas total sulfur specification. Although removing this COS from treated gas is technically possible with mixed metal oxide beds, it is in practice not feasible as it would require a high bed changeover frequency. Therefore, it was decided to investigate the use of molecular sieves for dehydration only. TABLE 5 summarizes the references for HP molecular sieve units with regeneration. Gas dehydration by molecular sieve is well established for pressures up to 70 bara, and some units are operating in the

dense phase (> 100 bara). However, no experience exists of operating these units at 160 bara with regeneration. The main concerns of operating at UHP are: 1. Increased coadsorption of hydrocarbons, leading to a longer mass-transfer zone and the increased possibility of coke formation on the bed (faster degradation and reduced bed life) 2. HP regeneration is inefficient 3. LP regeneration increases the regeneration time and requires compression 4. Increased vessel wall thickness means a longer heating/ cooling cycle. The following mitigations have been recommended by vendors for adsorption: 1. Longer bed depth (to counter longer mass-transfer zone) 2. More bed volume (to counter increased degradation and reduced bed life) 3. High-strength material and multiple vessels of smaller capacity (to reduce wall thickness). Regeneration at a pressure lower than 100 bara has two advantages: 1. Application of a well-established process for subsequent treatment of regeneration gas 2. Approximately 30% reduction in regeneration gas flowrate. 2.4 Proprietary Peng-Robinson package Jerinić Jerinić extrapolation

2.0

TEG in gas, ppmv

These valves operate for 12–18 months between overhauls. Operation. Notes on the operational condition of the amine unit included: • Historic measurements gave an H2S concentration of around 10 ppmv in the treated gas. • The absorber has a chimney tray with a large disengagement space, but no mesh pad to prevent carryover of amine. However, little in the way of amine carryover is observed downstream. • No foaming is observed.

1.6 1.2 0.8 0.2 0.0 0

50

100 Pressure, bara

150

200

FIG. 5. Saturated TEG content in gas at 25°C.

TABLE 5. References for HP molecular sieve dehydration units Project

Operator

Operating pressure

St. Fergus, Scotland

Mobil

116 bara (adsorption/regeneration) Forcing valve used for regeneration Operated well since the early 1990s, no major valve/ mechanical issues Direct feedback illustrates excellent operational safety performance

St. Fergus, Scotland

Shell

110 bara

Norway

Statoil

105 bara (adsorption)/ 70 bara (regeneration) Gas Processing | JULY/AUGUST 2016 33

GAS TREATING

TABLE 6. UHP references for TEG units Project

Operator

Operating pressure

Patricia Baleen Gas Plant

Santos

160 bara 80 MMscfd Structured packing Dry gas water content: 65 ppmv Initial high glycol losses due to high operating temperature in the absorber Improvements reduced TEG losses

Frigg (offshore)

Elf Petroleum 160 bara Norge Trays

Lan Tay, Vietnam (offshore)

BP

140 bara Originally designed to process 350 MMscfd of gas Increase in TEG loss after the capacity was increased to 450 MMscfd

CATS

BP-Amoco

105 bara 600 MMscfd Structured packing Dry gas water content: 2 ppmv TEG observed downstream

However, the disadvantages are: 1. Increased cycle time (maximum depression rate 2 bar/min–3 bar/min) 2. Requirement of a recompression facility. During the course of the investigation, it was identified that molecular sieves with trace amounts of H2S in the feed can cause spikes of H2S on regeneration. To minimize the adsorption of H2S, a 3A molecular sieve instead of a 4A sieve should be selected. The pores of the 3A molecular sieve are too small for the H2S molecule to enter; however, a small amount of H2S is still likely to be adsorbed on the surface of the molecular sieve (the total weight adsorbed on the bed will be on the order of 100 g). This small quantity of H2S is sufficient to increase the concentration of H2S to approximately 100 ppmv in a “spike.” This spike will blend with the feed gas to put off the specification of the treated gas. To mitigate this problem, it will be necessary to install beds of mixed metal oxides to remove the H2S spike from the regeneration gas. The mixed metal oxide beds should be sized to be changed out every three years in conjunction with the molecular sieve. Only one supplier was able to estimate H2S levels on which to base the design; therefore, it was not possible to confirm this value from experience or from other suppliers. This means that the beds could be undersized, which would result in additional changeouts. TEG dehydration. Dehydration of TEG is a proven method of

dehydration at UHP conditions. As the process requires only a single UHP vessel, it is consequently less expensive than the molecular sieve process. In addition, the gas specification (33 ppmv water) is achievable by TEG dehydration with gas stripping. TABLE 6 summarizes the TEG UHP references. As high TEG losses were observed on the Patricia Baleen, Lay Tay and CATS projects referenced in TABLE 6, it was decided to 34 JULY/AUGUST 2016 | GasProcessingNews.com

investigate the solubility of TEG in natural gas at UHP by finding experimental data and validating a proprietary simulation program using Peng-Robinson thermodynamic package against the experimental results. Little published data for TEG solubility in hydrocarbons exists.5 This data goes up to only 90 bara, although it shows the “bowl shape,” with solubility increasing at higher pressures. The referenced article5 explains how this is due to a retrograde phenomenon that occurs when the gas approaches the cricondenbar. Depending on the extrapolation of the experimental data, a threefold to fivefold increase in TEG solubility will be seen from 50 bara to 160 bara. TEG losses (due to solubility in gas only) at 160 bara and 25°C, at a gas rate of 260 MMscfd, will be in the range of 40 kg/d–50 kg/d. This TEG loss will increase the operating expense of the unit and operator intervention. Retrograde condensation of liquid TEG in the export pipeline would also result from the reduction in solubility of TEG with pressure. Literature warns: “TEG will accumulate as a liquid slug, causing significant plugging of the flow cross-sectional area.”5 Dehydration conclusions. Molecular sieve dehydration is unproven at UHP in regeneration service. The H2S spike uncovered during the analysis, although not connected with UHP operation, sparks concern over the ability to design an appropriate molecular sieve system for this application. TEG dehydration technology, with experience at operating at UHP, should be better suited to this application. However, concern exists at the high level of TEG loss at UHP. This poses an operational problem in providing makeup to a small inventory system. It also raises the concern that this would lead to TEG condensation in the export gas pipeline as the pressure is reduced (i.e., retrograde condensation). GP a

NOTE BASF’s OASE Gas Treating Excellence amine treating technology was used in the generation of the experimental results.

