Waste Heat Boiler Deskbook by V. GanapathyContents Preface. Acnowiedgemens spor Brequntly Used Abbreviations. iiotactas xiii CHAPTER 1: Heat Recovery Systems HRSGs for gas turbines Natural versus forced circulation boilers. Auxiliary firing Computing fuel requirements. HIRSG system efficiency. Fresh air firing... eRe design Features). Tinned surfaces and design. Steaming economizers... Emissions of NOx and CO... Methods of reducing pollutants. Bypass dampers. Recent trend.nnnnmen STIG and Cheng cycle systems. Enhanced oil recovery applications Reciprocating engine heat recovery... Hydrogen plant waste heat boilers... Boilers for sulfuric acid plant Incineration and heat recovery Solid waste incineration, RDF firing. Fluid bed combustors for MSW Hazardous waste incineration .. High temperature corrosion.Heat recovery boilers. Incineration of wood wastes, tires Incineration of liquids, fumes, VOCS CHAPTER 2: Fire Tube Boilers. Guidelines for fire tube boilers.. Design procedure. Determination of tube side coefficient. Example of design. “Effect of tube size on design. Simplified approach to design. Predicting boiler performance. Simplified approach to predicting performance ‘Checks for fouling Effect of scale on boiler performance. Hydrogen plant boilers.. 7 Gas bypass flow calculations. Determining heat losses from boiler... References. . Nomenclature. CHAPTER 3: Water Tube Boilers. Guidelines for water tube boiler Heat transfer calculations. Convective heat transfer coefficient Determination of tube side coefficient ‘Non-luminous heat transfer coefficient Gas pressure drop calculations... Inline versus staggered arrangement Design of evaporator: Performance calculations.. Selecting designs with low pinch and approach points wv. Comparison of bare versus finned erapOratOreecnnnennl67 Radiant heat transfer. :Design of superheaters Design procedure Performance calculations... NTU method of performance calculations Metal temperature calculations. External radiation seen Flow in parallel streams Minimizing tube wall temperatures. Steam temperature control. Designof economizers. - Performance of complete HRSG. References. Nomenclature.. CHAPTER 4: Simulation of HRSG Design and Performance Importance of HRSG simulation Design and performance calculations Design temperature profil Guidelines for selecting pinch and approach points... Example of design. Performance calculation procedure. Software for HRSG simulation - COG2N Supplementary firing and HRSG efficiency. Improving efficiency of HRSG Deaeration steam calculations. Steam turbine calculations. Optimizing temperature profiles usirg COCEN... Using field data to simulate HRSG pesformance Malti-pressure HRSG design and | performance simulatio: References. Nomenclature CHAPTER 5: Specifying Waste Heat Boilers. Application or system design Space and layout guidelines viiGas data... Boiler duty.. Auxiliary fuel data Emission data... Feed water analysis, low down. Surface area, fin configuratio Cost data for fuel, electricity and steam. Dram sizing, References. APPENDIX A: Finned Tubes: Heat transfer calculations. Fin efficiency and effectiveness 278 Gas pressure drop .. Tube wall and fin tip temperature. Design example... ‘Comparison of bare versus finned evaporator. ‘Comparison of in-line versus staggered arrangement see Fin configuration and performance 290 Importance of tube side coefficient... 291 Effect of fouling factors. 292 Surface area and duty. 300 Nomenclature 304 References. 306 APPENDIX: Low Temperature Corrosion. ‘Causes and cures. Methods of avoiding cold end corrosion Condensation on surfaces. Corrosion is stacks, duc ‘Heat loss calculations through multilayer insulation.. Hot casing design Nomenclatur References. evens 320APPENDIX C: Heat Transfer Equipment Vibration, APPENDIX D: Gas Turbine Data... APPENDIX E: Gas and Steam Properties... Specific heat, wiscosity, thermal conductivity of gases.. Enthalpy of gases... Estimating flue gas properties Effect of pressure on heat transfer... Converting.% volume to % weight. Properties of stearnand:compressed water. APPENDIX F: Tube Thickness Calculations.. Tubes and pipes subject to internal pressure.. Designing vessels and tubes subject to external pressure. APPENDIX.G: Conversion Factors... INDEXPreface ee During the past 20 years I have had the opportunity of engineering a wide variety of industrial boilers and Heat Recovery Steam: generators. During the past 7 years at ABCO Industries in particular -have had the pleasure-of custom designing-over two hundred’ fire:tube and. water tube, waste: heat boilers, each with different gas/steam parameters, which are in operation in the USA and abroad; these units were built for diverse heat recovery applications such as gaseous, liquid, solid waste and hazardous waste incineration systems, gas turbine exhaust, effluents from chemical plants such as sulfuric acid and hydrogen plants, petro- chemical plants, cat crackers in refineries and for effluents from clean as well as dirty. processes; the gas flow ranged from 2000 to 1.5 million pounds per hour, which implies a wide variety of boiler configurations and design features as you will see in the text. Energy management programs are wital to the economic life of any industry and heat recovery boilers play a dominant role in those projects which otherwise waste energy from hot flue gases. | decided to write this book after reviewing hundreds of specifications for heat recovery boilers prepared by consultants and would be users of the equipment; unfortunately several of them are poorly written without emphasis on the process aspects and ‘optimization of installed plus operating costs, with a result that the end user or the owner gets an equipment which perhaps meets the budget requirement but which could incur significant operating costs in the form of higher gas pressure drop or fuel consumption o lower steam production year after year. The book addresses various aspects of heat recovery boilers, such as engineering, specifying, system design, optimization. and performance evaluation. Hence engineers and managers involved in several disciplines of energy management including plant operation will find the book useful and informative. aAcknowledgements ee I would like to thank ABCO Industries for their encouragement and support in the preparation of this book and for the use of several ABCO ‘illustrations and photographs. I would'also like to thank the following publications for permitting me tovuse my articles, which originally appeared in them: Power Power Engineering Chemical Engineering Oil and Gas journal Hydrocarbon Processing Heating, piping and Air-conditioning Sci-Tech Publications Pennwell Books Marcel Dekker Inc. T would also like to thank ESCOA Corp for permitting me to use their correlations for éxtended surface heat transfer calculations. Several readers from various continents have been writing to me regarding my publications during the past several years, which has been indeed been motivating and I would like to thank them for their interest. IV. GanapathyList, of Frequently Used Abbreviations ABMA - American Boiler Manufactures Association ‘ASME - American Society of Mechanical Engineers ‘CO-Carbon Monoxide EOR- Enhanced Oil recovery FTB- Fire tube boiler GTE - Gas turbine exhaust HRSG - Heat Recovery Steam Generator MSW - Municipal Solid Waste NIMBY - Not in my back yard NOx- Nitrogen oxides WL - Normal water level PPB - Parts per billion PPM - Parts per million RDF - Refuse Derived Fuel SCR - Selective Catalyst Reduction System STIG - Steam Injected Gas Turbine TDS - Total dissolved solids WHB - Waste Heat Boiler WTB - Water tube boiler VOC - Volatile Organic CompoundsIntroduction The book is aimed at engineers, consultants and managers involved in specifying, operating, engineering, marketing and procuring waste heat boilers (WHBs) or heat recovery steam generators (HRSGs). It offers valuable information on not only the heat recovery systems in chemical plants, gas turbine cogeneration and combined cycle’ plants, solid waste, liquid and gaseous incineration systems and flue gas heat recovery in general, but also provides the characteristics of each system such as gas analysis, fouling and slagging tendencies, high and low temperature corrosion, potential and the impact of these on design and performance aspects of HRSGs. During the past 15 years and particularly during the last 7 years at ABCO Industries I have had the opportunity of engineering, a wide variety of fire tube and water tube waste heat boilers for different types of applications as mentioned above. Having designed over two hundred boilers with gas flows varying from 2000 t0 1.5, million pounds per hour and steam flows varying from 2000 to 250,000 pounds per hour, I feel that custom designing HRSGs is an art as well as a science, as there are numerous configurations possible depending ‘on economics, cleanliness of gas, gas and steam parameters and layout considerations. I have had also the opportunity of authoring four books and over 175 articles on heat recovery boilers and steam plant systems in journals such as Power, Power Engineering, Chemical Engineering, Heating Piping Air-Conditioning, Oil and Gas Journal, Hydrocarbon Processing, Plant Engineering; the feéd back from the readers has been very encouraging, which prompted me to bring out this work. Another reason was that in the course of reviewing specifications on HRSGs from various consultants and engineering organizations, I felt that less emphasis was being placed on process and optimization aspects, which is very important in the long run to the owner of the plant. Due to lack of knowledge on HRSGs and their performance aaspects, several of the specifications are poorly written and do not furnish adequate information to engineer an economically and technically sound design. Many engineers also lack knowledge or do not know how to evaluate alternate design options. For example if you read Appendix A, several examples are given to show that with finned tubes one can have a lower surface area and still transfer more energy by proper choice of fin configuration. Several engineers and purchase managers still purchase HIRSGs for critical applications based on Surface area and are of the view that more the surface area the better and a design with a lower surface area would not perform, ‘Also, I have come across several specifications which do not place emphasis on: HRSG operating’ costs; during the life time of the HIRSG, the cost of moving the gas through the system due to high gas. pressure drop or the cost of fuel which is required to generate a desired quantity’ of steam may be very significant. While the consultant I6oks af the initial cost alone, to the owner of the plant ‘who has to pay for the fuel and electricity for years to come, the life cycle cost of the HRSG is important. Heace addition of secondary heat recovery surfaces such as condensate hyeater or economizer may, have to be looked into, though the initial cost may be slightly more. A few examples on the subject of evaluating operating and life cycle costs are discussed in the book. This book offers useful information on design and off-design performance aspects of Heat recovery systems and components such as superheaters, evaporators and economizers, which are elaborated by over 65 fully worked out examples. You will find quantitative answers to commonly asked questions on heat recovery boilers and systems; some of them are: ‘+ How can one improve'the efficiency of a HRSG system? + What isthe effect of auxiliary fring on system efficiency? How to compute the fuel requirements and oxygen consumption for gas turbine exhaust boilers? + How to select pinch and approach inlet conditions? ints? How do they vary with gas What is the effect of scale on boiler performance and tube wall temperatures and heat ux? + How to compute the dew points of hydrochloric acid, sulfuric acid, hydrobromic acid, nitric acid?+ Which is better arrangement for bare an¢ finned tubes, indine or staggered? ‘+ How do boilers with finned tubes compare with bare tube design for the same duty? + How to compute the gas temperature at the SCR at off-design conditions. ‘+ How to avoid high and low temperature corrosion problems? + How to use field data to predict off-design performance or fouling of HRSGS? + With finned tubes can you transfer more duty with less surface area? What is the effect of fin configuration? + How to size and predict off-design performance.of fire tube and bare/{finned water tube boilers, superheaters, economizers? + How'to compute tube wall and fin tip temperatures? + How to compute thickness of tubes: subject to internal on external pressure? ‘+ How much gas should be bypassed for gas temperature control? + What is the effect of gas pressure on heat transfer? ‘+ How to evaluate HRSGs for possible noise and vibration problems? ‘The first chapter deals with heat recovery systems. HRSGs are Used in various applications such as gas turbine exhaust, incineration systems, chemical plants and refineries to mention a few. In order to design a HRSG for any application, thie characteristics of the yas stream are important. For example, auxiliary firing in gas turbine HIRSGs'is discusse¢ in depth with examples on computing fuel Tequirements, oxygen consumption, impact on system efficiency and emissions. Features of boilers such as natural or forced circulation, single or multiple gas pass design, insulated casing or fully water cooled ;membrane wall construction are discussed, along with ‘methods of minimizing steaming concems in economizers. Various aspects of WHEs in Municipal Solid Waste (MSW) applications, Refuse Derived Fuel (RDF) fired units and other incineration systems are discussed with emphasis on type of boilers, ‘whether fire tube or water tube, fouling and slagging concerns and high and low temperature corrosion potential. Methods of minimizing these concerns through boiler design and selection of steam parameters are addressed. ‘The second and third chapters deal with Design and off-design Performance calculation procedures for fire tube and water tubeboilers with bare and extended surfaces. Plant engineers can use the simplified procedures described in the text for instance to check for fouling, estimate bypass flow for gas temperature control or estimate the gas temperature at the Selective Catalytic Reduction system SCR) at different load conditions. Effect of tube size and arrangement ‘whether in-line or staggered on design and performance is elaborated ‘quantitatively. Examples are also given on how to compute the tube wall temperatures, including the effect of scale. Arrangement of headers on flow mal-distribution in superheaters and the effect of tube configuration on direct radiation to tubes:are also discussed. Circulation aspects’are also discussed along with various:configur- ations available for superheaters, evaporators and economizérs. Simulation of single or multi-pressure unfired or fired HRSGs can be performed using the methodology described in chapter 4. Guidelines on selecting pinch and approach points are discussed. One can predict the performance of complex unfired and fired multi- pressure HRSGs under different load conditions without actually designing the unit. Such studies would be helpful to consultants in simulating the entire combined cycle or cogeneration plant behavior and economics. Methods of improving the efficiency of HRSG systems through addition of condensate heater, deaerator ‘coil or heat exchanger are addressed. Examples illustrate how one can also optimize the temperature profiles biy rearranging the heating surfaces. Methods of computing deaeration steam requirements and power output from steam turbines are discussed with examples. The software COGEN which is used in HRSG evaluations is recom- mended to those involved in engineering combined cycle and cogeneration projects, as on can simulate complex HRSG systems without actually designing the plant,saving a lot of engineering time Chapter 5 shows how one should specify waste heat boilers from the process view point. Adhering to the guidelines will save a lot of time for both the boiler designer and the purchasing manager responsible for evaluating alternate bids. Advantages of extended surface over bare tube is discussed in detail with examples in Appendix A. Btfect of arrangement of tubes ice. inline versus staggered and the selection of fin configuration are elaborated. Examples also show how one can transfer more energy with less surface area with finned tubes. The effect of tube sidecoefficient and tube and gas side fouling factors on tube wall and fin tip temperatures are illustrated with examples. Appendix B cites the causes of low temperature corrosion and suggests methods of minimizing the problems. Dew points of hydrochloric, sulfuric and hydrobromic acid may be computed using the correlations given. Heat losses through casing may be evaluated using the program déscribed. One can evaluate an HRSG design for possible noise and vibration due to vortex shedding using the methods discussed in Appendix C. Gas turbine based HRSGs are widely used in cogeneration and combined cycle plants. Appendix’D gives the exhaust gas data for several widely-used machines. Gas and steam properties are provided-in Appendix E, with’ correlations for saturated and superheated steam. Example illus~ trates how one can compute gas mixture properties. The effect of gas pressure on heat transfer inside and outside tubes is also addressed. Appendix F shows the method of computation of thickness of tubes subject to internal and extemal pressures according to recent ASME code procedures. In sum, over sixty five examples from real life situations are worked out covering design and off-design performiince aspects of various types of waste heat boilers and systems; in addition, elaborate matter of fact discussions on systems and equipment should make this book indispensable to engineers involved in various disciplines of heat recovery. This book in the authors view would be an invaluable addition to the library of engineers and consultants involved in operation, maintenance, engineering, specifying or purchasing waste heat boilers. Since no single book can cover all of the aspects of the subject, the author suggests that serious professionals involved with heat recovery systems and waste heat boilers should acquire the other books’ and the software COGEN written by him; for more information, please contact the author at : V. Ganapathy, P.O. Box 673, Abilene, Texas 79604, USA.Chapter 1 Waste Heat. Boilers ony Waste. heat: boilers (WHBs) or Heat Recovery, Steam Generators: (HRSGs) as they are often called are used’ to recover energy from waste gas streams such as:those encountered in sulfuric acid or hydrogen plants, refineries, solid, liquid and gaseous incineration systems, power plants and in cogeneration systems using gas turbines and reciprocating engines. With rising fuel costs and limited supply of premium fuels, it is prudent to maximize the energy recovered from waste gas streams whenever possible. ‘Basically HRSGs can be classified into two broad categories: 1. Those which are required to’cool gas streams to a desired ‘temperature range from process considerations; examples could be found in lydrogen or sulfuric acid plants} in these plants, the energy recovery aspect is of secondary importance; the exit gas temperature, from the boiler has to be controlled within a narrow range of ‘temperatures for further downstream process purposes and methods such as gas bypassing would be used to achieve this objective. 2. In the other category of waste heat boilers, the objective is to maximize energy recovery compatible with considerations of high o low temperature corrosion and economics. Examples could be found in gas turbine based combined cycle or cogeneration systems, incin- eration plants and flue gas heat recovery in general. ‘There is no standard design methodology or procedure for engineering of waste heat boilers, since one comes across a wide range of gas temperatures, gas analysis, pressures and steam parameters Table 1-1 shows some of the gas streams encountered in the industry. If the gas pressure is high, a fire tube boiler is preferred, Figure 1-1.‘Waste Heat Boiler Deskbook 4 _pampuco mys et sem, ung-z] seweounin onset PojeovOT “TL sc}eRUDH AISA AEMPN OL UIQ BHD '§ ONY Joys US 8 se sss eee pinauos 9 8 mumps Kaepuanes'g see any oop Lies y Sse onOnIN sOuNALDD wos fO5"Z see MS my we oe ez és se seat soos st osbrone oF el vst us 1 5 “s * se ors ose swe erot es te sem 7 ve bret ors oem ot ors 00 S070 aoe oF soeo SOTO For ovo ovat 07-95 st osteo v0s002 ores Fro sret sere osse ooo't00 zoe e0 64 89 wor 008 oor't oe ore et stot woe 1 onr't-a0e £S Ora O18 2039 ore — ase-osT 29 Orso trol wos 1 osast ore meee 1 Ooo'T00E Y% Of ON IN me 3, "dinay PH OCHN ZH Sf MD 09 ,t0D Sos 5 ‘amasasd ‘3605 260M NVEAL jo woptsoduen f VramgeL,Waste Heat Boilers 3 Figure i- Hevated drum fire tube boiler [curtesy ABCO Industries] If the gas flow is large and several levels of steam pressure are used, ‘a water, tube boiler as shown in Figure 1-2 may be appropriate. More discussions on fire tube versus water tube type could be found in Chaptets 2 and 3. Another important aspect to the type of boiler is the cleanliness of the gas stream; if clean, extended surfaces could be used and the boiler may be made compact as ir applications; on the other hand if the gas stream is dirty as in municipal solid waste systems, the tube surfaces should be bare, with Provisions for cleaning and ash removal. A large water cooled membrane wall radiant section may be required to cool the gases below the fusion points of eutectics before entering the convection section. Ample consideration should be given to high and low temperature corrosion aspects(sxusempur oa Astanan‘auyqam ee 28x04} osu psy seed opis sunssoad-nIMA Zt OMB ‘Waste Heat Boiler Deskbook| | | Waste Heat Boilers 5 HIRSGs FOR GAS TURBINE EXHAUST Gas turbines are widely used prime movers in both the low (less than 10 MW) and in the high end (150 MW) of the power spectrum. ‘They have several advantages such as high efficiency in simple cycle mode, low installed cost per kilowatt compared to fossil or nuclear power planis, quick startup capabilities, smaller space and low cooling water requirements. Exhaust gases from gas turbines are usually clean as they burn premium fuels with high excess air. The energy in the gas stream may be used for several applications such as heating of water or air or heat transfer fluids’ such as Therminol or glycol and.in most cases, for generating steam for process or power using HRSGs. ‘Combined with HRSGs, gas turbines can operate in combined cycle or cogeneration mode, Figure 1-3 and 1-4, thereby improving the efficiency of the overall system compared to the operation of the gas turbine alone. In combined cycle mode a large amount of power can be generated if a fired HRSG is used. Several non-operating nuclear power plants. in the USA are being replaced with gas fired gas turbine based combined cycle systems. In cogeneration mode, gener- ally low pressure steam is generated and used for process heating or cooling applications. i Turbine: steam | Heater Fvel| |eneator| ‘Trae Figure 1.3. Combined cycle plat,‘ Waste Heat Boller Deskbook _—— ————— Figure 14, Cogeneration systems [Pennell Books}. NATURAL VERSUS FORCED CIRCULATION BOILERS Due to the large mass flow associated with gas turbines, water tube boilers are generally used for heat recovery. Fire tube boilers Ihave been used occasionally with very small machines. HIRSGSs for gas turbines could be of natural or forced circulation type, Figure 1-5 and 1-6. In natural cireulation units, the boiler tubes are vertical and the thermal head diiferential between water and steam-water mixture is responsible for the circulation through the system. The circulation ratio is arrived at by balancing the system resistance to flow and the available thermal head. It could fall in. the range of 6 to 30 depending on the system used. The heat flux inside the tubes, steam pressure and circulation ratio are all im rortant variables in anv circulation svstem and the conditions forWaste Heat Boilers 7 Departure from Nucleate Boiling (DNB) ate set by them. One has to bbe careful particulirly with evaporator tubes with extended surfaces as the heat flux inside the tube can be significantly larger compared to the heat flux while using bare tubes. + igure 1-5, Natural cirulation boiler. (couresy ABCO Industries} In forced circulation units, circulating pumps circulate the steam water mixture through the tubes of the evaporator to and from the drum. Since a pump is used to achieve circulation, one can size it to obtain any circulation ratio. It is typically in the range of 4 to 10. ‘Cost of pump operation and economics plays a role in pump sizing. In Europe, forced circulation designs are common. Advantages dlaimed are smaller floor space and quick startup capabilities However, natural circulation designs do not neeé circulating pumps to a6 Waste Heat Boiler Deskbook e146 Fee drton oe. 'Waste Heat Boilers ° maintain the circulation of steam water mixture through the evaporator tubes, thereby saving operating cost and concerns about pump failure or maintenance. The difference in cold startup time periods is not largely due to the simple fact that in the transient ‘heat up phase, the bulk of the time is spent on heating the metal and water of the evaporator module, which is nearly the same whether it is a natural or a'forced circulation HRSG. The tube side heat transfer coefficient does not impact the overall heat transfer coefficient, which is dependent on the gas ‘low and temperature alone. If there are numerous hot restarts, there is a small reduction in the hot restart period due to the circulation of hot water in forced: circulation, designs, but again, this is not a significant factor. to. cchangeithe design concept. In favor of natural circulation, one can say that the concerns about high heat flux and DNB are less due to the vertical orientation of evaporator tubes, which provide a natural path for the steam bubbles to move; ina forced circulation design, the - tubes are horizontal and the heat flux for incidence of DNB is much’ lower. Separation of steam bubbles from the mixture can occur dué to low tube side velocities, resulting in stratification and bumout. This "is more of a concern if the gas side flow non-uniformity ‘high, : ~ Which Can result in variations in heat flux aeross the width of the HIRSG. Fired HRSGs have to be critically evaluated for DNB. Jn summary both natural and foréed circulation boilers are widely used in the industry; while the natural circulation design has an edgé over the forcedas discussed above, the final choice is usually ._ Made by the end user based on his experience. | AUXILIARY FIRING i Since the exhaust gases contain a lot of oxygen, in the 14 to 16% by volume range, (if there is no significant injection of steam in the 28 turbine, see Table 1-1) additonal steam can be generated in the, boiler by increasing the exhaust gas temperature to the HRSG through the addition of fuel alone. HRSGs for gas turbine can be classified into three broad categories, depending upon the firing ternperature. 1. Unfired 2. Supplementary fired 3, Furnace fired aD Waste Heat Boiler Deskbook ‘The term supplementary firing is sometimes referred to as auxiliary firing, When the heat input is significant and a membrane wall furnace is used, then the HRSG is referred to as a furnace fired ‘unit. Note that one may use a membrane wall furnace design even for a low firing temperature situation by choice. Hence these terms are loosely used’ in: the industry and one should be aware of the implications of the firing temperature on the selection and design of ‘the fumace, firing duct and the HRSG. ‘Table 1-2 below gives an idea of the steam generation capabilities of each type. Note that no additional air is needed for combustion for any of the three types as:the oxygen in the gas stream: is generally adequate. If there is steam injection in the gas turbine, additional air; sometimes referred to as augmenting air, may be required? #9! : ‘Table 1-2 Inlet gas temperature and steam genenation for gas turbine HRSGs TumdSyiem —»«Gasin\Tema.cE Gas Steam ata Unfired ‘200 - 1000) 85107 ‘Supple. trod 1000-1700 251085 Fumace fired 1700-3200 321025 UNFIRED HRSGs When the plant steam requirements are such that the energy in the exhaust gases is adequate, an unfired HIRSG is selected. Typical gas temperature entering the HRSG ranges from 800° to 1050°F, depending on the gas turbine used. Figiire 1-7a shows a two-gas pass design with a superheater, evaporator and economizer. This design ‘occupies a small floor space. The evaporator tube bundle is divided into two portions by using a baffle plate. The gases flow across the superheater and the bottom half of the evaporator, make a 180 degree turn and then flow across the top half of the evaporator tubes into an economizer. Various gas inlet and exit configurations are possible; for instance the gas-can enter at the top and flow down- wards or exit at the side. Insulated downcomers are located at the turning section.‘Waste Heat Boilers n Figure 1-7a. A two-pass unfied HRSG with superheater, evaporator and economizer, ; Ieourtry ABCO Industri. ‘The|casing is internally insulated With 4 to 5 inches of mineral ber or insulation and is protected from the hot gas stream by a Stainless steel or corten or carbon steel liner, which is designed to move or expand in the plane of the’ casing. Lower grade liner ‘material is used in low gas temperature regions, ‘The economizer can be located at the gas exit as shown. Designs that help minimize steaming concerns are discussed later. Figure 1-7e shows a single gas pass design. The stack could be self staring or mounted over the turning section. a2 Waste Heat Boiler Desktook Figure 17>. A two-gas pass HRSG installation, [evuriesy ABCO Industries] AUXILIARY FIRED HRSGs Supplementary firing of exhaust gases is done to raise the temperature of the gas stream entering the boiler to a maximum of 1700°R, 50 that additional steam can be generated without major ‘modifications to the unfired boiler design. The 1700°F limit is set by the design of the casing, which consists of several layers of high temperature insulation inside the casing, varying from 6 to 10 inches thick, which is protected from the hot gases by a suitable high grade alloy linér material. The liner should cover as-much of the insulation fiber as possible. and yet be free to thermally expand in both the directions. Liner materials whatever their grade begin to warp above 1700°F, resulting in the insulation being blown off by theWaste Heat Boilers B gas during the course of time; hence the limit on supplementary firing temperature. Appendix B can be referred to for calculations of casing tem- perature! and heat loses through multilayer insulation. The easing design described above is used for the duct work leading to the HRSG. from the bumer and also in the HRSG portions having a high gas temperature, Carbon steel liners and lesser insulation thickness may bbe used in low temperature portions of the HRSG and economizer. ‘The firing transition is about 10 to 15 feet long, depending on the boiler size and the transition angle between the bumer and the HRSG. The'HRSG does not differ much in concept from the unfired design, except for the sizing of drum and other heat transfer surfaces and components to handle the higher steam generation, The HRSG. steam output can be easily doubled ag seen from Table'1-2 by the | additional fuel input. Figure 1-74 shows a supplementary fired | > HIRSG with a superheater. Bisure 17. Single pass unired HRSG with superheater, screen, SCR, evaporator and ‘conomlser. Iourtesy ABCO Industries]u Waste Heat Hoiler Deskbook “The burner is located between the superheater and evaporator due to design considerations. Due to the high firing temperature, the evaporator has to be designed with varying fin configuration. The first few rows are bare, followed by tubes with increasing fin density. ‘This is done to minimize the tube wall and fin temperatures and heat flux inside the tubes. Methods of evaluating the performance of cnfired afd fired HRSGs are discussed in Chapters 9 and 4. eee‘Waste Heat Boilers B Duct burners, Figure 1-8, are used to raise exhaust gas” temperdtures to about 1700°F. These fire generally premitim fuels such as natural gas or distillate oils and heavy os. Line bursere are similar in construction to duct burners but are located or mounted on the front wall of a water cooled furnace. These are used to raise the gas temperatures to nearly 2300°F. Due to the higher gas tem. perature an insulated casing design is not adequate. i Rn rum ans Sepp ta tb gt fon | esta) | | The HRSG for such an application is usually of membrane wall construction as shown in Figure 1-9. Since the enclosure is fully water cooled, there is no limitation on the firing temperature with these designs. | | The furnace length is determined. by the Jame length usually (on the order of 10 to 15 feet. The burner supplier should be contacted ‘in the case of special fuels. The furnace section is followed by the a6 ‘Waste Heat Boiler Deskbook convection section, which as before consists of a combination of bare and finned tubes. The superheater, if used, could be located within the convection bank. Fagus 133) HAIL fred HRSG [erawing'and photo courtesy ABCOWaste Heat Boilers v Line burners and duct burners typically have alow gas pressure drop, on the order of 03 in we. If steam is injected into the gas turbine either for NO; control or for increasing the power output as in the Cheng cycle system, the oxygen content in. the gas stream reduces. This fact along with high water vapor content in the gas stream forces one to use an augmenting air supply to ensure flame stability in the burner. In furnace fired HRSGs, register burners are used to fire the exhaust gas stream to the maximum possible extent, ‘namely adiabatic combustion temperature, depending on the oxygen content in the exhaust gas. The bumer resembles @ conventional bumer used in fired boilers, Figure 1-10 with windbox and throat for bumer clements. The ges pressure drop in the windbox can be as high as 4 to 5 in we and hence imposes a penalty on gas terbine power output. It may be noted that each 1 in we reduces the power output of the = turbine by nearly 0.25%. One of the advantages of furnace fired * designs is that it can even handle solid fuels. The gas turbine acts as a fan supplying hot combustion air to the boiler. The boiler has a fally water cooled radiant section of membrane wall design followed bya convection bank and economizer. {__\_,The superheater, if used, can be buried within the boiler bank "a5 shown. When the capacity of ie HRSG 15 large, 2 shop assembled D-type boiler may not be adequate and a field assembled HESG resembling a fossil fuel fired utility or industrial boiler may be required. The cu off steam flow is approximately 150,000 pp Another important aspect that should be considered in the clesign of the firing system is the maldistribution or non-uniformity in the velocity profile across the duct cross section as the gases exit the gas turbine, Also due to the size ofthe duct work between the gas turbine and the boiler (nearly 30 to 50 feet high and 10 to 15 feet wide) and the manner by which the gases exit from the turbine and enter the HRSG system, additional maldistribution can occur. One of the ways of dealing with ths situation is to perform a model study and incorporate turning or guide vanes in the duct work ahead of the HRSG and the burner; in addition, the transition angle of the Guctwork connecting the burner and HRSG must ‘be carefully reviewed. If it is too small, it minimizes the maldistribution but adds to the cost; some suppliers use a distribution grid, whichee waste Flat Boters » consists of a perforated plate with approximately 50% opening. This increases the gas pressure drop by nearly 05 to 1.00 in wand minimizes the maldistribution to some extent depending on the % opening. A combination of both vanes and distribution grid may be required in some cases. Model analysis is important and could help prevent serious maldistribution problems particularly in units with ‘burners and superheaters. FRESH AIR FIRING Invall of these firing: systems, atmospheric fresh’ air can be introduced to support combustion in case the gas turbine trips. In refineries and in critical applications where steam production should be maintained in the event of a gas turbine trip, this addition is suggested. Figure 1-11 shows the scheme. A double louvered damper with seal air fan prevents hot gases from going to the fan when itis not in operation and the gas turbine is running, A guillotine or.damper prevents air from leaking into the gas turbine duct work when the fan is running and the turbine is off, S | |—xRs6 LI ldiverter 2slide gate 3.Firing duct 4.burner S.Fan Samper w/fan 5 Figure Il. Scheme for fresh ar fring,20 ‘Waste Heat Boller Deskbook It should be noted that the burner duty increases significantly when itis on fresh air as it has to be raised from ambient temperature to the firing temperature. The sizing concept for the fresh air fan is discussed later, as it affects the system efficiency. ‘One of the concerns in several plants is the time for switching over from gas turbine mode to fresh air firing mode. This could be on ‘the order of a few minutes. It is possible to reduce the time duration between the gas turbine trip and start of fresh air supply by using an induced draft fan behind the HRSG, which operates all the time and handles part of the HRSG resistance when the gas turbine ‘operates and the complete HRSG resistance: when on fresh air. The fan curve and the HRSG performance must be studied in depth before selecting the system. (COMPUTING FUEL REQUIREMENTS Engineers have to frequently compute the energy required to raise the temperature of gas turbine exhaust to a desired level; also the concern sometimes is whether there is sufficient oxygen in the gas stream to achieve the objective without additional combustion air. Presented below is an analysis end a chart that may be used to perform quick estimates of oxygen depletion and fuel consumption. Supplementary firing of gas turbine exhaust gases using natural gas or distillates is frequently done to generate additional steam in waste hear boilers. With chart shown here, it is possible to estimate: 1. Energy that must be added in the burner system to raise the temperature of a given exhaust gas quantity Wy from Ty to Tr ‘The quantity of fuel required to do this in scf/hr and in Ib/r 3. Maximum fuel input that is possible with a given oxygen content of the exhaust gas (which usually varies from 15 to 19%). ‘Also, the chart may be used in reverse to determine the gas temperature after the burners if fuel input is known.waste Heat Boilers a a —. 5 supplementary fuel parameters Figure 1-12. Chart for supplementary fuel parameters (Oi and Gas Journal.) NOMENCLATURE A. = Amount of combustion air in turbine exhaust, Ib/h. C= Constant for combustion, depends on type of fuel, 1b/MMBtu, © = Oxygen content in exhaust gas, vol%. LHV and HHV = lower and higher heating values of fuels, Btu/Tb Qeand Qmaz = Supplementary fuel input and maximum fuel input possible, MMBtu/hr2 ‘Waste Heat Boiler Deskbook Wy = Fuel quantity, lb/hr for oil fuels and scf/hr for natural gas Wg = Turbine exhaust gas quantity, lb/hr TyandT2 = Exhaust gas temperature entering and leaving burner system, °F. jyand hz = Enthalpy of gas entering and leaving burner, Bru/Tb. ‘Theory. The energy required to raise We Ib/hr of turbine exhaust from Ty to Tp is given by: Qe= We (ha hy) ap Where hy and kz are the enthalpies of the gas at Tz and T2, respectively. With the chart, there is no need to look for lz and hz; itis adequate if Ty and Tz are known. We, the fuel quantity required, is given by: Wy= Qe/LHV (1-2) LEV dhe lower heating value of the ful is 18,000 Bts/Ib for istillate oils and 1,000 Btu/scf for natural gas. Hence Wg is in Ib/hr {or oils and in scf/hr for gas. Also, itis often desirable to know the maximum amount of fuel that can te fired for a given amount of oxygen in the gases, especially when the turbine exhaust is used as combustion air in conventional boilers. ‘This can be found as follows. Let © be the vol% of oxygen in the exhaust gas. The total combustion air available in Wg Ib/hr of gas may be shown to be: | | A= 100% Wig x 0% 32/(23 x 100x295) a3) Molecular weight of exhaust gases was taken as 29.5, that of | oxygen, 32. One million Btu of fuel fired requires a nearly constant amount of air. For oils, this is 745 Ib and for natural gas it is 730 Ib. So ‘the amount of fuel that can be fired with "A" Ib of air will be: Maximum = 108A/(C xHHV)waste Hest Boilers B where Cis the constant referred to earlier. Converting to fuel input in MMBtu/hr on an LHV basis, the maximum energy that can be input as fuel i: 08 xA xLHV)(C x HEV) (4) Qmax Now LHV/(C x-HHV) can be shown to be nearly 0.00124 for both oils and natural gas. Using this: Qmax= 58.4 xWy x0 x10. (1-5) Exampled: A waste heat boiler for turbine exhaust gases handles 530,000 Ib/hr of gas at 900°. To obtain additional steam, it is desired to raise the gas temperature through supplementary firing to 1400°F. Determine: 1, Fuel energy required to raise the exhaus: from 900 to 1,400°F. 2. Fuel quantities required to do this in Ib/hr (oils) and in sef/hr (natural gas) 3. Maximum fuel input possible if the oxygen content in the exhaust as it leaves the gas turbine is 15% by volume. Solution. Go up from Ty = 900 to eut Tz = 1400 and move right. Conmect with We = 530,000 and extend to cut Qz scale at 76 MMBtu/hr. (Use about 80 for sizing the’ burner system.) If distillate is used, fuel quantity is read off the same scale as 4,300 Ib/hr. If natural gas is used, it 18 76,000 scf/hr. ‘Connect oxygen = 15 with Wy = 530,000 and extend to cut Omar scale at 465 MMBtu/hr. By working in reverse, T can be obtained if Qz is known. The oxygen depletion is obtained by proportion. For 76 MMBtu/h, the oxygen depletion = 76 over 465 x 15 = 10.86%. BASIC PROGRAM COMPUTES FUEL INPUT, FIRING TEMPERATURE For accurate computations one may use the program presented below. Supplementary fuel firing of turbine exhaust gases is frequently utilized in combined cycle and cogeneration projects (Figure 1-13) to generate additional steam from heat recovery boilers.a Waste Hest Boiler Deskbook Wy Te bay sumer eet opr Barer i Ta exhaust haz Duet bume ster Figure 1-13. Burner erangement showing the mass and energy balance. ‘Typical turbine exhaust has 14:16% free oxygen before entering the combustion system. Part or all of it may be utilized to increase the final combustion temperature to desired levels. Duct bumers typically raise the exhaust gas temperature to approximately 1700°F maximum. In furnace-fired heat recovery boilers the final combustion temperature could be as high as adiabatic combustion temperature, which is the temperature obtained by consuming all of the free oxygen in the exhaust gas. Natural gas is the most widely used fuel. In order to analyze the performance of the burner system and the boiler—and to calculate the cycle efficiency—the following data are required: 1, Fuel required to raise the temperature of the incoming turbine exhaust gas to a particular value. Sometimes fresh air is used instead of turbine exhaust gas. 2. Given a particular fuel input, the combustion temperature (reverse of case 1) that results. 3, Fuel data such as density and heating value in Btu/Ib and Btu/sct.waste Heat Bollers B 4, Flue gas analysis, molecular weighé, 5. Maximum fuel input possible to consume all the oxygen. With the BASIC program presented here one can perform all the calculations with ease. The program runs on IBM PCs and compatibles. Figure 1-142, Progam for computing fuel input, combustion temperature, VeLsney ore 2 RORATTHOR Wcananaryy 5 cis nay ote an cosSom a sem i faronresvoscsrer? IS reasece ns maatawese 24k egnas ras a5, hain nana ahvaseanasmnisasiassaa ART eS asbata aisbsts isa mcsntisiné Seseeci time shagsia ney tsaannysssanin eaeeasmaeieees eats aobataeudssizaeasusnnsne sae gaan cass ras, hea easinnexri e785 504.0432 @PCLEPANT FUEL GAS ANALY VOL ~pRINT™ = Sa Fon ito serait jensen xr eslivinadainicaneatargae Segveagzreatlowat Parvo tt POMMFHAEDe"aihnBDURT-A2~AISnARDe AALS YSE QA 2ateatien 1Oabe AHA DIE AS INUT-ExWAUST Gus ANALYSI-VOL OF CO2.H2D.N2,02-"CORANAA.N2A,OEA:PRLNT™ * IMT SOREN DUTY SLA HAL APF amon frets SUAS BTA SNEE ASTM AY wevatnas.7- PAITO92 08) BDO ENST” anata 51 625 FLOKABRTENESWAIATAIR Sto wave Ts amanee-coza-ntewancdeenone2-O0at0 119 CozaN£023~ ATW ah- WAR” MWASN2AW-MDA* 2zWE:C2AN-OzA* 2aneWA ‘a coats igosracnvarenaazrataranhaaseeuesarne em ee eee Tasers a.m saoiorewe iidintsrieata(atanot We (alt erae soe a eae S5- SCSSOTPABIAte eae wera | | vols rai scat tgtoataaraal Swe | ubcnanronain.nsoscoureonsare-oarecnatOaOe oy pe Ttuecrarco-cozawaenan NODE CHF WAM -WAIROZFGeO2AM HAIR DIRAEE =26 Waste Heat Boiler Deskbook Figure 1-14 (cont'd) Eso ir rremreo Ten 205 BF ausevesrovre ven 200 a nov dena. ceg-seMirTeMTEMPeRO:COTD We 3B Ropcicovee wareeonrneniab-Soseb:co2VrcOUMSEE:R2OV-WA/SUN: D018: ‘evvsOK sow noezatven Sumi Diz-soaVsnR0W/ 4068 ‘oun ortensv otrrEOY="8-OXYW"SEAIV-2HEOND Ps cesconint: goiLTs oF COMBUSTION GALZILATIONS'FRINT” Frovaietuey DEna¥Lesct™"aenseeLNvaTUREP=ANVEr LAVA save rain FLOW Ho" HAIR MOLWTAAPSVOL OR EODA: be Herre OVDLCHRA OREO" HOVOD ORV: {-S0aeSONV-MOL Te SNES aeornint™ 3 SFT awa remperTewe-purva er Une" ;°10°6rMAX FUL ATU" OMAK: = {1S PRINT-FUEL GAS ANALYEIGSVOL"RINT™ © TODFRInT Haass To LOCATE 7-124PRINTEASSHDAN Biker sovocare 2218.49 ‘gir reur dan fen Mowe sn 260070 100 {ess Tem oscar es tara osus 20086 )on2 {fob oaee sus 2onerczons an hee seOxatnoznetPwcycoTe 10 ‘as Meccosten te avemea eyo 5882-7427 EMP ODI SEBS 20°TEMP Hoe OKs) ease rear 00) S0" the CTEMP EEO soo RETURN Hise Re-nrOx-wUneM-n-wA-ciIrce-BUS8 ame neToRN ‘Theory: The enthalpy of flue gas after combustion is obtained from: (Wy gy + Wy xLHV)/(Wy + Wd a6) f f Wy = flow rate of turbine exhaust gas ot fresh air to burner, Ib/h hg1 = enthalpy of incoming air or exhaust gas, Btu/Ib Wy = fuel input, Ib/h = Q/LHY Q._ = fuel input to bumer system, Btu/h (LHV basis) LHV = lower heating value of fuel in Btu/Ib hhg2 = enthalpy of final products of combustion, Btu/Ibste Heat Boilers az ‘hg and hg are functions of the air /flue gas analysis and temperature. Based on combustion constants (Ref 1), the oxygen required, nitrogen, carbon dioxide, sulfur dioxide, and water vapor produced are computed and the flue gas analysis for any given fuel input is | obtained. A. subroutine calculates the enthalpy. Also, the heating value of the fuel is computed based on the analysis. Up to 14 constituents can be handled by the program, as seen in Figure 1-14b. Figure 1-14b. Inputs and results for Example FUEL ANALYSIS el damian: 295 n 2am 925 1 Bpenme 10 ” = | Sinton 70 Balrog ns pease 20 Swerve 7 Tope 20 Tehyaeee aide 71 EXWAUST GAS FLOW:L:N.TEMP.F=? 15000800 EXHAUST GAS ANALYSISAVOL OF €0220,N2.02-? 4 5:3285,82 [BURNER DUTY.ATUIM FINAL TEAP.Finpet zr forte unknomale?0,3600 [RESULTS OF COMBUSTION CALCULATIONS | ain Tein so FLoMLim 138000 MOL WT= 282718 wvOL-COZ 4¥H20~ 53 02152 N2“75S FLUE GAS FLOWLave 1383281 vOL-COZ= 550) n20» 86878602" 1150723 N2= 7421282 sSuz- 7160526 62oU w= 24 S088? FINAL TEMP 153,112 DUTYMM BTU.Me 28.5 MAX FUL INATU/E 1252603608 FUEL GAS ANALYSIS-SWOL shras @ | Seba 0 I irope- 15 Spenane 0 wateraper 0 Thaperaoe- 0 drogen side 1 If the fuel input to the burner in Btu/a is known, the fuel quantity is obtained and the. flue gas analysis and..quantity are, computed. The final enthalpy is obtained and, through an iterative process, the corresponding temperature is found. a a28 ‘Waste Heat Boiler Deskbook If the final desired temperature is given, the fuel input is assumed and iterated to yield the final enthalpy, after computing the flue gas analysis. Quick converging techniques are used. ‘The program also estimates the fuel input to consume all of the ‘oxygen in the incoming air or turbine exhaust gases. Caleulation Examples ‘Two examples illustrate the versatility of the program. Example 2: 135,000 Ib/h of turbine exhaust gas at 900°F, having the analysis CO2 = 4%, HzO = 53%, Nz = 755%, and 02 = 15.2% all by volume, enters duct bumer and has to be raised to 1600°F using riatural gas having the following analysis: methane = 95%, ethane = 2.5%, nitrogen = 1.5% and hydrogen sulfide = 1.0%, all by volume. Determine: 1. Fuel heating value in Btu/Ib ard in Btu/sct. 2. Fuel quantity required in Ib/h and in Bu/h. 3. Flue gas analysis and molecular weight. 4, Fuel input to consume all ofthe free oxygen. Step-by-step instructions for running the program: 1. Key in the program from steps 1 102010 as shown in Figure 1-14a. Key the letters RU, N (RUN mode) and press the “enter” key. 3. You will see the 14 gases printed on the screen, one below the other, starting from methane and ending at hydrogen sulfide (see the output in Figure 1-I4b for the various gases, 14 in all. 4, Then the question is asked at the bottom of the screen, "What is the maximum number of gas?" By this is meant the number of the constituent in the fuel gas that has the maximum number (1 to 14). Iffor instance, we have methane, propane, and nitrogen in the fuel, nitrogen has the number 12 (see the list of gases in the results, Figure 1-14b) and is the largest. Hence one would enter 12, in this case. 5. Once this number is entered, the list of gases up to the maximum number appears on the screen, dne by one. Enter the volume % of ‘each as shown in Figure 1-14b. Ifa particular gas is not present, center zero for its volume.Waste Hat Boilers 2 6. Once the gas data are fed in, the next question asked is "Exhaust gas flow and temperature.” Input the gas flow and the tem- perature, separated by a comma, as seen in Figure 1-14, 7. Next the gas/air analysis is fed in. Volume % of CO, HzO, Nz, 2, are inputted, each separated by a comma, as seen in Figure 1- 1b. ; 8. The next set of data to be fed in are: “duty, final temperature.” Now, input zero for the unknown, and the known value, each separated: by a’ comma, as shown in Figure 1-14b. If the final temperature is 1600 and the ‘duty is to be calculated, input 0,1600. If the duty is, say, 2,000,000 Btu/h and final temperature is to be found, input 2000000,0. This ends the data inputting. The results appear as shown in Figure 1-14b. ‘Gas analysis and other data can be seen. ‘The fuel gas and flue gas analysis, final temperature, maximum possible fael input, and heating values of fuel are all printed out. ‘The fuel input is 28.35 MMBtu/h and maximum possible fuel input is 1233 MMBtu/h. Flue gas quantity is 136,339 Ib/h and 115% oxygen is still available in the exhaust gas. Lower heating value in Btu/Ib is 20,640 Btu /Ib. Example 3: 100,000 Ib/h of fresh air at 80°F enters a burner and 100 MMBtu/h of fuel is inputted. Determine the final temperature. Fuel has 95% methane, 3% ethane and 2% propane, all by volume. Solutior Procedure similar to Example 2 is followed. The analysis, used for airs 79% nitrogen and 21% oxygen by volume. It is seen from the printout that the final temperature is 3150°F and maximum possible fuel input is 126.1 MMBtu//h. See Figure 1-14. Analysis of the flue gas and other data may also be obtained from the printout.20 ‘Waste Heat Boiler Deskbook Figure 4 Input and els for Brample FUEL ANALYSE" Vmahine 785 Ream 27 Somes 22 EXHAUST GAS FLOW.3i,TEMP-=? 10000080 [EXHAUST GAS ANALYSIS4VOL OF C02,H20,n2,027 087821 [BURNER DUTY-ETU.H FINAL TEMP. putz for unkoown?T0000000 [RESULTS OF coMBUSTION CALCULATIONS . FUEL DENSITY.L asc? .So22506-2 LHVATUCE® 864375 LAV-TUILEe 2169828 {IR TEMP IN 9 FLOW-LEe 10000 MOL WT=2624::VOLCO?® 0 H20- 00262182678 FLUE GAS FLOW-LE Me 1sGH843VOL-CO2=T752647 H2OH 16.1457 OD 491802 NBe 72.08 $02 8MOL WT=27538%9, ‘ FINAL TEMP> 2168582 DUTY.MM BTU/H=TOOMMAX =UL INATUIM 12512206-38 FUEL GAS ARALYSISSVOL Boerne Styzogn:@ ‘Weabon monoid. 0 Veahon sense 8 iRaloopen waar oper 9 Thopeoaee 0 ebybrgen slide SUPPLEMENTARY FIRING AND SYSTEM EFFICIENCY The efficiency of the HRSG system improves with firing. This can be seen from the example given below. The reason is that with the same oxygen content entering the bumer, more fuel is being fired thus reducing the excess air leaving the stack; also, with an increase in inlet gas temperature the exit gas temperature from a HRSG with an economizer usually decreases. This is due to the significantly larger ratio of water to gas flow in the fired mode compared to the ratio in the unfired mode. The gas flow remains nearly constant, while the steam production and the water flowing through the economizer increases, depending on the extent of firing. This fact is partly responsible for the improvement in efficiency. More information on HRSG heat balances and temperature profiles and the use of the software COGEN can be seen in Chapter 4.‘waste Heat Boilers at Example 4: A supplementary fired boiler generating 200 psig steam operates as shown in Table 1-3. (the data have been obtained from COGEN software described in Chapter 4.) Determine the system efficiency using ASME PTC 4.4. Table 1-3:-Data for Supplementary Fired Boiler 1. case pl perhan 1 2 3 2. gastlow, pph 180,000 150,000 150,000 3. inlet gas temp, °F 900 ; 900 ‘900 4, fring temperature 900° 1290 1718 5. bummer duty, LHV MMBtu/h: 0 17.30. 37.60 6. steam flow, pph 22,780° 40,000 60,000 7. steam pressure, psig 200 200 200 8, feed water tomp, °F 240 240, 240 8, exit gas tome, F 327 315. 310 40. steam duty, MMBtuh 2287 39.90 59.90 11. system efficiency % 68.7 79.2 84.90 Note: gas analysis in :% vol COz =3, HO = 7, No = 75,02 =15 blow down =3% Solution: Using COGEN program and the methodology discussed in ‘Chapter 4, the design and off-design performance at vatious loads were obtained; the results are shown in Table 1-3. Gas enthalpy at 220 Btu/Tb (from Appendix E). Efficiency = output /input = 39900000 /(17300000 + 150000 x 220) 79.20 % for case 2. ‘As seen from the above table, the efficiency increases with additional fuel input. The additional fuel input generates nearly 90 to 95% of its energy equivalent as steam. FRESH AIR FIRING AND EFFICIENCY Fresh air firing is often used in gas turbine plants in case the gas turbine fails or is shut down and steam should still be generated. How does one select the fresh air fan capacity? The efficiency of the HRSG system increases with a decrease in stack gas temperature and mass flow, as will be shown in Chapter 4 a2 Waste Heat Boiler Desbibok Hence an obvious method of improving the system efficiency is to select as low a mass flow as possible for the fan compatible with fir- ing temperature considerations. That is if the capacity of the HRSG can be maintained at a lower fresh air fan flow without exceeding, the firing temperature that may call for a different type of HRSG (membrane wall versus insulated casing) then that is the best choice. In order to illustrate this, let us se2 how the HRSG in Example 4 performs with fresh air fan while generating 40,000 pph of steam. Example 5: There are two fan selections—one with 130,000 pph’ capacity and’ another with 150,000 pph. Determine’ the system efficiency, when 40,000 pph at 200 psig is generated, feed water temperature is same as before, Solution: Using COGEN the results are obtained for both the cases, see Chapter 4. ‘Table 1-4: Fresh Aix Firing Performance 1. airflow, ppb 130,000 150,000 2. inlet temp, °F 60 60 3. firing temp, °F 1424 1204 4. ont gac tomp, F 208 ae 5. steam flow, pph 40,000 40,000 8. burner duty, MMBtu/h 48.72 50.92 7. efficiency, % B18 79.25 ‘The following points may be seen: 1. As the capacity of the fan reduces, the efficiency improves as the heat loss from the system is lower. 2. The stack gas temperature is also reduced due to the higher firing temperature with the lower air flow, adding to the improvement in efficiency. 3. The cost of the fan will be lower as the capacity and hence the {gas pressure drop are lower. 4, The firing temperature is higher but if the design can handle it ‘without a major change, lower fresh air flow is preferred. 5. The burner duty is naturally more as the air is raised from the ambient temperature to the firing temperature, compared to gas turbine operation.Waste Heat Boilers 2 HRSG DESIGN FEATURES Finned Surfaces and Design Fined surfaces (solid and serrated) are extensively used in HIRSGs for gas turbine applications. Their use makes the HRSG compact and weigh less; the gas pressure drop, which is an operating penalty, is also reduced with finned surface design. Appendix A discusses the method of calculation with finned tubes and also com- pares a HRSG design with and without finned tubes. When natural gasis fired in the gas turbine or burner, the clean gases permit one to use as high’a'fin density as 5 or even 6'fins/in for evaporators or economizers. Tubes may be arranged in line or staggered fashion. With distillate fuels, the fin density should be lower say 4.5 or 5 fins/in. Fin height ranges from 5 to .12 in., depending on fin tip - temperature calculations. Higher the fin density or ratio of external to tube internal area, higher will be the heat flux inside the tubes, tube wall and fin tip temperatures and the gas pressure drop. Superheaters on the other hand have a low tube side heat transfer coefficient and hence it is prudent not to use-a high fin density as discussed in Appendix A; if so done, the tube wall and fin tip = temperature will be higher; the surface area will appear to be high with’a high fin density, but due to the low overall heat transfer coefficient, the duty transferred could be the same or even less. Hence ‘one has to be careful while evaluating the performance of finned surfaces and not go by surface area alone, as it is misleading. The Product of surface area and overall heat transfer coefficient determines the duty and not the surface area alone. More information on finned tubes and their optimization can be found in Appendix A. ‘When the gas turbine burns a dirty fuel such as heavy oil, the fin density should be lower, say not more than 2 or 3, depending on the presence of slagging constituents in the fuel ash; more clearance Should be allowed between tubes; inline ‘arrangement is preferred as soot blowers could be more effective. If the fuel oil is extremely dirty, a bare tube design, though expensive, may be the only choice With fired HRSGs the gas temperature along the gas flow Path will change, Hence the evaporator should have combinations of bare and finned tubes of varying fin densities so as to minimize the hheat flux and tube wall and fin tip temperatures. Usually the tubes, aePy Waste Heat Boiler Deskbook at the front end would have a low fin density followed by high fin density section. The thermal design procedure should account for the different fin combinations. Ifa superheater is used, itis preferable to bury it within the convection section when the firing temperature is very high. This helps lower the metal temperatures and hence increase its life. Methods of minimizing superheater metal temperatures are discussed in Chapter 3. Usually HRSGs for small ges turbines come in packaged. designs, while for large machines (gas flow above 250,000 ph) each surface’ suchas superheater, evaporator or economizer is built in a separate module for shipping purposes. STEAMING ECONOMIZERS ‘As will be seen in Chapter 4, one of the problems associated with low load or low ambient operation of a gas turbine is steaming in the economizer. Formation of steam in the steam water mixture obstructs the flow of the mixture when it flows in the downward direction and may result in vibration or water hammer concems. Iti preferable to have the tubes which are steaming to be oriented in such a way as to aid the flow of bubbles, namely upwards. Figure 1- 15 shows a horizontal tube economizer design that has multi-streams {in which the water in the last section flows from the bottom to the top header. Steam bubbles if any should be easily removed with this configuration. The economizer could also be designed as a two gas pass unit as shown in Chapter 3 with the water in the exit section having an upward flow. The other options are: 1. Reverse the flow direction of water using valves, which is cumbersome; in non-steaming mode the economizer operates in counter flow configuration, while in steaming mode, it operates in parallel flow configuration. 2. ‘The exhaust gas may be by-passed around the economizer to decrease its duty and thus prevent its steaming. This is a loss of energy. sis ne 3. Some boilers are designed so thet the gas flow to the boiler itself {s bypassed diiring steaming conditions; this is not recommended‘waste Heat Boilers Eo as it results in a significant loss of energy by virtue of the evaporator not handling the entire gas stream. , 4, Bypass a portion of the economizer surface on the water side so that the surface area participating in heat transfer is reduced and hence the duty or enthalpy rise decreases thus avoiding steaming. ‘The author prefers the design shown in Figure 1-15 as it does not resultin loss of energy in any mode of operation. Figure 1-15. Arrangement of HRSG with horizontal gas flow economizer. (courtesy ABCO Industries) Emissions of NO, and CO Firing of fuel in the turbine and auxiliary burner introduces pollutants such as‘ NOx and CO into the exhaust gaé stream. Stringent pollution regulations dictate the amount of NOx and CO that can leave the stack of present day boilers. Pollutants are often% Waste Heat Boiler Deskbook specified in ppmvd (parts per million volume dry) with reference to 15% oxygen in the gas for emission monitoring and control purposes, while bumer suppliers often suggest their contributions in mass flow units. The procedure outlined below may be used by engineers for relating the mass flow rate of NOx end CO to ppmvd and vice versa. Emission-Data Conversions Monitoring combustion emission products—NOs and CO in particular—is increasingly important, due to tightening regulations. Gas: turbine cogeneration or combined-cycle plants, for example, produce both of these emissions, and regulations pertaining to them are very strict. Because of this situation, it is desirable to have a quick and easy way of converting emission data from mass flowrates (Ib/h) to ppm(o), dry basis, which is the usual way that they are reported to regulatory bodies. An oxygen concentration of 15% is the usual reference in industry. With the enclosed chart it is possible to accomplish this purpose. The chart provides a conversion factor, F, which allows both the Ib/h-to-ppm(o) conversion and the reverse. Nomenclature V_— = volume or volumetric retio; Va, volumetric ratio ‘of NOx (ppm); Vc, volumetric ratio of CO (ppm) w = flowrate of a constituent gas in the exhaust stream, Ib/h W_ = total exhaust stream flowrate, Ib/h F = conversion factor MW__ = molecular weight of exhaust stream, Ibs %H20 = percent volume of water in the wet exhaust stream %02 = percent volume of oxygen in the exhaust stream Underlying Theory If w, in units of Ib/h, is the flowrate of NOx (usually reported as NO?) in a turbine exhaust stream of W Ib/h flowrate, the following equation gives the volumetric ratio on a dry basis. (The molecular weight of NO? is 46.)Waste Heat Botlers a _ V-= [@o/46)/(WIMW)] x 100/(100 -%6#20) (17) In this equation, %H20 is the volume of water vapor in the exhaust stream, and MW is the molecular weight of the gases. V_ must further be corrected for the 15% oxygen, dry basis, as follows: 7 10° V 21-15) Ve lin ppm(v)} = ——_10V@h=15)_ 21—[100/(100=96EEOT x %O2 (18) { In this case, #02 is the’percent volume of oxygen in the wet exhaust gas: Similarly, the CO emissfon'rate in ppm(v), 15% Op, dry basis, is: (1-9) Figure 1-16 solves for the concentfations of both NOxand CO; — the conversion factor, F, is obtained from it as a function of %H2O and %O2, A molecular weight of 28.2 was used for the construction of the chart. Then, (@ n/W) x10 xF L642 x (ef W)X 106% F 10) Vn lin ppv) Velin ppm(v) Sample Problem | Determine the NO and CO concentrations, in ppm(v), 15%, dry basis, if25 Ib/h of NOx and 15 Ib/h of CO are present in 550,000 Ib/h of turbine exhaust gas. The percent volume of H20 is 10, and the Percent volume of Opis 11 inthe wet exkaust gas. } Solution: From the figure, F at 10% HzO and 11% Op is 046. Then, Vn equals (25/550,000) x 106 x 0.46 = 21. Also, Ve equals 1.642 x (15/550,000) x 108 0.46 = 206. " ‘Methods of Reducing Pollutants Gas turbine combustor modifications such as staged combustion, premix burning and proper distribution of air and fuel often referred to.as dry NO, control methods are being done toreduce NOx and CO. Siemens, a European supplier of gas turbines is introducing a i5 ‘Waste Heat Boiler Deskbook: nginering) Figure 1-16. Chart for converting NO,, CO emissions. (Che EMISSION-DATA CONVERSION CHART | 2 5 Conversion factor, F 1% 12108 64 H,0, vol. % ‘The Intersections of the vertical and diagonal Imes yleld conversion factors for calculating the volumetric ratios (In ppm) ef CO and NOsWaste Heat Boilers 9 machine guaranteeing 9 ppmvd NO; with no steam or water injection with'CO at less than 8 ppmvd. Some other suppliers have come out with machines with emissions-less than 25 ppmvd NO. However the popular method is steam or water injection in the gas turbine. Adding steam or water reduces the flame temperature and suppresses the formation of NOx by up to 70%. However water injection can increase the CO and UHC (unburned hydrocarbons) especially at low loads, increase the heat rate nearly 5% to make up for the heat of vaporization lost to stack and possibly shorten the life of the. turbine components. Steam injection does not exact as high a heat rate penalty as water injection but the quantity of steam injected is nearly 50% more. However the electrical power output increases as the mass flow and the specific heat of the exhaust gas increases. These methods typically reduce NO;xto 40 to 50 ppmivd range. Note that the addition of water or steam affects the exhaust gas analysis and hence the duty of the HRSG. Gas turbine suppliers can furnish more information on the characteristics of their machines with steam or water injection. The water used for injection should be of high quality, preferably demineralized. If steam is used, it should have a good purity and.as low total dissolved solids as possible. In the Cheng cycle system discussed later, a combination of drum internals plus an external steam separator was used to achieve the desired steam purity of less than 50 parts per billion of solids. If the steam or water is not of good purity, the solids can deposit on the turbine blades resulting in poor performance and even turbine failure. With NO, requirements in the range of 6 to 10 ppmvd in several locations, SCRs (Selective Catalytic Reduction Systems) are seen as the only proven solution to meeting these levels of pollutants, even though they are expensive. Figure 1-17 shows an arrangement of a HRSG with SCR for NOx reduction. The base metal catalyst (vanadium) is sandwiched between modules in an operating temperature regime of 600 to 750°F (for some catalysts). This may require splitting up of the evaporator module so as to obtain the window of temperature at different loads and gas flow conditions. An example to compute the gas temperature. at the SCR at different gas inlet conditions is given in Chapter 3. Ammonia is injected upstream of the catalyst and is mixed with the gas stream0 ‘Waste Heat Boiler Deskbook RSG col Zar annontn tale Aeontraller Sone thrkine 1. HRSG coll 2. air 3. ammonia tank 4. controller 5. gas turbine 6, duct burner 7. ammonia grid 8, SCR or CO catalyst 9. stack Figure 17-a eb, Arengement of catajts for’ NO, and CO reduction, before it reaches the catalyst, where the NO; is converted to Nz and 20. The excess of ammonia called ammonia slip should be minimized by proper controls. Base metal catalysts are suitable for higher temperature operation in exhaust gas streams containing higher levels of SO,. In oil fired ERSGs the excess ammonia can react with $03 to form ammonium sulfate which can deposit on surfaces at low temperatures and cause fouling and corrosion. One has to check the HIRSG performance at different loads and ambientWaste Host Boilers a conditions to ensure that the temperature window for proper opera tion of the catalyst is maintained. Typically catalysts have 80 to 90% removal efficiency. Catalyst for CO removal could be located in a higher temperature zone, even ahead of the superheater. If there is a bumer, itis desirable to shield the catalyst from the radiation by locating it downstream of a screen secticn or superheater One supplier offers'a precious metal catalyst located in’a low temperature zone namely 430 to 530°F for removal of both NO. and CO. The advantage is that the catalyst can be located beyond the evaporator and ahead of the economize and splitting up of the evaporator is avoided. The gas pressure drop can also be lower due to the use of a single catalyst versus two. It is also claimed that the excess ammonia or slip reacts directly with’ the precious metal catalyst to form nitrogen, water and nitrous oxide, which is not a pollutant; however it is a greenhouse gas. An additional gas pressure drop of 2 to 3 in we due to the SCR fas to be considered in the overall evaluation of the system performance as it affects the gas turbine power output. As mentioned elsewhere, every 4 in we reduces the gas turbine output by nearly 1%. With machines reaching the 150 MW range, the loss in power cannot be ignored. BYPASS DAMPERS One of the major auniliaries used in gas turbine plants is the bypass damper, Figure 1-18, The objective of using this is to be able to folate the boiler from the gas turbine and also modulate the gas flow during startup of the boiler so that the steam pressure in the Poller can be raised slowly. However there are several large gas turbine HRSGs which are connected directly to the turbine. If seal air fans are not used, there is likely to be leakage from the gas side to the atmosphere, resulting in loss of energy; this could be 0.5 to 1.0% of the energy in the gas stream, which is not insignificant, Sealing efficiency of a damper could be specified in two ways and itis important to note the difference. One way is specify on the basis of the area. If the leakage efficiency is 99.5% on this basis, it - ———2 Waste Heat Boller Deskbook means that .5% of the area is not sealed and gases can leak through this area. 4 Damper arrangement Bypass slack Damper To stack Waste heat boller ‘igure 118, Bypass damper scheme for gas turbine HRSG, igure 1-18. Bypass damper es ‘The amount of gas that can leak is a function of the differential ‘across the damper, the area of damper and the gas density. If the leakage flow is say 1000 pph when the total flow is 100,000 pph, the sealing efficiency on flow basis is: +100 (1000/100000) x 100 = 98%. A. simplified approach to estimating the leakage across dampers is given below. Considering the flow situation similar to that of an orifice we can obtain an expression for the gas velocity for the leakage flow as follows: Let us derive an expression for the leakage of gas across a damper, stating the assumptions made. Most of the dampers used for isolation of gas or air in ducts are not 100% leakproof. They have a certain percentage of leakage area, ‘which causes there to be a flow of gas across the area. Considering the conditions to be similar to those of flow across an orifice, we havease Heat Bolrs 2 Venu 4 28 a . 205 (4) where 828 velocity through the leakage area, fps Hy = differential pressure across the damper, in. WC density of gas and water, Ib/cu ft acceleration due to gravity, ft/sec? coefficient of discharge, 0.61 ‘The gas flow Win Ib/h may be obtained from vi W600 94 .000- Vs oe D0 (1-12) Where E is the sealing efficiency on an area basis (5). Most dampers have an E value of 95 to 99%. This figure is to be provided by damper manufacturers. A is the duct cross section, 2. Substituting for Ca= 0.61 for pg = 40/(460 + 1) and simplifying, we have W=288400000-) ,/ 460 0% 7 (1-13) where tis the gas or air temperature, °F, Example 6: A damper used in a HRSG system is claimed to have a sealing efficiency of 99%. If the gas flowing in is 120,000 ph at 1000°F and the differential between the gas and atmosphere at ‘he damper is 7 iP wc, determine the leakage flow and the loss in energy. Assume that the damper area is 10 sq ft. Solution: Using Equation (1-13) W = 2684 10 100-99) x YHA = 1720 pp ‘The sealing efficiency on flow basis = 100 — (1720/120000) x 100 ~ 98.56%,“ Waste Heat Boller Deskbook. ‘The loss in energy with reference to 60°F = 1720 x 260 = 45 ‘MMBtu/h. RECENT TRENDS AND THEIR IMPACT ON HRSG DESIGNS With improvements in firing temperature in the gas turbine to 2400-2500°F range and’ higher pressure’ ratio, the exhaust gas temperature is likely’ to be highe:, on the order of 1050 to 1150°F. This may necessitate a larger HRSG with additional heat recovery surfaces and possible reheaters, as in conventional utility boilers, ‘which can improve the combined cycle efficiency by 210 3%. Gas turbines are also being used with solid fuels such as coal and wood. The low Btu gas from the gasifier is fired in the gas turbine as shown in Figure 119 and 1-20 after being cleaned in a hot gas cleanup system. Pressurised fluidised bed combustion/ {gasification system may also be coupled to gas turbine combined cycle systems, Figure 1-21. In another concept, the hot coal combustion products transfer energy to air using a ceramic heat exchanger. The hot air in turn drives the gas turbine. The ceramic exchanger unlike conventional metal exchangers, can withstand 2000 degree C. The hot air can be raised to 1200 degree C, typical inlet temperature of a high efficiency gas turbine. The exhaust gases transfer energy to steam via a HRSG. The system is expected to be 50% more efficient than today’s conventional steam turbine plants. Landfill gas from refuse sites is being widely used in ‘cogeneration projects; reciprocating engines as well as gas turbines use this fuel, which typically has 40-50% methane, 35-45% carbon dioxide, some nitrogen and water vapor. Due to the lower heating value, their combustion temperature is lower and hence generate lower NO, levels. Sulfur compounds and chlorinated fluorinated hydrocarbons present in the gas can cause low temperature corrosion. in parts exposed to the gas. Figure 1-22 shows a typical system for collecting and burning LFG gas in an engine.viaste Heat Boilers Stack Compressor Gas bine FS coat ‘Stew turin = t___ DIRECT COAL-FLELED TURBINE Pai em S {° L_ Economica} = HH = ‘ADVANCED PRESSURIZED FLUIDIZED BED COMBUSTION Figure 1-19, 120, 41-21. Coal based combined eyce systems (Pouer]| ‘yaate Heat Boilers ” | HIRSGS FOR STIG AND CHENG CYCLE SYSTEMS Steam injected gas turbines (STIG) are widely used to improve the power production capabilities of gas turbine systems. This requires modifications to the gas turbine combustion chamber. Unlike steam injection for NO, reduction, STIG systems inject a significant amount of steam into the gas turbine. This alters the mags flow and gas analysis significantly. One of the popular systems which is used. to balance the thermal load and electrical power is the Cheng cycle, Figure 1-23. In this cogeneration system, the saturated steam that is, generated is consumed for process, while low pressure superheated steam is injected into the gas turbine to increase the power output; variations in process steam requirements are handled by varying the fuel input to the duct burner located between the superheater and evaporator. This system offers more advantages compared to a combined cycle system. [oS | Figure 1-223, Cheng cycle heat recovery system. 1. Equal or slightly more electrical power output compared to a combined cycle system with the same gas turbine. 2. Absence of major auxiliaries such as steam turbine and condenser, which results in less complexity and cost. 3. Very little cooling water requirements compared to a combined cycle system.al ‘Waste Heat Boiler Deskbook Figure 1-28b, HRSG for Cheng eyclesystem,[Courtegy ABCO Industris} 4. The flexibility of varying process steam and power output without wasting energy. In a conventional gas turbine cogeneration plant, a diverter would have to be used to bypass gas if the steam demand reduces, resulting a waste of energy. When less steam is required for process in Cheng cycle systém, the excess steam can bevaste Heat Boilers ~” superheated and injected into the gas turbine, thus increasing the electrical power. 5, Asa incidental advantage, the NOy emission is significantly lower in the injection mode. ‘The only drawback is the large amount of good quality feed water that is lost to the atmosphere along with the exhaust gases in the injection mode, ‘The superheater is designed so that it can ran dry if required; this mode would occur when maximum process stem is required. Four ‘modes of Operation are possible with this system, as seen in Figure 1- 2A. 6k e 2 2 suprtsnenrany = Se aI Shean 5 ierci 5 3 3s 7 2. |ieearne wim 2 [RE enarms, win f [inet SERDAR Finn Tae zaa70 STEAM OUTPUT, KG/H Figure 1-24, Operating modes for Cheng eye, 1, unfired, full injection; no process steemy this mode is attractive for plants which have no requirements for steam or have little cooling water but which would like to maximize the power output using a HRSG and gas turbine alone. Due to the increased mass flow and the higher gas specific heat, the gas turbine power output 's increased by more than 60%. 2, fired, full injection; maximum power and process steam are generated.50 ‘Waste Heat Boiler Deskbook 3. fired, no injection; maximum amount of process steam is generated. The superheater runs dry. 4. unfired, no injection; all the steam goes for process. ‘The HRSG can operate not only in the above four extreme points butalso in between with any combination of power and process steam. by varying the injection steam, and the fuel input to the HRSG. The electrical power output based on an Allison 501 KBS ‘machine varies from 35 to 5.5 MW, while the process steam can vary from nil to 50,000 pph as shown in Figure 1-24. The HRSG is designed to handle the’ significant variations in steam load, fring temperatures, gas analysis and exhaust mass flow; a feed water exchanger is located between the deaerator and economizer to preheat the make up water and hence improve the system efficiency. ‘As mentioned earlier, the superheater is designed to run dry. In order to minimize concerns with exfoliation of metal oxides into the turbine due to the varying wet end dry operation, the superheater ‘tubes are alonized on the inside. Table 1-5 shows the gas data with and without injection. This system is in operation in several plants which have varying steam and electrical load, Figure 1-7c. ‘Table 1-5, HRSG Performance in Various Modes a 8 c . 1 Gist Tanp, Fe) season) suas00) og) mT 2 GasTomprmsurec ic) asneasy —30(45) EAS) TORES S GusTempiotnp,Fi) —engs) SSNS) 10955 4 CaTomp wear wap gas) YI) 5. Suck as Temp.) eo, 17008) a0) 8. Ta Seam pp eh) zrmajea) —sroooee7o) —esoonn7e) —zarose75 7, Stam Press, Pa a) ssa 1 Sopot Swan seh fght 9800029) 98008009) 8 SuanTonp, FC) essusy —aseu7 a er a 1996) as) 1. GaP. Dap. (nme) SaI7S3) 8720 63160 sar) CASE A: ai Hn: Fd po Fr Osiwaste Heat Boilers ~ 5 OTHER APPLICATIONS OF GAS TURBINE EXHAUST ‘There are several used to which the exhaust from gas turbines are put such as direct drying of wood products or cement. The exhaust may be used as combustion air in fired heaters. Compressed air or heat transfer fluids such as therminol, glycol may be heated by the exhaust gases. Recuperators or air-heaters resemble HRSG modules except for the fact that compressed air flows inside the tubes, not steam/water. Due to the low tube side heat transfer coefficient, a low fin density is recommended as explained in Appendix A: ENHANCED OIL RECOVERY (EOR) ‘One'of the methods of enhancing the output of oil is through injection of steam into the wells. This is called secondary. oil recovery. Wet steam of 80% quality at high pressures, on the order of 2500 psi, depending on the depth of the well, is generated using gas. turbine exhaust in EOR (enhanced oil recovery) applications. The water available for these applications could be of poor quality having on the order of 1000 to 5000 ppm tatal dissolved solids. The HRSG resembles an economizer module or a forced circulation evaporator and is generally of the once through type. Drum type units have not been used due to their stringent water Quality requirements and also the poor response to load changes. Water enters the coils at one end and come out as wet steam at the other end. Wet steam ensures that solids in the water, which may be of poor quality, are not deposited inside the tubes but dissolved and carried along with the wet steam. One has to be concerned with heat flux inside the tubes Particularly with fired units. Hence selection of fin configuration along the gas flow path has to be made with great care. The tubes at the hot gas end are bare or have a lower ratio of external to internal area and the fin density can be increased at the lower gas temperature regions. With small capacities (on the order of 50 to 100,000 ph), the design is of single stream; multiple streams or Parallel passes are avoided so that concerns with two phase flow i52 ‘Waste Heat Boiler Deskbook instabilitiés are minimal. In large units, methods to improve two phase flow stability, such as increasing the inlet restriction by using. an orifice in each stream, are employed. The units could be unfired or fired. Low temperature systems sich as Kalina cycle, in which the medium inside the HRSG tubes isa varying mixture of ammonia and ‘water are being developed as also organic cycles. RECIPROCATING ENGINE HEAT RECOVERY, Generally the exhaust from reciprocating engines has a lower temperature and hence is used for heating hot water or for generating, low pressure saturated: steam. At full load a reciprocating engine converts about 33% of the available energy into useful work and the remaining 67% is lost to jacket water, to the exhaust gas and loss to atmosphere. The effect of ambient temperature on power output is not as significant as with-a gas turbine, whose output could drop by nearly 3% to 5% with a 10°F increase in ambient temperature. With larger gas turbines evaporative coolers are used to alleviate this concern, but this is an added cost. Reciprocating engines also have a higher part load efficiency and require fuel gas at a much lower pressure. Though they are more ‘expensive, they find applications in systems requiring higher power to heat ratios, hot water or low pressure steam. Figure 1-25 shows a system for generating low pressure steam on the order of 10 to 15 psig, from jacket water and exhaust heat. In an ebullient cooled engine, water is allowed to boil under controlled conditions inside the engine jacket The resulting steam water mixture goes to an external steam separator where the steam is separated and water circulates back. The exhaust gas heat recovery system operates in parallel with engine jacket as shown. WASTE HEAT BOILERS IN HYDROGEN PLANTS. In the, process of manufaccure of hydrogen. through steam reforming of natural gas or napatha several waste gas streams areprste Heat Boilers 2 Engine Woste heat boier Condensate receiver "Figure 125, Heat recovery scheme with reciprocating engine. generated. The steam reforming process converts a mixture of hydrogen, methane, carbon dioxide, carbon monoxide and water vapor in the presence of a catalyst. See Table 1-1 for the gas analysis and temperature ranges. Figure 1-26 shows the scheme for manufacture of hydrogen or ammonia. Reforming process requires that the reaction products at a high pressure be maintained at a high temperature in a reaction furnace heated by flue gases. The flue 405 exiting the furnace at temperatures of 1600 to 1800°F are used to generate high pressure steam in a waste heat boiler. Maximum heat recovery compatible with considerations of high and low temperature corrosion is the objective of the flue gas waste heat boiler. The reaction products in the furnace called the reformed gas, on the other hand, are cooled in a reformed gas boiler from abut 1650°F to 650°F. The exit gas temperature should be controlled within a narrow range at all loads so that the gas stream can be used for further processing; this boiler is called the reformed 8 boiler; steam generation is of secondary imoortance. Figure 1-27 shows a bOuc. ov. sis app-catv With internal Bas bypass control At low gas flow conditions, the exit gas temperature reduces (see Chapter 2 for methods of evaluation and method of computation of bypass flow). Hence a portion of the gas has to be bypassed and mixed with the cooler gases to obtain the desired exit gas temperature. Since the gas pressure is high, a fire tube boiler is appropriate for, this service. The materials of Construction have to be chosen considering the hydrogen damage Possibility.s Waste Heat Bole Deskbook. = or comme —E] mn, As Figure 1.26, Scheme of a hydrogen plant Figure 1-28 shows the Nelsons chart which is used as a guide to ‘material selection. The presence of hydrogen increases the tube side heat transfer coefficient nearly 4 to 5 times that of a flue gas stream or air; refer to Chapter 2 for methods of sizing boilers. Due to the high heat flux associated with the presence of hydrogen, on the order of 80 to 120,000 Btu/sq fth, an elevated drum fire tube boiler in Which the tube sheet can be kept cool with water-steam mixture is preferred over a single shell boiler which has an integral steam space, Figure 1-29.Figure 1-27. Fire tube boiler with intemal gas bypass for a hydrogen plant. [euriesy ABCO Industries} Ferrules are also used, Figure 1-30, to transfer thé heat flux away from the tube shel region Inty the tubes, which are cooled by the circulating steam-water mixture. The front tube sheet is also Protected by refractory/This reduces the temperature differential across the tube sheet. Design of refractory is critical at the front end. This boiler operates by itself in a small plant or could be connected to 2 common steam drum catering to several other boilers as in a large hydrogen or ammonia plant. | Flue gas boilers are generally of water tube design in large Plants generating high pressure high temperature steam for steam turbines and process use; they are of fire tube design in smaller Plants, In some plants the steam drum is made common to the flue gas and reformed gas boilers through a system of downcomers and risers. Figure 1-31 shows a boiler in a small hydrogen plant in which the flue gas stream is cooled in a water tube boiler, while the | seformed gas and converted gas streams are'eSoled it fife tibe Boilers. The steam drum is common to all the boilers. A large hold up time of 5 to 8 minutes from normal level to empty is used in these plants, so |emp auospen-9z-t mt itera asoucaH om on eo cot Ce ° \Weste Heat Boiler Deskbook oxmes 104 wauvtiune wvazaz0v4¥0s ——ite Fist Boies 9 that in the event of feed pump failure there is adequate time to shut down the system or take corrective measures. Figure 1-29, Single shell fre tube boller for incineration heat recovery. [courtesy ABCO Inductriea c BOILERS FOR SULFURIC ACID PLANTS In the manufacture of sulfuric acid through contact process sulfur gases resulting from the combustion of raw sulfur are cooled from about 2100°F to 750°F in a waste heat boiler before being sent to a converter. The exit gas temperature from the boiler has to be maintained within a narrow range of temperature generally by using a gas bypass system, which could be external or internal to the boiler. The bypass duct is refractory lined. Figure 1-32 shows a water ‘ube boiler for a large 2000 tpd sulfuric acid plant. Due to the size, an ternal steam drum with downcomers and risers was used. The casing may be of membrane wall design which maintains the casing at the saturation temperature and hence alleviates the need for a hot casing, which is generally used for this corrosive a58 Waste Heat Boiler Deskbook ‘Figure 1-30. Arrangement of ferrules, inlet vestibule for high gas temperature spplieation. [courtesy ABCO Industries] Par es Figure 11. Combination fire tube, water tube boiler for a small hydrogen plant. l leourtery ABCO Industries)paste Heat BoersPs ‘Waste Heat Boter Deskbook service More information on hot casing design is given in Appendix B. Extended surfaces are used to make the design compact. Due to the high gas inlet temperature and resulting heat flux, the design consists of a few rows of bare tubes followed by tubes with low fin density and then tubes with high fin density. This varying fin configuration also minimizes the tube wall and fin temperature. ‘The gas pressure in some sulfuric acid plants are high, on the order of 3 to 5 psig, which calls for a heavily reinforced casing design; alternatively one could locate the heat transfer surfaces in a pressure vessel. If economizers are used; the feed water temperature has to be-very high, on the’ order of 350°F to minimize dew point concerns, particularly if sulfur trioxide is present. Carbon steel tubes with extended surfaces have been used sitccessfully in several plants in the USA. Some plants, particularly those in Europe and Asia use cast iron gilled surfaces. This design probably offers more protection from corrosion due to its heavy thickness. SULFUR CONDENSERS Sulfur condensers are waste heat boilers generating low pressure steam used for recovering energy from the combustion of hydrogen sulfide gas in a reaction furnace. In these boilers, high temperature gases with a typical analysis as in Table 1-1 enter the boiler, usually a fire tube type at very high temperature, in the range of 2300 to 2600°F. A two pass design is used. The turn around section is designed for a gas temperature of about 900 to 1200°F and the exit gas temper- ature is about 600°F. Some sulfur is removed as liquid at the end of the first pass through drains, which are Kept hot. The boiler is slightly inclined to facilitate draining. The boiler duty has sensible as well as latent heat components due to the condensation of sulfur. Due to the presence of sulfur, sulfur dioxide and water vapor care is taken to design the réfractory at the inlet and exit end of the boiler so as to keep the casing at a high temperature to avoid dew point corrosion concerns. External insulation to forma hot casing (see Appendix B) may also be needed. Figure 1.33 shows a two pass elevated drum fire tube boiler for this service. Several different gas streams may also enter and leave a sulfur reclaimer.§ & 2 4 2 a "si / he Be |@ ‘Waste Heat Boiler Deskbook FLUIDISED BED CAT CRACKER HEAT RECOVERY Heat recovery boilers are used to recover energy from waste gas streams from fiuidised bed catalytic crackers used in refineries. Since the gas stream has particulates, bare tubes are preferred for the heat transfer surfaces unless the particulate concentration is very low. Fig. 1-34 shows a boiler for this application.The gas inlet temperature is on the order of 1400°F. Superheated steam at 600 psig, 750°F is generated. As the gas inlet comes from a far away location with sev- eral tums, a flow modelling study was done to ensure that a non-uni- formity in gas velocity profile across the cross section was minimal. Figure 1-34, Large boiler with supecheaer, evaporator and economizer for fluidised ‘bed cat cracker application. [courtesy ABCO Industrie} INCINERATION AND HEAT RECOVERY Solid Waste Incineration Various types of wastes such as MSW (municipal solid waste), pathological wastes, liquid wastes and sewage sludge are generated by communities and organizations. It is estimated that eachWaste Heat Bouers a individual will be generating a ton of waste per year, which has to be disposed of. In addition about 250 metric tons of hazardous waste are generated each year by chemical, electronic and manufacturing industries. Several methods have been used in the past to handle the waste disposal problem: 4 landbills b compostng e. recycing d. incineration Fhe age old method of using landfills for burying the waste, though in vogue, cannot handle all of the volume; due to various State and local regulations and the NIMBY syndrome (not in my back yard), the Tandfills have to be located very far away from communities with the result that itis expensive to transport the waste; hauling and tipping fees are on the increase; landfill sites ane becoming scarce; these aspects have forced several cities to look for other options to get rid of various wastes, One ofthe methods is composting or size reduction. This applies sage eats such 2s leaves, grass, tree trimmings, MSW and sewage Sudge. Composting consists of three steps: material separation att Sizing, aerobic decomposition and curing. The recyclables are femoved and the remainder is shredded, screened and allowed to decompose in air controlled environment at 150°F, sustained for thine days to destroy pathogens—followed by several weeks of curing, Recycling/ materials recovery is economically feasible for most Incineration is another option widely used for disposing of aSyeral types of wastes. An advantage of incineration is that there is fio to 90% reduction in volume of waste which reduces the cost of tipping and landfill space requirements... In addition, energy is Siailsble from the products of combustion for generating hot water or | Steam for power or procésé: Steam may be used for generating "The overai heat trantor caefficiont Ui is obtained from UUp = (dofhid) + Ug + (dol24Ky) In (dol) + ffi (old) + fly (2-2) | The boiling heat transfer coefficient hp outside the tubes will be very high on the order of 2000 Btu/sq ft h°F and hence the smaller tube side coefficient hj influences Up. DETERMINATION OF fj The tube side coefficient hj consists of two components namely the convective heat transfer coefficient he and the non-luminous, ln; hence hy = he + hn: usually the non-luminous coefficient hy is very Small for flow inside tubes but-may. be computed by-the method | Shown in Chapter 3 using tube inner diameter as the beam length; hi1 ‘Waste Heat Boiler Deskbook is nearly equal to he, the convective heat transfer coefficient if the partial pressure of triatomic gases such as CO2, H,0 is small ‘The tube side coefficient is obtained from the basic equation of Dittus-Boelter: Nu =.023 Re 4 p, 04 2-3) Nusselt Number Ny = hi di/12k 4) Reynolds Number Re= 15-2,/(ud) (2-5) Prandtls Number P, = 11 Cp/ke 2-6) The fluid properties 4 , Cp,k are evaluated at the average bulk temperature of the gas. Substituting the above in Equation (2-3) and simplifying we have: he = 248 w 4 Fy; 1S 2-7) where Fy = (Cpls) 446 (2-8) Factor Fy for air, flue gases are shown in Table 2-1. Gas Data and Fy Factors for Air and Flue Gas Hk BR G wok 05369 01878 1657 2570 .o«s2 0174 1702 0632 02211 11756 2647 .0587 0211 1805 0809 0287 1865 2800 0763 0288 “1991 0968 0350 Z01S “2047 0622" 0358 2159 1103-0412 2198 ‘3080 1089 0429 2314 1232 0873 2205 (3190 .1188 0490 2456 (flue gas analysis used above: % vol CO? = 12, HzO = 12, Nz = 70,02 =6) A few examples will illustrate the use of the above table.crete Btls 103 SP pamplet: E200 pph of air at atmospheric pressure flows inside a tube of 5 inner diameter 1.7 in at an average gas temperature of 600°F. [> © Determine he, Tube outer diameter is 2 in. [== Solution: Fy from above table at 600°F is 0.1810. Substituting in (2-7), he= 2.44 x 2008 x .1810/1.7 18 = 11.78 Btu/sq fthF gas velocity V = .05 wold? 29 density p = 1/0 = 492 x MW x Pg/{(359 x (4660 + T,) x14.7) he gas pressure decreases by nearly 3.2% for every 1000 feet increase altitude. MWis the gas molecular weight, P= 28.90 492/(460 + 600)/359 = 03728 Ibjeu fe (effect of elevation or gas pressure was neglected in the above ‘computation of density.) Hence» =26.82 cu ft/Ib and =928 fps 05x 200%26.2/1.7/1.7 e The effect of gas pressure on he is not significant at low = Pressures say up to 500 psig. See Appendix E for more information on computing i at high pressures. SEaEERCanE A DESIGN EXAMPLE Example 2: Determine the size of fire tube boiler required to cool 100,000 = PPh of flue gas from 1300°F to 474°F; gas analysis is % volume CO? = 12, HO = 12, No = 70, Op = 6 and gas pressure is atmospheric. team Pressure is 150 psig and feed water is at 220°F. Tube size used is 2 x 177 in, (design pressure = 200 psig). Assume that the fouling factors on gas side = .002 and steam | Side = 001 sq fth F/Btu and the metal conduccivity = 25 Btu/feh F. Steam side boiling coefficient is assumed to be 2000 Btu/sq ft h F. Heat loss including margins may be assuumed to be 2%.104 Waste Heat Boiler Deskbook Solution: From Table 2-1, the gas specific heat at the average gas temperature of 5 x (1300 + 474) = 887°F is 0.283. Note that computation of the tube thickness of tubes subject to extemal pressure is an involved procedure. Appendix F deals with the subject. Let us assume that the size 2 x 1.77 in. is adequate for the design pressure of 200 psig. ‘The duty = 100000 x .283 x .98 x (1300 - 474) = 22.91 MMBtu/h; steam flow using 5% blow down is: We 22.91 x 108/{(1195.5-188) +05 x (338 - 188)] = 22570 pph. Enthalpy of saturated steam, feed water were obtained from steam tables, Appendix E. Let us compute hg froin Table 2-1, factor Fy= 2027. A starting point in the selection of tube count N is the mass flow per tube, which ranges from 120 to 200 pounds /hour for a 2 in. tube. Let us use 600 tubes and proceed with the design. w= 100000/600 = 167, Using Equation(2-7), A 44 1678 2007/1.77 13 = 10.62 Btu /eq fthF ‘The method described in Chapter 3 may be used to compute lin, the non-luminous coefficient, It is small for fire tube boilers as the beam length is the tube inner diameter. Partial pressures of CO and HO are 0.12 each; the beam length = 1.77 in. = 1077 ft; for the average gas temperature of 887°F or 1347 R, from Equation(3-17), of Chapter 3, €g=.058. The metal temperature may be assumed to be 20°F higher than the steam temperature; this will be verified later but the effect of this is insignificant, Ji = 5B 0.173 x9 x (13.474 - 8.44) /(1347 - 840) 49 Btu/sq fth °F Hence fy = 10.62-+ 49 = 11.11 Btu sqft h°F AU = 2/(.77 x 11.11) + 002 x 2/1.77 + 001 +1n (2/.1.77) x2/(24 x 25)‘The various resistances (in sq ft h °F/Btu) are: due to gas side heat transfer = 1017 05 side fouling = 00226 etal resistance 0008 outside fouling 001 outside heat transfer = 0005 Hence Up = 9.45 Btu/sq ft h°F Jp order to compute U; based on tube inner surface area, use the I] the log-mean temperature difference 4 fe (1300-474)//ts (1300-366) (474-366)) = 383°F «the saturation temperature at 150 psig is 366 °F) E The required Surface area $1= 22.91 x 106/(383 % 1068) = 5560 5q ft = 14x 1.77 x 600 x L/12; hence the length L = 20 ft Jn order to evaluate the gas pressure drop, the following BX 1S x Fx(L +5 di) x wold (2-10a) The factor 5é;is used to account for the inlet and exit losses; a nd uacutate method would be to compute the losses based on inlet and exit gas velocities. If ferrules are used at the gas inlet the gas velocity at the ‘rule will be very high and the inlet loss is tken ae times the locity head; the exit loss from the tube is taken ae i velocity bead, which is computed as follows. - Velocity head VH = p V2 x 12/(62.4x2x 32} (2-108) ‘The frieion factor f in tubulent flow depends on tube inner Hameter, see Table 2-2 below. .106 Waste Hest Bolle Deskbook ‘Table 2-2: Friction Factors Versus Tube Inner Diameter 37 1 1 2 28 8 4 § 1 028.0245 023.021 0198 018 0175 0165, .018 Density at the average gas temperature of 887°F and at atmospheric pressure ist p= 28:96 492/359 x (B87 + 460)] = 0294 Tb/eu ft;2 = 344 ca t/TB f=-02, w= 167, L=20, di 77; substinating, in (2-10) AP} = 93x10 (167)? x 02 x (20 45x17) x344/177 = 2.96 in we Computing the losses using inlet and exit velocities, we have the following: ‘Loss through the tube = (20/2855) x2.96 = 2.05 in we Gas density at inlet = 28.96 x 492 /(359 x 1760) = .02255 Ib/cu ft YY =.05 x ld? = .05 x 167 / (02255 x 17 x 1.7) = 1045 #t/s 5 VH = 5x 104.5 x 104 .5 x. 02255 x12/(624 x 64) = 37 in we ‘The gas density at exit = .0425 Lb/cuft The gas velocity = 62.7 ft/s VH= Sinwe Total loss through the tubes 2.92 in we To this one must add the turning loss due to inlet or exit duct work connection, ‘Once the tube count and length are determined, the shell diameter may be found for elevated drum boilers. Tubes may be in triangular or square pitch. For triangular pitch it can be shown that Dz = 16rd Qap and for the square pita D. = 125PuN (2-12)aT Boilers w ‘The above formulae do not take into account the space required "for refractory or steam separation. Depending on the thickness of fractory used, the shell diameter could be arrived at. If a single shell boiler were used the shell diameter is = In order to compute the tube wall temperature, the heat flux ‘must be known. Again this may be based on tube inner or outer surface. [> As the steam water mixture is on the outside and critical heat flux or at flux’to- avoid DNB (departure from nucleate boiling)’ refers to’ the steam side, let us compute the heat flux on outer surface basis, qo. fo = Un x (Tg- by) (2-13) One could compute qo along the gas path or at the gas inlet; local Uy and T have to be used in these calculations. Average qo = 945 x (887 ~ 366) = 4923 Btu/aq ft ‘The temperature drops across the various films are computed based on their resistances; see the computation of Up: across the gas film = 4923 x .1017 = 50°F ‘across the inside fouling = 4923 x 00226 = 11°F | across the metal wall 4923 x 0004 = 2°F hence the tube outer wall temperature = 887 —500~11-1.=374°F There is no need to compute iim again as she effect of this wall ‘temperature is insignificant. The heat transfer coefficient Up at the gas inlet would be higher-due to the higher temperature, velocity and gas properties. Hence the heat flux would be higher. However considering the fact Dm108 ‘Waste Heat Boiler Deskbook the heat flux to avoid DNB is in excess of 100,000 Btu/sq ft h, itis not a concern in this design. With high tube side coefficients as in hydro- gen plants, itis pradent to compute Up and qo at the tube sheet inlet. ‘One of the important variables in any design is the operating cost associated with moving the gas through the tubes. Assuming ‘that an induced draft fan is used at the boiler exit, the fan power consumption could be found as follows: P= 1.96 x 10-6x W;4 Pil p) (2-14) where. Wj 100,000, 4P;= 322: The fan drive efficiency 71is assumed t0 be 0.60. The gas density at the exit is: p= 28.96 x 492/[359 x (A7A + 460)] = substituting in (2-14), we have: 0425 Ib/cu ft P 21.96 x1 x3.2/ (0825 x.6) = 24.6kW Using the operating cost and the installed cost, which is impacted by the tube count and length, one could evaluate the life eycle cost; several alternates could be studied by varying the tube size or velocity and one could arrive at various configurations. Life cycle costing could be then used to arrive at the optimum design. EFFECT OF TUBE SIZE ON DESIGN Example 3: Determine the effect of varying the flow per tube or velocity and the tube size forthe same duty. : Solution: Using a computer program, several alternatives were’studied and are presented in Table 2-3 below. The gas inlet velocity was varied from 98 to 162 fps. The following conclusions may be drawn: 1. As the gas velocity increases the surface area reduces and gas pressure drop increases significantly. 2, Smaller the tube, shorter the boiler and smaller the surface area for the same gas pressure drop. This conclusion is important1 ‘Table 2-3: Effect of Tube Size and Gas Velocity 4.751.521 20x177 25x 2238 Zvelips | 98 120 163 98 123 162 98 123 162 tubes 1000 800 600, 750 600 450 470 375 280 15.75 16.7518 18.75 20 21.5 24.75 26.0 28.50 62695003 4298 6519.55584480 «681257104673 19.47 11.08 1370 9.07 10881319 8.73 10.29 12.72 raPi. nwo 2.053.346.23 1.973.206.00 1.95 3.16 6.00 KN -15.625.4474 —18,024.4.45.5 14.9 24.0 456 hen one wants to fit.a boiler into a small’ space. By going in-for = smaller tubes (if the cleanliness permits), one could probably fit it in. [= Note that the length increases to 26 ft from 1675 ft for an [approximate 3.3 in. we pressure drop when we increase the tube size ~ from 1.75 in to 25 in. Surface area for the smaller tube is 5333 sq ft _versus 5710 for the larger tube. 3. Alternatively, higher the gas:velocity results in a longer boiler for the sane size. The length increases frcm 18.75 to 21.5 for a 2 in. tube when the inlet gas velocity changes from 98 to 162 fps. Note that a smaller shell, lesser tube.count could, mean a less ‘expensive boiler though the length and the gas pressure drop could 'be more. The operating cost has to be considered in making the decision as it increases significantly from 15 to 45 kW. 4. In general it could be said that the % variation in length has | more effect on the performance compared to the same % change in tube count. SIMPLIFIED APPROACH TO DESIGN A simplified approach has been developed by the author to quickly arrive at the design of fire tube boilers. A few equations have been developed based on the assumption that the tube side ‘efficient is equal to the convective heat transfer coefficient and ‘hat the’non-luminous heat traiisfer i insignificant? Whe analysis is ‘*tdequate for engineering purposes. \nn Waste Heat Boller Deskbook 1. Ina fire-tube boiler, gas flows inside of the tubes; steam- ‘water mixture flows on the outside. The gas heat-transfer coefficient is small, about 10-20 Btu/(ft2) (hi (°F), compared to the outside coefficient of 2,000-3,000 Beu/(fi2} () CF).The metal resistance is also small and, hence, the gas-side coefficient governs the overall coefficient and the size of the equipment. Since the inside coefficient governs U, we can rewrite Equation (2-2) as follows (neglecting lower-order resistances, such 25 fo, ‘metal resistance, and fouling factors, which contribute to about 5% of, wD: U=095 hy dj /do (2-15) Combining Equations (2-1) , (2-2) and (2-7) , we have, after substituting $ = 3.14d}1N/12, and for low per tube w= Wi/N; QNATF, Wi9S) = 0.606 LNO2/4;08 (2-16) This simple equation relates. several important variables. Given Q, A7, Wand Fi, , one can try combinations of Ldj and N to arrive at a suitable configuration. Also, for given thermal data, (LN 92/4; 5) is constant in Equation (2-16). Fis shown in Table 2-1 for fluegas and air. For other gases, Fy may be computed from Equation (2-8). ‘When a phase change occurs, as in a boiler, AT is written as: ar. (Ty = 3) Tp ~ tf In (1 ~tg) ~ (T2.-te) 2.17) Combining Equations (2-1), (2-16) and (2-17), and simplifying wwe arrive at the following expression: Inl(Ty ~ t(T2— 19] = 0.606 (Fy / Cp) N°2 109402 4508 Factor Fy /Cpis given in Table 2-4.Ege obe Ballers m Equation (2-18) relates the major geometric parameters to jhermal performancé. Using this method, one need not evaluate wat-transfer coefficients. Now, consider pressure diop. The equation that relates the AP} = 9.3% 10S fWyN) 2 (L + 5d) vfaj5 =9.3 x10-5(WYN)2K 20 (2-19) where Kos fll +5 da/dj> (2-20) Combining Equations (2-18) ~(2-20and eliminating N= InfTy~t)/AT2—t o)] = 0.24 F/G) Ky v9 /aP;01 2 ee Kye +5 dg tyag13 (2-22) Kz and K2 appear in Tables 2-1 and 2-6, respectively, as a function of tube LD. and length. In the turbulent range, the friction Factor for cold:drawn tubes is a function of LD. Using Equation (2-22), one can quickly figure the tube diameter [> length that limit tube presoure-drop to desired value, Any two of the three variables NV, L and dj determine thermal performance, as well as gas pressure drop. Let us discuss the conventional design j= procedure: ‘ 1. Assume w, calculate N. Calculate U, using Equations (2-15) and (2-8) . Calculate L after obtaining $ from Equation (2-1). Calculate AP; from Equation (2-10). If the geometry or pressure drop obtained is unsuitable, repeat Steps 1-4. This procedure is lengthy. Some examples will illustrate the simplified approach. The " Preceding equations are valid for single-pass design; however, with minor changes, one can derive the relationships for multipass units (eg, use length = 1/2 for two-pass tinits). -: fn ‘Waste Heat Boiler Deskbook : Table 2-4, Factor Fy /Cp, FaiCp for Air and Fluegas Temperature, °F Fx/Cp FolCp 100 0.6660 0.3730 200 0.8870 0.3945 300 0.7088 o.at4o 400 0.7205 0.4308, 600 0744s 0.4591 1000 0.7680 0.4890 7 1,200 0.7760 0.5030 i Huegas ‘ Temperature,"F, FulCp FxiCp Ho. 200 0.8590 0.3698 og 300 0.6780 9.890 ek 400 0.6920 0.4041 i 600 o.7140 0.4300 200 0.7300 o.a4ea 4,000 07390 0.4696 i 1,200 07480 0.4773 (Fiuegas is assumed to have 12% water vapor by volume) Table 25. Values of Kz as a Function of Tube Diameter and Length, Ai, in 1.00 1.25 1.50 1,75 200 225 250 275 3.00 8 7.09 593 422 345 292 252 220 1.95 1.75 10° 899 675 534 438 370 317 278 248 221 42 10.92 820 648 531 448 385 998 200 2.67 14 12.89 966 7.69 625 527 453 395 350 3.14 18 1488 11.14 8.80 721 607 521 455 402 361 Lf 18 16889 1265 998 B17 688 591 515 456 4.10 20 1892 1416 11.17 914 770 660 576 510 456 22.2098 15.70 12:38 10.12 852 731 837 564 5.05 24 2905 1724 1359 11.11 935 802 699 6.19 5.54 26 25.13 1880 14.81 12.11 10.19 8.74 761 6.74 6.03 28 2724 2037 16.05 19.11 11.00 9.48 B74 730 652fe Tube Bolles om : Table 2-6. ‘Valites of K as a Function of Tube Diameter and Length on no tas os 200228 amas 00 02990 aynz7 00128 gaze 0109 a0 0087 wares ane 1 0m50 arm agiee oozz9 oor acer ened . nate? aoe 12 eam10 a4ats o48S9 e252 on1oH ars OMS nated ce 14 4s70, ous 00385 0277 ats ante o9m9 nmnat omnes tym eae axes» v0H80, 00%2. cores 9008 css o00t9. cawee 20 05789. oxte2- noo. asso. gow. ooret’c1onso” acon coves 22 ome came noms aoa7s ores corte acne aos acer = a8 asero’ azte0 nse70 ac4co naar aor's 008? g0ie ones 2 a7iso ese 90526 a04zs. onzi9, cart oomrt 9404S. a0nse 2 075i o246s ass2 047 one 027 o8m7S 80047 amon Example 4: A firetube waste-heat boiler will cool 65,000.1b/h of fluegas, rom 1,160.to'440°F. Saturation temperature is 350°F. Molecular weight is 28.5, and gas pressure is atmospheric. If L is to be limited to [= 20 ft due to layout, determine N and AP; for wo tube sizes: [1.2 L.77in, @ in. O.D.1.77 in. LD, and (2) 1.75 x'1521 in Solution: ‘ ‘Use Equation (2-18)to find N, Use 2 in. tubes. (F1/Cp) From ‘Table 2-4 is 0.73 for fluegas at the average gas temperature of 0.5 (1,160 + 440) = 800°R. In[1,160 —350)/ (440 -350)] = 2.197 2.197 = 0.606 x 0.73 x N 92 x 20/(66,000 92x 1.77 98) = 0.6089 N°2, N= 611. Compute aP; using Equation (2-19) . From Table 25, Kp is 0.03. Compute the gas specific volume. Density (p) = 285 x 492/[359 (60 + 800)} = 0.031 Ib/ft3 or v = 32.25 ft9/Ib. Substituting in Equation 219 we have: AP; = 9:3 x 105 x (66,000/611)2 x 0.035 32.25 = 1.23 in. HzOm ‘Waste Heat Boiler Deskbook Repeat the exercise with 1.75-in. tubes; length remains at 20 ft. From Equation(2-18), we note that or the same thermal performance and gas flow, N02 L/4;08 = a constant. The above concept comes in handy when one wants to quickly figure the effect of geometry on performance. Hence: 61102 x 20/1.7708 = NO-2 x 20/1,52108 or N= 333. With smaller tubes, one needs fewer for the same duty. This is due to a higher heat-transfer coefficient; however, the gas pressure- drop would be higher. From Table 2-5, K = 0.076 for 1.521-in. tubes. From Equation (2-19): AP; = 9.3 x 10° x (66,000/333)? x 0.076 x 32.25 = 8.95 in. Hz. Example 5 Size heat exchanger for 2.0 in: tubes, with a pressure drop of 3.0 in. HzO. For the same thermal performance, determine the geometry. Solution : Using the conventional approach would take several trials to arrive at the right combination. However, with Equation (2-21), one can determine the geometry rather easily In{(1,160 —350)/(440—350)] 197 = 0.24(F Cp Kyo1/aP P41 From Table 2-4, (F3/Cp) = 0.73; AP;= 3,0 = 32.25. Then: In{(1,160 - 350)/440-350)] = 2197 = 024K; 32.2504 x 0.73/3 05 = 0222Ky , Ki = 9.89. From Table 2-5, we can obtain several combinations of tube diameter and length, that have the same Kz value, and that would yield the.same thermal performance. and pressure. drop, For.the 172... in. LD. tube, L is 21.75 ft. Use Equation (2-18) to calculate the number of tubes: 2.197 606 x 0.73 x 21.75 x N0-2/(66,0000? x 1.708), or N = 402.ns “Thus, several alternative tube geometries can be arrived at for ie same performance, using the preceding approach. One saves a lot time by not calculating heat-transfer coefficients and gas properties. Se iLIFE-CYCLE COSTING Such techniques determine the optimum design, given several iternatives..Here,the’ major operating cost is from moving, the-gas igough the’ system, and the installed’ cost is that of the equipment: and auxiliaties, such as the fan. The life-cycle cost is the sum of the apitalized cost of operation and the installed cost. Lee = Cog +e 2-23) ‘The capitalized cost of operation: Ce = Ca¥(1-¥7 (1 ~ ¥) (2-24) whoro :¥=(7 + e141) . (2-20) ‘The annual cost of operating the fan is es:imated as: Cu = 0.001 PHC, (2-26) where the fan power consumption in kWis: P= 19x 10° WA Pill pn) (2-27) ‘The above procedure is used to evaluate Lec, The alternative with the lowest Lee is usually chosen if the gecmetry is acceptable. Example 6: Evaluate the three options int Examples 4 ahd’5;‘andt détéfmine’ the best one. All three provide the same thermal performance (ga flows, and inlet and exit temperatures are the same). However, theus ‘Waste Heat Boiler Deskbook pressure drops and tube geometries are different. Use the following data: €= 0.06, i= 0.10, T = 10 yr., Ce= 25 mills /kWh, fan efficiency is 0.60, and the system operates for 6,000 h/yr. The installed cost of each is shown in Table 2-7. Calculate Lee for the first case: 4j .77 in., N= 611, L = 20, AP; = 1.23 in. H20, 0 = 32.25 40/1. P19 10-6 x 66,000 x 1.23 x32,25/0.6 = 8.29 kW Ca= 0.001 8.29 x 6,000 x25 =$1,245/yr Y=1.06/11 9636 Ceg= 1,245 x 0.9636 x (1 — 0.9636"°/0.0364 = $10,212, Hence Lee= 10,212 + 75,000 = $85,212. Table 2-7 summarizes the results. Altemative 3 has the lowest Lecand hence should be selected, if the geometry is acceptable. Table 2-7. Life-Cycle Costing Finds The Most Economical Alternative Alternativ 1 2 3 ‘Tube size, in. 2x1I7 1.78%1821 2x 1.77 Number of tubes et 323 402 Length, tt. 20 20 21.75 Gas pressure drop, in, HzO 1.23 2.95 3.0 Power, KW a3 603 20 Ca Sir 1,285 9,085 3,000 Cog 8. 10,212 84,410 24,609 1S 075,000... 50,000. 58,000 Leg 85,212 134,410 82,609pce Tobe Boilers ww PREDICTING THE PERFORMANCE ‘OF A GIVEN BOILER ___ Once a boiler has been designed, its geometry is fixed. Often a plant operates under different load conditions which results in different gas flow, inlet temperature and gas analysis or even steam. Sure. PRS How does the given boiler perform under these conditions? What is the new exit gas temperature or duty or steam production? ‘There is another important need to predict the boiler performance [_._ under different gas or steam conditions, That is to check if there is |, fouling or-accumulation of deposits'on the cutside or inside boiler tubes which can reduce the duty transferred. If the exit gas temperature is higher than predicted, the reason could be fouling on the gas or steam side. The following example | illustrates the procedure to predict the performance of a given boiler. Example 7: The boiler designed in Example 1 has to operate under the “following conditions: Gas flow = 75000 ph; inlet temperature = 1400°F; Steam pressure 175 psig; feed water temperature = 20°F. Determine the exit gas temperature, duty and steam production. Let us use the 2x 1.77 tube size, 600 tubes 20 ft long design. Assume | that the fouling factors and gas analysis are same as before. Strictly speaking, a trial and error procedure is required to solve for the performance; it consists briefly of the following steps: 1. Assume the exit gas temperature; calcvlate the assumed duty Qa. 2, Determine U at the average gas temperature. 3. Determine the log-mean temperature AT. 4, Compute the transferred duty Qy= USAT, 2 ~°5. FQ, and Qj are equal the iteration stops else steps 1 to are Tepeated using a revised exit gas temperature based on Qt ius Waste Heat Boiler Deskbook However an iteration can be avoided as seen below: Qe= WoC x(T-Ta)xhr= So AT ue (2-28) but AT = ((Ty - Tey (Tae Toyll (04 Le) M(Te-Teaee (2-29) combining (2-28) and (2-29) and simplifying, we have: Woe Cp hi = Uo Sol In (Ty=te)1 (Ta -te)] OF Inf( Ty) CTs tl] = Vo Sat WoCp ha (2-30) In the above equation T2is the unknown; Uy is computed assuming that the average gas temperature is same as before and then corrected; since variations in average gas temperature of up to 30 to 40°F do not affect Uo significantly, an iteration can be avoided. Solution: Saturation temperature at 175 psig = 378°F. Assume that the exit gas temperature is 450°F. For the average gas temperature of 925°T, Fy = .024, Cp = .2845; w = 79000/600 = 125 Ib/h. em 2.44 x 125x204 1.77!" = 8.48 Bru/sq ft h F Let us assume that ly does not change, which is reasonable. fin = 49 Hence , 1/Up~ (2/(1.77 x8.97) + 012726 + 0004 + 001 + .0005 = 13 Up = 7.68 Btu/sq ft HF (as otter resistances do not change) ‘The external surface area Sp = 3.14 x 2 x 600 x 20/12 = 6280 sq ft Log-mean temperature difference = e {1400 - 378) - (450 -378)]/In|(1400 - 378)/(450 - 378) = 358°F ‘Assumed duty (Qa = 7500.98 x .2845 x (1400-450) 9.86 MMBeu/hns = 7.68 x 6280 x 358 = 17.27 MMBtu/h ‘It appears that the boiler cannot transfer the assumed duty but something less than that. Another iteration may be performed with Flower Qa. Now let us use Equation (2-30) and see how iteration is avoided. In [0.400 - 378)/ (Ta - 378)] = 7.68 x 6280/(75000 x 2845 x $8) = 2.3065 or (1400 - 378)/(T2 - 378) = exp (2.3065) = 10.04 AT = {(1400-378) ~(480-378)]/bx (1022/102) = 399 (Qt = 399 x 6280 x 7.68 = 19.23 MMBtu/h and hence close. ‘The enthalpy addition to steam is: . (1197.6-188) + .05 x (351-188) = 1017.75 Btu/Ib. Hence the steam production = 19.23 x 10/ 101775 = 18895 pph. |) one can revise the éalzulations based on the new average gas temperature but the variation woul be margin SIMPLIFIED APPROACH TO PREDICTING PERFORMANCE One can simplify Equation(2-30) further and arrive at the Sllowing equation in order to predict the performance offre tube oilers. ‘A few assumptions have to be mad |. 1.-The fouling factors and the non-luminous-heat transfer coefficient are not high and that the overall heat transfer Coefficient is proportional to the convective coefficient he. De19 ‘Waste Heat Boiler Deskbook 2. The gas temperature profile does not vary significantly in the boiler so that the effect of variations in gas properties can be neglected. The above assumptions are fairly reasonable. Let us apply them to arrive at the following expression, From (2-30) assuming that Upis proportional to W;-® and treating all other terms on the right hand side as a constant for a given boiler, we have: In IC Tete)! (To to] = UoSe (WX Cp ht ) = KW? 31) Example 8: ‘Use Equation (2-28) to arrive at K factor from Example 1 and predict the performance of the boiler for conditions in Example 5. Let us compute K factor. Using the design data of Example 1: Wy = 100,000 pph, T= 1300°F 5 = 66°F, Ty = 474°F 1m [(4300 - 366) /(474-366)] = In (8.648) = Kx 100000" or K = 21.573. 1573 = Let us use the new conditions: Wy = 75000; Ty = 1400°F, t; = 378°F. Using (31), In ((1400 - 378)/(T -378)) =.21.573 75000 0 Tz = 482°F, which is close to what we obtained earlier. 2.285 ‘Thus Equation (2-31) gives reasonably gooid results for purposes of quick estimates. The pressure drop may be computed as before. Checks for Fouling: A boiler may deteriorate in performance due to accumulation of deposits inside-or outside of tubes over a period of time. This may be~ due to the nature of the gas, which contains dusty or fouling components, which cannot be removed on line or due to the build up of salts outside the tubes due to poor water chemistry.ja Tube Boilers m 2 One has a tendency to use poor water chemistry in fire tube = Spoilers as they are low pressure boilers. If one analyzes Equation (2- ©), one can see that fouling factors do influence the overall heat = ansfer coefficient and hence the performance. Poor water chemistry “ands to the formation of scale and sludge ~ Scale is a relatively hard and adherent deposit, while sludge softer and less adhérent, The buildup of boiler sealeis most severe those areas in which the maximum heat transfer occurs. Seale buildup is associated with compounds whose solubilities decrease with increasing temperatures. Conversely sludges are precipitated ~ dizectly from the boiler water when their solubilities are exceeded. Scale and sludge increase the resistance to heat transfer and decrease the overall heat transfer coefficient and most importantly increase the tube wall temperature The thermal conductivities of some stale materials are given in Table 2-8 below. Table 2-8 ‘Thermal Conductivities of Scale Materials Material. Thermal Conductivity, Beu/sq ft h f/in. Aoaleite aa Calcium Phosphato, 8 Calcium Sulfate 18 | Magnosium Phosphate 8 | Magnetic ron Oxide ’ 2a | Silcato'seale (Porous) 6 Boiler Stoel 319 Firobrick 7 Insulating brick 7 EFFECT OF SCALE ON BOILER PERFORMANCE Example 9: The boiler in Example 2 is assumed to have a .03 in thick Silicate scale. Determine the performance. when 100,000.pph of. gas at 00 om 1300°F enters the boiler. Steam pressure as before is 150 psig and feed ‘water temperature is 220°F. bm —CS1 Waste Heat Boiler Deskbook Solution: ‘The outside fouling factor ffo is obtained first. ffo = thickness /conductivity = .03/.6 = .05 sq fth F/Btu. Let us assume that hiedin remain unchanged, though they would be slightly more due to the higher average gas temperature in the boiler. Compute Up using the other resistances as before. 1/U = 2/(1111 x1.77) +.05 + 00226 + 0005 + 0004 = 1532; Up = 6.53. Using Equation (2-30) Inf(4300-366)/(T2-366)} = 6.53 x-6280/(100000 x.2845 x98) = 147 oF (1300-366)/ 4349 = (Tp - 366) or Tz = 580°F.Hence Q; = 100000x 2845 x (1300-580) = 20 MMBtu/h. Average heat flux qo = 6.53 x (940-366) = 3748 Btu/sq fth. Drop across outside heat transfer film =.0005 x 3748 = 2°F. Drop across scale = .05 x 3748 = 188°F. Hence outer tube wall temperature = 366 + 188 + 2 = 556°F. This is significantly higher than before. Fouling inside the tubes on the other hand will also result in lower Up, lower duty and possibly increase in gas pressure drop, which can be monitored. For example in the above example when the gas flow was 75000 pph and inlet temperature was 1400°F, if the exit gas temperature had been measured to be 500°F versus the 480°F computed, surely one has to analyze the performance closely and see if the fouling is due to the gas side or water side. A shut down cleaning may be necessary. Fouling is an operational cost and affects the duty and steam production and even the cost of moving the gas through the boiler and must be attended to. (One can also compute the fouling factor by using the following procedure. The expression for U may be written as: 1/U= K/Wy8+ A +f (2-32) Compute A,K and ff based on three sets of performance data -~-bver a period of time-Keep.zecording data and continue.to.estimate.ff-~ overa period of time using K,LI and A. Ifit shows an increasing trend, there is fouling and a shut down cleaning is warranted.Spe Tube Boers rey “ fYDROGEN PLANT BOILERS AAs discussed in Chapter 1, fire tube waste heat boilers are used |g hydrogen plants to cool reformed gas and maintain a desired exit as temperature at all loads, generally through internal gas bypass, FRgore 2-4. The high tube side heat transfer ecefficient results in a high overall heat transfer coefficient and hence heat flux at the ube sheet inlet. The following example illustrates the fact. Example 10: In a hydrogen plant, 50,000 pph of reformed gas at 1650°F is | cooled ina boiler to 650°F. Gas analysis is: % vol CO2 =5, HO =30, CO'= 10, Hy =52, CHg=3, N2=.1. Gas and steem pressures are 285, psig and 635 psig respectively. Determine the heat flux and boiler performance when 190 chrome moly steel tubes of size 15 x 1.14in.. snd 20 feet long are used. Feed water is in at 400°F. Fouling factors on ‘gas side = 003 and steam side=.001. © Solution: ‘The gas properties must be evaluated. Using the methodology ‘discussed in Appendix E, the following data were arrived at: Table 2-9 Data For Process Gas Temp,"F Cp a k 16507517 0891315 | 1150 7080 0731081 | 0.6640 05350751 Let us estimate hc, w = 50000/190 = 263. Hence he = 2.44 x 2638 »(708/.0723)4 x 1031-6/1.1415 = 106 Bru/sq ft MF. hy may be shown to .88 using methods described earlier. 1/Up = (1'5/1.14/106.88) + 001 + 003 x 1.5/1.14 + .0005 + .00069 = .0123 + .001 + .003947 + .0005 + [°00069:=01845. (steam side’coefficient = 2000 Btu/sq fr h'Pand metal” ~~ an esistance = 00068). Hence Up = 54.2 Btu/sq fth F. lm1 ‘Waste Heat Boller Deskbook Substituting in (2-30), In [(1650-495)/(T2-495)) = 54.2 x 1492/(60000 x .99 x .708) = 2.307 (heat loss = 1% and Sp = 3.14 x 1.5% 190 x 20/12 = 1492 sq ft. Hence Tz = 610°F; since the desired exit temperature was 650°F, it appears that the boiler is oversized and hence gas has to be bypassed. ‘The average heat flux is: Up x (fy-t5) = 54.2 x (1150-495) = 35500. As the inlet it can be shown that the clean Up = 78 Btu/sq ft h F, neglecting the fouling factors. Hence the maximum heat flux at inlet = 78 x (1650-495) = 90000 Btu/sq ft h. Heat fluxes of up to 100,000 to 150,000. Btu/sq ft-h can be permitted depending upon the sieam: pressure: and cleanliness of water. It should be noted that the tube wall temperature will be higher if scale forms outside the tubes. ‘The average tube wall temperature is estimated as follow: Drop across the steam-water mixture = 35500 x .0005 = 18°F. Drop ‘across steam side fouling layer = 35500001 = 36°F. Hence outer tube ‘wall temperature = 495 + 18 + 36 = 549°F. If there is scale formation, the: tube: wall temperature will be much higher as shown in-an earlier example. GAS BYPASS CALCULATIONS In process boilers such as those used in hydrogen or sulfurié acid plants, the exit gas temperature from the boiler has to be maintained within a range at all loads. This is usually achieved by bypassing a portion of the gas around or through the boiler. ‘When gas is internally bypassed, it transfers little energy to the steam water mixture as the bypass pipe is usually protected inside by a liner material. Hence the bypass gas remains at a high temperature close to the gas inlet temperature; the rest of the gas is cooled through the boiler tubes and is at a low temperature; the hot bypass gas then mixes with the cooler boiler gas and achieves the desired. exit. gas temperature. The. bypass. flaw-is decided. by: a. temperature controller at boiler exit. Figure 2-4 shows a typical arrangement. The exchanger diameter increases due to the inclusion of the bypass pipe within the tube bundle.Tube Bolles 125 When an external bypass is used, it works in the same way ‘with the difference that the bypass pipe is external to the boiler; yhe exchanger diameter is smaller; however the bypass pipe jperates at a much higher gas temperature and has to be protected th refractory; the damper is also expensive. es. Sil -pample 1: = Determine the amount of gas to be bypassed in Example 10 to achieve 650°F exit gas temperature. olution: Let the gas flow through’ boiler = 45,0C0. pph and 5000 pph ‘through bypass. Estimate'T2 using (2-30): he = (45/50)5 x 106 = 97.4; | hing = 88; Up = 51 Btu/sq ft h F. In [(1650-495)/(T2-495)] ='51 x {1492,/(45000 x.99 x 708) = 2.408. Hence T = 599°F. The approximate _, mixed temperature = (45000 x 599 + 5000 x 1650)/'50000 = 704°F; this is higher than 650°F required. (A more accurate method of estimating the mixed temperature is by using the gas enthalpies; however the above is for illustration purposes only). Since the exit gas temperature is higher, more gas must be “cooled in the boiler. Try 48000 pph through boiler and 2000 pph through bypass. Up = 52.73; T2 = 606°F; mixed temperature = 648°F. This is acceptable. Hence the bypass flow is 2000 pph and has to be sized accordingly. One must also check the boiler performance at part loads and determine the bypass flow for each load and see if the control mechanism can handle the flow based on the resistance of main and bypass paths. DETERMINING HEAT LOSSES FROM BOILER ‘We have been assuming a heat loss on the order of 1 to 2% in the boiler. However the loss could be computed more accurately as follows. Example 12: ae ‘A fire tube boiler is 6 ft in diameter and 2) ft long and operates at 150 psig. Estimate the heat loss from the shell. 3 in. of mineral126 ‘Waste Heat Boler Deskbook fiber insulation is used. Ambient temperature is 70°F and wind velocity is 100 fpm. Casing emissivity is .9. Boiler duty is 23 MMBIu/h. Solution: Using the program described in Appendix B, for a saturation temperature of 36°F it maybe shown that: heat loss from casing 29.6 Btu/sq ft h and the casing temperature is 86°F. See printout below, Figure 25. The total'loss from the casing: = 3:14 x 6x 20 x 29.6 = 11155 Btu/hi The heat loss'= (1155/23 x 106) x 100 = .05%. If the inlet and exit vestibules are considered along with an extemal drum, downcomers, risers, the heat loss will be more and can be computed by summing up the heat loss In each section. The total heat loss used in the calculations may include a margin depending upon the performance guarantees. ‘REFERENCES 1. V Ganapathy, "Applied heal wansfer", Pennwell Books, Tulsa, 1982. 2. V Ganapathy, “Simplified approach to designing heat transfer equipment", Chemical Engineering, April 13, 1990. 3. V. Ganapathy, "Evaluate the performance of waste heat boilers", Chemical Engineering, November 16, 1981, Pg 291. 4. V. Ganapathy, "Steam plant calculations manual", Marcel Dekker, NY, 1984. NOMENCLATURE Ca —Annual cost of operation, $/yr Ce =Cost of electricity, mils/kWh Cp Specific heat of fluid, Btu/Ib F aes Ceo — Capitalized cost of operation, $ doi ~Tube outer and inner diameter, in. @ ~Escalation ratews ~ Friction factor = ffi fo ~ Fouling factors, inside and outside, sq ft h F/Btu F1_ Factor defined in Equation (2-8) HH ~Period of operation, h/yr AH, ~Enthalpy absorbed by steam, Btu/Ib he ~ Convective heat transfer coefficient, Btu/sq ft h F hy ~ Heat transfer coefficient inside tubes, Btu/sq ft h F iq ~Non-luminous heat transfer coefficient, Btu/sq ft h F hy ~Heat loss factor; 1% loss equals a iy of 99 hg ~ Heat transfer coefficient outside tubes, Btu/sq ft h F Te ~Installed cost, $ i ~Interest rate 52a Kz. ~ Factors defined in Equation (2-22) and(2-20) k —Thermal conductivity of fluid, Bu/ft h F Km ~ Thermal conductivity of metal, Btu/ft hE L_ —Length of tube, ft Lec ~Life cycle cost, $ MW. ~ Molecular weight of fluid No =Total number of tubes Nu ~Nusselt number P ~ Electrical fan power consumption, kW AP —Pressure drop, in we ' Pr —Prandt! number 4 ~ Heat flux; subscripts ind o stand for inside and outside tubes, Btu/sq ffh Q ~ Duty, Btu/h; subscripts a and ¢ stand for assumed and transferred Re ~Reynoldsnumber S. ~Surface area, 6q ft; subscripts and o stand for inside and ‘outside tubes T ~Life of equipment, yr Ty,Tz_ —Gas temperature in and out, F AT ~Log-mean temperature difference, F £ — Average gas temperature, F t, ~Saturation temperature, F U_ — Overall heat transfer coefficient, Btu/sq ft h F: ‘Subscripts i anid 0 Stahd for iisideé and outside ©" ~ Specific volume, cu f/Lb: 128 ‘Waste Heat Boiler Deskbook Flow per tube, Lb/h, Total mass flow, Lb/h Gas density, Lb/eu ft ~Gas viscosity, Lb/fth —Efficiency of drive, fraction Srv se i economizer superheater —__| ine tube bolter ‘Figure 2.3. Fre rabe boiler with supeshester and economizer.GAS DUT 5 aL BOILER TUBES y DAMPER BYPASS PIPE Gas aul es boiler (—) bypass pipe damper Figuré24, Arrangement of internal and externa gas by pass scheine.130 Waste Heat Boiler Deskbook seounssensestounccesesuceneavaxencnscususseeeracaasuuunassastitssctee Project: examle soe SHICK-IN TEP-F TEMp2_— cont reap? con cesing 0.09 85.63 0.00 0.000100 0.00 singe 3:00 365100 200.00 0130 400.00 naa HEAT 1088 -Bro/etzn= 29.55123 muber of layers of insulation: 1 2M TROP 70 WIND VEL-fpe 100 ERISSH .9 MAK LOSS-BTU/FTZM 170.664 as sesesesetacaseeusctansssaacereTcaesaseoEese seaeeeesete Figure25. Printout of casing heat loss,‘hapter 3 ‘ater Tube Boilers Water tube boilers are more difficult to'design’ compared to fire = tube boilers-due to the complex arrangement or disposition of heating "surfaces such’as superheaters evaporators and economizers. Figure 3- shows the various configurations that are feasible for a single = pressure system: = __ The superheater may be located ahead of the evaporator or as “discussed in Chapter 1 when the gas inlet temperature is very high, / itmay be located between a screen section ard an evaporator. When __ the degree of superheat is very small, it may be located even behind ‘the evaporator. Also, the supetheater could be in several stages, ‘ach with a different fin configuration or mechanical arrangement to ensure appropriate tube wall and fin tip temperatures and degree of - superheat. ‘The evaporator bundle could have varying fin configurations or tube pitch along the’gas flow direction to handle the gas temperature or slagging concerns. It could also be built in two parts with a SCR Gelective Catalyst Reduction system) in between them as in the case of gas turbine applications. If we add to this a multi- pressure HRSG situation or the presence of a radiant section with different operating parameters and fuel firing, the design procedure ‘would be indeed difficult without a computer. ~ This chapter discusses the methodology of design for various types of heating surfaces such as superheater, evaporator and economizer, which are the building blocks of a complex water tube HRSG; simplified procedures are also suggested, followed by Procedures for evaluating their performance under different gas inlet ~ conditions. oes eeeWaste Heat Boller Deskbook ee a ee «fEvaP] [SH |[EvaP] [ECON (@[EVAP} | SH | [EVAP @) SH | [EvaP| [ECON (@[EVAP] [SH ECON ©) SH | [EVAP (®[EVAP] [SH ” EVAP| [ECON () EVAP ) [ECON | (10) SH (a) PSH ECON Key EVAP = Evoporator ECON = Econonizer SH = Superheater Figure 3-1, Possible Configurstions for « Watertube HRSG.ter Tbe Boilers is much smaller and will weigh lese, large, say exceeding 100,000 ph. 3. Various types of fuels can slag lone'is to use a BatP8ss design with different tube sises no each pass. On-load ‘ming is difficult in fie tube boilers,14 ‘Waste Heat Boiler Deskbook also to be wary of water tube boilers with extended surfaces, which operate at a much higher heat ‘lux and tube wall temperature compared to bare tube designs or fire tube boilers. 9. Due to the higher heat transfer coefficients associated with {gas flow over the tubes, water tube boilers require less surface area and hence the gas pressure drop can be lower than in a fire tube boiler. 10. A water tube boiler will be more expensive in the smaller gas flow range say 50,000 pph or less compared to fire tube type but less expensive for larger mass flows. For some situations such as gas. turbine exhaust where the ratio.of gas to steam flow is high and the pinch point is low, water tube boilers with extended surface may be the only choice as the fire tube equivalent will have a very high gas pressure drop and will be extremely large. Figure 32 Weter-Cooted Membrane Wall D Type Boller Burning Waste Fuels [Courtesy ABCO Industries).Qe USAT = Wor Cpx (T1-T2) x he We Hy 4) = The overall heat transfer coefficient for. finned tubes may be discussed in Apperdix A. The overall MUo= Tig + Ofhiydalds + ff Caley Ho + (AI24K rn) Indole) 6-2) PPS 82S Side coefficient hy = he-+ yy (3-3) The non-luminous coefficient hn can be obiained by the feeods discussed Iter. There are sever correlations for lective Heat ‘Transfer Coefficient ~ ‘The Grimson’s Equation is widely used for determining he. Mus BReN 3-4) Coctticients B and N for bare tubes in inline’ or” staggered epas ras Biven in Table 3-1. The gas propertios ae evaluated 08 film temperature, which may te kee oe average of ‘nd fuid temperatures. The maximum gas velosne eee Dt16 Waste Heat Boller Deskbook used while determining the Reynold’s number. Figure 3-3 shows inline and staggered configurations. It may be noted that in the case of staggered tube bundles the maximum velocity may occur in the : Giagota or tmnaveray Plane copending on te tonmvese and i Tonga pick Tor the manda velo occur n the agonal plan, the ( siogoal pth nce be ony hl ofthe tonevere ch, (Sp-@) L=2 (xed) Lor x= (St + 2 q from the right angled triangle ABC, (52? + 5% = P= [Sp= MPP oF $= [25,4 + 21572 / If Sp=41in,d = 2 in, then Sj should be less than 23 in, for this to occur. oe S1 is usually on the order of 3 in, or more and hence the maximum velocity occurs in the transverse section. Reynolds Number Re= GafI2 3-52) Nusselt Number Nu = he do 12k (3-5) where G = We XI2I[ N(SrdiL) (3-6) Table 3-1. Grimson's Values of B and N Suid 125 Sud a ON OB N Staggered 325 0518 0556 0505 0554 0519 0588 0522 0.562 150 0451 0868 0460 0562 0452 0568 0489 0.568 20 doe 0572 0436 0.568 0482 0555 0449 0.570 30 03:0 0552 0.356 0500 o4¢ 0562 O47 0.574 Inline 125 0388 0592 0.275 0.608 9.100 0.708 0.0633 0752 350 0367 0588 0250 0.620 0101 0702 0.0578 744 20 oie 0570 0259 0.602 0229 0832 0198 O.sAt 30 0200 06010357 0.584 0374 0561 0286 0.50aa ‘Waste Heat Boller Deskbook used while determining the Reynold's number. Figure 3-3 shows inline and staggered configurations. It may be noted that in the case of staggered tube bundles the maximum velocity may occur in the diagonal or transverse plane depending on the transverse and longitudinal pitch. For the maximum velocity to occur in the diagonal plane, the diagonal pitch x need be only half of the transverse pitch. (Sp-@L=2 (xed) L or x= (S¢+ D2 from the right angled triangle-ABC,. (Sy2)? + $% = x = [Sp = AQP or Sy= 2S ed+ PI/2 If S;=4in, d =2.in, then S} should be less than 23 ii for this to occur. 5; is usually on the order of 3 in, or more and hence the maximum velocity occurs in the transverse section. Reynolds Number Re = Gai/I2 (3-52) Nugsolt Number Nu = hed, /T2k (3-56) where G = Wo xI2/{ No(SrdJL] (3-6) ‘Table 3-1. Grimson’s Values of B and N Syd 125 15 2 3 Sud BN 8 N 2B N B Staggered 125 Sie 05598 0505 0554 0518 0555 0822 082 150 0451 0868 0460 0562 0452 0565 0488 0.568 20 0404 0572 0416 0568 0482 0555 0419 ©0570 30 0310 0592 0356 0580 04s 0882 0421 057 Inline 125 0388 0582 0278 0.908 0100 0704 0.0833. 0.752 150 0.367 0586 0250 0620 0101 0702 0.0578 0.744 20 0418 0570 0.299 002 0229 0832 0.108 O.GsE 30 0290 0.6801 0357 0ses 0373 Ose 0286 O.60F17 Staggered Banks 12541520 2000 106 /106 197 10 121 118 108 046 8000, 10: 105 103 98 1m oss sz, oss woo 10 10. 10 095 105 105 102 049 2000 095 085 103 103 147 14s 1081.02 Boog: O98 095 101 101. 140, 106. 100. 096 zoom’. 095° 09s 10 088 1% 02. O98). Ons 2000: 073 073 098 108 1m 128 xia Loe 3000 083° 083 10 1o2 12 120° 10s 102 zooco os 19 10 20 109 307 101 37 2000 068 056 099 10 125 126 WB tad 8000 ost ost 102 102 125 ias 1m 108 zoooo 091 ost ior 10 114 113 1t0 302 "Another widely used Equation is that of Fisherden and Saunders. 35FRe4 Pr i. 3-6) i where prandll number Pr = nCylk 7) Factor F correcting for inline and staggered arrangements is E given in Table 3-2. It will be seen later that with bare tubes, inline arrangement is preferred as the increment in est transfer ccefficient increase in gas pressure drop E Example 1: 150,000 pph of flue gases having an analysis of % CO2= 12, H2O= 12, No = 70 and 02 = 6 by volume flow over a bank of boiler | tubes of size 2 in,, outer diameter with a transverse and longitudinal E pitch of 4 in., and arranged in inline fashion. Tubes /row = 16, length = 10 ft; assume that number of tubes deep exceed 10.‘either a8 Waste Heat Boller Deskbook é a] su. Fieure 3.3 In-Line and Stagsered Tube Bundles,oer Te Boles 19 If the saturation temperature inside the tubes is 353°F and the erage gas temperature is 700°F determine the convective heat igansfer coefficient using both the correlations. tution: From Appendix E on gas properties, at the film temperature of '5(053 + 700) = 526°F, Cp = 2695, pr= 0642 and k= .02344, G = 150000 42/118 x10 x(4-2)] = 5000 Tb sq ft h, Re = 5000 x 2/12/,0642 = 12980. From Table 3-1, B = .229 and N = 632, Nu =.229 x 12980632 = 91 or he 2[(12X 02344) = 91-Hence he = 12.80 Btu/Sq it hF. Letus now determine fc using Equation (3-6). From Table'3-2, F 0642 x :2695/.02344 = 738, Nu = 35x 12980 x-7389 = 93.85 182 Btu/sq fth F. We will see the effect of tube pitch and “configuration on heat transfer later. jetermination of Tube Side Coefficient hy ‘The tube side coefficient for the flow of steam, water or boiling = steam-water mixture plays a role in the estimation of overall heat transfer coefficient U, as seen from Equation (3-2). The basic equation tor obtaining h; for single phase fluids is that of Dittus-Boelter, see reference 1. . NU = 023 Re? Prt (3-8) The above equation may be simplified as shown in chapter 2 and hi may be written as: hex 2.44 W* Cid 3-9) Factor C is given for steam in Table 3-3. For water flowing through tubes at a temperature T in Degree’F, C= 10/4318 + 21410} (3-10) Example 2: Determine hj when 5000 pph of superheated steam flows inside ‘tube of inner diameter 1.78 in. in a superheater. The steam pressurea Waste Heat Boiler Deskbook: {s 1000 psia and the average temperature is 80°F. What is hi when the steam is in saturated condition? Solution: | From Table 33, C = 345. Hence; = 24450008 x 345/1.7818 271 Btu/sq ft h F. Ifthe steam is saturated, then C = 49 from Table 3- 3. Hence hy = 385 Btu/sq ft h F. The mass velocity inside the tubes Gi = 5000 x 144 x 4/314 x 1.78? = 289480 Lb/Sq ft h. Mass velocity of steam inside superheater tubes is dependent on pressure, temperature gad steam pressure, drop, to. ensure uniform flow through parallel CGreuits. It could vary from 250 to 600Lb/sq ft h. ‘in the case of two-phase flow or boiling inside tubes a. value of 2000 to 3000 Btu/sq ft h F may be used for bj. Its effect on U as explained above is not significant. Example 3: ‘Determine h; when 10,000 ph of 500 psia water at an average temperature of 400°F flows in an economizer tube of inner diameter 1.738 in, Solutio From (10), C= 10-1518 214In400) = 921; hj= 2.44 x 100008 x 921/1.738)-8= 1316 Btu/sq ft h F. Note that typical water velocity in ‘economizer tubes ranges from 2to 6 ft/s. From steam tables, Appendix E, the specific volume of compressed water at 400°F = 01185 cu ft/Lb 7 water velocity = .05 x Wo]d2 =..05 x 10000 x .01855/1.738? = 3.07 ft/s. NON-LUMINOUS HEAT TRANSFER COEFFICIENT Due to the partial pressures of triatomic gases present in the fue gases such as carbon dioxide, sulfur dioxide and water vapor, significant amounts of energy can be transferred between the gas stream and the boiler surface due to non-luminous radiation heat transfer, particularly if there is a large amount of steam injected into the exhaust eas stream.ner Tubo Boilers oy ‘The variables that influence the non-luminous heat transfer coefficient are: pO 1. Partial pressure of CO2, HzO, $02, $03, = 2.Beam length L, which is dependent on the pitch and jangement of the tube bundle. 3. Temperature of the gas stream and the surface temperature of She bundle. A number of factors enter into the calculation of non-luminous jadiation heat-transfer coefficients between hot gases and tube | pundles. Net.interchange of radiation between gases and tube bundles given by the basic equation. QA» ofe;t-a Tt} oan For quick practical estimates, a good assumption that [simplifies the above equation is €,= G;-. Then QIA = &; ofFy-TET (3-12) E- From this, the non-luminous heat-transfer coefficient hy may | be obtained as follows: hy = “ ip —Te) (3-13) Where €1=€:+7ew—de o14) The emissivity of hot flue gases is dependent upon the emissivity of carbon dioxide, ¢, emissivity due to water vapor, €w and correction terms 1) and Ae. ‘Thesé values may be obtained from the well known Hottels charts, Figures 5a, 5b, 5c, 5d, once the partial pressures of carbon dioxide and water vapor, the beam length of the heat receiving surface and the gas temperature are known. The beam length, L ofa tube bundle is given by: L = 1-08(SrSi- 785d? Vd (3-15)a Waste Heat Boiler Deskbook Figure 3-4b gives L if S7,S1 and d are known. Hence it may be noted that an estimate of hN Tequires information on partial pressures of CO2 and H20 in flue gas, which in turn implies availability of results of combustion calculations of fuels. This is a lengthy procedure. However, a simplified procedure has been developed based on detailed calculations with various fuels. Equation (3-15) is rewritten, as: hw = Key 3-16) egis calculated for various fuels at normally used excess air factors and presented in Figure 3-4a along with K. The accuracy of this figure has been checked through detailed calculations and a ‘maximum error of 5 percent was observed. Hence for quick engineering estimates of hy, Figures 342 and 3-4b may be used. The advantages are: * Detailed combustion calculations for fuels need not be carried out and thus a lot of time is saved (sometimes even the fuel analysis may not be readily available). * Reference to Hottels’ charts is not necessary, again saving time. * Quick comparisons of € gas a function of fuel used may be made-This is an important factor as furnaces of boilers, fired heaters etc., have to fire diiferent fuels during their operation, and performance with alternative fuels must be known, An example is worked out to illustrate the versatility of the charts. Example 4: A boiler is fired with No. 2 oil at 10 percent excess air. Determine the non-luminous radiation heat transfer coefficient over 1 tube bank when the average flue gas temperature is 1650°F and tube wall temperature is 400°F. ST=5.0in,, S,=35and d=20in,13 5 with Sp=5 and extend to cut line Connect with d = 2 and extend to cut L scale at 7.8 in In Figure 3-4a go up from tg = 1650°F to cut fo = 400°F and move Fight to cut K scale at;26. Extend the vertical from ty = 1650 to cut. [-18and move left to cut the reference line. Connect with point 3 (No. 2 oil and extend to cut eg scale at 0.105. Hence hy = 0.105 (26) = 2.74 » significant compared to the increase in gas pressure drop. The choice {-=0f tube spacings is based on considerations of slagging or bridging, Aistribution of external radiation from flame or cavities, gas pressurer 15 ‘waste Heat Boller Deskbook drop, heat transfer and finally the mechanical ligament efficiency fn case of convection sections using boiler drams. “se the longitudinal pitch increases, #0 will the gas pressure drop and the length of the section. The longitudinal pitch should be saw enough to result in a low gas pressure drop but large enough fo yield a good ligament efficiency. The following example compare Jaime corsus staggered arrangement and analyses the effect of longitudinal pitch. seRRF -p Example 9 750,000 phi of fue gases at 9(0°F are cooled to 500°F ina boiler generating 125 psig steam. Neglecting the effect of ‘non-luminous: aerifcent, study the effect of using longitudinal itches of 3, 4 and 6 coe Rabe dvameter is 2 in, and transverse pitch is 4 in. Boiler duty is 16MMBtu/h. ame Solution: Using the procedure given in example 3, hewas computed for the various configuration. Nor-luminous radiation was not con- Rferoa. The number of rows deep was estimated assuming that the Sther resistances were negligible, The gas pressure drop was com puted for each case using the equations given above. he resus su shown in Table 3-4. ‘Table 3-4. Study of In-line Versus Staggered Arrangements 3 sila 15 20 3.0 Indine Stagg Inline Stagg In-line Stagg, hgGrimson 10.96 1379 1281 13:13 W291 1268 he Fishenden 11.90 1435 1320 1356 13.30 13.20 Friction Factor 0396 0607 04928 0807 0686-0807 Ie Average 1150 1407 13 1335 13.10 12.85 NoofRows 49 4000 430 42 #3 8S, APpinwe 132 212 144 2.28 2.00 238re Tube Boers 155 je following points may be noted: 1. The, staggered arrangement does not have a significant ease in hte over the inline when the longitudinal pitch to meter ratio exceeds 2. It appears attractive below 1.5. However tio is not widely used in industry due to the fact that the | igament efficiency is affected if a drum is used. Also one has to look the bridging of tubes by ash particles cr slag and cleaning siderations if tubes are so close. 2. The gas pressure drop is much higher for the’ staggered arrangement. There may be a trade off if Si/d is less than 1.5 but above this ratio, the heat transfer. coefficients are nearly’ the’same Sor both inline and staggered: but the gas’ pressure. drop with © "There is some increase in beam length with a larger $1/d ratio “but the overall benefits are minimal. Hence an inline configuration th a Sy/d and Si/d ratio of 2 is generally used unless there are her reasons such as external radiation or bridging of tubes with fholten particles from ash, which may require a large Sand . DESIGN OFEVAPORATORS =, ‘The starting point in the design of an evaporator bundle, Figure 3-11, is the estimation of overall heat transfer coefficient. The cross Sectional data such as the number of tubes wide, spacing and length Of tubes are assumed. Based on the mass velocity, Lis estimated. The surface area is then determined followed by the number of rows deep. ‘The gas pressure drop is then evaluated. This is only a feasible design. Several options are possible depending on the size and operating cost, which is influenced by the 888 pressure drop. Optimization is then resorted to, preferably using 8 computer program. The following example illustrates the design approach. Example 10: - 150,000 pph of flue gas is required to be cooled from 900 to 520°F ‘nan evaporator generating saturated steam at 125 psig with 230°F156 ‘Waste Heat Boiler Deskbook feed water. Blow down is 5%. Gas analysis is as follows: % volume C022 12, HzO = 12, Nz = 70 and Oz = 6. Heat loss from the casing = 1.0%. Fouling factors on steam and gas side = .001 sq th F/Btu. Solution: ‘Let us assume the following: Tube size = 2 x -105 in. Ginner diameter = 1.77 in.); number wide = 18; tube length = 10 feet; transverse and longitudinal pitch are 4 in., each. Average gas temperature = 5 x (000 + 520) = 7I0°F = 650 K. Fluid temperature inside tubes = 353°F. Hence film temperature = 5 x (353 + 710) = 531°F. The'gas properties from Appendix E are: Cp= 270, 1 = 0645 and k= .02345. The gas specific heat at the average gas temperature = 277. Duty = Q= 150,000 x99 .277%(900— 520) = 15.60 MMBtu/h. The steam enthalpy change = (1193-198) + 05 x G25-198) = 1001.4 Bru/Lb Hence steam generation = 15.60 x 108/1001.4 = 15,580 Lb/h. {G = 150000 x12/{18 x 10 (4-2) = 5000 Lb/sq ft h, Re = 5000 »2/ (42 x .0645) = 12920. Using, Grimsors Equation, for a spacing of 4 in. inline, Nu © 229 x 12920482 = 908 = hex 2/ (12x 02345). Hence 12.78 Btu/sq fth F. z Let us Compute tte ion-luminous coefficient; partial pressures of CO and HzO = -12, Beam length L = 1.08% (4 x 4 ~.785% 4/2 = 6.95 76 m, K = (B+ 16 X 12) (I~ .38 x 650) x .28/ (24 x 176) = 872, €g = 9x (1-e 872 X 176) = 128, Assume that the surface temperature is 00°F. A reasonable assumption is 30 to 40°F above the average fluid temperature. fy = 173%9X.128 x (1174 8.64)/ (170 860) = .85. (An additional factor of 9 was used to account for the emissivity of the surface.) “Using a hi value of 2000 Bta/sq fth F and fouling factors of 001 for both inside and outside the tubes and a tube metal conductivity of 25 Btu/fth F, 1/U = 1/(85 + 12.78) + 001 + 001 x 2/1.77 + (1/2000) x 2/177 +1 x 1m Q/1.7A)/ 24 x 25) = 07326 + 001 +0011 + .000565 + (0004 = 07633. Hence U = 13.1 Btu/sq fthF. ‘Logrmean temperature difference = {(900-353) ~ (520-353)]/In ~ GAT/I6N) = 320°F. Required surface area = S = 15600000/(320% 13.0) = 3721 sq ft= 3.14 x (2/12) x 18 x 10Naq j oF Nd = 39.5 use 40. Provided surface area = 3768 5q ft. Ro —CC‘CG cs kvl hie ira laa ie #8 scabies cite3 les 157 ie Tube Boll Let us compute the gas pressure drop. Gas density = (MW/359) x 392/(460 + Tg) = 29 x 492/359 x 1170) = 034 Lb/cu ff = 12920--15 [0M .08 x 2] = .0493. APo= 9:3 x 10-10 x 5000? x 40 x .0493/.034 = 35 in we. The average heat flux-q (based on inner diameter) = Ux jg t) = 19.1% (710-959) x2/1.77 = 5285 Buu/sq th. —.” The temperature drop across the various resistances are 13.38 x 669 = 35.72 MMBtu/h. Since the discrepancy is large, and the transferred duty is higher than the assumed duty, another trial is required. Try 650°F as exit gas temperature. There is no need to compute U again as the difference will be marginal. Assumed duty Qa = 130000 x .99 x .289 x (1500 ~ 650) = 31.61 MoBtu/h. aT = 629; U= 13.38; $ = 3768. ‘Transferred duty Q; = 629 x 13.38 3768 = 31.71 MMBtu/h. This '5 dose enough. Hence the new duty is 31.71 MMBtu/h and the ‘oresponding steam flow = (31.71/15.60) x 15,580 = 31670 Lb/b. Gas pressure drop is computed as before. Average gas ‘emperature = 1075°F = 1535 R; P= 29 x 492/359 /1535 = .0259 Lb/cu ft; } f= 9954-18 x (044 + .08 x 2) = 0512: APs = 9.3 x 10-20 4333? x 40% 4512/.0259 = 1.39 in we. Tube wall temperature and heat flux may be ‘mputed as before.16 ‘Waste Heat Boiler Deskbook SIMPLIFIED APPROACH TO PREDICTING EVAPORATOR PERFORMANCE: Using the approach discussed in Chapter 2, Equations (3-25) 10 (3-27), we have: mn [Ty ~te)[(T2-t9)} = U SI(Wo% Cp (3-36) Substituting the values, we have: pe f€500 ~353)/T 353) = 13.38% 3768/(130000 » 289 > 99) = 1.355, Hence T = 649°F. If the effects of non-luminous hieat transfer and fouling are not significant and the gas analysis or temperature do not chang ‘much Tewarrant changes in gas specific heat and other properties, ne (Nominal) . UL Surface area-evap, Sq ft <— 13227 —> <— 16534 —> ~Eco, sq ft <— 5948 ——> <—9922_> 12.Pinch point, F 4 3 BR 13. Approach point, F 15 29 uw 26 COMPARISON OF BARE VERSUS FINNED EVAPORATORS Appendix A compares the design of an evaporator bundle for the same gas parameters with and without finned tubes. The Conclusions are that a finned bundle is more. compact, has fewer rows “tubes resulting in lower gas pressure drop and operating cost and ‘Weighs less. However the heat flux is much higher with finned tubes td care should be taken to ensure that DNB conditions are notre ‘Waste Heat Boiler Deskbook: reached. This can be handled by designing a few rows at the high jas temperature portions with a lower fin density and increasing the fin density as the gas gets cooler. ‘Note also that as discussed in Appendix A, the effect of fouling con tube side is magnified several fold witha finned bundle compared toa bare tube bundle. RADIANT HEAT TRANSFER SURFACES Radiant sections using partially or fully water cooled membrane wall designs are used to cool gas streams at high gas temperatures, Figure 3-12. They generate saturated steam and may operate in parallel with convective evaporators if any. The design procedure is simple and may involve an iteration or two. The higher the partial pressures of triatomic gases, the higher will be the non- luminous radiation’and hence the duty. ‘If a burner is used as in the radiant section of a furnace fired HIRSG, the emissivity of the flame has also to be considered. As explained in Chapter 1, radiant sections are necessary to cov! the gases to below the softening points of eutectis if present so a5 to void bridging or slagging at the convection section. They are also required to cool gases to a reasonable temperature at the superheater if itis used. Example 1 "150,000 ph of flue gases at 1700°F has to be cooled to 1500°F in ‘a radiant section of cross section 8 feet x 10 feet. Saturated steam at 150 psig is generated by the gases. Determine the length. Gas analysis is: % volume of CO2 = 12, #0 = 12, N2= 70and O2= 6, Solution: ‘Let the length = 25 ft. Beam length (see above) = 3.4 x -volume/ surface = 3.4%8 x 10 x 25/(8X 25 x2 + 10 25 x22 108) = Ga ft = 2m. Average gas temperature = 1600 F = 1144 K. Using (18), K (B + 1.6% 12) (1-38 X L148) x.24/(24 x2) = 1945 eg = 9x (1-212) = 29.er Tube Bellers 16 Let average surface temperature of radiant section = 400°F. “surface area for heat transfer = (8 + 10) x 2x 25 = 900 sq ft. ransferred energy = Qy = .173 x 9 x 29 x (20.64 8.68) x 900 = 7.1 15 MMB /h. =. Since there is a large discrepancy and the required duty is more, ‘larger surface area is required. As the furnace length does not significantly change the beam length and hence the gas emissivity, “ye can assume that the transferred duty is proportional to the surface area, Hence the required furnace length = 9.15/7.1) x 25 = 32 eee ee run Nl radiant section ‘dewnconer i GAS t jp OUT menorane wall Pigure 3.42. Radiant Section,0 ‘Waste Heat Boiler Deskbook: HRSG Configurations and Circulation Figures 3-13a-i shows various types of evaporators and the configurations they can assume depending on availability of space, ayout considerations, gas and steam parameters and cost. Figure 3-13a shows a two gas-pass design. In this type of design, the gas makes a 180 degree tur after passing through the first pass. ‘The bottom and top half of the evaporator tubes are separated by a baffle plate. This design offers a lot of flexibility in terms of gas inlet and exit locations. The gas inlet can be from the bottom or side or even from the top and the outlet can be at'the'top or side and bottom. This design occupies a small floor space. When the gas, temperature is high as in fired gas turbine exhaust applications, the first half can consist of bare or low-finned tubes, while the top half can have tubes with a higher fin density. The economizer is located at the gas exit section. The superheater can be located at the gas inlet. ‘The unheated (insulated) downcomers are located at the gas turning section. These pipes carry the water to the mud drum and are responsible for the circulation. Figure 3-13b is a single gas pass cross flow design. The height has no limitations but the width depends on shipping clearances The downcomers are located on either side of the casing external to the gas path. The boiler tubes act 2s risers. In larger units the steam ‘drum can be external to the headers and external downcomers and risers should be used to promote circulation. Figure 3-13c is a widely used design, particularly in D or A or O type boilers or evaporators for diry gas applications, where a large number of bare tubes are required, The last few rows of the evaporator act as downcomers and the rest as risers, No external circulation system is required unless this module is in parallel with another section. In Figure 3-134 several evaporator modules are connected in parallel through a system of downcomers and risers to the steam rum, This design is used when a sreen section is neded as discussed in Chapter 1. A superheater may be Sandwiched between, the to vaporator modules. Also, if a large number of evaporator rows are feeted and each module can be built with only say 22 to 24 rows deep, multiple modules may be required.m saosin so ene noveg y-e ec ots Z ° P : | suewoDunop Suauo>uKop svayo>unop \weer iy Wieste Heat Boiler Deskbook m Figure 313g, Single Gas Pass Evaporator (ADCO Industries}Figure 3.13h, Two Gas Pass Evaporator [ABCO Industries]ma ‘Waste Heat Boiler Deskbook Figure 3-13e shows a forced circulation design with external downcomers and risers and circulation pamps to ensure flow of steam water mixture through the evaporator tubes. Figure 3-13f shows a radiant section with external downcomers and risers. ‘Figure 131, Horizontal Gas How Evaporator (ABCO Industries) CIRCULATION CALCULATIONS The objective of circulation calculations is to ensure that the heated riser tubes are adequately cooled and whether the heat flux inside the tubes is low enough to prevent condition known as Departure from Nucleate Boiling (DNB). Reference 3-1 shows the calculation procedure to evaluate circulation and DNB conditions. In order to perform circulation calculations, which is an iterative process for natural circulation boilers, the static thermal1% i avallable to move the steam water mixture through the system ced agains the various losses such as: 1. Friction losses in down comer pipes. © 2. Friction losses in riser pipes and boiler system. 3. Gravity losses in boiler and riser system. | 4::Acceleration losses due to phase change. 5: Losses in drum internals. To’ start: with, a circulation’ ratio" (CR) is assumed and the fous: losses’are determined and balanced, against the thermal id available. If they don’t: balance, another iteration is ranted “ Once the CR is determined, the heat flux and the vapor quality [dryness fraction in the riser tubes are determined along its length Gf at critical points and compared with allowable dryness fraction tions. ‘The system is safe as long as the actual heat flux is lower than Pf allowable. CR in natural circulation botlers ranges from 5 0 20 Ppending on the pressure and layout. A system need not be designed BP a particular CR as long as the actual heat flux is lower than the Ellowable for the existing steam conditions E-_In forced circulation’ units, the system can be designed for a icular CR say 4 to 10 as the losses in the evaporator tubes are sally large compared to natural circulation units and the thermal dis inadequate to handle the losses. ‘The circulation pump can be selected to handle the additional $ses. The allowable heat flux inside horizon‘al evaporator tubes is Bech lower than that inside vertical tubes. Hence one has to be ul in'selecting the fin configuration in fired HRSGs. The tube le velocities should also be such that the vapor does not separate $5 the bulk water-steam mixture. These factors and the tube Higuration eventually determine the CR and pump sizing in forced f foulation HRSGs,176 ‘Waste Heat Boiler Deskbook DESIGN OF SUPERHEATERS Superheaters are used in waste heat boilers to superheat the steam that is generated in the waste heat boiler; sometimes addi tional steam from other boilers may also be superheated. Saturated Steam may often be taken off for process from the waste heat boiler evaporator drum and the balance could be superheated. ‘These situations arise in chemical plants and refineries. Hence one hhas to understand the process requirements before designing the ‘equipment. Unlike utility boilers, the steam temperature and pressure are generally low, varying from 600 to 1500 psig and 600°F to 900°F. If ream is used fr injection into a gas turbine or for use in steam turbine, hhigh purity steam is required, on the order of 50 to 200 ppb solids in steam. Hence good steam purifying equipment may be required. ‘Depending on the type of waste heat boiler used and the space or layout considerations, several types of superheater designs are feasible. Figure 3-14a-f shows a few types. Figure 3-14a_ shows a superheater with vertical headers and horizontal tubes. with the gas flow direction being horizontal. This Gesign fits well in a. short but very wide boiler, such as the two- gas pass design shown in Fig. 3-13a. The headers are drainable. Batfles could be used in the headers to match the number of parallel streams. Figure 314b shows a unit which has horizontal tubes with horizontal headers and is designed for horizontal gas flow. The headers are baffled to account for the stream requirement; the unit is drainable and finds application in short boilers which have layout limitations. Figure 3-Mde_ shows a drainable inverted loop superheater; this {s used for small degree of superheat in units where space is a concern. “The unit could be buried within a convection bundle or space. Figure 314d shows a design that may be used within a convection bundle. The advantage is that it could be located in a convenient gas temperature zone so as to obiain a bandwidth of steam temperature without the need for control; an example would be 2 fired D-type or O-type boiler. One can also avoid incidence of alc eee e8887 tal tubes but vertical gas flow direction. F Figure'3-14f shows a superheater for tall, less wide units, as in ‘ine applications. The tubes are vertical and drainable using at| the bottom of each tube. By splitting the header, E In addition, these designs could be arranged in parallel, counter xed flaw configuration. The unit could bbe split up into two-or -modules for steam temperature:controt purposes and could have ‘or multiple stream depending on steam sidepressure drop and be. 2 20} 0000 10.0 0 /0| 0000 20/0) o000 cob woog8 este Heat Boller Deskbook Figured dg. ‘A Drainable Supethester TABCO Industries] Figure 3.14h. A Large Superheater Under Erection TABCO Industries}Re Tube Boilers 1 Beir location, superheaters could be classified as radiant, semi- Bi snt or convective. In general, ifthe steam temperature is not say below 900°F, supetheaters are designed as convective units. ‘Avsuperheater receives energy from: three ‘sources, namely mab radiation, convection and non-luminous radiation. B Ob=O, + Qc+ Qn = Ws AHy = We xCyx(Ty-T2) xhp (3°38) External radiation Q, may be due to presence of a flame in the ce or due to a cavity. Later we shall see how to estimate the Bprtribution from this source and the effect of tube pitch on energy Q-Qp= Qe + Qn= USAT (3-39) Uis computed as shown earlier after computing he, hy and hj. steam side pressure drop is obtained from: EAP, =336 flew? v/a (3-40) A superheater has to be designed to superheat 30,000 pph of [uated steam at 635 psia to 782°F using 150,(00 pph of flue gases at RF. