• Author / Uploaded
• Fu_John

Citation preview

-

I- FA E.

CEMENT I_ A NI 1-

•

OPERATIONS HANDBOOK 1 Second Edition July 1998

The concise guide to cement manufacture Philip A Alsop

TERNATIONAL •

Revie w w

CEMENT PLANT OPERATIONS HANDBOOK For Dry Process Plants Philip A Alsop, PhD

Second Edition July 1998

Tradeship Publications Ltd

ACKNOWLEDGMENTS I would like to thank the legions of grey men with whom 1 have worked and from whom 1 have learned over the years. I acknowledge with thanks the help of Mr James Post with whom I worked on the first edition. For their help with this revision, I wish specifically to thank Dr Hung Chen, Mr James Korthals, Mr Mark Mays and Mr Herman Tseng who have reviewed all or part of the manuscript and have made a number of valuable suggestions and corrections. I am particularly grateful to Mr Andrew Jackura for writing the chapter on Reliability and Maintenance and for substantial contributions to other sections. Southdown, Inc has kindly encouraged the revision of this book. The Company has not, however, reviewed the text and is in no way responsible for factual errors or contentious suggestions. As before, the author retains proprietary rights over all expressions of ignorance and opinion. Finally, I am indebted to Mr David Hargreaves of international Cement Review for his continued interest and support in the re-issue of this small work.

Philip Alsop Houston, TX June 1998

PREFACE The first edition of this handbook appeared to find some use within the Industry and we are encouraged to revise and expand the material. As previously noted, while there are a number of excellent books covering plant design, process engineering, and cement chemistry, there is a fairly sparse literature addressing cement plant operations and little to serve the hapless novitiate. Again for brevity, the objective has been constrained, and whole areas of operations technology and management have been omitted as being inappropriate to address in so limited a compass. It is also appreciated that regulations, specifications, and even operating practices are not universal, and our observations should be discounted accordingly. The scope attempted comprehends: A consideration only of cyclone preheater kiln technology which comprises more than 80% of world production and virtually all kilns installed since 1970. The use only of metric units. A review of major plant sub-systems with a proposed list of data which should be available to plant and corporate management, and some suggestions regarding problem areas and possible solutions. A summary of cement types and concrete problems. A collection of process formulae. A selection of reference data and notes. An outline of plant assessment and plant valuation. References to review articles and a limited bibliography. Addresses of pertinent organizations.

Cement Plant Operations Handbook • iii

The assessment of cement plant equipment and operations involves numerous terms and numbers, many of which are prone to varying definitions. We would like to offer the following comments and suggestions towards standardization: Cyclone preheater kiln - There is much confusion in terminology regarding air-suspension preheater kilns with and without secondary firing or precalcination. The term preheater does not distinguish air-suspension from grate preheaters, and preheater is frequently used in distinction from precalciner. Dry process, of course, also refers to long dry kilns and fluidized beds. We would, therefore, advocate "cyclone preheater" and "cyclone preheater with precakiner". Plant capacity - Annual capacity can relate to, various assumptions for kiln operation and for cement intergrinding. A reasonable standard is the designed, or established, daily clinker production assuming 85% annual run factor and 5% cement additives: Annual cement capacity = Clinker tlday x 365 x 0.85 / 0.95

Some elaboration is still required for the plant which produces 3,000t/day of clinker and 6,000t/day of cement, but it is better understood as a 980,000t/year plant than 1,860,000t/year. Kiln run-factor - Various definitions have been encountered including fire-on time, and running-time exclusive of planned shut-downs. Feed-on time is suggested to be the most significant parameter and should be expressed as a percentage of 8760 hours per year.

Cyclone stage numbering

This is a divisive subject. Uncomfortably for those of us who have always numbered from the top, it must be asserted that, with the proliferation of cyclone preheaters of other than 4 stages, numbering from the bottom allows more meaningful correlation from kiln to kiln. -

iv • Cement Plant Operations Handbook

OBITER DICTUM Joseph Aspdin of Leeds patented Portland cement in 1824 for a ground mixture of calcined limestone and clay.

m wri

The designation "Portland cement" was originally due to a resemblance in colour and character to limestone found on the Isle of Portland off the south coast of England.

rn

The recognition that high temperature clinkering (C3S formation) greatly improved performance is attributed to Isaac Johnson in 1845. Joseph Ransome patented the rotary kiln in England in 1885 and the air-suspension cyclone preheater was patented by VogelJorgensen in Czechoslovakia in 1932. "In the old days their monstrous red-hot bodies revolved with a cosmic roar and howl, belching hellish flames...and over all the acrid stench of cement." Fyodor Gladkov; Cement (1925)

Cement Plant

Operations Handbook • v

Publisher: David Hargreaves, International Cement Review Layout & Design: Mary Flack, Paul Benewith Copyright © Philip Alsop All rights reserved. No part of this publication may be reproduced, stored in a retrieval system, or transmitted in any form or by any means, electronic, mechanical or otherwise, without the prior permission of Tradeship Publications Ltd. This publication is intended solely for use by professional personnel who are competent to evaluate the significance and limitations of the information provided herein, and who will accept total responsibility for the application of the information. Printed by Bishops Ltd, Portsmouth, United Kingdom Front cover: The CBR cement works in Lixhe, Belgium Photographer: Mr Serge Brison

Section A — Process Summaries

1 INTRODUCTION

1

2 RAW MATERIALS

3

1 Raw Materials 2 Reserves 3 Crushing 4 Drying 5 Preblending

3 RAW MILLING & BLENDING

13

1 Raw Milling 2 Blending 3 Kiln Feed

4 BURNING & COOLING

21

1 Kiln Burning 2 Control Systems 3 Kiln Control 4 Start-up & Shut-down 5 Refractories 6 Insufflation 7 Kiln Bypass 8 Preheater Cleaning 9 Kiln Fuels 10 Coal Firing 11 Clinker Cooling 12 Kiln Mechanical 13 Emergency Power

5 FINISH MILLING

61

1 Clinker Storage 2 Finish Milling 3 Separators 4 Cement Storage 5 Cement dispatch

Cement Plant Operations Handbook • vii

1 M7-1WI.X

CONTENTS

6 POLLUTION CONTROL

75

1 Dust Collection 2 Pollution control

7 QUALITY CONTROL

85

1 Chemical Analysis 2 Sampling 3 Physical Testing 4 Process Control 5 Microscopy 6 Cement Strength 7 Cement Types & Specifications 8 Cement Intergrinds 9 ISO 9000 10 Concrete Problems 11 Domestic Water Treatment

8 MAINTENANCE

107

1 Failure Modes 2 Computerized Maintenance Management Systems 3 Reliability Centered Maintenance 4 Cost Management 5 Organization 6 Role, Planning & Control 7 Mobile Equipment Maintenance

9 ACCOUNTING

115

Cost Accounting 2 Investment Justifications 3 Project Cost Estimation 4 Profit & Loss Statements

10 MISCELLANEOUS

121

11 PLANT REPORTING

125

1 List of Reports 2 Equivalent tonnes 3 Downtime Reporting 4 Daily Production Report 5 Process Summary 6 Equipment Downtime Report

viii • Cement Plant Operations Handbook

135

1 Power

136

1 Specific Power Consumption 2 Power Conservation 3 3-Phase Electricity 4 Power Generation 5 Co-generation

2 Fans & Air Handling

139

1 Fan Laws 2 Impeller Build-up 3 Gas Specific Gravity & Specific Heat 4 Plant Air Distribution 5 Orifices, Pitots & Venturis 6 False Air 7 Dew Point 8 Stack Draft 9 Spray Cooling 10 Abrasion Resistance

147

3 Conveying 1 Comparative power consumption 2 Fuller Kinyon Pump 3 Bucket Elevator 4 Belt Conveyor 5 Screw Conveyor 6 Air-slide 7 Drag Conveyor 8 Tube Belt Conveyor 9 Sandwich Belt Conveyor 10 Pneumatic Capsule Conveyor 11 Water pump

151

4 Milling 1 Sieve Sizes 2 Circulating Load 3 Separator Efficiency 4 Tromp curve 5 Critical Speed 6 Charge Loading 7 Grace Factor 8 Mill Power

