• Author / Uploaded
• Eduardo Paulini Villanueva

• Categories
• Combustión
• Gas natural
• Par termoeléctrico
• Presión
• Dom

Citation preview

Training Services

Flare Systems

*EDS

2004/Flare Systems-1

Flaring is a volatile organic compound (VOC) combustion control process in which the VOCs are piped to a remote, usually elevated, location and burned in the open air using a specially designed burner tip, auxiliary fuel, and steam or air to promote mixing for nearly complete (>98 percent) VOC destruction. Completeness of combustion in a flare is governed by flame temperature, residence time in the combustion zone, turbulent mixing of the components to complete the oxidation reaction, and available oxygen for free radical formation. Combustion is complete if all VOCs are converted to carbon dioxide and water. Incomplete combustion results in some of the VOC being unaltered or converted to other organic compounds such as aldehydes or acids. Flares are generally categorized in two ways: (1) by the height of the flare tip (i.e. ground or elevated); and (2) by the method of enhancing mixing at the flare tip (i.e. steam-assisted, air-assisted, pressure-assisted, or non-assisted). Elevating the flare can prevent potentially dangerous conditions at ground level where the open flame (i.e. an ignition source) is located near a process unit. Further, the products of combustion can be dispersed above working areas to reduce the effects of noise, heat, smoke, and objectionable odors. The primary function of a flare is to dispose of toxic, corrosive, or flammable vapor safely, under relief conditions, by converting them to less objectionable products by combustion.

1

Table of Contents

„

Flare Systems – – – – – – – – –

Purpose Flare selection Support structures Combustion theory Radiation theory Regulatory compliance Equipment design Flare design requirements Incineration *EDS

2004/Flare Systems-2

Flare operation and design have become more complicated than simply burning waste gas. The major components of any flare facility are the knockout drum, flare stack, continuous purge gas, flare tip and ignition system. Normally the flare system will use some additional utilities such as steam or air to make the flare smokeless. In designing a flare, considerable effort should be devoted to meeting the prime objective, which is the safe disposal of waste gas. The overall characteristic of the flame is a function of many variables, such as exit velocity of the gas, air-to-gas ration, wind speed, gas composition and flame temperature. Almost 35% of the flare design depends on the heat radiated by the flame. The less heat radiated, the more desirable the flare. The major sources of thermal radiation are free carbon, water vapor and carbon dioxide, but the principal source of these three are fee carbon. Free carbon is formed from the pyrolysis of the hydrocarbon gas. Noise is another type of environmental pollution. There are two types of noise emitted by the flare. Normal noise amounts mainly form the gas combustion, the second type is shock noise generated by the gas velocity gets. In any flare system, waste gas shall be sent to a knockout drum to knockout any hydrocarbon condensate. As shown in the table of contents, we will be covering a wide range of topics of flare systems.

2

Disposal System „

Disposal of vapors and liquids discharged

Types of Systems „ Open System –

„

Discharge directly to atmosphere

Closed System –

–

Discharge to a collection header Dispose to a flare *EDS

2004/Flare Systems-3

Line sizing for relief and flare headers requires the use of compressible flow equations. UOP has developed a computer program for this purpose. The calculation method usually starts at the flare itself where the outlet pressure is atmospheric. The program uses design flows through the system (typically the general electrical power failure case). Flare tip velocities are designed up to 0.5 MACH and, therefore, a pressure drop can be determined. The pressure drop through the water seal, molecular seal, and knockout drum must also be included. The next step is to determine the equivalent lengths of the piping this requires one to determine the strength length of pipe, elbows, and other fittings. Try to limit the velocity throughout the system to 0.7 MACH. Never exceed MACH since this will cause a discontinuity in the flow pressure drop and a lot of noise. Next, the properties will best estimated for MW = W/(MW/W), for temperature T = WT/W. Next, the inlet pressure for each section is calculated by knowing the pressure drop for that section. One must continue calculations back to the relief valve. Next, one needs to check to see if the allowable pressure is less than the actual pressure if not then one will have to increase the relief header size to reduce pressure drop in the system. The equations used assume isothermal flows in the relief header. This is adequate and conservative for most applications. There are only a number of relief systems that are allowed to atmosphere directly.

3

Typical Relief System

*EDS

2004/Flare Systems-4

Discharge pipe size for direct spring operated valves is critical. If it is improperly sized, it may cause valve failure. Pressure losses may occur in discharge headers causing excessive back pressure and excessive back pressure may cause the relief valve to close. When the valve closes, the back pressure in the discharge header decreases, and the valve reopens. After determining the maximum load on the system, it is necessary to decide on the location of the flares and the size of the headers and flare lines. Location and height of the flares must consider radiation heat and emissions. This requires fixing the maximum back pressure for the system and choosing between conventional, pilot operated, or balanced pressure relief valves of the various relief stations. UOP typically divides the type of relief valve based on set pressure. Set pressures less than 250 psig are conventional type and set pressures greater than 250 psig are balanced type relief valves. Depending on the plot plan, the range of equipment design pressures, desirability of isolating certain streams, etc., it may be desirable to have tow or more flare systems. The maximum loads for each system needs to take into account general electrical power failure, cooling water failure, and fire cases.

4

Relief Header Sizing - Equations to Use V = M * SV / A / 60 „ SV = 379.5/MW * (14.7/(Pout+14.7)*((460+T)/520))) „ SonicV = 60 * (32.17*1.1*1546*(460+T)/MW)^0.5 „ MACH = V/SonicV „

„ „ „ „ „ „ „ „

V = velocity in ft/min M = flowrate in lb/hr SV = specific volume in ft3/lb A= pipe area in ft2 MW = molecular weight Pout = outlet pressure in psig T = temperature in F SonicV = Sonic Velocity in ft/min *EDS

2004/Flare Systems-5

Relief header sizing is a very calculation intensive. The two main criteria used in sizing relief headers is to keep the velocity below 0.7 MACH in any segment and to limit the velocity at the flare tip to 0.5 MACH. The second main requirement is to maintain the pressure in the header below the maximum allowable pressure for the relief valves.

5

Relief Header Sizing - 7 Steps „ „ „ „ „ „ „ „ „

Step 1: Start at flare tip and calculate the pressure drop across the flare tip at 0.5 MACH Step 2: Determine equivalent lengths for all segments Step 3: Limit relief header velocity to less than 0.7 MACH Step 4: Establish properties of the gases T=Σ(wi*Ti)/Σwi where T=temp and w is flowrate lb/hr MW = Σwi/Σ(wi/MWi) where MW is Molecular Weight Step 5: Calculate pressure drop Step 6: Review allowable pressure against actual pressure for each segment Step 7: Review velocity is below 0.7 MACH in each segment *EDS

2004/Flare Systems-6

The above slide describes the seven steps to relief header sizing. When the maximum vapor relieving requirement of the flare system has been established and the maximum allowable back pressure has been defined, line sizing reduces to standard flow calculations. Since vapors in the flare headers are relieved from a high pressure system to almost atmospheric pressure, there is an appreciable kinetic energy change throughout the pipeline. The flow condition is that of compressible flow. The nature of compressible flow in the case of flare headers may be assumed to be isothermal since flare lines are normally long and not fully insulated.

6

Purpose of Flare „

Define Loadings to be Handled –

–

–

Calculate loadings for all contingencies Geographic location of each source Calculate maximum load (power failure,fire case) • Fire case limited to a ground area of 230 - 460 square meters • Calculate maximum back pressure

*EDS

2004/Flare Systems-7

Selection of a disposal method is subjected to many factors such as the geographic location of each source, maximum loading for each case (i.e. General Electrical Power Failure, Cooling Water Failure Case, etc.). The purpose of the flare and relieving system is to conduct the relieved fluids to a location where it can be safety discharged and burned. The disposal system of flares consist of the relief valves, piping, drums and some type of combustion system. In the past, many relief valves were discharge directly to the atmosphere. This is still acceptable if environmental regulations permit such discharges. Therefore, most steam relief valves or ones containing air are relieved directly to the atmosphere. The relieving vapors from different relief valves and control valves are normally collected in individual relief subheaders located near each process area. These headers are interconnected and lead to localized knockout drums.

7

Major Factors Influencing Flare Design „ „ „ „ „ „ „ „ „ „

Gas Composition Flow Rate Gas Pressure Available Initial Investment Operating Costs Gas Temperature Energy Availability Environmental Requirements Safety Requirements Social Requirements *EDS

2004/Flare Systems-8

Flaring is a volatile organic compound (VOC) combustion control process in which the VOCs are piped to a remote, usually elevated, location and burned in an open flame in the open air using a specially designed burner tip, auxiliary fuel, and steam or air to promote mixing for nearly completed (>98 percent) VOC destruction. Completeness of combustion in a flare is governed by flame temperature, residence time in the combustion zone, turbulent mixing of the components to complete the oxidation reaction, and available oxygen for free radical formation. Combustion is complete if all VOCs are converted to carbon dioxide and water. Incomplete combustion results in some of the VOC being unaltered or converted to other organic compounds such as aldehydes or acids. The flaring process can produce some undesirable by-products including noise, smoke, heat radiation, light, Sox, NOx, CO, and an additional source of ignition where not desired. However, by proper design, these can be minimized. Major factors influencing flare design include gas composition, flow rate, gas pressure available, initial cost, operating costs, gas temperature, energy availability, environmental regulations, safety requirements and social requirements. The single largest factor depends on the site. In the United States, environmental requirements drive what, if any, type of flare can be installed.

8

Elevated Flare System Flare Tip Steam Ring Dry Seal Knockout Drum Pumpout Pump

Flare Knockout Drum

Flare Stack

PI TI Instrument Air Vent Emergency Gas Purge

Switch

LIAH

LGR

Solenoid Valve (With Manual Reset)

RO RO

Purge Gas

Gas To Pilot

PI

TAH

Grade

M

Pilot Ignition Systems Locate At Flare Knockout Drum

Normal Gas Purge Steam

Pressure Relief From Process Units

Slop To Slop Tank

PI

PC

Fuel Gas

Plant Air

*EDS

2004/Flare Systems-9

The above drawing is the typical elevated flare system one finds in a refinery. The relief valves are collected from the process units and are sent to a flare knockout drum. In the knockout drum, the liquid is removed and the gases are sent to the flare. The flare is elevated in this case to keep all the burning occurring above the ground level where radiation and toxic compounds may collect. At the top of the flare is the flare tip where all of the burning occurs. The steam ring is shown for smokeless operation, otherwise there would be a lot of smoking occurring. The dry stack is provide to reduce the purge gas required. Purge gas are required to maintain a positive pressure in the flare system. If there was not a positive pressure in the flare system, then a vacuum condition might occur and air could get into the system. Air is not wanted in the system because an explosive condition might occur in the header. In the flare knockout drum, the liquids are separated and collected. Since there is some value to the liquid, the liquid is normally sent to either the slop oil tank or to the crude oil tanks for reprocessing. The pilot ignition system is used to keep fuel gas supply to the pilots and the flame front generator is used to light the pilots.

9

Elevated Flare System (Knockout Pot in Stack)

*EDS

FS-R00-02 2004/Flare Systems-10

The above drawing shows and elevated flare system with the knockout drum in the flare stack. The advantage of this scheme is that it takes up less plot space. In addition, the knockout drum and the flare stack/tip can be all produced by one vendor. The disadvantage of the knockout drum in the stack is that the flowrate of the flare gas is limited because of the limited knockout drum cross-sectional area. The stack is also limited to a self-supporting type or a derrick type. The emergency gas purge allows for addition fuel gas to be injected into the relief header during an upset condition. The emergency gas purge is required if a very hot release happens. This is because after the release the volume in the knockout drum and the relief header will decrease because of the cooling occurring in the lines. Therefore, the emergency gas purge needs to be designed to handle the cool down of the gases say from 500°F to 100°F in a one hour time period.

10

Elevated Flare System

(Water Seal and Knockout Pot in Stack) Title Guide

Steam Ring Flare Tip

Flare Stack

Flare Knockout Drum And Water Seal

Knockout Drum Pumpout Pump Switch Instrument Air Vent

Emergency Gas Purge

PI

TAH

M

Water Seal

Solenoid Valve (With Manual Reset)

RO RO

Pressure Relief From Process Units

LIAH Gas To Pilot

Grade

Pilot Ignition Systems Locate At Flare Knockout Drum

Slop to Slop Tank Normal Gas Purge

Purge Gas

LGR

PI PC Steam

Water Fuel Plant Gas Air *EDS

2004/Flare Systems-11

The above drawing in an elevated flare system with the knockout drum and the water seal within the flare stack. The water seal is a critical part of the flare system because it provides a flash back protection in the system. In addition, since the inlet gas pipe discharges into the water seal at approximately 4 inches below the top of the water, a back pressure is produced in the relief header and therefore a positive pressure will be produced in the flare system. The seal drum can also be used to set up a simple staging system for burners. The water holdup should be sufficient to provide a 10-foot slug of water in the inlet line to the drum if an explosion occurs downstream of the drum. A seal dam is normally provided in each drum to prevent the burner back pressure from varying the seal level. The seal loops from the drum to the sewer should provide a seal depth of at least 10 feet. Each drum is provided with a seal water rate of approximately 25 gpm. The sumerged ends of the inlet lines to the drums are notched to prevent cyclic release of the gas which accumulates from normal leakage.

11

Ground Flare System

Flare Knockout Drum

Knockout Drum Pumpout Pump

PI TI

LGR

LIAH

Switch

PI

Instrument Air Vent

Emergency Gas Purge

Main Header PC Burners

Solenoid Valve (With Manual Reset)

M

Stage Header

PO PO Normal Gas Purge

Purge Gas

Ground Flare Retention Dike Grade

Pressure Relief From Process Units

Slop To Slop Tank

Gas To Pilot

PI

Pilot Ignition Systems Locate At Flare Knockout Drum PC

Fuel Gas

Plant Air

FS-R00-05 *EDS 2004/Flare Systems-12

The above drawing shows a ground flare installed in Mexico. As one can see from the photograph, the ground flare is installed in a dike enclosure. The person shown is on top of the dike and is feeling the effects of the radiation. The ground flare in the photo is at peak rate as can one can see from the total number of burners in operation. The ground flare system is different from the elevated flare system in a number of important areas. One naturally is that the ground flare is located at grade and the elevated flare is located some distance above grade. The second difference is that the ground flare requires more pressure at the burner tip. While the elevated flare system requires only about 1 psi at the flare tip for pressure drop reasons, the ground flare requires 7 to 10 psi at the burner tip. The high pressure required is that a ground flare does not require steam to make it smokeless. The ground flare is smokeless because of the higher pressure at the burner tip causes a lot of turbulence at the burner tip. This turbulence causes the air and fuel to mix extremely well to produce a clean burning flame.

12

Two Stage Flare System (Elevated/Ground) Flare Tip

Seal Flare Stack Flare Knockout Drum

Knockout Drum Pumpout Pump

PI TI

LGR

LIAH

Switch Instrument Air Vent

Water Seal

PI

Solenoid Valve (With Manual Reset)

Enclosed Ground Plane

Gas To Pilot

M

Pilot Ignition Systems Locate At Flare Knockout Drum

Emergency Gas Purge RO RO

PI Normal Gas Purge

Purge Gas

Grade

Pressure Relief From Process Units

PC

Slop To Slop Tank

Water

Fuel Gas

Plant Air

*EDS

FS-R00-06 2004/Flare Systems-13

The above photograph shows a two stage flare system. The first stage is the enclosed ground flare and the second stage is the elevated flare system. The advantage of this two stage flare system is that the main flaring can take place in the enclosed ground flare. In this manner, neighbors and people within the refinery are usually not aware that the refinery is burning any material. This is highly advantageous for social reasons, environmental reasons, and safety reasons. The two stage system works by the principle of using two water seals. The water seal for then enclosed ground flare is typically set at about 4 inches of water column and the water seal for the elevated flare is set at about 10 inches of water column. Therefore, the normal flowrate of flare gases is to the enclosed ground flare because of the lower pressure drop in that system. When the flow rate becomes large, the pressure drop of 10 inches is exceeded and the flow goes both to the enclosed ground flare and to the elevated flare.

13

Two Stage Flare System

*EDS

FS-R00-04 2004/Flare Systems-14

The above drawing shows a two stage flare system which utilizes an air assisted flare for the first stage and the elevated emergency flare for the second stage. The advantage of this system is cost compared to the two stage system which utilizes the enclosed flare. The operation is similar to the two stage flare system which utilizes the enclosed ground flare. The main flare flow gases will go to the elevated air-assisted flare by means of the water seal. The water seal will be utilized only for the elevated emergency gas flare. The air-assisted flare will be not utilizing the water seal and the flow of flare gas will go directly to that flare. Once the water seal differential is exceeded in pressure drop, the emergency gas flare will open to allow both flare to be in operation during peak flow rates. The advantage of this system is that the main burning of flare gases occurs in the air-assisted flare and not the emergency flare. Large diameter flares (such as the emergency gas flare above) have a difficult time burning low flows. The low flow results in burning inside the flare tip which can result in premature failure of the tip. By allowing the main burning to occur in the small diameter air-assisted flare, premature failure of the tip normally does not occur.

14

Flare Stack

„

Structure Self Supporting – Guy Supported – Derrick Type –

*EDS

2004/Flare Systems-15

For safety reasons, a stack is used to elevate the flare. The flare must be located so that it does not present a hazard to surrounding personnel and facilities. Elevated flares can be self-supported (free-standing), guyed, or structurally supported by a derrick. Self-supporting flares are generally used for lower flare heights but can be designed up to 250 feet. Free-standing flares provide ideal structural support. However, for very high units the costs increase rapidly. In addition, the foundation required and nature of the soil must be considered. Derrick-supported flares can be built as high as required since the system load is spread over the derrick structure. This design provides for differential expansion between the stack, piping, and derrick. Derrick-supported flare are the most expensive. Guyed wire supported flares is the simplest of all the support methods. However, a considerable amount of land is required since the guy wires are widely spread apart.

15

Support Structure Self-Supported

„

„

Stack less than 100’ (30M) Tight plot area

SelfSupported

*EDS

2004/Flare Systems-16

The self-supporting structure is usually the best construction for under 35 meters. It requires little plot area, is relatively simple to erect, and has a reasonable capital cost. For higher stacks, the self-supporting has increased material thickness and will greatly increase foundation requirements and normally raise the cost above the alternatives. Self-supporting stack should be used as follows: • • • • •

Stacks less than 100 feet Tight plot space Height less than 350 feet When liquid carryover is likely When an integral knockout drum is specified.

16

Self-Supportive Structures Description

When to Use

When Not to Use

Self-supported flare stack is utilized for structures from 20 to 350 feet tall (6 to 100 meters). Usually this design has the lowest installed cost and requires the smallest plot area.

• Stacks less than 100 feet (30 meters) • Tight plot area • When liquid carry-over is likely • When integral drum is specified

• Stacks taller than 350 feet (100 meters) • Cost sensitive applications greater than 100 feet (30 meters

*EDS

2004/Flare Systems-17

The self-supported flare stack is utilized for structures from 20 to 350 feet tall. Usually this design has the lowest installed cost and requires the smallest land area. Self-supporting stack should be used as follows: • • • • •

Stacks less than 100 feet Tight plot space Height less than 350 feet When liquid carryover is likely When an integral knockout drum is specified.

Self-supporting stacks should not be used for: • Stacks taller than 350 feet • Cost sensitive applications greater than 100 feet

17

Support-Structure Derrick „

Stacks over 250' (75M)

„

Tight plot area

„

Gas temperature over 450°F(232°C)

Derrick

*EDS

2004/Flare Systems-18

When plot space is a concern, a derrick structure is normally used. In addition, when thermal expansion is a problem, derrick structures are also used. Derricks typically require increased capital investment and foundation requirements compared to a guyed stack. The derrick structure will typically be double the price of guyed wire support. They also require more erection time because the derrick must be built on-site. Derrick supported flare stacks are used for structures from 150 to 550 feet tall. They are used in the following application: • • • • •

When plot space is tight On stacks over 250 feet tall When gas temperatures are over 450°F When the customer strongly prefers the derrick structure With offshore systems

18

Derrick Structures Description

When to Use

When Not to Use

Used for structures from 150 to 550 feet tall (45 to 266 meters). Relatively easy to erect and has superior strength when assembled.

• Plot space is tight • Stacks over 250 feet (75 meters) • Gas temperatures are over 450°F (232°C) • With Offshore systems

• Stacks less than 250 feet (75 meters) • Cost sensitive applications

*EDS

2004/Flare Systems-19

Derrick supported flare stacks are used for structures from 150 to 550 feet tall. They are used in the following application: • • • • •

When plot space is tight On stacks over 250 feet tall When gas temperatures are over 450°F When the customer strongly prefers the derrick structure With offshore systems

They should not be used for the following conditions: • Stacks less than 250 feet tall • Cost sensitive applications. They are usually modular in design for ease of erection. One vendor uses 39 foot sections which can be assembled at grade then lifted for erection. The three leg design reduces the number of pieces which reduces the erection costs.

19

Support-Structure (Guyed)

„

Stacks over 250' (75M)

„

Low capital cost

Guyed

*EDS

2004/Flare Systems-20

The simplest and generally most economical design is the guyed support. Capital cost and foundation requirements are minimal. However, the guy wires can cause some problems. A significant plot area is often required to accommodate the wires. The radius is generally equal to two-thirds the height for most guyed flares but can approach 80% of the height for very tall flares (on the order of 150 meters). Thermal expansion of the stack is another problem to consider. A high design temperature such as 300°C will require some method to accommodate thermal expansion. Some vendors can provide this through innovative stack and guy wire design. However, a more practical solution is to enclose the flare pipe in a structure that can support the horizontal load yet allow freedom to move vertically.

20

Guyed Structures Description

When to Use

When Not to Use

Guy wire support flare stacks are typically the lowest material cost system, but they require the largest plot area. Used in systems from 100 to 700 feet tall (30 to 213 meters).

• Stacks over 100 feet (30 meters) • Radius equal to stack height available for guy wires • Low capital cost is required • Liquid carry over is unlikely

• Stacks less than 100 feet (30 meters) • Tight plot area • Liquid carry over likely

*EDS

2004/Flare Systems-21

Guyed wire supported flare stacks are typically the lowest material cost system, but they require the largest plot area. Used in systems from 100 to 700 feet tall. Guyed supports are normally used as follows: • • • •

Stack height greater than 100 feet Radius equal to stack height available for guy wires Low cost important Liquid carryover is unlikely

Guyed supports should not be used in the following application: • Stacks less than 100 feet • Tight plot spaces • Liquid carryover is likely

21

Radiation Effects

Solar Radiation on Earth ~ 300 Btu/hr-ft2 *EDS

2004/Flare Systems-22

Most of us learned about radiation from the effects of the sun. We found that by staying in the sun we feel relatively warm and by going into shade we will get cooler. Some have even learned the hard way about getting a sun burn with is impart due to the solar radiation of the sun. A flare stack’s height reduces the intensity of radiant heat at grade level, to protect personnel and equipment. This intensity is often calculated by assuming that heat radiates uniformly and spherically from the center of the flame. Emissivity is defined as the fraction of the available chemical energy (calculated as the mass flow times the lower heating value of the flared gas) that is radiated as heat. As the flame, about 0.4 of this energy actually radiates as heat. However, as this energy radiates spherically, it decays inversely with the square of the distance it travels, and some of it is absorbed by the atmosphere. After the heat has radiated 100 meters, the apparent emissivity is more like 0.15, assuming spherical radiation and neglecting all other factors. The wide range of measured radiation efficiencies (from 0.05 to 0.35) and diverse statements from vendors reflect the complex dependence of emissivity.

22

Radiation Theory L2 (ft2) = (t)*(f)*(R)/(4*PI*K) Where: t f R K PI L

= = = = = =

fraction of heat intensity transmitted fraction of heat radiated net heat release (Btu/hr) allowable radiation (500 Btu/hr-ft2) 3.14159 minimum distance from flare tip

*EDS

2004/Flare Systems-23

The radiation source is generally though of as being located halfway up the flame. A common approach to the prediction of the flare flame radiation to a point on or near garde is simplify the geometric problem by assuming the flame has a single radiant epicenter and to use a simple factor to cover a number of radiative heattransfer variables. The radiant heat intensity to a unit area aligned for maximum reception can be expressed in the above formula. Most of the published radiant intensity prediction methods use the lower heating value (LHV) of the gas in calculating the heat release rate, Q. In order to determine the line-of-sight distance form the radiant center to the point of interest, you need to locate the center. This requires an understanding of flame length and the path of the flame follows as it leaves the flare burner. API 521 “Guide for Pressure-Relieving and Depressing Systems” presents a method of calculating fraction of heat radiated. Flame length is determined as a function of heat release. Flame lean is determined by a momentum relationship between the jet velocity and the wind velocity. The radiant center is given a somewhat unrealistic location at the center of a straight line drawn between the end of the flame and the flare tip. An intensity of 4.73 kW/m2 (1500 Btu/h-ft2) is recommended for areas where emergency actions lasting several minutes may be required.

23

Heat Release from a Flare R = Heat Release (Btu/hr) W = Flare Gas Flow Rate (lb/hr) B = Net heating value (Btu/lb) R = W*B

*EDS

2004/Flare Systems-24

The above formula determines the heat release from a flare. Factors to consider regarding thermal radiation levels is that clothing provides shielding, allowing only a small part of the body to be exposed to the full intensity. In the case of radiation emanating from an elevated flare, standard personnel protection, such as hard hats can reduce thermal exposure. A level of 5000 Btu/hr-ft2 (15.77 kW/m2) is the heat intensity on structures and in areas where operators are not likely to be performing duties and where shelter from radiant heat is available such as behind radiation shields. The conservative design approach used mostly ignores wind effects and calculates the distance assuming the center of radiation is the base of the flame (at the flare tip), not in the center. The total emissivity of the flame is the sum of the emmissivities of the gases and solids present in the flame. Combustion of hydrocarbon gas results in soot and smoke. If soot is burned completely within the flame, it becomes Carbon dioxide. If it does not burn completely, it leaves the flame envelope as smoke. To determine gas emissivity of a flame is difficult. For gases within an enclosure such as a fired heater, one can make a fairly accurate estimate of the gas composition, thickness and temperature. For a flare it is more difficult

24

7 6

Radiation Theory

5 4

Exposure Times Necessary to Reach the Pain Threshold

3

Threshold of Pain

2

Safe Limit

440 Btu/(hr) (ft)2

1 0

10

550 740 920 1500 2200 3000 3700 6300

30

40

50

60

Exposure Time, Sec.

Radiation Intensity Btu/hr-ft2

20

Kilowatts per M2

Times to Pain Threshold (Seconds)

1.74 2.33 2.90 4.73 6.94 9.46 11.67 19.87

60 40 30 16 9 6 4 2

*EDS

2004/Flare Systems-25

What level is acceptable for radiation. An acceptable level under one set of conditions may not be acceptable under other conditions. Frequently, one radiant intensity level is allowed for people, whereas another higher level is permitted for equipment. When discussing the allowable level for people, it is common to tie the level to some time period of exposure. Different levels are usually specified for plant personnel and the public. A heat intensity of 4.73 kW/m2 (1500 Btu/h-ft2) for several seconds (16 seconds). This 16 seconds allows for situations where a worker is infrequently in the flare area or must go into the flare area briefly to take some action. This level is also acceptable when the exposure is more general but the maximum flaring event itself is of limited duration. The chart above shows the exposure times necessary to reach the pain threshold. The point where there is a safe zone is at 440 Btu/h-ft2. A related issue is how much importance should be given to solar radiation in establishing the design radiant heat intensity levels. To be additive, the worker must be aligned with the flare and the sun in such a manner that the two exposures are truly additive. The above chart does not take into account solar radiation.

25

Contours of Radiant Heat Intensity Safe Boundary (440 Btu/Hr/Sq.Ft.) Boundary for Radiant Heat Intensity (1500 Btu/Hr/Sq.Ft.) - Normally Fenced in with Warning Signal Protection Required for Personnel

Protection Required for Equipment Boundary for Radiant Heat Intensity (3000 Btu Hr/Sq.Ft.)

*EDS

2004/Flare Systems-26

It is important that the emissivity of the flame not be confused with the fraction of heat radiated. The emissivity of a flame is a function of the combined emissivities of the gases and solids in the flame at a certain temperature. The fraction of heat radiated, on the other hand, is an overall characteristic of a flame that accounts for gas composition, flame type,l state of fuel/air mixing, soot and smoke formation, quantity being burned, flame temperature and flare burner design. The fraction of heat radiated is determined empirically and must be used in the same manner that it was determined.

26

„

„

„ „ „

„

„ „

Environmentally acceptable combustion Tips normally proprietary in design Flame Stability Ignition Reliability Exit Velocity 1 to 600ft/s (.3 to 183 m/s) Exit velocity at 50% of sonic velocity Multiple Pilot Burners Surrounding Windshield

Flare Tip *EDS

2004/Flare Systems-27

The flare tip assembly shall be complete with pilots, pilot flame sensors, noise muffler, steam manifolds and stem jets. The flare tip shall also include all required pilot gas manifolds and steam distribution headers, rings, manifolds and runners for complete smokeless operation. The upper 10 feet of the flare tip shall be 310 stainless steel minimum. The lower half shall be 304 stainless steel minimum. The flame stabilizers shall be 310 stainless steel. The noise muffler shall be refractory lined carbon steel. If refractory lined tips are furnished, the detail of the lining and method of attachment to the flare tip shall be shown. Flare stack diameter is generally sized on a velocity basis, although pressure drop should be checked. A maximum velocity of up to 0.5 Mach for peak, short-term, infrequent flow, with 0.2 Mach maintained for the more normal and possibly more frequent conditions for low-pressure flares. Pressure drops of 2 psig (14 kilopascals) have been satisfactorily used at the flare tip. Too low a tip velocity can cause heat and corrosion damage. The burning of the gases becomes quite slow, and the flame is greatly influenced by the wind.

27

Flare Tip Design

„

Flare Tip Design Considerations Design for maximum flow rates – Design for maximum temperatures – Design for wind conditions – Design for minimum flow rates –

*EDS

2004/Flare Systems-28

The above flare tip is one that was placed in the UK with very high wind loads. This specific flare tip only lasted 3 years before it had to be replaced. In general, flare tips last up to 10 years before they need to be replaced. With respect to life of the flare tip, a failure mode which have been observed involves cracking of the shell of the flare burner. This cracking invariably propagates from the end of a fillet weld on brackets and attachments welded to the shell of the flare tip. During bucking of the flare tip which results from the uneven temperature differentials of flame impingement, the stresses concentrate at the stress riser cause by the weld and propagates a crack through the shell. To avoid this type of problem, plug-welded brackets on the flare tip are used. This complete circle plug-welded approach eliminates any stress riser with its potential for cracking.

28

High and Too Low Relief Flow Rates Can Cause Flame Instability Air

V Air Aspiration Air Intrusion

a. Flame Dip

Methane D

Pipe

b. Flame Blowoff

c. Analysis Of Flame Dip

*EDS

2004/Flare Systems-29

The exit velocity, heat radiation and sonic velocity for flares is critical. Sonic Velocity = 233(Heat radiated,(Btu/hr-ft2)*Gas temp(R)/MW)^0.5 Exit velocity f/s = (MACH)*(Sonic Velocity) Flame stability is extremely important in flare tip design. A flameout does not appear to be a real problem at either low or high venting flowrates. As log flow rate velocities air intrudes into the top of the stack. If the flow rate is sufficient to produce a flame visible from the ground, air intrusion usually is not significant. At low velocities, combustion can take place inside the flare. When the slow upward flow of a lighter-than-air gas permits air to flow downward along the stack wall, a flame dip occurs. A diffusion flame propagates into this region and is quenched at the wall, then air flows downward again, causing another flame dip, and the cycle repeats it self. If the velocity of the relief gas rises until it exceeds the flame velocity at every point, the flame will be lifted to another stable position above the flare tip. This is called blowoff with flame extinguishment is called blowout. Turbulence promoting flame holders and pilots effectively stabilize flames at high gas velocities.

29

Flare Efficiency „

Efficiency of flare depends on the following – Type of fuel – Flow rate of fuel – Wind velocity – Ambient turbulence – Height of the stack – Presence of HC droplets – Presence of water droplets *EDS

2004/Flare Systems-30

The use of the term “flare efficiency” can be somewhat misleading because of the wide variety of definitions that have been used in the past. The most rigorous and universal definition of efficiency is the “Combustion Efficiency” which for a hydrocarbon flare, is simply the mass of carbon in the form of carbon dioxide which is produced by the flare divided by the mass of carbon in the form of fuel entering the flare. The “Hydrocarbon Destruction Efficiency” can also be useful in characterizing flare performance. It is defined in a somewhat different manner as amount of hydrocarbons entering the flare in the flare gas minus the amount of unburned hydrocarbons leaving the flare all divided by the amount of hydrocarbons entering the flare. This is the fraction of hydrocarbons destroyed by the flare. It should be noted that this definition does not consider the form of the chemicals into which the hydrocarbons are converted via combustion.

30

Pilot and Ignition Systems

„ „

Continuously burning pilots Flame front generator –

–

–

Fuel gas and air admitted to the ignition pipe in a combustible ratio Gas is ignited by an electric spark Flame travels through the pipe

*EDS

2004/Flare Systems-31

Even the best designed pilots will occasionally go out, some method should be available for remote lighting of the pilots. In the past, methods such as a flaming arrow, a burning rag wound around a rock, and flame guns were used with moderate success. Modern techniques include flame front generators, electrical “spark plug” type igniters, and aspirated air igniters. Continuous pilot burners on the flare tip ensure ignition. The pilot burners themselves are commonly ignited by a flame front generator system which is activated manually from a remote location. Fuel and air are mixed upon entering the igniter tube. A spark ignites the mixture and a flame front rushes up the tube to light the pilot. To ensure ignition regardless of wind direction, three continuous pilots are usually spaced equally around the flare tip. Weather shields on the pilot nozzles prevent blowout by strong wind. Because it is difficult to see whether the pilot burners are lit in the daytime, thermocouples on the pilot activate an alarm system to warn of a pilot flame failure. The fuel gas supply to the pilots and igniters must be clean and reliable. Low pressure alarms are often installed to warn operators of the loss of fuel to the pilots. Prevention of hydrate formation in the long small-diameter fuel piping needs to be prevented. A knockout pot is normally installed after that last pressure reducer.

31

Pilot Burners

„

Automatic systems may be activated by: Thermocouples – Infrared Sensor – Ultraviolet Sensor (ground flare application) –

*EDS

2004/Flare Systems-32

EPA regulations require the presence of a continuous flame. Reliable ignition is obtained by continuous pilot burners designed for stability and positioned around the outer perimeter of the flare tip. The pilot burners are ignited by an ignition source system, which can be designed for either manual or automatic actuation. Automatic systems are generally activated by a flame detection device using either a thermocouple, an infra-red sensor or, more rarely, (for ground flare application) an ultra-violet sensor.

32

Installation of Thermocouples *Correct

Installation

*Incorrect

Installation

*EDS

2004/Flare Systems-33

The above picture shows a correct and incorrect installation of a thermocouple. The correct installation has the thermocouple at the interior of the pilot where as the incorrect installation is on the outside. By placing the thermocouple on the outside, there is a chance of an incorrect reading. The pilot flame shall always be present at the burner tip. Each pilot shall have a thermocouple attached to determine the flame temperature. Pilot flame failure can be sensed by local or remote monitoring. Thermocouple failure can cause a problem if the flare is unavailable for maintenance over long periods of time. Sometimes two thermocouples are mounted on each pilot with separate leads running down the stack. This allows for switching of the active thermocouple by simply swapping the leads.

33

Pilot Windshield „

„ „

Allows pilot to operate at wind speeds greater than 100 mph Should always be specified Prevents misreading of the thermocouples

*EDS

2004/Flare Systems-34

Pilots shall be able to withstand the effects of rain, snow, and wind speeds greater than 100 miles per hour. They must also stand up to the high temperature caused by flaring. Pilot windshield design, although often overlooked, is extremely important. Without an effective design the risk of an inoperable flare is a real danger. Even if the flame is not extinguished by high winds, it may be directed away from the flare and rendered useless. To prevent this hazardous condition from arising, the number of pilots used on a flare should be increased as the size of the flare is increased. Large flares require the use of several pilots to assure ignition in any wind direction. The size and number of pilots is determined by the size, design, and function of the flare, and the heat level of the waste gas.

34

Flame Front Generator Ignition System

F

Air

B

D

A To Pilot #1

H J Gas

To Pilot #2 To Pilot #3

E C

Gas To Pilots *EDS

2004/Flare Systems-35

The above drawing show a flame front generator ignition system. They work by introducing a flammable mixture of air and gas into a 1 inch pipe leading up to the pilot. Once this mixture is distributed throughout the length of the pipe, it is ignited by an electric spark plug. A flame front will then travel through the pipe up to the flare tip, lighting the pilot. Many flame front generators are designed to light a flare as far as one mile away. The following is a typical operating instruction for using flame front generator ignition system: • • • • • • • •

Open the air valve and set air pressure at 10 psig Open the gas valve and the gas pressure at 10 psig. Purge lines for two to three minutes, depending on distance from flame front generator to flare pilot tip. Spark to light the mixture. If nothing happens, purge and spark again. If nothing happens, reduce the air pressure, and repeat purging and sparking. If still nothing happens, lower the air pressure further in several steps and purge and spark again. If still nothing happens, try higher air pressures.

Pressure settings depend on the gravity of the fuel gas.

35

Flare Control Panel „

Flare Control Panel includes the following: Pilot Gas – Steam Control – Pilot Ignition System –

*EDS

2004/Flare Systems-36

The above photograph shows a flare control panel which contains the pilot gas control the steam control and the pilot ignition system for the flare system. Flare system control can be completely automated or completely manual. Components of a flare system which can be controlled automatically include the auxiliary gas, steam injection, and the ignitions system. Fuel gas consumption can be minimized by continuously measuring the vent gas flow rate and heat content (Btu/scf) and automatically adjusting the amount of auxiliary fuel to maintain the required 300 Btu/scf for steam-assisted flares. Steam consumption can likewise be minimized by controlling flow based on vent gas flow rate. Steam flow can also be controlled using visual smoke monitors. Automatic ignition panels sense the presence of a flame with either visual or thermal sensors and reignite the pilots when flameout occurs.

36

Pilot Gas Requirement „

The average pilot gas consumption based on an energy-efficient model is 70 scf/hr. The annual pilot gas consumption (Fp) is calculated by: *

*

Fp (Mscf/yr) = (70 scf/hr)*(N)*(8,760 hr/yr) Fp (Mscf/yr) = 613*N

Flare Tip Diameter (IN) N can1-10 be calculated 12-24 30-60 >60

„

Number of Pilot Burners (N) 1 from the following table: 2 3 4

*EDS

2004/Flare Systems-37

Pilot gas is critical for flare application. The average pilot gas consumption based on new energy efficient model is approximately 70 scf/hr. Therefore, as shown in the above table, the annual pilot gas consumption in Mscf/yr is based on 613 times the number of pilot burners. The table shows that under a ten inch flare tip only one pilot burner is required. From 12 to 24 inches, two pilot burners are required. Over 24 inches to 60 inches, three pilot burners are required. Over 60 inches, at least four pilot burners are required.

37

Multiple Pilots

„

Multiple pilots allow one pilot to fail

„

Most flares have two to four pilots

„

Equally spaced around the flare

*EDS

2004/Flare Systems-38

The above drawing shows a three pilot arrangement. This allows one pilot to fail and have the other two in operation. The pilots in this case would be equally spaced around the flare. Therefore, they would be space at 360 degrees divided by 3 or at 120 degrees from one another. EPA regulations require the presence of a continuous flame. Reliable ignition is obtained by continuous pilot burners designed for stability and positioned around the outer perimeter of the flare tip. The pilot burners are ignited by an ignition source system, which can be designed for either manual or automatic actuation. Automatic systems are generally activated by a flame detection device using either a thermocouple, an infra-red sensor or, more rarely, (for ground flare application) an ultra-violet sensor.

38

Safety Aspect

Fuel

Oxygen

„

Ignition

Two of the three elements for explosion are always present in a flare system

*EDS

2004/Flare Systems-39

Stacks shall be purged with fuel gas to ensure safe flare operation against explosion and detonation. A question asked is how much purge gas is required? The answer is important because of the cost to purge a flare which may run a refinery $50,000 per year. Combustion can be defined as the rapid chemical combination of oxygen with the combustible elements of fuel. There are just three combustible chemical elements of significance - carbon, hydrogen, and sulfur. Sulfur is usually of minor significance as a source of heat but it can be major significance in corrosion and pollution problems. The objective of good combustion is to release all of this heat while minimizing losses from combustion imperfections and superfluous air. The combination of the combustible elements and compounds of a fuel with all the oxygen requires temperature high enough to ignite the constitutes, mixing or turbulence, and sufficient time for complete combustions.

39

Purging „

Flare purge gas –

Any gas which cannot go to dew point under any condition of operation • • •

–

Fuel Gas Inert Gas Nitrogen

Purge Rate •

•

Flare Stack — Linear velocity 1FPS to 5FPS (.3 to 1.5 m/s) Flare stack with molecular seal — 0.10 FPS to 0.20 FPS (.03 to 0.06 m/s) *EDS

2004/Flare Systems-40

Air or oxygen in a flare system is extremely dangerous. For this reason, flare systems should be purged. Dry seals are used to reduce the rate of purge gases required to keep air out of the system. The total volumetric flow to the flame must be carefully controlled to prevent low flow flashback problems and to avoid flame instability. Purge gas, typically natural gas, N2 or CO2, is used to maintain a minimum required positive flow through the system. If there is a possibility of air in the flare manifold, N2 or another inert gas must be used to prevent the formation of an explosive mixture in the flare system. To ensure a positive flow through all flare components, purge gas injection should be at the farthest upstream point in the flare transport piping. The minimum continuous purge gas required is determined by the design of the stack seals, which are usually proprietary devices. Modern labyrinth and internal gas seals are stated to require a gas velocity of 0.001 to 0.04 ft/s (at standard conditions). Using the conservative value of 0.04 ft/s and knowing the flare diameter (in), the annual purge gas volume, Fpu, can be calculated: Fpu (Mscf/yr) = (0.04 ft/s)(πD^2/4/144 ft2)(3,600 sec/hr)(8.760 hr/yr) = 6.88D^2 (Mscf/yr)

40

Purge Gas Requirements

„ „ „ „

Prevents flashback problems Flare operates at positive pressure Purge all subheaders (upstream) .04 feet per second to 1 feet per second (.01 meters per second to 0.33 meters per second)

F (Mscf/yr) = (0.04 ft/sec)*((PI*D^2/4)/144 ft2))*(3600 sec/hr)*(8,760 hr/yr) F (Mscf/yr) = 6.88*D^2

*EDS

2004/Flare Systems-41

There is another minimum flare tip velocity for operation without flashback or instability. This minimum velocity is dependent on both gas composition and diameter and can range from insignificant amounts on small flares to 0.5 ft/s on greater than 60-inch diameter units. Purge gas is also required to clear the system of air before startup, and to prevent a vacuum from pulling air back into the system after a hot gas discharge is flared. (The cooling of gases within the flare system can create a vacuum.) The purge gas consumption from these uses is assumed to be minor.

41

Dry Seals

„

Molecular Seals

„

Double Seals

„

Fluidic Seals

„

Airrestors

*EDS

2004/Flare Systems-42

Air in the flare stack is a serious problem. There are two common types of flare dry seals: • Diffusion type - (Molecular Seal, Double Seal) • Velocity Seal - (Fluidic Seal, Airrestor) These are located just below the flare tip and are used to reduce the amount of purge gas required to prevent air from getting back into the stack. The molecular seal and double seal uses the difference in molecular weights of the purge gas and the air to form a gravity seal which prevents the air from entering into the stack. An inverted bucket forces the incoming air through two 180 degree bends before it enters the flare tip. If the purge gas is lighter than air, the purge gas will accumulate in the top of the seal and prevent the air from infiltrating the system. The velocity seal, which includes the fluidic seal and the airrestor, works under the premise that infiltrating air enters through the flare tip and hugs the inner wall of the flare tip.

42

Molecular Seal

Flare Assembly

„ „ „ „

Molecular Seal

Prevents explosions Prevents entry of air Reduces purge gas Performs silently with small pressure drop

Liquid Drain

*EDS

2004/Flare Systems-43

The molecular seal has a baffled cylinder arrangement which forces the incoming air through two 180 degree bends one bend up and one bend down before it can enter into the flare stack. If the purge gas is lighter than air, the purge gas will accumulate in the top of the seal and prevent the air from getting into the system. If the purge gas is heavier than air, the purge gas will accumulate in the bottom of the seal and prevent air from getting into the system. The molecular seal reduces the purge gas velocity required through the tip to 0.04 feet per second. Some purge gas composition the rate will limit oxygen levels to below 0.1%. These low purge rates do not prevent burnback inside the flare tip, which will result in short tip life. This effect will deteriorate the metal wall of the flare tip of the molecular seal and is hidden until the flame burns through the tip or seal. This will result in the tip or seal to be replaced and cause shutdown for immediate maintenance. Molecular seals are purged at 0.4 feet per second to keep the flame out of the flare tip and insure proper flare life.

43

Double Seal Outlet To Flare Burner

Clean-Out

*EDS

2004/Flare Systems-44

When air enters a stack, convection (not diffusion, which is much too slow) is almost certainly the controlling process. How much purge will prevent the entry of air depends on the type of dry seal installed. The above picture is that of a double seal. The double seal is similar in design principle as that of the molecular seal. Both designs work on the basis of density differences. The flare gases are required to go up the double seal and then make a 180 degree reversal then go down a distance and then make another 180 degree reversal before being sent to the flare tip. These 180 degrees reversals prevents the air from getting back into the flare headers. If air does get into the double seal, most of the air will be stuck at the point near the clean-out zone.

44

Fluidic Seal Air

Flare Tip

Air

Flow Path Of Flare Gas *EDS

2004/Flare Systems-45

The velocity seal or fluidic seal works under the premise that incoming air enters through the flare tip and hugs the inner wall of the flare tip. The velocity seal is a cone shaped obstruction with single or multiple baffles as shown in the above picture. These baffles force the air away from the wall where it encounters the focused purge gas flow rate and is swept out of the tip. This seal reduces the purge gas velocity through the tip to 0.04 feet per second. This rate keeps oxygen concentrations below 4 to 8% oxygen. Without the velocity seal, the velocity in the stack would have to be increased.

45

Smokeless Flare Operation Smokeless Operation

Smoking

*EDS

2004/Flare Systems-46

Smoke is generated from fuel that has a carbon to hydrogen ratio greater than 35% by weight. The principle of the steam-assist flare is to reduce the formation of smoke by reducing the carbon to hydrogen ration. This can be achieved by using a cented steam feed and steam nozzles at the flare tip. The center feed is used to provide a wide range of flaring rates while the steam nozzles agitate the gas turbulently to achieve a good mix with air. The minimum allowable heating value for all types of assist flare is 300 Btu/scf. The majority of today’s flares are steamassist, because of their significant advantages. First, they have been proved to be successful in many years of operation. Next, when plants undergoes expansion, the system can be expanded as well without having difficulty to maintain smokeless emissions. Steam-assisted flare are extremely useful in variable flow rates of flaring. Some disadvantages are that there is a cost of steam supply as well as the capital cost involved. Air assisted flares employs the same principles as the steam assisted flare but is normally more costly. The design of the air assist flare is complicated because the design features as stack inside a stack. Waste gas goes inside the outer stack, while air is force through the inner stack. The amount of operational purge gas required for air assist flares is typically 50% higher than the steam assisted flare.

46

Steam Requirements and Smoke Suppression Methods „

In general, the following equation can be used: Wsteam (lb/hr) = Whc (lb/hr) * [0.68-(10.8/MW)]

„

Smoke Suppression Methods – – – –

Steam injection High pressure gas injection Low pressure air Internal energized flare

*EDS

2004/Flare Systems-47

Smoke free operation of flares can be achieved by various methods, including steam injection, injection of high-pressure waste gas, forced draft air or distribution of the flow through many small burners. The most common type involves steam injection. The amount of steam required for smokeless burning will depend on the maximum vapor flow at which smokeless burning is to be achieved and the composition of the mixtures. The composition involves both the percentage of unsaturated and the molecular weight. The higher the molecular weight of a hydrocarbon, the lower the ratio of steam to carbon dioxide and the greater the tendency to smoke. The amount of steam that must be injected to maintain a constant steam to carbon dioxide ratio as molecular weight increases can be calculated in the above formula. The steam can be injected through a single pipe nozzle located in the center of the flare or through a series of steam/air injectors in the flare through a minifold located around the periphery of the flare tip or in any combination of those. The steam injected into the flame zone to create turbulence and aspirate air into the flame zone by the steam jets. This improved air distribution combined with the steam gas shift interaction reactors more readily with the flare gases to eliminate fuel rich conditions which results in smoke formation.

47

Percent of Carbon Escaping as Black Smoke

Tendency to Produce Black Smoke 50

40

30

20

10 0.4

0.2 0.3 H/C Ratio by Weight

0.1 *EDS

2004/Flare Systems-48

As can be seen from the above chart, the lower the Hydrogen to Carbon (H/C) ration is, the higher the percent of carbon escaping as black smoke. For example, a H/C ratio of 0.2 results in 19% of carbon escaping as black smoke. Water or steam reduces smoke formation by the way the steam separates the hydrocarbon molecules, thereby minimizing polymerization and forms oxygen compounds that burn at reduced rates and temperatures that are not conductive to polymerization and cracking. Other theories project that the water vapor reacts with the carbon particles to form carbon monoxide and carbon dioxide and hydrogen thereby removing the carbon before it cools and forms smoke.

48

Automatic Steam Control Field Of View Steam Nozzles

Steam Control Valve

Monitor Flux Density Signal Controller Control Scheme *EDS

2004/Flare Systems-49

Automatic control is the optimum way to minimize steam consumption, supplying only the required amount to keep the flame smokeless at a particular waste gas flowrate and also at atmospheric conditions. Excessive steam can produce burning back into the flare tip, besides increasing flare noise and wasting steam. Steam injected into the combustion zone also cools the combustion temperature and at a high excess condition can reduce combustion efficiency. The picture above shows that steam is controlled by flame appearance. An optical unit is calibrated to a particular frequency in the infrared spectrum, ensuring smokeless flaring over the range of flowrates. The system provides a continuous output signal for control for the steam valve. The sending head detects the changes of hot carbon flux density and generates a signal, modulating the automatic steam control valve flowrate only allowing the right amount of steam to make it smokeless. With this fast response system, the detector is remote and not in contact with the flare gases. The control is independent of the gas flow rate. In addition, the system is very convenient since the components are located at grade.

49

Automatic Steam Control „

Minimizes steam consumption

„

Controlled by the flame appearance

„

Calibrated to a particular frequency in the infrared spectrum

*EDS

2004/Flare Systems-50

Manual control of steam involves remote operation of a steam valve by operating personnel assigned to a unit from which the flare is readily visible. This method is satisfactory if short-term smoking can be tolerated when a sudden increase in flaring occurs. With a manual arrangement, close supervision is required. Another method of controlling steam is using television monitoring with manual control. A feed forward control system pressure, mass flow or velocity can be used to control steam. By measuring the amount of flare gas flowing to the flare, the steam rate can be automatically adjusted to compensate for rate changes. This system would not be acceptable if there was a lot of changing in the molecular weight say from going from butane to paraxylene. The feedback system using an infrared sensor as the one shown above. Infrared sensors are used to detect smoke formation in the flames and automatically adjust the steam control valve to compensate. One disadvantage of the feedback system is that the infrared waves are absorbed by moisture, and the resultant feedback signal is reduced in raining or foggy conditions.

50

Less Than 50 psig Steam Wind

*EDS

2004/Flare Systems-51

Steam provide is usually MP steam (150 psig), special designs are available for utilizing slightly lower pressure steam. The major impact of lower steam pressure is a reduction in steam efficiency during smokeless turndown conditions. In very cold climites condensation may enter the flare header and collect and may freeze therefore, drainage of any condensate which may collect is critical. Many refineries because they have a lot of LP steam (50 psig) they believe that by switching from MP to LP will be better. This is not the case and can cause a lot of problems. By using LP steam, any wind that is at the flare tip will cause the steam not to be injected in the correct location. Therefore, much more steam will be required to make the flare smokeless. In addition, depending on the way the wind is blowing, the steam may not even reach the flare tip. Always try to use MP steam for smokeless operation for flare tips. If the steam is introduced at pressures below 10 psig, the desired turbulence or air entrainment will not be achieved due to insufficient momentum. The use of too much steam can also cool the flame too much and extinguish it. The flame may, in these event be reignited by hot zones and extinguishment by the steam again, at millisecond intervals.

51

Flashback Protection „ „

Flame Arrestors Liquid Seals

*EDS

2004/Flare Systems-52

Flame arrestors are passive devises with no moving parts. Liquid seals are active devices with water as the protection. The flame arrestor prevents the propagation of flame from the exposed side of the unit to the protected side by the use of wound crimped metal ribbon type flame cell element. This construction produces a matrix of uniform openings that are carefully constructed to quench the flame by absorbing the heat of the flame. This provides an extinguishing barrier to the ignited vapor mixture. Under normal operating conditions, the flame arrestor permits a relatively free flow of gas or vapor through the piping system. If the mixture is ignited and the flame begins to travel back through the piping, the arrestor will prohibit the flame from moving back to the gas source. Flare arrestors have specifically designed heat transfer characteristics for slow moving flames and low to medium pressure fronts. But flames moving at higher velocities and carrying higher pressure fronts can pass through a standard inline flame arrestor.

52

Liquid Seals „

Flare vapor piping submerged approximately 4 to 12 inches below the water level

„

Effective means to stop a flame front

*EDS

2004/Flare Systems-53

The most effective seal is the liquid seal. A liquid seal is simply a head of liquid (usually water) within the gas collection header that physically prevents air from intruding into the system. The liquid head prevents air from going into the system by preventing a vacuum. The liquid seal does require maintenance and utility requirements. Make-up water must be available at all times. In addition, freezing must also be considered. Normally, an overflow is provided to minimize the effects of hydrocarbon which settles out of the vapor. Depending on the effluent, a surge drum and pump may be necessary at the flare base to dispose of the seal liquid properly. The seal drum should be located between the stack and the other header drums and as close to the flare stack as possible. A variation of a seal drum is sometime incorporated into the base of the flare stack as shown in the above diagram. Continuous removal or intermittent skimming of hydrocarbons that may accumulate should be considered. If the hydrocarbons are not removed, there is a potential that the inlet piping may get plugged because of the hydrocarbons.

53

Liquid Seal To Flare

Vent Water Level

Water Supply

FI

4" (10 cm)

Submerged Weir Welded On End Of Flare Line

6" (15 cm) To Sewer Baffle

Sewer Seal Should Be Designed for a Minimum of 175% of Drum’s Maximum Operating Pressure

Try Cocks For Checking Hydrocarbons

10 Ft. (3M) Minimum

Flare Header

Drain *EDS

2004/Flare Systems-54

The above liquid seal is of the horizontal type. A minimum of 10 feet is required from the water level to the centerline of the outlet of the flare header. The problem of surging in seal drums can be minimized by the use of a submerged weir with Vnotches on the end of this pipe. The V-notch provides an increased flow area to the increasing gas flow. This V-notch type principle is similar to that of the bubble cap trays. Some design details of the liquid seal include anti-swirl or anti-vortex baffles on the liquid outlet lines. In addition, internally extended liquid outlet nozzles should be used because sediment will settle out in the drums and not in the low spot in the lines. This is shown in the above diagram with a 6 inch space between the top of the sewer outlet and the bottom of the drum. Designing the vessel nozzles, attachments, supports, and internals one should consider shock loading that result from thermal effects, slugs of liquid or gas expansion. Try-cocks as shown in the above diagram are useful in liquid level detection instead of level gauges. Instruments should be the simplest and most rugged available and should be easily maintained (externally mounted and valve). The use of seals instead of valves and of valves instead of traps is preferred, primarily because of the nature of the materials handled and the conditions under which these components must operate.

54

Flame Arrestor

„

Stop flame propagation within a piping system by means of breaking the flame into very small flames via a crimped wound metal grid thus quenching the flame by means of heat transfer and dissipation

*EDS

2004/Flare Systems-55

Extended lengths of pipe allow the flame to advance into more severe states of flame propagation such as high pressure deflagrations and detonations. Bends in piping, pipe expansions and contractions, valves or flow obstruction devices of any kind, cause turbulent flow. Turbulent flow enhances the mixing of the combustible gases greatly increasing the combustion intensity. This can result in increased flame speeds, higher flame temperatures, and higher flame front pressure than would occur in laminar flow conditions. High pressure deflagrations and detonations can occur more easily at higher system operating pressures than at near atmospheric levels. Elevated pressures condense the ignitable gas giving the flame more matter and energy to release thereby boosting flame heat intensity,A critical concern in flame arrestor installation is the possibility of a flame stabilizing on the face of the flame cell element. A flame that continuously burns against the flame arrestor element for a period of time can heat the element above the autoignition temperature resulting in flame propagation through the element. The time period varies with the type of element, mixture of air and gas, type of gas and velocity at which the gas stream is moving.

55

Liquid Seal Versus Flame Arrestor Product Name

Liquid Seal

Flame Arrestor

Function

Liquid seal is designed to stop flame propagation

The arrestor is designed to stop flame propagation

Product Type

Active Device

Passive Device

Testing Protocol

Not available

FM USCG API

Maintenance

Switches and cleaning Switch malfunctions, liquid freezing

Cleaning of flame cell elements Corrosion

Failure Modes

*EDS

2004/Flare Systems-56

The above chart shows the advantages of a liquid seal and the advantages of a flame arrestor. Liquid seals are much more common in flare system then are the flame arrestors. Liquid seals are also designed for detonation thereby the location of the installation is not as critical as that of the flame arrestor. In the initial state of flame propagation, the gas is ignited and the flame has propagated a short distance in the piping system. Flame front velocity is below the speed of sound thus defined as deflagration, and pressure piling is very low which makes it a low pressure deflagration. The intermediate state the flame has been allowed to propagate further down the piping system thus increasing both the flame velocity and pressure piling. The flame front velocity is just below the speed of sound and the pressure piling effects have increased tremendously resulting in a high pressure deflagration. The transformation form deflagration to detonation transition occurs when the flame front pressure and velocity reach an undefined limit. This explosive transformation causes the flame to accelerate to sonic, supersonic or hypersonic. In the advanced flame state, the flame has been allowed to propagate further down the piping system allowing detonation to occur. The flame front is moving at supersonic velocity.

56

Knockout Drums „

Principle Features –

– – – –

Complete removal of either slugs or mists of liquid (300 microns to 600 microns) Recovers valuable condensed hydrocarbons Ends maintenance difficulty caused by “Wet” gases Used as the base for the flare riser Ends “Wet Gas” control problems

The allowable vertical velocity in the drum may be based on the necessity to separate droplets from 300-600 microns in diameter. *EDS

2004/Flare Systems-57

Knockout drums are used to remove any liquid flowing to the flare. The knockout drum is also designed to disengage entrained liquid droplets. Normally the recommended particle disengagement is 300-600 microns in diameter. Although largely vapor, hydrocarbon relief streams from process plants often contain some liquid, even with out liquid carryover due to plant upsets, because droplets form as the vapor is cooled below the dewpoints. Knockout drums collect such liquid before it reaches the flare not only to reclaim it but also to prevent burning liquid from dropping around the base of the flare. Relief systems have been over loaded due to process upsets. When liquid upsets happen, if the knockout drum is not designed correctly then a problem of burning rain down the flare happens. Burning droplets can be carried by wind, starting fires in remote areas. Sizing this knockout drum is one of the most important aspect to flare systems. The maximum liquid flow to a knockout drum that could occur under any circumstance must be identified. If the liquid relief is huge then one might consider a separate relief system or may have many process unit knockout drums located in the process area before the main knockout drum at the flare stack. Because of the size, most knockout drums are horizontal. The one in the picture above is a vertical knockout drum and probably be designed for lower flow rates.

57

Hazards of Burning Rain

„

Injury to Personnel

„

Damage to Equipment

„

Source of Fire

*EDS

2004/Flare Systems-58

Separation of the entrained liquid in the knockout drum is required to prevent the burning rain. Burning rain occurs if the separation of the liquid is not provided adequately in the knockout drum. A separation of 300 microns is required in the flare knockout drum before the vapors are sent to the flare. If this separation does not occur, then there is the potential for burning rain. Burning rain is extremely dangerous because it can injury personnel, damage equipment and start localized fires. The flare knockout drum needs to be monitored to make sure that the level does not rise too high. This is because if the level rises and there is a major refinery release, say an emergency power failure then the effective separation zone is reduced. Therefore, the knockout drum should be checked at least daily and material which collects should be pumped out to the slop oil tank or to the crude oil tank depending on the composition of the material. Burning rain can also destroy the flare tip and flare stack. The design of the knockout drum can prevent this from happening.

58

Design Considerations Separation of Gas & Liquid

*EDS

2004/Flare Systems-59

Hydrocarbon relief streams are mainly vapors, but they may carry some liquids that condense in the collection line. Therefore, material entering the knockout drum will be a mixture of vapor and liquid. A particle that is 300 microns or less can be burned in a flare with out hazard. Larger particles shall be removed in the knockout drum. The liquid is pumped out from the bottom of the knockout drum either for reuse or for disposal in a slop oil tank. Knockout drums are either horizontal or vertical. They are available in a variety of configurations and arrangements that include horizontal drum with the vapor entering at one end of the vessel and exiting at the top of the opposite end with no internals. The drum can also be a horizontal drum with the vapor entering at each end on the horizontal axis and a center outlet. The drum can also be a horizontal drum with the vapor entering in the center and exiting at the two ends on the horizontal axis. The drum can also be vertical as the one shown above with the vapor entering with a tangential nozzle. The drum can also be vertical with the vapor entering at the top of a certain diameter and provided with a baffle so the flow is directed downward.

59

Design Considerations Liquid Holding Capacity

*EDS

2004/Flare Systems-60

Selection of the knockout drum arrangements depends on economics. When large liquid volume storage is required and the vapor flow is high, normally a horizontal drum is more economical. A split entry or exit reduces the size of the drum for large flows. As a rule of thumb, when the drum diameter exceeds 12 feet the split flow arrangements is normally economical. UOP has a number of programs to help size knockout drums. Liquid particles dropout when the vapor velocity traveling through the drum is sufficiently low. The drum must be of sufficient diameter to effect the desired liquid-vapor separation. It is general practice to assume a liquid holdup time between 10 to 30 minutes.

60

Illustrative LRGO Arrangement

First Stage Air Assisted Flare

Programmable Controller Rupture Disk

LRGO Tips

Pressure Transmitter

Stage 2 Stage 3

Flare Header Stage 4 Control Valve

Stage 5

Radiation Fence *EDS

2004/Flare Systems-61

The above drawing in an illustrative Linear Relief Gas Oxidizer (LRGO) Arrangement. The flow rate of gases will flow to the first stage air assisted flare. The reason is the low pressure at that flare tip (they run at atmospheric pressure). Once the air assisted flare gets overloaded, the back pressure builds in the system up to a point of approximately 7 psig. At this point, stage tow of the LRGO tips open. If the flow still increases and the back pressure is still increasing, at say 8 psig, stage 3 will open. If the flow still increases more and the back pressure increases more, say at 9 psig, stage 4 will open. If the flow is still increasing, say at 10 psig, stage 5 will open. The above arrangement is a 5 stage LRGO system. Therefore, the maximum capacity of this flare is up to whatever can be released when the last stage is open. The stages are opened when the pressure controller sends a signal to the control valve at each of the 4 LRGO stages. In case the control valve fails closed, there are rupture disks which will break allowing the flare gas to by-pass the control valve and go to the LRGO stages. In addition, there is a radiation fence to reduce personnel exposure to the LRGO.

61

LRGO - Ground Flare - Critical Location

*EDS

2004/Flare Systems-62

LRGO can be installed in almost any location. The one above was built near the gulf on a sandy surface. The one in the photo above is a four stage LRGO system. Open ground flares, as shown in the next few slides, come in a variety of configurations. Some have many small burners such as that shown on two slides later than this slide and some have large burners such as the one shown above and on the next page. The many small burners that together can flare larger quantities of relief gas than an enclosed flare. Their capacity is limited by the available plot area to distribute the burners. The individual burners are fully smokeless, using the kinetic energy of the flare gas to entrain enough air for complete stoichiometric combustion. The drawback of the open ground flare is that the required pressure at the flare tip is about 7 to 10 psig, which is much greater than an open elevated flare. Rows or sections of burners are than staged using control valves, so that the proper pressure is maintained at the burners. As flow rate increases, the header pressure increases and more burners are made available.

62

LRGO - Designed for the Tundra of Alaska

*EDS

2004/Flare Systems-63

The above is a picture of a ground flare or what is known as a LRGO - Linear Relief Gas Oxidizer. The LRGO was installed in the tundra of Alaska to minimize the effects of radiation on the tundra. For smaller plants, open pit ground flares can sometimes be made large enough to replace an elevated flare. They can also be built with a refractory fence to reduce radiation effects. The ground flare (LRGO) can have the burner stages arranged in a number of different configurations depending on the plot plan layout. The one above is shown as a half-semicircle which allows more plot space for other requirements. Even with an LRGO, a knockout drum is required just prior to the burner tips as shown in the yellow shelter above.

63

Ground Flare Designed in Mexico

*EDS

2004/Flare Systems-64

Open ground flares can have a variety of configurations. Some have many small burners that together can flare large quantities of relief gas, such as the one shown above. Their capacity is limited by the available plot area to distribute the burners. The individual burners are fully smokeless, using the kinetic energy of the flare gas to entrain enough air for complete stoichiometric combustion. The disadvantage of ground flares is that they require pressure at the flare tip of about 7 psig. This is much greater than the typical elevated flare. Rows of burners are staged using control valves, so that the proper pressure is maintained at the burners. As flow rate increases, the header pressure increases and more burners are made available.

64

Ground Flare - Operating *EDS

2004/Flare Systems-65

The above photograph shows a ground flare in operation. Note that because of the thermal stresses the burner tips actually bend. After the release they will return to their original position. The ground flare tip design uses gas stream pressure to promote mixing at the burner tip of the ground flare. With sufficient gas pressure, pressure assisted or ground flares can be used where steam or air assist was previously required. Multiple burner heads are staged to operate based on the amount of gas being flared. Flare gas properties determine size, design, number and group arrangements of burner heads. Gas streams with 10 psig or greater available pressure are normally selected. The problem with ground flare is that most refineries have a crude unit/vacuum unit which have relief valves set at 50 psig (3.5 kg/cm2g) and therefore there is not enough pressure available for this type of flare. The refinery could break their flare system into a high and low pressure system where they may send all relief valves with a set pressure greater than 250 psig to the high pressure flare system and send all relief valves with a set pressure less than 250 psig to the low pressure flare system.

65

Advantages of Ground Flare „ „ „ „ „ „

No structural support required Erection is relatively straightforward Maintenance is easy Operating costs are negligible Flame of flare not visible Fairly quite system

*EDS

2004/Flare Systems-66

There are many advantages to the ground flare. No structural supports are required and therefore the cost may be less. The erection of the ground flare can take place at grade level so erection is relatively straightforward. If maintenance is required, the ground flare be located at grade can be fixed easily. The operating cost of the ground flare because no steam or air is required to make the flare smokeless is negligible. A big advantage of the ground flare is that the flare is not visible to neighbors or others. The system is fairly quite and, therefore, more socially acceptable.

66

Disadvantages of Ground Flares Must be well isolated from the rest of the refinery „ Requires considerable space and long interconnecting piping „ Combustion takes place on ground „ Concentration of toxic gas at grade may remain high „

*EDS

2004/Flare Systems-67

The disadvantage of ground flare is that they must be isolated from the rest of the refinery. In elevated flares, depending on the radiation profile, the flare may be located near the process units. As can be seen from the above photograph, the ground flare requires considerable space and long inter connecting piping resulting in a lot of capital cost. A big disadvantage is that the combustion takes place on the ground so therefore the concentration of toxic gas at grade may remain high for some time after a release. Drainage must also be considered.

67

Air-Assisted Flare „

Select the proper blower requirements

„

Blower to provide air flow and gas in proportion to each other to properly mix

„

Should be provided on the outside of circular air riser *EDS

2004/Flare Systems-68

The key to successful smokeless burning of flare gases using an air assist flare is to properly design the burner system and to integrate that system with a properly selected blower. Of particular importance in an air-assisted flare is obtaining a zone to promote mixing. The burner design must be executed to provide air flow and gas flow in proportion to each other to properly mix. There must be more gas flow near the outside of the circular air riser than towards the inside because the bulk of the air flow are in a circular plenum is contained on the outside of the plenum. In addition, the burner design must provide for stability of the waste gas under greatly varying flow conditions and compositions. This design is more easily achieved by using a spider-type burner whose center hub acts as a stability point for the burner insuring stable combustion through a wide range of compositions and turndowns.

68

Quiet Flare (Low Noise) „

Qualitative sense decibels (dB) describe the loudness of sound and noise. Whisper = 20 dBA – Conversation = 65 dBA – Food Blender = 88 dBA – Motorcycle = 100 dBA –

„

OSHA requires equipment to have 1000

Maximum Velocity Vmax (ft/sec) 60 log10(Vmax) = (Bv + 1214)852 400

It is standard practice to size the flare so that the design velocity of flow rate Qtot, is 80 percent of Vmax: Dmin (in) = 12*[((4/PI)(Qtot/60sec/min))/(0.8*Vmax)]^0.5 Dmin (in) = 1.95 * (Qtot/Vmax)^0.5 „ Where: Qtot = Q + F (measured at stream temperature and pressure) „ Dmin should be rounded up to the next largest available commercial size „

*EDS

2004/Flare Systems-82

Rearranging gives F (scfm) = Q * (300 - Bv)/(Bf - 300) The annual auxiliary fuel requirement, Fa, is calculated by: Fa (Mscfm/yr) = (F scfm)(60 min/hr)(8760 hr/yr) Fa (Mscfm/yr) = 526F Typical natural gas has a net heating value of about 1,000 Btu/scf. Automatic control of the auxiliary fuel is ideal for processes with large fluctuations in VOC composition. These flares are used for the disposal of such streams as sulfur tail gases and ammonia waste gases, as well as any low Btu vent streams.

82

Improve Flare Burner Life Vector Diagrams F

W

F

W

W

L G

W= F = LP = G = S =

High Exit Wind Velocity Flare LP Zone Gas Jet Supplementary Energy

G

L

F S

G

L

Low Exit Low Exit Velocity Velocity With Secondary Energy Source

*EDS

2004/Flare Systems-83

The above diagram shows how steam (supplementary energy) injected into the flare tip can improve the flame tilt. Pressure drops as large as 2 psig have been satisfactorily used at the flare tip. Too low a tip velocity can cause heat and corrosion damage. The burning of the gases becomes quite slow, and the flame is greatly influenced by the wind. The low-pressure area on the downwind side of the stack may cause the burning gases to be drawn down along the stack for 10 feet (3 meters) or more. Under these conditions, corrosive materials in the stack gases may attack the s

83

Incinerator Design

„

Mechanism –

„

Oxidation reaction

Factors –

Time, temperature and turbulence

*EDS

2004/Flare Systems-84

Incinerators are called thermal oxidizers, combustors and thermal reactors. The function of an incinerator is the destruction of organic and inorganic wastes using an oxidation reaction. The products of combustion are Carbon monoxide Nox, Sox, acid gases, particulate and unburned hydrocarbon. The carbon monoxide are generally a function of burner operation and the unburned hydrocarbon and SOX and acid gases generally a function of incinerator operation. There are many permit requirements such as tons per year that can be emitted and ppm that can be discharged. The equipment shall be designed for the worst case scenario if limited by an hourly average ppm level. A tons per year emission limit may allow designing for a normal case, while merely handling an infrequent or intermittent surge. The waste characterization determines the type of incinerator. The burner selection is dependent on waste type and special consideration such as low NOX. The incinerator design for introduction of some waste must provide for retention time. Flue gas treatment may be necessary to meet the required emissions levels if particulate or acid gases are present in fuel gas.

84

Principles of Combustion „

Burners –

„

Chamber –

„

Ignite the fuel and organic material Appropriate residence time for oxidation process

Three T’s of Combustion Temperature – Time – Turbulence –

*EDS

2004/Flare Systems-85

Some process design consideration are low waste gas pressure drop of 4” w.c. and NOX requirements because of high fuel rate. The mechanical design consideration for natural draft and for forced draft must be considered. Natural draft is less expense, forced draft provides better turndown, and forced draft is better for adding heat recovery . The layout of the incinerator must be considered. Is horizontal better than vertical? The cost of the vertical unit is less expensive, the vertical unit requires less plot space. The horizontal is easier to access and add downstream equipment. If the heat release is large, then the horizontal may be required to eliminate impingement. The refractory of brick castable or blanket must be considered. Highly abrasive particulate requires special castable or brick. The blanket has limited velocity capability. The blanket provides ability for on/off operation and heat up and also costs less.

85

Coupled Effects of Temperature and Time on Rate of Pollutant Oxidation Pollutant Destruction, %

100 80 60 40

1 sec 1.0 sec 0.01 sec

0.001 sec

Increasing Residence Time

20 0 600

800

1000 1200 1400 1600 1800 2000 Increasing Temperature, °F

Residence time of gases in combustion chamber calculated from: t = V/Q t = Residence Time (s) v = Chamber Volume (ft3) Q = Gas volumetric flow rate at combustion conditions (ft3/s) *EDS

2004/Flare Systems-86

The above chart is the coupled effects of the three T’s of temperature, time and turbulence on the rate of pollutant oxidation. By increasing temperature only and holding the retention time, the same increases the pollutant destruction efficiency. By increasing the retention time only and maintaining the same temperature increases the pollutant destruction. Depending on the percent pollutant destruction required determines what temperature is required in the incinerator and how long of a residence time is required. Therefore, for the same gas volumetric flow at combustion conditions, a higher temperature and lower retention time or a lower temperature and higher retention time may be the more optimum design for the incinerator.

86

Schematic of a Thermal Incinerator Fume

Fuel

Exhaust

Combustion Air (Fume)

*EDS

2004/Flare Systems-87

The above drawing is a schematic of a thermal incinerator. As you can see from the drawing, the waste gas or fume is introduced at the top left hand corner. Combustion air and fuel are added to ignite the mixture. The time it takes for the fume to enter and leave the incinerator as exhaust is know as the retention time. The temperature in the incinerator is the number used in calculating the efficiency. You can also see the turbulence when the air, fuel and fume are combined within the incinerator. Using low NOX burners or 2 stage furnace (reducing then oxidizing), catalytic NOX reduction or special scrubber may be required after the exhaust to maintain an NOX ppm limit which is set by the local law and governments such as EPA.

87

Typical Marine Vessel Loading System Product Loading Arm Product from Storage Tanks Vapor Arm

Natural Gas/ Inerting Gas Enriching Gas Detonation Analyzer Arrestor

Vapor Mover

Hydrocarbon Vapor to Control Device

Knockout Drum(s) Discharged Vapors Sump Pump

Ship or Barge Dock Facilities

Condensate to Tanks

Shoreside Facilities

*EDS

2004/Flare Systems-88

To achieve high destruction efficiencies through the loading cycle requires that provision be made to keep the combustor at a minimum temperature prior to introducing loading vapors and during high vapor flow rates when the majority of the vapor is air, nitrogen or carbon dioxide depending on the ship being loaded. Achieving high destruction efficiencies requires that the oxygen in the combustion air be given adequate opportunity to come in contact with hydrocarbons which can be a very low percentage of an inert vapor stream. Mixing is critical to making this happen. The more time the vapor stream is at temperature in turbulent contact with combustion air, the higher the likelihood of complete combustion which means no carbon monoxide or hydrocarbons. The less combustion air that is required (excess air) to achieve complete combustion, the less auxiliary fuel is required to maintain temperate. This is especially important when dealing with very low Btu content vapors associated with inert vessels or non inert ships that use inerting techniques to meet the Coast Guard safety criterion. The design of the combustion equipment, which accounts for approximately 20% of the total marine terminal equipment cost, is influenced by the emission rules. Due to the cyclic nature of the use pattern of combustors in marine service, ceramic fiber is generally the refractory system of choice due to its low heat capacitance and resistance to thermal shock. These units are routinely brought from ambient to 1500°F in less than 5 minutes.

88

Storage Tank and Tank Truck Loading Hydrocarbon Concentration Profile

Total Hydrocarbon Vapor Concentration in Vessel (Volume Percent)

70 60 Gasoline Vapors 50 40 30 Crude Vapors

20 0

0

25

50

75

100

Percent of Storage Tank or Truck Filled *EDS

2004/Flare Systems-89

The loading occurs at the transportation loading rack, the vapor combustion system is in a standby mode with no pilot flame, the vapor block valve is closed, and the air assist blower is off. Automatic startup of the vapor combustion system is initiated by an electrical signal from the loading rack that product loading will occur shortly. The startup sequence consists of a short air purge using the air assist blower to purge the air plenum of the open flare or the enclosed flare of any combustibles prior to pilot ignition. This brief air purge is followed by automatic electronic ignition of the pilot. After pilot ignition, product loading begins at the loading rack and an air vapor mixture begins to flow from the transports being loaded to the vapor combustion system. Flow through the vapor combustion system first consists of the air vapor mixture from the lading rack bubbling through a liquid seal. As soon as sufficient flow is available, it will be detected by the pressure monitoring controls which will automatically open the vapor block valve allowing the air vapor mixture to flow thorough the flame arrestor to the burner, where the combustible vapors are ignited by the pilot and burned. The air assist blower provides partial combustion air and mixing energy to the burner tips to assure smokeless combustion. As the loading operation at the loading rack is completed, vapor flow to the combustion system decreases. The pressure monitoring system closes the vapor block valve when the vapor flow is insufficient to maintain minimum burner velocity. The pilot and air assist blower remain on for a brief time period after loading is complete. If no further loading occurs, the combustion unit will shut down in the standby mode to await automatic restart. 89

Marine Vessel Loading Hydrocarbon Concentration Profile 70 Total Hydrocarbon Vapor Concentration in Vessel (Volume Percent)

60 50 40 30 Gasoline Vapors

20

Crude Vapors

10 0

0

25

50

75

100

Percent of Marine Vessel Filled *EDS

2004/Flare Systems-90

As shown in the above chart, the hydrocarbon vapors are in higher concentration during the final portion when the vessel is being filled. The vapors immediately upon being transferred to the loading facility, are analyzed and conditioned, if necessary, to assure they are well outside the explosive limit. The combustion system must be designed to cope with air, nitrogen, and carbon dioxide vapors with hydrocarbons concentrations varying from 0 to 100 percent. The application must also deal with flow rates which are wetted, loaded and topped off. The large turndown (up to 100 to 1) encountered in marine terminals, satisfying the low end velocity criteria makes it impossible in a single burner configuration to have stable and efficient combustion at the high flow rates. To accommodate the wide range of vapor flows, multiple burners are used and are brought in and out of service to assure the flame arrestor tips are always kept cool and that exit velocities never exceed the stability limit of the burner

90

Typical Lean Oil Absorption Vapor Recovery System

*EDS

2004/Flare Systems-91

Economic consideration as well as environmental considerations have led to vapor recovery systems as the one shown above. Since a vapor recovery system is used for extreme emergencies one must not compromise the integrity of the flare system. The above drawing shows a lean oil absorption vapor recovery system. Hydrocarbon vapor is introduced into the absorption column where lean oil is absorbed by the vapor. The treated vapor after being “scrubbed” can be released into the atmosphere. The rest of the section is the absorber regeneration process. Instead of adding lean oil continuously, most of the lean oil can be regenerated. This involves separating the absorbed material from the lean oil used to absorbed the material.

91

Typical Refrigeration Vapor Recovery System Vapor Condensing Unit (Evaporator)

Precooler

Thermostatic Expansion Valve

Hydrocarbon Vapor Inlet

Treated Vapor

Blower

Freon Return

Compressed Refrigerant CompressorCondensor Power Input

Power Input Condensate Collection Tank

Recovered Product

*EDS

2004/Flare Systems-92

The above drawing is a refrigeration vapor recovery system. The hydrocarbon vapors are precooled and then sent to a vapor condensing unit or what is know as an evaporator. The hydrocarbons will condense once it fall below its dew point. The product that condenses is stored in a condensate collection tank before being sent back to the process unit for reprocessing. The treated vapors from the condensate collection tank are sent to the atmosphere by a blower. The refrigeration vapor recovery system is similar to that of a refrigerator or an air conditioning unit which uses some type of freon depending on the temperature the stream needs to be reduced to. The disadvantage of the refrigeration vapor recovery system is the high cost of electricity required for the system.

92

Typical Catalyst Oxidation System Hydrocarbon Vapor Inlet

Thermal Zone

Treated Vapor Out

Burner Heated Air or Process Stream

Catalyst Bed

Feed Air or Process Stream

Oxidation Zone

Preheat

Optional Heat Recovery

Stack

*EDS

2004/Flare Systems-93

The above drawing is that of a catalytic oxidation system. The hydrocarbon vapors are fist preheated and then sent into the thermal zone of the oxidation system. After being heated to over 1000°F, depending on the catalyst, it is sent through a catalyst bed and the oxidation zone. The vapors are then preheated against the hydrocarbon vapor air going into the thermal section. The vapor are still going to be hot and then can be either sent back to the process streams or an optional heat recovery unit can be installed. After the heat recovery unit, the cooled treated vapor will be sent to a stack for final disposal to the atmosphere. The advantage of using both thermal and catalytic system is that the temperature does not have to be increased to get the same destruction/recovery efficiency.

93