ACKNOWLEDGMENT This paper was prepared with support from Dr. Ralf Notz, senior technology manager of OASE Gas Treating Excellence at BASF SE, whom the author also wishes to thank. LITERATURE CITED Carroll, J. J., F. Y. Jou, A. E. Mather and F. D. Otto, “Solubility of methane and ethane in aqueous solutions of MDEA,” Journal of Chemical and Engineering Data, University of Alberta, Canada, July 1998. 2 Jan Addicks, J. and G. A. Owren, “Solubility of carbon dioxide and methane in aqueous MDEA solutions,” Journal of Chemical and Engineering Data, Norwegian University of Science and Technology, Trondheim, Norway, May 2002. 3 De Koeijer, G. and E. Solbraa, “High pressure gas sweetening with amines for reducing CO2 emissions,” Proceedings (Elsevier) from IEA GHGT-7 Vancouver 2004, Statoil ASA, Research and Technology, Trondheim, Norway. 4 Buller, A. T., O. Kårstad and G. de Koeijer, “Carbon dioxide—capture, storage and utilization, Statoil research and technology memoir No. 5,” Stavanger, Norway, 2004. 5 Jerinić, D. et al., “The measurement of the triethylene glycol solubility in supercritical methane at pressures up to 9MPa,” Elsevier B.V., March 2008. 6 Aspen Technology, Aspen HYSYS v7.3 online manual. 1

PAUL ROBERTS was formerly principal process consultant within Advisian, the independent consulting business line of WorleyParsons. He is now an independent consultant. Mr. Roberts graduated with a BSc degree in chemical engineering from Birmingham University and is a chartered engineer.

GAS TREATING

Manage activated carbon effects on MDEA solution foaming

D. ENGEL, S. WILLIAMS and A. HEINEN, Nexo Solutions, The Woodlands, Texas

In the oil and gas industry, activated carbon (AC) is used in many applications, both as an adsorbent and as a support media for chemical reagents. One of the most common uses is in amine units. The occurrence of foaming episodes in amine units is perhaps the single most common problem leading to operational losses. The AC has the function of removing soluble contaminants from the amine solvent, thereby reducing foaming tendency. However, no systematic study exists of the relative effect of AC adsorption with respect to foaming reduction in amine solutions. The work in this article focuses on the contact times and amounts of AC affecting contaminated methyldiethanolamine (MDEA) amine solutions, and the effect on foam stability and foam-reduction (break) kinetics. The work was carried out using contaminated MDEA samples from a US refinery with considerable foam stability. The results indicate that contact times of at least 15 minutes (min) are necessary for proper foam reduction and solvent cleaning. Increasing proportions of AC were also found to reduce foam tendency, with an almost linear correlation up to 50 wt%. Surface tension experiments also confirm that contaminated MDEA samples, with stable foam formation, can be purified to a state that is nearly foam-free after proper treatment with AC. Gas treating and amine units. Amine units are employed in gas processing plants and petroleum refineries to remove acid gases (H2S and CO2) from gas streams, liquefied petroleum gas (LPG), recycled gases and refinery offgases. An amine unit is also used in CO2 sequestration, metals production and syngas production, among others. The amine unit (FIG. 1) generally consists of an absorber

or contactor tower, a regenerator tower and ancillary equipment, such as heat exchangers, filtration systems, pumps, valves and instrumentation. The active liquid medium in an amine unit is an alkanolamine solution (i.e., methyldiethanolamine), typically in a 20%–50% concentration in water. The amine solution recirculates within the unit. In the absorber, the lean amine solution reacts with the H2S and CO2 via direct or indirect reactions and absorbs H2S and CO2 from the gas stream (known as sour gas) to produce a sweetened or treated gas. The rich amine solution, high in H2S and CO2, exits the absorber at the bottom of the absorber tower. The rich amine solution is routed into the regenerator, generally passing through a flash tank to reduce the pressure and remove offgas and light hydrocarbon, if present, and a heat exchanger to heat the rich amine and cool the regenerated lean amine stream. The regenerator reverses the reaction that took place in the absorber and strips the H2S and CO2 from the amine solution. The stripped gases are then sent to

a number of processes for proper disposal or recovery. The stripped lean amine solution is sent back to the absorber after cooling and conditioning with filtration and AC adsorption. The main chemical reactions taking place in the amine unit are: H2S + CH3 N (C 2H 4 OH)2

(1)

+

CH3 N H (C 2H 4 OH)2 + HS– CO2 + H2O + CH3 N (C 2H 4 OH)2 +

CH3 N H (C 2H 4 OH)2 + HCO3–

(2)

The amine unit is a high-efficiency system that operates under stringent specifications, and any downturn in performance can lead to products out of specification, solvent losses and high operational costs. Contamination in amine units is very detrimental to plant operations. To enable processing plants to run with minimal instabilities, increased capacity and high reliability, it is necessary to condition process streams using proper contamination Acid gas

Effluent separation Sweet gas

Lean amine filtration Cooling Lean amine Pump

Surge tank Regenerator

Lean/rich exchanger

Contactor

Sour gas Inlet separation

Reflux accumulator

Rich amine

Flash tank

Reboiler

Flash gas

Lean amine Rich amine filtration

Lean amine

FIG. 1. Process flow diagram for a typical configuration of an amine unit. Gas Processing | JULY/AUGUST 2016 35

GAS TREATING control methods. A variety of new and old technologies can remove certain contaminants efficiently; yet, the complexity and misinformation associated with many removal options have led to disconnect among the needs of end users, recommendations from suppliers and specifications from engineering companies. Proper knowledge of feed gas and amine treating systems is a vital component of unit design and operation. Feed gas should be conditioned to remove solid and liquid contaminants before it enters the amine absorber, and recirculating amine streams should also utilize correct filtration and coalescing technologies. Lean amine streams, in particular, must be conditioned by filtration, as well as by AC adsorption before re-entering the amine absorber to prevent foaming, fouling and a number of other problems. AC adsorption removes dissolved contaminants from the amine stream and is a critical—yet often overlooked—system for efficient and reliable amine unit performance. AC adsorption. AC is an inert solid adsorbent material commonly used to remove a number of dissolved contaminants from water and process fluid streams. AC is a porous, inexpensive and readily available adsorbent that provides a large surface area for contaminant adsorption. It is an extremely effective material for dissolved contamination removal related to color, odor and foam-promoting species, among others.

The removal process takes place via an adsorption phenomenon based on surface interactions of the contaminant and the C grain surface. The interactions occur by weak and reversible Van der Waals forces and dipole-dipole forces. As a result, the separation is generally effective for organic components. It is important to mention that AC beds are not intended to be used as filters for suspended solids removal or for the removal of emulsified liquids. The operation of an AC bed is specific to removing dissolved contaminants only. Therefore, AC beds should not display any meaningful differential pressure increase across the bed. The properties of AC are associated with the source of the C and its configuration (FIG. 2). Several different origins of AC exist with inherent properties, such as pore structure and size distribution, as well as different sizes and production methods. AC can be made from coconut husks, lignite, coal, bitumen and wood, among others. The source will determine both the adsorption capacity and the size distribution of the contaminants it can adsorb. As far as configurations, the AC can be powder or granular (most common in amine units), or extruded in forms such as blocks and pellets. AC bed systems are commonly used to remove impurities so that the amine solution can be properly utilized for effective contaminant removal. If the impurities are not properly removed from the amine solution, then foaming, corrosion and other

problems may occur in the plant, leading to considerable negative technical and economic effects. In general, AC of the bituminous type is chosen for amine units because of its balance of small, medium and large pores within the C grain. This distribution is often the most suitable for amine streams with a wide distribution of molecular-size contaminants. A typical AC bed system is shown in FIG. 3. The vessel arrangement comprises a pre-filter for protecting the bed from suspended solids, the AC bed itself, and the post-filter for capturing the C fragmentation residues. In general, the AC system is installed in the lean amine circuit after cooling. Processing into the bed is usually anywhere from 10% to 50% of the total amine flow; a minimum of 25% of the total flow is recommended. As far as design is concerned , most C beds are vertical in orientation with a top-to-bottom flow and a minimum of 15 min residence time. To better understand the functionality of AC, a series of experiments were performed to measure its effectiveness in removing impurities that primarily cause foaming of the amine solution. Contaminated MDEA samples with stable foam formation were used for the experiment. The amine solution was taken from a US Gulf Coast refinery. Additional experiments were also performed with the objective of determining any potential correlation between the amount of AC that an amine solution is exposed to, and foaming tendency. It is important to note that it was not the intention of the work to mimic the exact conditions in an amine unit, as this would pose considerable challenges in a laboratory scale. The testing that follows was performed under controlled laboratory conditions that resemble the unit operation and allow for correlations to be concluded. Materials and methods. Ten 20-mL

FIG. 2. General schematics of the different types of ACs showing their associated pore structures.

36 JULY/AUGUST 2016 | GasProcessingNews.com

glass vials were used to contact the contaminated lean amine with bituminous AC (8 × 30 mesh). Eight of the vials contained 1.5 g of AC, and the other two vials did not contain any C. The eight vials with the C were then mixed with 15 mL of the lean MDEA amine solution. The MDEA in each vial was in contact with the C for different increments of time (5 min, 10 min, 15 min, 30 min, 1 hr, 2 hr, 4 hr and 8 hr).

HARD WORKING ENGINEERS We have turned things around before. No matter what the oil price is: The world needs energy. It is at times like these new technology and new processes see the light of day. Together we have turned things around before – by innovating, reducing costs, and working in a smarter way. We are all adjusting once again, to make sure we can succeed in the new market reality. And we know we can do what it takes. Now is the time to get together and prepare for the future. ONS 2016 provides you with the latest insights, the new technology – and future business opportunities. This is the place to be. Welcome to the leading energy meeting place.

www.ons.no New conference feature: Build your competence at our new Technical Sessions!

GAS TREATING The amine was then filtered into a clean vial to completely remove the AC from the amine. The ninth vial contained 15 mL of the contaminated lean MDEA that had not been exposed to the AC, and the tenth vial contained pure MDEA at 50% in distilled water. The 10 vials were then lined up on a shaking rack (FIG. 4). The rack was mechanically shaken in a consistent manner for 90 sec to impart energy for foam formation. Pictures were taken at 10 sec, 30 sec, 1 min, 2 min, 5 min, 10 min, 30 min

and 1 hr after shaking was ceased. Control samples of the untreated MDEA solution and a pure MDEA in water (50%) solution were also included in the test, analyzed for their interfacial tension and compared. Once the shake-induced foam test was completed, a second experiment was performed to understand potential effects on foaming as the amine was contacted with increasing amounts of C. Five different samples were made by soaking 15 mL of the MDEA in 0 g (0%), 0.75 g (5%), 1.5 g (10%), 3.75 g (25%) and 7.5 g (50%) of

FIG. 3. A typical lean amine AC adsorption system with pre-filtration and post-filtration systems.

FIG. 4. Shaking rack set up before agitation.

Effect of AC on amine solvent (MDEA) foaming MDEA foam height vs. break time bottle test

4.0 3.5

Control lean MDEA solvent from a US refinery with stable foam tendency

MDEA foam height, cm

3.0 2.5

Time exposed to AC 0 min 5 min 10 min 15 min 30 min

1 hr 2 hr 4 hr 8 hr

2.0 1.5 1.0 0.5 0.0 0

500

1,000

1,500

2,000 Time, sec

2,500

FIG. 5. Graph showing the effect of AC on a lean amine solution.

38 JULY/AUGUST 2016 | GasProcessingNews.com

3,000

3,500

4,000

AC. The MDEA and AC were contacted for 20 min, and the amine was then filtered to remove any C residues. The vials were then subjected to the shake-induced foam test for 90 sec. Pictures were taken after 10 sec, 30 sec, 1 min, 2 min, 5 min, 10 min, 30 min and 1 hr after shaking was ceased. Control samples of the untreated lean MDEA solution and a pure MDEA (50%) solution were also included in the test, analyzed for their interfacial tension and compared. Results. After the shake-induced foam was completed, the pictures were evaluated and the foaming height was measured using computer software. The foam height over time was then plotted (FIG. 5), using the pictures taken. The graph shows the foaming tendency and foaming stability of nine of the 10 vials. The tenth vial, pure MDEA, did not show any foaming tendency. FIG. 5 shows that the longer the contaminated lean amine solution was contacted with AC, the more rapidly the foaming was reduced. Although the foaming was never completely removed, a substantial difference in the amount of foaming reduction over time was seen as the amine solution was contacted with AC. It could also be observed that contact times of 10 min, 15 min, and 30 min gave similar results. Increasing the contact time increased the foam-reduction kinetics. Extended contact (8 hr) of the contaminated MDEA with AC eliminated foaming in 30 sec, while the untreated contaminated MDEA displayed foam for up to 4 hr. As part of amine unit best practices, it is recommended that the contact time of an amine solution be a minimum of 15 min, consistent with the above obtained laboratory data. FIG. 6 shows a marked color change from the contaminated lean MDEA that was not exposed to AC, as compared to the vial that contained the MDEA that was exposed to AC for 8 hr. This is likely because the AC collected the impurities and clarified the MDEA solution. A change in viscosity was also observed but not measured. The pictures taken during testing show that foaming of the contaminated MDEA that was contacted with the AC dissipated rapidly compared to the untreated sample. This can also be observed in FIG. 7, corresponding to the experiments that generated the plot in FIG. 5. The different vials contained the contaminated lean amine contacted with AC for increasing periods

GAS TREATING of time (from right to left). The vial on the left is the untreated lean amine with the highest foam formation, and the vial on the right is the pure amine solution (50%) control that exhibits no foam formation. Several of the samples were analyzed for interfacial tension after the foam-induced shake test. The analysis was carried out using the pendant drop technique. The surface tension is one of the contributing factors to foam formation because it reduces the molecular interaction forces in the amine solution, enabling liquid to be released more easily from the bulk solution into the above head-space volume. The interfacial analysis results are summarized in TABLE 1. As can be observed, the interfacial tension of the pure MDEA (50% in distilled water) solution is significantly higher than that of the contaminated lean MDEA sample. The lowering of interfacial tension in the contaminated MDEA sample was caused by the dissolved contaminants, surfactants and amine decomposition residues present in the solution. Upon exposure to AC, it was observed that the interfacial tension increased, but never reached the levels of the control pure amine solution. It can be interpreted from the results in TABLE 1 that AC plays an important role in reestablishing the surface tension of a contaminated lean amine solution and reducing foam tendency. Extended contact times are also necessary for proper contaminant removal. The extent of increasing the interfacial tension with longer contact times seems to taper off asymptotically after 15 min, as little difference in interfacial tension was measured in samples after 5 min and 15 min of exposure to AC. This correlation should be further confirmed by more comprehensive testing with other contact times between 5 min and 8 hr. It should also be noted that interfacial tension of the contaminated lean

MDEA sample did not reach that of the pure MDEA solution (50% in water) even after 8 hr of contact time. This aspect implies that a larger proportion of AC is likely needed for complete purification. A second experiment was conducted by exposing the contaminated amine sample to different amounts of AC. After the amine was contacted with different percentages of AC, the foam-induced shake test was performed. The images from the different vials were then analyzed, and the foaming height was determined for each vial in each picture. The foaming height was then plotted over time, as shown in FIG. 8. The graph shows that contacting the contaminated lean MDEA in increasing amounts of AC decreased foam height and eliminated foam rapidly. The foam from the sample that was contacted with 50% AC (on a weight basis) was eliminated within 60 sec. The foam on the other amine samples exposed to decreasing amounts of AC displayed a slower rate of foam reduction. This is interpreted in terms of the faster contaminant adsorption kinetics when larger amounts of AC are utilized for contacting the amine solu-

tion. Several of the samples from the above experiment were analyzed for interfacial tension. The results of these tests are summarized in TABLE 2. The analysis was performed using the pendant drop technique. It can be observed from the results that increased proportions of AC in contact with the contaminated lean MDEA sample restores the interfacial tension of the sample. The extent of the increase in interfacial tension with increasing proportions of AC seems to be nearly linear. This correlation should be further confirmed by testing different samples with other proportions of AC between 10% and 50%.

FIG. 6. Vials before (right) and after (left) contact with AC and filtration. Foam in the right vial was stable for up to 4 hr.

FIG. 7. Vials after shake test. The top rack shows vials 10 sec after the shake test. The bottom rack shows vials 5 min after the shake test. The control MDEA is at far right (clear solution).

TABLE 1. Interfacial tension analysis results of lean MDEA samples after varying contact times with AC Test 1 surface tension, mN/m

Test 2 surface tension, mN/m

Test 3 surface tension, mN/m

Average surface tension, mN/m

Pure 50% MDEA in water

48.29

48.28

48.3

48.29

0.01

Contaminated lean MDEA

31.63

31.62

31.6

31.62

0.02

Lean MDEA after 5 min contact with 10% AC

35.26

35.25

35.27

35.26

0.01

Lean MDEA after 15 min contact with 10% AC

35.79

35.8

35.78

35.79

0.01

Lean MDEA after 8 hr contact with 10% AC

41.14

41.11

41.11

41.12

0.02

Sample

Standard deviation surface tension, mN/m

Gas Processing | JULY/AUGUST 2016 39

GAS TREATING

TABLE 2. Interfacial tension analysis results of lean MDEA after contact with varying proportions of AC Test 1 surface tension, mN/m

Test 2 surface tension, mN/m

Test 3 surface tension, mN/m

Average surface tension, mN/m

50% MDEA

48.29

48.28

48.30

48.29

0.01

Lean MDEA

31.63

31.62

31.60

31.62

0.02

Lean MDEA after 20 min of contact with 5% AC

32.03

32

32.02

32.02

0.02

Lean MDEA after 20 min of contact with 10% AC

35.4

35.38

35.41

35.40

0.02

Lean MDEA after 20 min of contact with 25% AC

43.04

43.03

43.04

43.04

0.01

Lean MDEA after 20 min of contact with 50% AC

52.1

52.1

52.09

52.1

0.01

Sample

Foaming height vs. time

3.5

Blank 5% 10% 25% 50%

2.5 Foaming height, cm

contact time. On a more simplistic note, the study shows that AC adsorption plays a fundamental role in amine solvent foam reduction, and, by implication, most (if not all) amine units should properly utilize AC beds for foam prevention. GP

AC

3.0

2.0 1.5 1.0 0.5 0.0 0

500

1,000

1,500

2,000 Time, sec

2,500

3,000

3,500

4,000

FIG. 8. Graph comparing foaming height over time of MDEA soaked in different amounts of AC.

The contaminated lean MDEA contacted with 50% AC was measured to have a higher interfacial tension than that measured for the pure MDEA (50% in distilled water) solution. This is probably because the contaminated lean amine solution did not have a 50% concentration in water. The concentration was likely near 40% in distilled water, which is common in amine units with liquefied petroleum gas (LPG) feed streams. Concentrations above 40% will cause excessive emulsification, leading to excessive amine solvent carryover losses. Takeaway. The various experiments

described here were run under lab conditions, but close to actual process conditions, so that correlations can be established. The data shows that AC, in fact, assists in the removal of contaminants that cause foaming and foam stabilization. The data also shows a foam-reduction

40 JULY/AUGUST 2016 | GasProcessingNews.com

Standard deviation surface tension, mN/m

correlation between increasing contact times of the amine solution with the AC. The contaminated lean MDEA solution shows less foaming tendency as contact time increases. Contact times of 15 min (minimum) were effective in reducing foam at an acceptable rate. Foaming was never totally removed during the experiments, but a significant difference was noticed in terms of foam break times. A correlation was also observed between the weight of AC and the foamreduction kinetics. The contaminated lean MDEA solution exposed to larger amounts of AC displayed substantial reduction in foaming tendency compared to a contaminated lean amine that was not exposed to AC. It also resulted in an increase of the solution’s surface tension. Results suggest that lean amine solutions contacted with ≥ 25% AC by weight will reduce foam almost completely at ambient conditions after only 5 min of

DAVID ENGEL has more than 20 years of industrial experience in a variety of technical areas. He is the inventor in 17 US invention patents and the author of a number of technical and scientific papers. Dr. Engel has developed business and technology for Eastman Kodak, Eli Lilly, Pentair, General Electric and Sulphur Experts globally. He has presented a number of seminars and technical courses on a variety of process engineering and chemistry subjects. Recently, he has specialized in advanced process systems and multicomponent separation methods for removing or mitigating contaminants in process streams. Dr. Engel is the cofounder of Sulphur Experts—Filtration Division and managing director of Nexo Solutions. He holds a BS degree in industrial chemistry and a PhD in organic chemistry. He is a member of the American Chemical Society and the Gas Processors Association, president of the American Filtration and Separation Society (Southwest Region), a GLC Consulting member, and a board member (editor) for Elsevier and Genesis BioHealth. SCOTT WILLIAMS is a process engineer at Nexo Solutions. He has industry experience in many projects, and has been instrumental in providing solutions in oil and gas, petrochemical, chemical and water treatment applications. As part of the engineering group, Mr. Williams is responsible for technical design and solutions development in engineering and technology applications, and he also provides support for Nexo’s analytical and specialized service projects. His latest focus is in the area of contaminant removal using novel systems and chemistries for H2S and mercaptans removal from gas and liquid streams. Mr. Williams has also recently worked on projects involving oil-based drilling mud characterization, inlet separation and coalescer evaluations, and back-washable metal-based media for NGL feed filtration systems. He holds a BS degree in chemical and biological engineering from the University of Colorado at Boulder.

PLANT DESIGN

Prevent hydrate formation with high-pressure deethanizer design C. C. CHEN and Y.-S. LIU, Wood Group Mustang, Houston, Texas

Conventional dehydration: Dry desiccant beds. Typically,

molecular sieve or silica gel beds are used to dehydrate light liquid hydrocarbons, such as liquefied petroleum gas (LPG) or natural gas liquids (NGL). As shown in FIG. 1, a dry desiccant system is batch-operated and consists of adsorption and desorption (regeneration) beds with a complex switching valve arrangement. The regeneration system includes a regeneration gas heater, regeneration gas cooler, regeneration gas separator, regeneration gas compressor and other miscellaneous items. The process of periodical switching between adsorption, heating and cooling cycles is complicated. However, it can be managed by an automatic timer control. Glycol contactor. The use of a continuous-operated glycol dehydration unit is simpler than that of a dry desiccant unit. However, a glycol unit is commonly used for gas dehydration. To operate a glycol unit, a vapor stream is drawn from and returned to the deethanizer after it is dehydrated. This means that the deethanizer column requires an extra length for side vapor/ liquid draws and returns (FIG. 2). Regen gas from cooling recycle Hot regen gas

Filter Cold regen gas

Regen gas heater, cooler, separator and compressor

Wet NGL feed

Regen gas from heating cycle

Dehydrated NGL

Molecular sieve #2 regenerating

NGL fractionation design. Natural gas and water can form a crystalline solid compound commonly known as gas hydrate. The hydrocarbon gas molecules (e.g., methane, ethane, propane and carbon dioxide) are trapped in a rigid, cage-like lattice of water molecules.1 Gas hydrates can cause pipeline plugs, which lead to safety and operational issues in the oil and gas industries.2 Furthermore, gas hydrates can plug tower trays and valves, resulting in tower flooding.3 Gas hydrates can form in high-pressure, lowtemperature conditions where free water is present. Therefore, the three existing methods to prevent and mitigate hydrate formation are: 1. Dehydration 2. Maintaining an operating temperature higher than the hydrate formation temperature 3. Operating at a pressure lower than the hydrate formation pressure. In an NGL fractionation plant, a refrigerated deethanizer condenser is used to separate ethane from heavier components. If the NGL feed is wet or saturated with water, and if high-purity ethane is to be produced, then hydrates can form in the top

section of the deethanizer and in the chilled deethanizer condenser, where the operating temperature may be lower than the hydrate formation temperature. The conventional method of suppressing hydrate formation in the deethanizer system is to remove water content in the NGL feed using either dry desiccant beds or a glycol contactor.4 A better approach may be to eliminate the requirement of a dehydration unit by optimizing the operating conditions.

Molecular sieve #1 adsorbing

To produce high-purity ethane from a wet NGL feed, a highpressure deethanizer design has been developed to prevent hydrate formation without using a dehydration unit. Formations of hydrates, which are crystalline solids composed of water and light hydrocarbon molecules, can result in issues with operation and safety. If high-purity ethane is desired as a product, then the deethanizer condenser and the top tray temperatures may be lower than the hydrate point. This can cause hydrate formation at these locations if the wet NGL is not dehydrated. Conventional methods used to prevent hydrate formation include removing the content water in a glycol or a dry desiccant dehydration unit. Either way requires a complicated dehydration system. Here, a design method to eliminate the requirement of a dehydration unit is discussed. The deethanizer condenser temperature increases rapidly while the hydrate formation temperature increases slowly with the pressure. Using these characteristics, a high-purity ethane can be obtained by increasing the deethanizer pressure to force the condenser temperature to surpass the hydrate point. This way, the dehydration process is not required. The economic analysis and the limitations of this approach are discussed in detail. This method has been used successfully in a commercial setting.

Water Drain

FIG. 1. Molecular sieve dehydration. Gas Processing | JULY/AUGUST 2016 41

PLANT DESIGN Additionally, a dehydrated gas compressor or a side draw liquid return pump may be needed to compensate for the pressure drops in the draw piping, return piping and the glycol contactor. A glycol dehydration unit consists of a glycol contactor and a glycol regeneration system. The glycol regeneration system usually includes a minimum of a flash drum, a glycol reboiler/ still column, a glycol/glycol exchanger, a high head lean glycol pump and a glycol cooler. Injecting stripping gas to help regeneration is also necessary, in some cases. An alternative approach to dehydrating the vapor stream outside the deethanizer is to integrate the glycol contactor into the deethanizer (FIG. 3). The lean glycol dehydrates the hydrocarbon vapor leaving from the bottom chimney tray. The diameter of a standalone glycol contactor section is usually smaller than that of the deethanizer due to the minor amount of glycol C2 and lighter to condenser Dry gas

Reflux

Glycol cooler Lean glycol Glycol contactor

Glycol regeneration

Rich glycol Wet gas Liquids Wet NGL feed

used. However, in this alternative design, the diameter of the glycol contactor is typically the same as that of the deethanizer. These larger dimensions increase capital cost and may also reduce the mass-transfer effect caused by low loadings. High-operating-pressure approach without dehydration. The conventional dehydration methods described in the previous section utilize a complex dehydration unit that not only requires higher CAPEX, but also more plot for the additional equipment. To simplify the design, a high-pressure deethanizer that produces 95 LV% ethane from wet NGL without the need for a dehydration unit has been proposed. Typically, a deethanizer operates between 200 psig and 400 psig. FIG. 4 shows that when the deethanizer condenser is operated at low pressure, the condenser temperature is substantially lower than the hydrate formation temperature. In this case, gas hydrates will form if the stream is not dehydrated. If pressure increases to approximately 390 psig, then the initial condensing temperature is roughly equal to the hydrate formation temperature. At approximately 490 psig, the overhead vapor can be totally condensed without forming gas hydrates. One engineering firma has capitalized on this principle by designing an alternative approach for producing 95 liquid volume percent (LV%) ethane from wet NGL by operating the bottom of the

Water

TABLE 1. Case 1 results

Inlet scrubber

Deethanizer

Reboiler return C3 and heavier to reboiler and depropanizer

FIG. 2. Deethanizer with glycol dehydration. C2 and lighter to condenser Reflux

Bottom pressure, psig

350

400

450

500

Top tray temperature, °F

57.1

65.9

73.9

81.1

Condenser temperature, °F

46.5

55.9

64.5

72.2

Hydrate formation temperature, °F

50.4

56.8

59

60.5

Dehydration required

Yes

Yes

No

No

Reboiler temperature, °F

207

222

236

249

Condenser duty, MMBtu/hr

31.4

32.6

34.2

37.1

Reboiler duty, MMBtu/hr

70.9

76

81.4

88.1

DeC3 reboiler duty, MMBtu/hr

57.5

55.4

53.3

51.4

32

41

50

57

2,909

2,634

2,392

2,304

Refrigerant temperature, °F Refrigerant compressor, brake horsepower 80

Lean glycol

60 40

Wet NGL feed

Deethanizer/dehydrator

Temperature, °F

Rich glycol

20 0 Hydrate formation temp. Initial condensing temp. 50% condensing temp. Total condensing temp.

-20 -40

Reboiler return C3 and heavier to reboiler and depropanizer FIG. 3. Alternative design of deethanizer with integrated glycol contactor.

42 JULY/AUGUST 2016 | GasProcessingNews.com

-60 90

190

290 Pressure, psig

390

490

FIG. 4. Deethanizer condensing temperatures vs. hydrate formation temperature. Composition: C1 = 2 LV%, C2 = 90 LV%, C3 = 8 LV%, saturated with water.

Hosted by：

Diamond Sponsor:

4-7 APRIL 2017

80%

Makuhari Messe Chiba, Tokyo, Japan

EXHIBIT AT THE WORLD’S LARGEST GAS EVENT

exhibition space sold

Gastech is geared to energy professionals who are operating on a technical and strategic level across the global gas and LNG sector. By exhibiting at Gastech 2017, you will be part of the world’s largest and most influential gathering in the natural gas and LNG industry.

54,000+

25,000+ 2,500

750+

600+

350+

Hosted Meetings

Regional and International Exhibitors

CEOs, Government & Ministers

Square Metres Exhibition Space

International Attendees

200+

70+

30

14

4

Speakers from Leading Energy Specialists and Companies

Countries representing the global gas production, supply, service & technology value chain

New Technical Conference Sessions

International Country Pavilions

Days packed with high quality Commercial Conference content

Commercial and Technical Delegates

Who exhibits? Exhibiting companies represent the following sectors: IOCs, NOCs, EPCs & FEEDs, Shipbuilders, Marine Engineering, Gas Processing and LNG Technology & Service Providers.

To discuss your participation please email [email protected] or call

Organised by:

LONDON Damian Howard T: +44 (0) 203 772 6038

JAPAN Atsuhiro Furuya (Tony) T: +81 (3) 6890 0783

www.gastechevent.com/gp3

SINGAPORE Shunker Goel T: +65 6422 1489

PLANT DESIGN deethanizer near 500 psig (FIG. 5). This commercial-scale plant is in operation and has been running successfully for several years. The drawback of using a high-pressure deethanizer is that a higher-pressure vessel requires a thicker wall, and a high-pressure component separation requires increased duties for the condenser and reboiler compared to a conventional design. The operating pressure is approximately 225 psi below the critical pressure at the deethanizer top and approximately 150 psi below the critical pressure at the deethanizer bottom; therefore, phase separation is still effective. To demonstrate the selection of optimum operating conditions for different ethane purities, two case studies were conducted for a deethanizer processing 90 Mbpd of NGL containing 50% C2. Case 1: Desired C2 product is 90 LV% C2. As shown in TABLE 1, the temperatures of the condenser at 350 psig and 400 psig are below the hydrate formation temperature. Therefore, NGL dehydration is required when the operating pressure of the deethanizer bottom is below approximately 410 psig. Case 2: Desired C2 product is 95 LV% C2. As shown in TABLE 2, the temperatures of the condenser at 350 psig, 400 psig C2 and lighter to condenser Reflux

Wet NGL feed

Deethanizer

Reboiler return

and 450 psig are below the hydrate formation temperatures. In addition, the temperatures of the top tray at 350 psig and 400 psig are also lower than the hydrate formation temperatures. Gas hydrates will form in the condenser at a pressure of 450 psig and lower. Similarly, gas hydrates will also form at the top tray inside the column at a pressure of 400 psig and lower. Comparing the results of Cases 1 and 2, higher C2 purity is associated with lower temperatures of the top tray and condenser at the same pressure. Therefore, the likelihood of gas hydrate formation increases. In Case 2, NGL dehydration is not required only when the deethanizer bottom pressure is near 500 psig. Economic analysis. For a high-pressure deethanizer with a

bottom pressure of 500 psig, the capital cost of the deethanizer system is significantly higher. However, savings from not needing a dehydration system, as well as from the reduced capital cost of the smaller refrigeration unit using higher-temperature refrigerant, exceed the extra capital cost of the higher-pressure deethanizer system. TABLE 3 lists the estimated capital costs that are impacted by the operating pressure of the deethanizer. The total estimated capital cost of a high-pressure deethanizer system is approximately $2 MM lower than that of a conventional deethanizer system with a dehydration unit. OPEX impacted by deethanizer operating pressure. The main increase of the utility costs for a higher-pressure deethanizer at 500 psig bottom pressure is due to its higher reboiler duty. Although the condenser duty is also increased, a highertemperature refrigerant can be used for the warmer condenser. As a result, the hp of the refrigerant compressor is lower, as shown in TABLE 2. Since the outlet temperature of the deethanizer bottoms becomes higher, the duty of the downstream depropanizer reboiler is reduced. TABLE 4 lists the estimated major differences of utility consumptions that are impacted by the operating pressure of the deethanizer. Some minor differences of utility consumptions TABLE 3. Capital costs impacted by operating pressure

C3 and heavier to reboiler and depropanizer

Capital cost, $MM

FIG. 5. Deethanizer without dehydration.

Deethanizer bottom pressure 400 psig Glycol dehydration unit

TABLE 2. Case 2 results

Differential capital cost, $MM

500 psig 500 psig–400 psig

12

–

–12

Deethanizer unit

33.3

46.5

13.2

Bottom pressure, psig

350

400

450

500

Refrigeration unit

26.7

23.5

–3.2

Top tray temperature, °F

43.6

52.7

60.9

68.5

Total

72

70

–2

Condenser temperature, °F

36.5

46.2

54.9

62.9

Hydrate formation temperature, °F

56.5

58.1

59.4

60.7

Dehydration required

Yes

Yes

Yes

No

Reboiler temperature, °F

202

217

231

244

Condenser duty, MMBtu/hr

40.6

42.6

45.6

50.1

Deethanizer bottom pressure 400 psig 3,818

4,495

677

2,718

2,514

–204

72

–

–72

TABLE 4. Yearly utility costs impacted by operating pressure Yearly cost, $1,000/yr

Reboiler duty, MMBtu/hr

77.9

84

90.6

98.9

DeC2 reboiler

DeC3 reboiler duty, MMBtu/hr

62.2

59.8

57.5

55.3

DeC3 reboiler

Refrigerant temperature, °F Refrigerant compressor, brake horsepower

22

31

40

48

4,460

4,156

3,843

3,705

44 JULY/AUGUST 2016 | GasProcessingNews.com

Glycol dehydration unit

Yearly differential cost, $1,000/yr

500 psig 500 psig–400 psig

Refrigerant compressor

1,240

1,105

–135

Total

7,848

8,114

266

PLANT DESIGN are not included. For example, the hp change of the deethanizer reflux pump, flow change of cooling water, and glycol makeup cost are excluded in the OPEX comparison. The natural gas price for glycol regeneration was assumed to be $3/MMBtu, the electricity price for the refrigerant compressor was assumed to be $0.05/kWh, and the steam cost for reboilers was assumed to be $5/1,000 lb. The total estimated yearly utility cost of a high-pressure deethanizer system is approximately $266,000/yr higher than that of a conventional deethanizer system with a glycol dehydration unit. Based on the estimated capital cost savings for the high-pressure design, the payback period of using the conventional system with a dehydration unit is approximately 7.5 years, which is marginal for the economic benefit consideration. The economic analysis results are strongly influenced by individual equipment design assumptions, and the conclusions are impacted accordingly. As previously shown, the difference in capital investment for the two options is small, so the actual incremental economics may not be compelling or even exist. The main benefit of the high-pressure deethanizer design is the ability to eliminate the extra complexity introduced by a dehydration unit. Takeaway. A high-pressure deethanizer process for producing high-purity ethane product without dehydration has been designed and proven to be a successful alternative in commercial operations, compared to conventional designs. This process

eliminates the need for the installation of a complex molecular sieve or a glycol hydration unit, providing simplicity for both installation and operation. GP a

NOTE Wood Group Mustang developed the design discussed in this article.

LITERATURE CITED Gas Processors Suppliers Association, Engineering Data Book, 13th Ed., Tulsa, Oklahoma, 2012. 2 Hammerschmidt, E. G., “Formation of gas hydrates in natural gas transmission lines,” Industrial Engineering & Chemistry, Vol. 26, Iss. 8, 1934. 3 Kister, Henry Z., Distillation Troubleshooting, Wiley Publishing Inc., Hoboken, New Jersey, 2006. 4 Kohl, A. L. and R. B. Nielsen, Gas Purification, 5th Ed., Gulf Publishing Company, Houston, Texas, 1997. 1

CHYUAN-CHUNG (C. C.) CHEN is a process manager at Wood Group Mustang in Houston, Texas, and a registered professional engineer in Ohio. He has more than 40 years of engineering experience in oil and gas, gas processing, refining and other areas. Dr. Chen holds a BS degree from National Taiwan University and a PhD and MS degree from the University of Rochester, all in chemical engineering. YEN-SHAN (AMY) LIU is a process engineer at Wood Group Mustang in Houston, Texas, and is a registered professional engineer in Louisiana. Prior to joining Wood Group Mustang, she was an assistant professor of chemical engineering at the University of Louisiana at Lafayette, and a senior consultant at ioMosaic Corp. Dr. Liu holds a BS degree from Mississippi State University and a PhD from Texas A&M University, both in chemical engineering.

SALES OFFICES—EUROPE FRANCE, GREECE, NORTH AFRICA, MIDDLE EAST, SPAIN, PORTUGAL, SOUTHERN BELGIUM, LUXEMBOURG, SWITZERLAND, GERMANY, AUSTRIA, TURKEY

Catherine Watkins, Publisher Phone/Fax: +1 (713) 520-4421 E-mail: [email protected] www.GasProcessingNews.com

SALES OFFICES—NORTH AMERICA IL, LA, MO, OK, TX Josh Mayer Phone: +1 (972) 816-6745 Fax: +1 (972) 767-4442 E-mail: [email protected]

AK, AL, AR, AZ, CA, CO, FL, GA, HI, IA, ID, IN, KS, KY, MI, MN, MS, MT, ND, NE, NM, NV, OR, SD, TN, TX, UT, WA, WI, WY, WESTERN CANADA Ryan Akbar Phone/Fax: +1 (713) 520-4449 Mobile: +1 (832) 691-6053 E-mail: [email protected]

CT, DC, DE, MA, MD, ME, NC, NH, NJ, NY, OH, PA, RI, SC, VA, VT, WV, EASTERN CANADA Merrie Lynch Phone: +1 (617) 357-8190 Fax: +1 (617) 357-8194 Mobile: +1 (617) 594-4943 E-mail: [email protected]

DATA PRODUCTS J’Nette Davis-Nichols Phone/Fax: +1 (713) 520-4426 E-mail: [email protected]

Jim Watkins Phone: +33 (0)6 76 35 11 52 [email protected]

ITALY, EASTERN EUROPE Fabio Potestá Mediapoint & Communications SRL Phone: +39 (010) 570-4948 Fax: +39 (010) 553-0088 E-mail: [email protected]

UNITED KINGDOM/SCANDINAVIA, NORTHERN BELGIUM, THE NETHERLANDS Michael Brown Phone: +44 161 440 0854 Mobile: +44 79866 34646 E-mail: [email protected]

SALES OFFICES—OTHER AREAS CHINA—Hong Kong Iris Yuen Phone: +86 13802701367 (China) Phone: +852 69185500 (Hong Kong) E-mail: [email protected]

INDIA Catherine Watkins Phone/Fax: +1 (713) 520-4421 E-mail: [email protected]

JAPAN—Tokyo Yoshinori Ikeda Pacific Business Inc. Phone: +81 (3) 3661-6138 Fax: +81 (3) 3661-6139 E-mail: [email protected]

ADVERTISER INDEX Air Products & Chemicals Inc. ...................16 BCCK Engineering, Inc. ...............................15 CB&I ..................................................................14 Gastech .......................................................9–10 Gastech ...........................................................43 GE Oil And Gas ............................................... 5 Gulf Publishing Company Events—GasPro ..........................................31 Events—WGLC ...........................................47 Software ........................................................ 8 US Gas Processing Plant Directory......20 Jonell, Inc ........................................................19 Merichem Company ...................................... 7 ONS .................................................................. 37 Pentair .............................................................48 Prosernat .........................................................12 RedGuard ......................................................... 2 SNC-Lavalin ...................................................24 This index and procedure for securing additional information are provided as a service to advertisers and a convenience to our readers. Gulf Publishing Company is not responsible for omissions or errors. Gas Processing | JULY/AUGUST 2016 45

NEW IN GAS PROCESSING TECHNOLOGY BOB ANDREW, Technical Editor

Expanded gasphase filtration services

Maintenance, repairs and unplanned downtime in manufacturing can be costly, and the corrosive effect of polluted or contaminated air in production facilities is often the culprit. Demand for comprehensive gas-phase filtration is rising, and increasingly sophisticated solutions are required to protect critical systems. Corrosion can wreak havoc on critical systems: an oxidized printed circuit board or a defective plug connection can cripple entire production processes. Optimized filtration can prevent such damage. High-quality gas-phase filtration systems are especially effective in contamination control. Freudenberg Filtration Technologies is now offering all of its gas-phase filtration services from a single source: from onsite contaminant analysis, to the selection of filter media and filter units, to permanent monitoring. Freudenberg’s Viledon gas-phase filter product line improves indoor air quality for the protection of personnel as well as provides corrosion control for sensitive products, processes and equipment. The filtration technology, which includes both ChemControl filter pellets and Freudenberg’s Versacomb honeycomb media, is designed to protect all sensitive process areas, including computer systems and switchgear, compressors, motor controls and other essential equipment. Viledon solutions meet all relevant international quality and performance standards, including the International Society of Automation (ISA) 71.04-1985 standard for corrosion levels on electronic and electrical equipment. www.freudenberg-filter.com

Modular concept for CHP plants At Power-Gen Europe 2016, MAN Diesel & Turbo presented a new modular concept for combined-heat-and-power (CHP) gas engine power plants. The company is responding to increasing European demand for highly efficient, yet flexible, technologies for power generation. The modular design allows individual units to be switched on or off, depending on the power demand. This ensures consistently more efficient operation and increased load compared with non-modular concepts. It also means that maintenance and overhaul activities can be carried out without shutting down the power plant. In the gas engine variant, the MAN 35/44G four-stroke gas engine can be specified with either single-stage or two-stage turbocharging. With the two-stage engines, a low- and a high-pressure compressor are coupled in series, thereby increasing the efficiency of the engine. The single-stage, turbocharged, 20-cylinder version has a mechanical power output of 10 MW, while the two-stage version has an output of 12.4 MW. It is also available in a 12-cylinder V version with a mechanical power output of 7.4 MW. dieselturbo.man.eu

New PLC aids device integration Honeywell Process Solutions has launched its ControlEdge Programmable Logic Controller (PLC), a new addition to Honeywell’s next-generation family of controllers designed to leverage the capabilities of the Industrial Internet of Things (IIoT). ControlEdge PLC, combined with Honeywell’s Experion Process Knowledge System, provides connectivity and integration to devices from multiple vendors. The PLC is said to offer easy configuration, efficient operations and reduced maintenance. It uses OPC UA protocol and built-in cyber security for integration to a range of instruments, equipment and software from multiple vendors. It is claimed to be the first PLC to offer Universal I/O, providing remote configuration and latedesign-change flexibility for improved project implementation. ControlEdge PLC is focused on process industries requiring discrete control for specific PLC applications, such as water treatment, balance of plant modular equipment, terminal automation and coal/ash handling. The PLC will be offered to end users; original equipment manufacturers; and engineering, procurement and construction companies. The emergence of the IIoT represents a digital transformation of manufacturing that shifts the source of competitive advantage away from physical machinery and toward information. ControlEdge’s IIoT-ready open platform is said to enable users to better leverage data across their assets. www.honeywellprocess.com

Mokveld commissions test bunkers The Netherlands-based Mokveld Valves BV has commissioned two new test bunkers for testing critical, highquality valve systems. The reasons for building the bunkers are the increasingly stringent demands of the industry in terms of quality and safety, such as fugitive emissions, PR2 or type approval testing. The bunkers allow pressure testing with nitrogen and helium up to 1,200 bar. Temperature testing is possible within a range of –196°C to 200°C. The bunkers are fully automatically controlled and meet the latest safety requirements. After the completion of these bunkers, Mokveld will further expand its testing facilities with the construction of two additional test bunkers in 2016. www.mokveld.com

US camera for NDT inspections Imperium Inc. recently announced the availability of its latest ultrasound camera and controller system, the AcoustoCam i700, to improve inspections on straight-beam applications. The AcoustoCam i700 is said to provide higher-resolution C-scan images than automated ultrasonic testing (AUT) or phased-array systems. It creates images in flat or curved materials up to 6 in. thick and is fully compliant with most industry UT codes. The camera produces sub-millimeter images of an entire field, rather than a single pinpoint, for better detection of pitting, cracking and other defects, while reducing false positives. The AcoustoCam i700 features Imperium’s new controller that is IP-66 rated and has a 12-in. LED display. In addition, the unit features integrated tools that support real-time video collaboration. www.imperiuminc.com

46 JULY/AUGUST 2016 | GasProcessingNews.com

Full-Time Students with a Valid Fall 2016 ID

2016 WOMEN’S

Get 50% off Your Pass to WGLC.

November 1–2, 2016

Contact Tranessa Hunt at [email protected] for information on how to apply.

Hyatt Regency Houston, Houston, Texas WGLConference.com/2016

Featured Speakers