Cement Plant Operations

Handbook • ix

1N O

Section B - Process Data & Process Calculations:

9 Ball weight & surface 10 Maximum Ball Size 11 Measurement of Wear

5 Kilns & Burning

157

1 Cement Compounds & Ratios 2 Coating index 3 Burnability factor 4 Heat of Formation 5 Heat Balance 6 Kiln Gas Velocity 7 Kiln Heat Loading 8 Kiln Retention Time 9 Kiln Volume Loading 10 Kiln Drive Power 11 Cooler Efficiency 12 Kiln Exhaust Gas Composition 13 Circulation of Volatiles

167

6 Fuels Data 1 Coal & Petroleum coke 2 Fuel Oil: 3 Natural Gas.

7 Materials Data

171

1 Bulk Densities 2 Specific Gravities & Grindabilities 3 Coefficients of Linear Expansion.

8 Conversion Tables

172

1 Length: 2 Area: 3 Volume 4 Density 5 Pressure 6 Energy 7 Weight 8 Miscellaneous

9 Miscellaneous Data

x • Cement Plant

Operations Handbook

173

1 Geometrical & Trigonometrical Formulae 2 Greek Alphabet 3 Atmospheric Pressure 4 pH & Normality 5 Laboratory Reagents 6 Sea Water Composition 7 Elemental Abundance 8 Hardness of Materials 9 World Cement Production 10 Ship Capacities

179

10 Statistics 1 Standard deviation 2 F Test 3 X2 Test 4 Linear regression

11 Hydration of Portland Cement

183

12 Other Kiln Types

185

1 Long Wet 2 Long Dry 3 Grate Preheater 4 Vertical Shaft

13 Technical & Process Audits

193

1 Kiln Specific Fuel Consumption 2 Cement Mill Specific Power Consumption 3 Other Systems

14 Plant Assessment Data List

197

15 Cement Plant Valuation & Construction Cost

205

References

211

1 Review papers 2 Books 3 Addresses of pertinent organizations

index

217

Cement Plant Operations Handbook • xi

xii • Cement Plant Operations Handbook

1 INTRODUCTION Cement is "a substance applied to the surface of solid bodies to make them cohere firmly" or, more specifically, "a powdered substance which, made plastic with water, is used in a soft and pasty state (which hardens on drying) to bind together bricks, stones, etc in building" (SOED). Portland cement is a calcined material comprising lime and silicates which is mixed with sand and stone and, upon hydration, forms a plastic material which sets and hardens to a rock-like material, concrete. Confusion between cement and concrete is endemic among the uninitiated. Portland cement is manufactured in a series of processes which may be represented as shown:

-

1,14 4

Limestone quarry

Clay Silica

Crusher

Iron

I

Raw mill Clinker silos

Limestone stockpile Blending silos

Gypsum

Additives

Clinker

Cement Cement mill

Cement sibs

4114

CEMENT PLANT SCHEMATIC PROCESS FLOW

Shipping

Limestone (calcium carbonate) and other materials containing appropriate proportions of calcium, silicon, aluminum, and iron oxides are crushed and milled to a fine flour-like raw meal. This is heated in a kiln firstly to dissociate calcium carbonate to calcium oxide with the evolution of carbon dioxide, and then to react calcium oxide with the other components to form calcium silicates and aluminates which partially fuse at material burning temperatures up to 1450°C. The reaction products leave the kiln as a black nodular material, clinker. The clinker is finally interground with a small proportion of gypsum (to control the rate of hydration) and the fine product is cement.

Cement Plant Operations Handbook • 1

r-

2 • Cement Plant Operations Handbook

2 RAW MATERIALS 2.1 Raw Materials The composition of portland cement varies from plant to plant due both to cement specifications and to the mineralogy of available materials. In general, however, an eutectic mix is sought which minimizes the heat input required for clinkering and the total cost of raw materials, while producing a cement of acceptable performance. An approximate analysis for raw mix on ignited basis, or for clinker, is: CaO 5102 Al203 Fe203 MgO Mn203 T102 503 K20 Na20

65-68% 21-23% 5- 7% 2-4% 1-5% 0.1-3% 0.1-1% 0.1 -2 % 0.1 -1 % 0.1 -0.5%

Note that, with a substantial proportion of the raw mix being CaCO3, heating either in a kiln or in a laboratory furnace evolves some 35% by weight as CO2; this results in a requirement of approximately 1.5t of raw materials to produce 1t of cement, and also requires that analytical data be clearly distinguished between "raw" and "ignited" basis. Cement mixes vary from "cement rock", a single component which, as mined, contains appropriate proportions of the required minerals, to 4 or 5 component mixes comprising one or two grades of limestone, a shale or clay, and one or more additives to augment Si02, Al203 or Fe203 levels. Kiln feed typically contains 78-80% CaCO3 so that limestone can only fall close to this level to the extent that it also contains the other ingredients. It is essential to have sufficient flux (Al, Fe, Mg, F) to promote fusion in the kiln, but MgO should not exceed 4-5% or the cement may be expansive. Excess alkalis (K, Na) affect both kiln operation (build-ups) and product quality (alkali-aggregate reactivity). Excess S causes kiln build-ups and limits the addition of gypsum which may

Cement Plant Operations Handbook • 3

result in setting problems. Conventional wisdom suggests that the stoichiomatric ratio of alkalis to sulfur should be kept between 0.8-1.2 (see, however, Sec. B5.14). Excess Cl causes serious problems for preheater operation. Materials, as mined, therefore, are typically proportioned: Limestone (CaO) Shale or clay (510 2, Al 203 & Fe2O3) Additives (5i02, Al203 or Fe 2O3)

85% 13% 860° > 900° > 1200° > 1280°

Evaporation of free water Evolution of combined water CaCO3 CaO + CO2 Reactions between Ca0 and Al203, Fe203 and 5102 Liquid formation Formation of C35 and complete reaction of Ca0

Cyclone preheater kilns have developed rapidly since the 1950s and have been virtually the only type of cement kiln installed over the past 25 years. The first units were 4-stage preheaters. Single string preheaters are limited to about 4500t/day (with up to 10MO cyclones) and larger kilns now have two- and even three- strings together with precalciners (secondary combustion vessels between kiln and preheater) allowing unit capacities in excess of 10,000t/ day. Heat recovery has also been improved, where heat is not required for drying raw materials, by using 5-and 6-stages of cyclones, and redesign of cyclone vessels has allowed pressure drop to be reduced without loss of efficiency (Hose & Bauer; ICR; 9/1993, pg 55). Exit gas temperatures, static pressures, and specific fuel consumptions for modern precalciner kilns are typically: 6-stage 5-stage

280° 310°

450mm 1-120 400mm

4-stage

350°

350mm

710kcal/kg (NCV) 725kcal/kg 750kcal/kg

Temperatures are 20-30° lower without precalciners and older systems are usually 20-30° higher than the above. Early 4-stage cyclone preheater kilns commonly have pressure drops of 700-800mm (higher if ID fans have been upgraded without modifying cyclones and ducts) and specific fuel consumptions of 850-900kcal/kg (Figure 4).

Cement Plant Operations Handbook • 21

••

_Ad

Figure 4

350deg C 750mm WG

r=93% 350deg C

Feed

550deg C 500mm WG

n=80%

700deg C 300mm WG

WOdegC

n=70% 11\

67GclegC

820deg C 150mm WG

000deg C 50mm WG

CYCLONE PREHEATER TYPICAL TEMPERATURE & PRESSURE PROFILE AND CYCLONE EFFICIENCIES

22 • Cement Plant Operations Handbook

In cyclone preheater kilns without precalciners, the feed is 20-40% calcined at the kiln inlet. Riser firing increases this, and addition of a caldner vessel allows up to 90% calcination before the meal enters the kiln. Precalcination should not exceed 95% as completion of the endothermic reaction would allow a dangerous material temperature rise before entry to the kiln with probable build-up and plugging. The major cyclone preheater configurations are shown in Figure 5. Other terms frequently encountered include: NSP (New Suspension Preheater) - Precalciner technology which was developed in Japan in the early 1970s. MFC (Mitsubishi Fluidized-bed Calciner) RSP (Reinforced Suspension Preheater) - Onoda design of precalciner vessel. AT (Air through) - Precalciner or riser firing using combustion air drawn through the kiln. AS (Air separate) - Precalciner using tertiary air.

4.1 Kiln Burning Kiln operation is monitored by: Production rate, tonnes/hour clinker Operating hours Involuntary downtime hours Specific heat consumption, kcal/kg Total fuel rate, tonnes/hour Proportion of fuel to precalciner/riser, % Secondary air temperature, °C ID fan draft, mm H2O Preheater exhaust gas temperature, °C Kiln feed-end 02, % Downcomer 02, % Kiln feed-end material — LoI, % — SO3, % Kiln drive power, kW A clock which integrates kiln feed-on running hours is useful. There are, of course, numerous other process parameters which should be logged, both to observe trends which may indicate problems, and to provide necessary mean data for process analyses such as heat balances.

Cement Plant Operations Handbook • 23

Figure 5

Calciner

Tertiary

SP - Cyclone Preheater with or without riser firing

ILC - In line Calciner fed by both kiln exhaust and tertiary

SLC - Precalciner fed by tertiary only. Can be single, 4' double, or triple string preheater

MAJOR CONFIGURATIONS OF CYCLONE PREHEATER KILNS

24 • Cement Plant Operations Handbook

Other kiln performance factors include: Primary air tip velocity, M/ sec Specific kiln volume loading, % Gas velocity in burning zone, M/sec Specific heat loading of burning zone, kcal/h per M 2 of effective burning zone cross-section area. Cooler air, NM3 /h per M2 grate area Cooler+primary air, NM3 per kg clinker Temperature, pressure and oxygen profile of preheater Excessive heat consumption should be investigated immediately and may be indicative of incorrect feed-rate measurement or feed chemistry, burner abnormality, insufficient or excess oxygen, air inleakage at kiln seals or preheater ports, low temperature of secondary air, and distortion or collapse of preheater splash-plates (Saxena et al; WC; 2/1996, pg 44). Clinker free-lime should be as high as possible to avoid the problems of hard burning, but safely below the onset of mortar expansion; typically between 0.5% and 2%. Having established the target, free lime should, if possible, be maintained within a range of about 0.5%. Variation of kiln feed rate or composition makes this control more difficult. It should be appreciated that overburning - a common solution to variable kiln feed chemistry or operator circumspection - wastes fuel, stresses refractories, increases the power required for cement milling, and reduces cement strength. Sasaki & Ueda (ICR; 8/1989, pg 55) found a l4kcal/kg heat penalty for each 0.1% reduction in free lime. A convenient supplement for free-lime measurement is the more rapid determination of litre-weight. This involves screening a sample of clinker from the cooler discharge to approximately +5/-12mm and weighing a standard 1 litre volume. Litre-weight is typically 1100-1300g/L (varying inversely with free-lime) but the target range should be determined with a minimum equivalent to the established free-lime upper limit. Secondary air temperature should be as high as possible in order to recover the maximum heat; usually 800-1000°C. Maximizing secondary air temperature involves optimizing clinker bed depth and air volumes injected to the first cooler compartments. Note that secondary air temperature is difficult to measure unless there is a hotgas take off from the hood for tertiary or coal mill air; an unprotected

Cement Plant Operations Handbook • 25

c

z $20 0 0 1-1

thermocouple in the hood is liable to erroneous and misleading measurement due to radiation. Precalciner kilns maximize the heat input to the calciner and, typically, 60% of fuel is fed to the calciner while 40% is burned in the kiln. This serves to minimize the size of the rotary kiln and its heat loading; it does not reduce specific fuel consumption. It has been widely found that preheater kilns without precalciner vessels can still benefit from feeding 10-20% of total fuel to the kiln riser. Kiln operation is noticeably more stable and brick life is extended. This is also a useful means of consuming low grade fuels or waste materials. The limit to fuel injection at the riser depends upon its size and consequent gas retention time, and upon fuelair mixing characteristics; overfuelling results in preheater operating problems, an increase in exit gas temperature, and CO in the exhaust. The vortex finders (dip tubes) of lower stage cyclones were for many years prone to collapse and, frequently, were not replaced. A new segmented design in high-temperature alloy is now available for original installations and retrofits (WC; 10/1994, pg 39). However, the effectiveness of the vortex finders should be carefully assessed by review of preheater temperature and pressure profile and of specific fuel efficiency both before and after the tubes are removed or fall out; in many cases there is scant justification for re-installation. For kilns with grate coolers, the burner tip should be in the plane of the kiln nose (hot) or slightly inside the kiln providing it does not suffer damage from falling clinker. The burner should normally be concentric with, and on the axis of, the kiln. Some operators prefer to hold the burner horizontal and even tilted into the load. Such orientation may result in reducing conditions and must be adopted with caution. It should be appreciated that both burner position and tip velocity are intimately related to hood aerodynamics and can not be considered in isolation. Kiln rings are sections of heavy coating, usually in the burning zone, though sometimes also near the back of the kiln, which can grow to restrict both gas and material flow and eventually force shut-down. Conversely, ring collapse causes a flush of unburned material. Ring formation in the burning zone is commonly attributed to operational fluctuations though a low coal ash-fusion temperature or high mix liquid phase will increase the risk (Bhatty; Proc ICS; 1981, pg 110). Early

26 • Cement Plant Operations Handbook

detection is possible with a shell scanner and rapid reaction is essential. Such ring growth may be countered by varying kiln speed or by small movements (10cm) of the burner in and out. Rings at the back of the kiln are usually associated with the volatiles cycle, particularly excessive sulfur at the kiln inlet. It is evident, though of little help, that rings are structurally more stable in small diameter kilns. Recurrence merits an investigation of cause(s) (Hamilton; ICR; 12/1997, pg 53). Certain plants have raw materials which contain significant proportions of hydrocarbons (kerogens), typically up to 3%, or may wish to dispose of oil contaminated soils. If fed conventionally to the top of the preheater, the hydrocarbons will tend to distill at intermediate temperatures and exit with the flue gas — if they do not explode in the EP (Ryzhik; WC; 11/1992, pg 22). To prevent the resulting pollution, and to make use of the heat potential, kerogen containing materials should be injected at higher temperatures; usually to a 1- or 2-stage preheater if the hydrocarbons are present in the limestone. The high temperature exhaust may then be used for drying or for power co-generation (Onissi & Munakata; ZKG; 1/1993, pg E7). If the hydrocarbons occur in a minor constituent, this component may be ground separately and fed to the kiln riser. Petroleum coke, or the residual carbon in fly ash used as raw material, being involatile, can be added conventionally with kiln feed and yield useful heat (Borgholm; ZKG; 6/1992, pg 141). Some fly ash, however, contains high and variable carbon (1-30%) and, unless preblended, can seriously destabilize kiln operation.

4.2 Control Systems (by Andrew Jackura, PE) Hard wired controls have largely given way to computerized systems. Relay logic for discrete (on/off) control tasks has for many years been handled by programmable logic controllers (PLCs) which also now have capability for analog control. Distributed control systems (DCSs) have likewise replaced control systems once made up of numerous electronic or pneumatic analog loop controllers. Recently, personal computers (PCs) have become available as man machine interfaces (MMIs or, of course, WMIs) working on both PLC and DCS platforms. The differences between systems lie mainly in their architecture.

Distributed Control Systems comprise a proprietary computer and software that performs supervisory control and data acquisition (SCADA), proprietary multiloop controllers for running the analog and

Cement Plant Operations Handbook • 27

w

c

2

szo 0 0

discrete control alogrithms, proprietary input/output (i/o) modules that interface loop controllers with field devices (eg pressure transmitters, damper operators), and proprietary software running on standard PCs for the MMI. Almost all DCS vendors (eg Honeywell, Rosemount, Bailey) design redundancy into the SCADA system and to multiloop controllers which yields very high reliability. DCSs also come with high level programming software which automatically takes care of common programming tasks and greatly facilitates system configuration and maintenance. All major PLC suppliers (AllenBradley, Siemens, GE/Fanuc) offer controllers which interface with DCSs, and a common architecture for DCSs employed in cement plants uses integrated multiloop controllers for analog control and PLCs for discrete control; with some 80% of cement plant control loops being discrete, this uses DCS controllers only for the few analog loops which require them while using the less expensive PLCs for discrete control. Such interfaced PLCs continue to be favored for discrete control due to speed, ease of programming, and reliability. Open Distributed Control Systems comprise SCADA software running on standard PCs, proprietary software running on proprietary PLCs for performing analog or discrete control alogrithms, proprietary i/o modules interfacing PLCs with field devices, and proprietary software running on standard PCs for the MMI. While a standard PC is used for both MMI and SCADA tasks, compatible software from a single vendor is used. The primary advantage of the PC system is the ease and economy of upgrading speed and memory. However, while hardware costs are lower than with proprietary DCSs, programming costs are usually higher because automated high level programming software is not yet available. Also, to obtain the same level of redundancy, additional PCs must be incorporated. The present cost savings for a PC system may be 10-15% less than for DCS (Feeley; Control Magazine; 11/1997, pg 40). The current trend is for DCS and PC systems to converge. Both DCS and PLC vendors are moving away from proprietary hardware and software to more open systems while it is increasingly common to find control systems running on PC platforms with software performing

28 • Cement Plant Operations Handbook

multiloop control functions as well as MMI and SCADA (Walker; WC; 3/1996, pg 68).

w

Various expert systems (Linkman, Fuzzy Logic, etc) can be overlaid to the computerized control platform and can give dramatic improvements in kiln stability, fuel efficiency, clinker quality and, consequently, in production rate. However, since process response to controlled variables changes overtime, such systems require constant attention. Adaptive programs are now being developed.

z

4.3 Kiln Control Kiln operation is a complex art of which the principal control variables are: 1 Burning zone temperature (pyrometer or indirectly from kiln drive power or NOx). 2 Feed-end temperature 3 Feed-end oxygen

cits

Typical Aim 1500°C 1000°C 2.0%

Control is effected by adjustments to kiln speed, fuel rate and ID fan speed. Whether normal operation is manual or automated, most kilns are still liable to upset periods due to ring building, coating loss, etc and, while every effort should in any case be made to minimize such perturbations, stability is a prerequirement for effective computer control. Kiln feed and speed are usually controlled with a fixed linear relationship and unilateral variation of kiln speed should, at most, be used only as a short-term expedient (eg to control a kiln flush). It has been asserted that for many kilns speed should be kept constant in the upper range of feed rates (Clark; WC; 3/1994, pg 43). Kiln speed should be such that volumetric loading is within the range 10-13% (Section B5.10). Typically cyclone preheater kilns rotate at 22.5rpm (50-70cm/sec circumpherential speed) and have material retention times of 20-40mins. Precalciner kilns rotate at 3.5-4.5rpm (80-100cm/sec) for similar retention times. Retention in the preheater is 20-40secs. It has been asserted by Scheubel (ZKG; 12/1989, pg E314) that CaO, upon calcination, is highly reactive but that this reactivity decreases rapidly so that slow heating between 900-1300°C can result in increased heat of formation of cement compounds. Keeping the same kiln retention time with increasing degree of calcination of the material

Cement Plant Operations Handbook • 29

O

entering the kiln results in extending this transition and there is evidence that the introduction of short, 2-pier, kilns for precalciner systems has led to the reduction of material residence time before entering the burning zone from some 15 minutes to 6 minutes with resulting improvement in clinker mineralogy and grindability. Kilns are frequently operated to the limit of the ID fan. In this case, low oxygen must be corrected by reducing both fuel and feed. Precalciner kilns burn fuel at the kiln hood using combustion air mainly drawn from the hot end of the (grate) cooler, and in the calciner using combustion air drawn from either the hood or the mid- section of the cooler via a tertiary duct. Most precalciner kilns have dampers in the tertiary duct, and some have dampers in the riser, to control relative air flows to the two burners in order to maintain the desired fuel split. Frequently these dampers fail and it is then essential to adjust the fuel flows to the actual air flows. This is effected by maintaining oxygen at the kiln feed-end at, say, 2%. The gas probe at the kiln feed-end should project inside the kiln to avoid the effect of false air inleakage at the kiln seal; this is a difficult location for gas sampling and an adequate probe is essential (ICR; 6/1995; pg 51). CO should, and NOx may, also be measured at the kiln entrance. The oxygen level required at the kiln inlet will depend upon kiln stability and combustion efficiency; with a good flame, 1-2% 02 should result in less than 100ppm CO while an unstable flame may yield in excess of 0.1% CO with 3% 02. In a cyclone preheater kiln without riser firing, the downcomer oxygen analyzer serves both as back-up to the kiln inlet unit and to monitor air inleakage across the tower; an increase in 02 of more than 2-3% suggests excessive inleakage. In a precalciner kiln, an additional gas analyzer may be installed in the exhaust duct from the bottom cyclone and, again, this should be operated at as low an oxygen level as is consistent with less than 100ppm of CO. Useful information on kiln operation can be obtained from frequent (2-hourly) analysis of clinker for 503, and periodic (8-hourly) sampling of the underflow from the bottom cyclone stage(s) for SO3 and alkali determination. Normal SO3 levels (typically 0.6% in clinker and 2-3% in underflow) should be determined and maintained. In precalciner kilns, retention time and heat loading are particularly low and alkalis (K,Na)

30 • Cement Plant Operations Handbook

tend to pass through to clinker while sulfur is volatilized and builds a cycle at the back of the kiln exacerbated by the deficiency of alkalis. If the kiln is burned too hot, this cycle increases excessively until build-up or cyclone plugging occurs. This is matched by an abnormally low 303 content in the clinker. Eventually, if the kiln is allowed to cool, this sulfur is released and transient high clinker SO3 results. Such variation in clinker SO3 will also give rise to varying grindability in the finish mill. It cannot be overemphasized that kiln stability, fuel efficiency, finish grinding power consumption, and cement quality all depend greatly upon the provision of a kiln feed and fuel with minimal variation both of chemistry and feed rate. Healthy suspicion should be nurtured towards both instrument signals and manually reported data. Particular areas for mistrust are: ■ False instrument signals of which pressure sensors and gas sampling probes are particularly liable to failure. ■ Short term variations masked by electronically damped signals. ▪ Feeder variations especially when the material is either sticky or fine and dry. • Chemical variations hidden by faulty analytical methods, statistical mistreatment, or outright fraud. Variations in kiln behavior always have a cause; any variations which cannot be explained by observed feed deviation or known operational disturbance should alert to the possibility of faulty data. For management scrutiny it is useful to have either a "read-only" CRT which can be interrogated without interfering with kiln operation, or a strip chart recorder which minimally shows kiln feed, kiln speed, and kiln drive amps. Automated kiln control seems, unfortunately, to have reduced operators' habits of looking in the kiln and inspecting the clinker produced. Modern kiln and cooler camera systems, however, are excellent tools (Prokopy; RP-C; 5/1996, pg 38) for observing flame shape and position of the load in the kiln (dark interface of unburned material), "snowmen" (build-up on grates below the hood), "red rivers" and excessive blow-through in the cooler. The appearance of clinker can also be instructive; preferably black with surface glitter, dense but not dead burned, absence of excessive fines, dark grey cores. Brown cores are

Cement Plant Operations Handbook • 31

a

m cto

0 0

usually due to reducing conditions in the kiln but can also be due to the decreased permeability of clinker resulting from high belite and sulfate concentrations which inhibit oxidation of ferrous (Fe 2+) iron to ferric (Fe3+) during cooling. This in turn is due to chemical variation of kiln feed and to low volatilization of sulfur in the kiln (Scrivener & Taylor; ZKG; 1/1995, pg 34). Other causes have also been proposed (Jakobsen; WC; 8/1993, pg 32). Brown clinker is associated with increased heat consumption, reduced grindability, cement strength loss, and rapid setting. Certain alarms on the kiln system are critical. Apart from normal mechanical alarms and the routine monitoring of kiln shell for refractory failure, the potential for explosion requires particular care. Gas analysis is conventional at the feed end of the kiln, at the downcomer, and at the dust collector entrance. CO above 1% should cause alarm, and above 2% should cause fuel, and EP if so equipped, to shut off. Flame detection is particularly vital during warm up of the kiln and fuel should be shut off by interlock if the flame is lost. When the kiln is up to temperature it is common to deactivate the flame detector but it should be impossible to start a kiln without this protection. The light-up of kilns is potentially dangerous as there is insufficient temperature in the system to ensure continuous combustion. Unburned fuel accumulates rapidly in the kiln and, if then ignited, will probably explode. It is important that ignition be achieved as soon as the fuel is injected and, if the flame fails during warm-up, the kiln should be purged with 5 times the volume of kiln, preheater, ducting, and dust collector (probably some 3-5 minutes) before reignition is attempted. A simple and reliable ignition system has been described by Davies (ICR; 9/1996, pg 77).

4.4 Kiln Start-up and Shut-down Detailed schedules should be provided to operators to ensure that what, one hopes, are infrequent occurrences do not result in undue stress to kiln components. follows agreement by production and maintenance management that all work is completed, that all tools and materials have been removed, and that all doors are closed. Work may, with discretion, continue in the cooler during warm-up but no workers should remain in

Warm up -

32 • Cement Plant Operations Handbook

Figure 6

Cement Plant Operations Handbook • 33

the cooler at the time of ignition. Commonly, warm-up from cold takes 24 hours from ignition to feed-on, but may be increased if extensive refractory work requires curing. A typical chart is shown (Figure 6) indicating the desired rate of increase in back-end temperature (this may also be set out in terms of fuel rate), the kiln turning program, the introduction of feed (usually 50% of full rate), and the increase of fuel, speed and feed to normal operation which should take another 8 hours from feed-on. For PC kilns, fuel is supplied to the calciner at the same time as, or soon after, feed-on. ID fan should be operated to approximately 10% 02 at the back of the kiln to feed-on whereupon the normal 02 target is adopted. For coal fired kilns, warm-up almost invariably employs gas or oil with switch-over to coal at the time of feed-on. If the coal mill uses hot gas from the cooler, there may be a delay before heat is available from the clinker. Before and during warm-up, equipment checks should be performed to ensure that each unit is ready to operate when required. Warm-up from shorter stops where the kiln is still hot, say stops of less than 24 hours, are conventionally accelerated to half the shutdown time.

Shut-down may be either: Emergency, in which case all equipment up-stream of the failure must be stopped immediately, or Controlled, in which case feed bin and coal system should be emptied, the kiln load run out as far as possible, and the cooler emptied. The burner pipe is withdrawn, or cooling air is continued through the burner, and the kiln is rotated on a standard schedule for about 12 hours until cold with the ID fan running at reducing speed.

• •

Suggested inching is as follows: 0 - 2 hours — continuous 2 - 4 hours — 1/4 turn every 15 minutes — 1/4 turn every hour 4 -12 hours If the shut down is for less than a day and does not involve entering the kiln or preheater, then heat should be retained either by stopping the ID

34

■

Cement Plant

Operations Handbook

fan immediately and shutting the preheater dampers after 2 hours, or (if there are no dampers) shutting down the fan after 2 hours.

4.5 Kiln Refractories A typical arrangement of brick types and Refratechnik's reported "average best service lives" in Japanese cyclone preheater kilns (without precaliners) is as follows: Discharge - 1 D 1D - 8D 8D - 10D 10D

-

feed end

8 months 70-85% alumina Basic, dolomite, or spinet 6-10 months 70% alumina 21 months 40% alumina 21-37 months

Kilns with precalcination average significantly longer brick life. A detailed historical record of refractory replacement and review thereof are important to minimize cost and service interruption. Typically, brick from the kiln nose to the back of the high-alumina brick section should be replaced if found to be 10cms or less in thickness, but such a rule-of-thumb is subject to much variation depending upon operating considerations. A useful practice is to drill through the brick every meter whenever the kiln is down and coating has been stripped (wider spacing and lesser frequency is adequate in the low alumina brick area). Such drilling requires discretion to locate the shell and to identify irregular circumpherential wear. Non-intrusive instruments to measure brick thickness are also available (eg Hoganas Linometer). The extent of coating should be observed whenever the kiln is entered and, roughly, basic brick should extend back to the top of the coated zone. Changes in fuels, feed, or burning conditions will affect the location of the burning zone. Coating location and refractory condition are usually monitored during operation with a shell scanner (eg Wulff; ZKG; 11/1993, pg E300). Kiln shells should also be inspected visually, particularly under tires where small hot spots may be concealed from the shell scanner. Warm areas of shell can be controlled by use of a fixed fan array or of movable fans which can be directed at the area. "Red spots", when the kiln shell reaches incandescence, should always be a cause for alarm and should not be allowed to persist for any length of time. If the hot

Cement Plant Operations Handbook • 35

spot is a dull red and is in the burning zone it may be possible to re-coat the area and continue operation. Specifically, a small sharp hot spot, relating to the loss of one or two bricks, occurring in the burning zone can be "repaired" by stopping the kiln for 2-5 minutes under the load with an air lance cooling the spot. However, response must be rapid and the long-term problems caused by warping of the shell should always be born in mind. Red spots on surfaces other than the kiln may be temporarily secured by building a steel box on the outside to cover the hot area and filling the box with castable refractory; the box should be cut off and permanent repairs effected during the next kiln shutdown. There is an extensive literature on kiln brick types and performance (eg Scheubel; ZKG; 1/1994, pg E22: Wright; WC; 12/1994, pg 2). Brick usage averages 850g /t of clinker produced for cyclone preheater kilns and 500g/ t for precalciner kilns (Scheubel & Nactwey; ZKG; 10/1997, pg 572). With considerable variation, installed brick thickness is related to internal kiln diameter: 5.2M

250mm

and brick specific gravities are approximately: Magnesite

3.05

70% Alumina

2.30

Spinel

2.95

40% Alumina

2.05

Dolomite

2.80

The two major bricking techniques are the epoxy method and the "ring-jack" method (Mosci; Brick Installation in Rotary Kilns; RefrAmerica 1995). Both have their place; the ring-jack is usually faster for long installations but does not allow turning of the kiln which may be important if other maintenance is to be performed on the shell, drive, or seals. Typically, installation after clean-out is at the rate of 0.5M/hour. In addition, monolithics, which comprise castable and plastic refractories, have various uses from the rapid gunning of large areas or

36 • Cement Plant Operations Handbook

complex shapes to the molding of burner pipes and distorted kiln nose rings (Fraser; Proc IKA; Toronto; 1992). Castables are concretes with refractory aggregate and a high-temperature resistant (high Al203) hydraulic binder. Castables may be "heavy" or "lightweight insulating" and are classed: ■ standard (>2.5% CaO) ■ low cement (1.0-2.5% CaO) ■ ultra-low cement ( sB) is equal to, or less than, the corresponding F value given in tables for a specified probability, then identity can be deduced.

F

X2 test (chi squared)determines fit of data to theoretical distribution. For present purposes this will be confirmation of normal distribution about a mean. Data are collected into sub- groups and the percentage of observed values falling within each sub-group (0) is compared to theoretical probabilities (E), eg: sub-group range

observed (0)

+la

theoretical (E) 15.9% 15.0 19.1 19.1 15.0 15.9

Then X2 = E (O-E) 2 /E is compared to tables to determine probable fit for the given degrees of freedom (for 6 groups, (0=6-1=5). Then X 2 should normally be less than 11 and greater than 1; greater than 11 indicates poorer than 5% probability of fit while less than 1 suggests too good a fit for natural data. Other numbers of sub-groups, other probability levels, and distribution other than Gaussian can be used. Comprehensive X2 tables are then required.

180 • Cement Plant Operations Handbook

Linear regression is the best fit of xy data to a line: y = MX C where: m = s xy /sx2 c = y - sxy .x/s2 sxy = E(x-x)(y-y)/n Covariance is a measure of the inter-dependence of two variables; independence gives a correlation (r) of 0, and perfect dependence gives 1 or -1. r = sxy /sx.sy

"t" Tables 0 1 2

80% 3.078 1.886

90% 6.314 2.920

3 4 5 6

1.638 1.533 1.476 1.440

7 8 9

1.415 1.397 1.383

2.353 2.132 2.015 1.943 1.895 1.860 1.833

10 20 30 40 120 =

1.372 1.325 1.310 1.303 1.289 1.282

1.812 1.725 1.697 1.684 1.658 1.645

95% 12.706 4.303 3.182 2.776 2.571 2.447

98% 31.821

2.365 2.306 2.262

6.965 4.541 3.747 3.365 3.143 2.998 2.896 2.821

2.228 2.086 2.042 2.021 1.980 1.960

2.764 2.528 2.457 2.423 2358 2.326

99% 63.657 9.925 5.841 4.604 4.032 3.707

99.9% 636.62 31.598

3.499 3.355 3.250

12.941 8.610 6.859 5.959 5.405 5.041 4.781

3.169 2.845 2.750 2.704 2.617 2.576

4.587 3.850 3.646 3.551 3.373 3.291

Normal Curve - Percentage of Values Within (tx6) of Mean t % (two tails) % (one tail) 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1.0 1.5 2.0 3.0 4.0

3.98 7.93 11.79

7.96 15.86 23.58

15.54 19.15 22.58 25.80 28.81 31.59 34.13 43.32 47.73 49.87

31.08 38.30 45.16 51.60 57.62 63.18 68.27 86.64 95.45 99.74

50.00

100.00

Cement Plant Operations Handbook • 181

"F" Tables (90% probability) 2 sA/sB 1 2 18.5 19.0 3 10.13 9.55 4 7.71 6.94 5.99 5.14 6 5.32 4.46 8 10 4.96 4.10 20 4.35 3.49 40 4.08 3.23 3.84 3.00 (98% probability) 2 1 SA/SB 2 98.5 99.0 3 34.1 30.8 4 18.0 21.2 6 13.74 10.92 8 11.26 8.65 10 10.04 7.56 20 8.10 5.85 40 7.31 5.18 6.63 4.61

3 19.2 9.28 6.59 4.76 4.07 3.71 3.10 2.84 2.60

3 99.2 29.5 16.7 9.78 7.59 6.55 4.94 4.31 3.78

4 19.2 9.12 6.39 4.53 3.84 3.48 2.87 2.61 2.37

6 19.3 8.94 6.16 4.28 3.58 3.22 2.60 2.34 2.10

8 19.4 8.85 6.04 4.15 3.44 3.07 2.45 2.18 1.94

10 19.4 8.79 5.96 4.06 3.35 2.98 2.35 2.08 1.83

24 19.5 8.64 5.77 3.84 3.12

19.5 8.53 5.63 3.67 2.93

2.74 2.08 1.79 1.52

2.54 1.84 1.51 1.00

4 99.2 28.7 16.0 9.14

6 99.3 27.9 15.2 8.47 6.37 5.39 3.87 3.29 2.80

8 99.4 27.5 14.8 8.10

10 99.4 27.2 14.5

24 99.5 26.6 13.9 7.31 5.28 4.33 2.86 2.29 1.79

7.01 5.99 4.43 3.83 3.32

6.03 5.06 3.56 2.99 2.51

7.87 5.81 4.85 3.37 2.80 2.32

"X2" Tables tp

99%

95%

1 2 3

0.02 0.11 0.30 0.55 0.87 1.24 1.65 2.09 2.56 8.26 22.16 37.48 70.06

4 5 6 7 8 9 10 20 40 60 100

0.10 0.35

10% 2.71 4.61 6.25

5% 3.84 5.99 7.81

0.71 1.15 1.64 2.17 2.73 3.33 3.94 10.85 25.51 43.19 77.93

7.78 9.24 10.64 12.02 13.36 14.68 15.99 28.41 51.81 74.40 118.50

9.49 11.07 12.59 14.07 15.51 16.92 18.31 31.41 55.76 79.08 124.34

••• 182 • Cement Plant Operations Handbook

1% 6.63 9.21 11.34 13.28 15.09 16.81 18.48 20.09 21.67 23.21 37.57 63.69 88.38 135.81

99.5 26.1 13.5 6.88 4.86 3.91 2.42 1.80 1.00

The two standard explanations of cement setting and strength development are Le Chatelier's interlocking crystal theory and Michaelis's colloidal gel theory (Benstead; WC; 8/1991, pg 21: Chen & Odler; Cem and Conc Res; Vol 22; 1992, pg 1130). Aluminate and, particularly, silicate hydration reactions are extremely complex, and many undoubtedly contribute to the setting and strength gain of cement. The subject is described in some detail by Lea (The Chemistry of Cement and Concrete). C3A is the most soluble of the major cement compounds and appears to dominate early hydration. C3A reacts with sulfate in solution to give ettringite (C3A.3CaSO4.32H20), the interlocking crystal growth of which may contribute to setting. Ettringite later converts to monosuiphoaluminate (C3A.CaSO4.12H20). In the absence of sulfate, C3A hydrates to C4A.14H20 and C2A.8H20 which crystallize to cause flash set. This explains the importance of gypsum as set retarder. However, it has been observed that setting is largely independent of C3A concentration and it is now believed that both setting and strength development are largely caused by hydration of C3S to tobermorite (C3S2.4H20), a gel of variable composition.

saA

11 HYDRATION OF PORTLAND CEMENT

0

O

Hydration of cement typically involves combined water of about 22% relative to clinker weight. The relative contributions to strength development are shown in the following diagram: Compressive strength development of individual Cement Compounds

rol

m Pozzolanic activity comes from the reaction of soluble Si02 from the pozzolan with CaO in solution. As free CaO will always be present in solutions in contact with hydrated cement, pozzolanic reactions provide "self-curing" of cracks in pozzolanic concrete.

••• Cement Plant Operations Handbook • 183

rn

184 • Cement Plant Operations Handbook

12 OTHER KILN TYPES

0

LONG WET, LONG DRY, GRATE PREHEATER & VERTICAL SHAFT.

Typical comparative data, with considerable variation, is as follows:

Shaft kiln Long wet kiln Long dry kiln Lepol kiln Cyclone preheater kiln Precalciner kiln

Maximum rating

Specific fuel

(t/d) 200

(kcal/kg) 900-1000 1200-1500 900-1200 800- 900 800- 900 700- 850

2,000 2,000

2,000 2,000

11,000

Length:Diameter

32 - 38 32 - 38

14-16 14-16 11-16

Cement Plant Operations Handbook • 185

'73

r 11 0 11V, r-3_Tra --.1

The earliest cement kilns were vertical shafts in which mixtures of raw materials and solid fuel were burned in a natural draft of combustion air. Ransome introduced the rotary kiln in the 1880s and this allowed more uniform heat transfer and controlled clinker burning. Initially rotary kilns used slurry feed - the wet process - as this facilitated raw material grinding and homogenizing. In certain areas of the United States, shortage of water led to a variant, the long dry kiln, which required, and resulted in, improved pneumatic blending systems. It was acknowledged that, while the rotary kiln was an excellent device for heat transfer and materials handling at clinker burning temperatures, it was inefficient for preheating and calcination. The first alternative approach to preheating was the Lepol, or grate preheater, system where nodulized raw materials are conveyed on a travelling grate permeated by hot kiln exhaust gas; with appropriate raw materials this process is successful. Ultimately, however, it was determined that the most efficient low temperature heat exchange and calcination can be effected in air suspension and this led first to the cyclone preheater and later to the addition of separately fueled precalcination. These last two systems now predominate and have been the substance of this book.

It has also been observed that the grindability of clinker differs significantly with kiln type. Relative power consumptions for clinker types are: Lepol kiln Cyclone preheater kiln Long wet kiln Long dry kiln

100 (softest) 107

112 117 (hardest)

12.1 Long Wet Kiln Long wet kilns were predominant until the appearance of cyclone preheaters in the 1950s. They are now obsolescent though they may still justify their existence where they are fully depreciated, where the market demands only a small production capacity, and where fuel is cheap. Wet kilns also avoid the need for drying of naturally wet raw materials and the homogenizing of slurry is still usually more effective than the blending of dry raw meal. Raw materials are milled with addition of, typically, 30-35% water by weight, to form a slurry which is stored and blended in tanks with continuous agitation (rotating rake augmented with air jets) before feeding to the kiln. Water is adjusted to produce a consistency which allows ease of conveying without segregation. As evaporation of the water involves a considerable heat penalty, use of water reducing agents may be justified; 1% water reduction is equivalent to about 15kcal/kg clinker. An approximate correlation of slurry density to water content is: 30% water by weight 32% 34%

= = =

1220kg/M 8 1160 1100

To enhance evaporation of water by increasing surface area for heat exchange, to facilitate the handling of feed as it transitions from slurry through plastic material, and to detrain dust from kiln exhaust gas, chain systems are hung within the kiln shell (Figure 10). A typical system would comprise one to two diameters of bare shell followed by one diameter of curtain chains as a dust curtain (curtain chains are lengths of chain about 75% the diameter of the kiln and attached at one end only in successive circles around the circumference of the shell). Next come some five diameters of spiral curtain chain to break up and convey the drying (plastic) feed down the kiln (Figure 10). While curtain chains are

186 • Cement Plant Operations Handbook

easier to manage, garland chains have been claimed to give better efficiency (garland chains are attached to the shell at both ends; the attachments should be 90° apart in a spiral down the axis of the kiln and the chain should hang slightly below the center line). Usually there is a second section of bare shell near the down-hill end of the chains to reduce circumpherential imbalance in gas temperature and material conveying. Duda recommends chain design parameters of 12% of daily clinker production for total chain loading, and surface area of 6-8.5M 2/M3 of chain section volume; de Beus (ICR; 12/1997, pg 41) suggests 15% and 610M2 /M3 respectively. Chain consumption is about 100g/ t clinker.

0

The feed material leaves the chain section at 5-10% moisture and proceeds to the preheating, calcining, and burning zones of the kiln. Total material retention time in a long wet kiln is approximately 3 hours and gas is discharged at 150-200°C.

z

Dust loss with exhaust gas should ideally be 8-10% but is often much higher as kiln production is increased with resulting increase in gas velocity. Return of dust to the slurry system is inadvisable as it frequently causes agglomeration and sedimentation. Up to 5% relative to clinker weight can be returned by insufflation into the kiln flame; beyond this quantity, flame cooling becomes unacceptable. Alternatives are separate slurrying in a vortex mixer and parallel injection with the main feed, and return using scoops which inject the dust slightly downhill from the chain section or into a bare section of kiln near the downhill end of the chains. The basic causes of high dust loss, however, are gas velocity and chain design and condition.

12.2 Long Dry Kiln Long dry kilns differ from wet kilns primarily in raw grinding and handling and in their lower specific fuel consumption. Within the kiln itself, dry kilns use only curtain chains as the requirement is for heat exchange and dust detrainment rather than for conveying. Usually, 6-7 diameters of curtain chain are employed below about 2 diameters of bare shell at the feed end; approximately half is hung in rings perpendicular to the kiln axis and the lower half is hung in a spiral arrangement. Chain loading is some 10% of daily kiln production. The gas discharge temperature of long dry kilns is typically in excess of 300°C and, if available, water is sprayed into the feed end to reduce gas temperature before dust collection.

Cement Plant Operations Handbook • 187

I

m

m

In both types of long kiln, chain fires can have catastrophic results and are at risk if the chain temperature exceeds about 1200°C under oxidizing conditions; corrective action should be taken if chain temperature reaches 1000°C.

12.3 Lepol (Grate Preheater) Kiln Polysius introduced this system during the 1930s and achieved a dramatic reduction in specific fuel consumption from the wet process. Nodulized feed is conveyed an a travelling grate through which the hot kiln exhaust gas is passed, originally once but, in a later development, twice. The material was preheated to approximately 900°C before entering the kiln while the exhaust gas was cooled to below 150°C, humidified for dust collection, and filtered by the material bed to a low dust concentration. Raw materials may be either wet milled and filterpressed to yield a cake of about 20% moisture, or dry milled and nodulized in an inclined rotating pan with a water spray to a moisture content of 11-15%. If the cake or nodules do not possess good mechanical and thermal stability — usually associated with clay ingredients there is excessive disintegration on the grate and loss of efficiency. The second (low temperature) exhaust gas pass through the grate dries and preheats the material. The first pass involves an initial gas temperature of about 1000°C and a final temperature below 500°C; this serves to condense volatiles exhausted from the kiln and it was found that if the gas were run through cyclones between the first and second pass, the collected dust contained a high concentration of volatiles, thereby providing an effective bypass. The material is discharged to the kiln inlet at incipient calcination and the short kiln thereafter operates similarly to a cyclone preheater kiln (Figure 10).

12.4 Vertical Shaft Kilns Shaft kilns originally constituted the only available technology from the beginnings of lime burning which can be traced at least to Greece in the 5th century BC. Since the beginning of the 20th century they have been largely superseded by rotary kilns. However, there remain areas where, due to lack of infrastructure, very small production units are appropriate and where relatively simple construction methods do not demand high cement quality. Such conditions can still favor shaft kilns and many are to be found in China, India, and in a number of developing countries.

188• Cement Plant Operations Handbook .

Figure 10

CURTAIN CHAIN

GARLAND CHAIN Gas duct Nodulizer

Grate

7-1: 1

Gas out

Atz „I

Kiln

Cooler

Bulkhead

GRATE PREHEATER KILN

Rotating discharge grate

Air injectors

VERTICAL SHAFT KILN

Cement Plant Operations Handbook • 189

Traditional shaft kilns were basically holes in the ground using mixtures of unground feed roughly mixed with solid fuel and burned in batches with natural draft. The lack of feed homogeneity together with non-uniform ventilation gave rise to widely varying temperature and oxidizing conditions so that quality was low and erratic. Considerable advances have been made and Rajbhandari (WC; 1/1995, pg 65) describes Spohn's black meal process as one of the most advanced shaft kiln technologies presently available (Figure 10). Practical unit capacities are 20-200t/d. Raw materials and solid fuel are ground together and nodulized (black meal process). Alternatively, but less effectively, raw mix and fuel can be ground separately and then blended and nodulized (white meal process). As with the Lepol kiln, stable nodules are important and usually require both a clay component and a solid fuel with less than 16% volatile content. The fuel may be coal, charcoal, coke, or petroleum coke. The kiln shaft is filled with the prepared mix and air is blown into the shaft at and near the base. The material is in turn heated, calcined, and burned at progressively higher temperatures as it moves down the shaft countercurrent to the combustion air. Near the base the clinker, with fuel already consumed, is rapidly cooled by the injected air and is discharged through a gate. Production is continuous with new feed added at the top to balance discharge. The process is, therefore, basically similar to that of rotary kilns. The principal difference is in uniformity; the rotary kiln ensures that the material is constantly agitated and that all material is subject to the same retention time and heat transfer. In the shaft kiln, however, there is a definite thermal gradient with the core material reaching a maximum temperature of ca 1450°C, some 200° higher than material at the walls. Differential melting of the material tends to increase air flow velocity at the walls which reinforces heat loss though the walls to exacerbate the difference. More sophisticated shaft kilns can compensate for this with increased peripheral fuel concentration and reduced wall heat loss. Retention time above 1250°C is typically 30 minutes. Increasing the air flow through the bed both increases production rate and clinker quality; the necessary air injection pressures (1000-2500mm WG) require an efficient air lock on the clinker discharge; either a triple gate or a controlled choke flow (seal leg). In practice, the seal leg is too dependent upon clinker bulk density and porosity to be effective and

190 • Cement Plant

Operations Handbook

the triple gate is preferred. In recent years there has also been a trend to increased diameter and reduced height to an aspect ratio of 2.5-3.0 but the diameter is usually limited to about 2.4M as increasing diameter makes uniform air distribution more difficult to achieve. Modern shaft kiln designs can be fully instrumented and PLC controlled.

•••

Cement Plant Operations Handbook • 191

192 • Cement Plant Operations Handbook

13 TECHNICAL & PROCESS AUDITS A technical audit is a review of all units of plant equipment to determine their operating capacities and efficiencies with relation to original design and to subsequent modifications. Current performance is compared to what is realistically achievable and for this purpose "rated" or guarantee capacities should not be taken uncritically. The purpose is to establish the efficiency with which the plant is being operated, to identify bottlenecks and to propose means of eliminating them (McLeod; ICR; 4/1996, pg 49). A process audit is an analysis of plant operations with the purpose of increasing production, reducing unit cost, and improving product quality by establishing and controlling optimum mix parameters. This program must be implemented on a comprehensive basis rather than by trouble- shooting individual deficiencies. Elements include but are not limited, to: ■ Establish optimum mix design with respect to cost, handling, grindability, burnability and product quality ■ Produce constant chemical and physical kiln feed characteristics ■ Operate kiln with minimum process variations and optimum free lime ■ Finish grind to a product having mid-market concrete strength with minimal variation (market position is, of course, a subject for management judgement but it is suggested that policy be based upon concrete performance, not that of mortar) A considerable amount of trial and error may be involved in establishing an optimum mix design but there are a number of rules of thumb: 58-65 (clinker basis) C3S LSF 5/R Liquid (1450°C)

0.92-0.97 2.2-2.6 23-26%

Large particle quartz in raw materials should be avoided if at all possible as grinding to reactive size, say ..... 0 4 j j

0 0. ..^.. .1, :.' co .— _. .0 M 0

h.,

.•

N

_.. N m cp 61) (.0

E

-Ti

co °

0,

CO''

3

0. o

Fo

CD M

OP CO "I ch (0

a CO C. Rj 0 Ca P" CO -• P ,,,.. p.) po _.. .P. _9 h.) i,„ —I c) 0r, .-4 15 cr M to ‘ v .4 IN cn co —• 0"

th

Ca

Ca -4 F. > E,' o 1.. CD 0, 2 co co L., „, b Z v zo -0 ." v) ,_. co a -.• N

00 in

Co n

TS 1:1

3 0

17; -4 ....1 a, U/ CD " bp o (1) -P C,

r.a.

b cn

1

F.

W Cn 0 co 0 °5 co

0, Fil to ,,, 5. 0 -L-4 PI° 11 U::. 43 C.n0 -I Ln A al ..■ oa 04.

x„ , .

0 ■-l• th g A

8

ca

-4 co 0") c,,

-,

T.2 1—

-4

A

Co

.

A

Co

-

Ca

ro

4..

Z— b 2 -L 2 ;.,, 7:. co N 33I 0, K.: X a. !-, pc co 10 w 0 , eri, 0 a (0 = cr■ E., m 4, zo, .., C0 t ■,, —

co --

Cement Plant Operations Handbook • 215

Kl

INDEX Page Abrasion resistance 145 Acoustic horns 50, 76 Addresses, organizations 214 Air distribution 142 Alarms 122 Alkali, S and Cl limits 39 Alkali-silica reactivity 104 Alkali:sulfur ratio 4 Audit, process 193 Atmospheric pressure 174 Bag filters 75 Bibliography 212 Blended cements 95 Blending 17 Blending ratio 11 Brinell hardness 176 Brown clinker 31 Budget, capital 107 Budget, operating 115 Bulk densities 171 Burnability factor 157 Cash flow 131 Cement big-bags 73 Cement bulk loading 73 Cement capacities, new 177 Cement comparative testing 89 Cement compounds & ratios 157 Cement dispatch 71 Cement hydration 183 Cement intergrinds 96 Cement milling 61 Cement packing 72 Cement particle size 67 Cement production, world 176 Cement storage 70 Cement strength 92 Cement types & specs 92 Chemical analysis 85

Page Chi squared tables 182 Chi squared test 180 Circulating load 151 Clinker cooling 48 Clinker grindability 62 Clinker heat of formation 158 Clinker storage 61 Coal data 167 Coal firing 44 46 Coal firing systems Concrete problems 100 Conversion tables 172 Conveying, belt 148 Conveying, bucket elevator 148 Conveying, drag chain 148 147 Conveying, FK pump Conveying, pneumatic capsule 149 149 Conveying, sandwich belt 148 Conveying, screw 148 Conveying, tube belt Conveying power 147 Cooler efficiency 162 Corrosion, kiln 58 Cost accounting 115 181 Covariance Co-generation 137 Crushing 6 Cyclones 75 Definitions Preface Delayed ettringite formation 104 Depreciation 207 Detached plumes 78 Dew point 145 Dioxins 81 126 Downtime reporting Drying, raw material 8 Drying, raw mill 13 Dust collection 75

216 • Cement Plant Operations Handbook

Dust suppression 78 Earth, elemental abundance 175 Electricity, 3-phase 137 Electrostatic precipitator 75 Emergency power 58 Equipment numbering 122 F tables 182 F test 180 False air 144 False set 67,104 Fan build-up 142 Fan laws 139 Flux, kiln feed 3 Fly ash 97 Free lime 25 40 Fuels Fuels data 167 Gas conditioning 76 Gas data 142 168 Gas fuels data Geometrical formulae 173 Greek alphabet 174 Grindabilities 171 Grinding aids 67 Hardness 176 Horomill 65 Insufflation 38 Interlocks, jumpered 122 Investment justification 117 ISO 9000 98 27 Kerogens Kiln, Lepol 188 Kiln, long dry 187 Kiln, vertical shaft 188 Kiln, long wet 186 Kiln alignment 57 Kiln burning 23 Kiln burning temperature 158 38 Kiln bypass

157 Kiln coating tendency 29 Kiln control 162 Kiln drive power 55 Kiln drives 162 Kiln exhaust gas 18 Kiln feed 19 Kiln feed:clinker ratio Kiln gas analysis 30 160 Kiln gas velocities 159 Kiln heat balance 161 Kiln loading, volume 51 Kiln mechanical 35 Kiln refractories 161 Kiln retention time 26 Kiln rings & build-ups 53 Kiln seals Kiln shell. 53 161 Kiln slope 194 Kiln specific fuel 29 Kiln speed 32 Kiln start-up & shut-down Kiln types 185 175 Laboratory reagents 171 Linear expansion coefts. 181 Linear regression 107 Maintenance 97 Masonry cement 82 Metals, toxic 90 Microscopy 153 Mill, Grace factor 63 Mill air flow 155 Mill ball size 154 Mill ball weights 152 Mill charge loading 155 Mill charge wear 152 Mill critical speed Mill diaphragm 64,155 154 Mill power 153 Mill retention time

Cement Plant Operations Handbook • 217

Mill specific power 65,196 Mobile equip. maintenance 113 Normal curve area 179 NOx 79 Oil fuels data 168 Oilwell cement 92 Opacity 78 Operators' log 122 Orifices 143 Packset 104 Periodic table End Petroleum coke 42, 167 pH & normality 174 Physical testing 89 Pitots 143 Plant assessment data list 197 Plant construction cost 208 Plant reporting 125 Plant valuation 205 Pollution control 75 Power conservation 136 Power consumption 136 Power generation 137 Power tariffs 127 Preblending 9 Preheater cleaning 40 Preheater configurations 22 Process flow Profit & loss 119 Project cost estimating 118 Quality control 85 Quarry operations 4 Raw material sampling 88 Raw materials 3 Raw milling 13 References 211 Reliability centered main. 109

Report, daily production 128 Report, downtime summary 132 Report, process summary 129 Reserves 5 Review papers 211 Rockwell hardness 176 Roll press 64 Roller mill, cement 64 Roller mill, raw 13 Safety lock-out 122 Sea water composition 175 151 Separator efficiency Separator types 68 Shift supervisor 122 Ship capacities 175 Sieves 151 121 Silo maintenance 97 Slag SO2 81 145 Spray cooling 145 Stack draft 179 Standard deviation 179 Statistics Stockpile inventories 121 Sulfur cycle 30 "t" tables 181 Trigonometric tables 173 143 Venturis Volatile cycles 38,164 116 Warehouse Waste fuels 43 Waste raw materials 4 149 Water pump Water treatment 105 121 Weigh Feeders X-ray diffraction 87 85 X-ray fluorescence

••*

218 • Cement Plant

Operations Handbook

Tradeship Publications Lt Old Kings Head Con 15 High Strel Dorking, Surre RH4 1A United Kingdoi Tel: +44 (0) 1306 740363/3E Fax: +44 (0) 1306 7406E E-mail: tradeship©dial.pipex.cm Website: www.cemnet.co .0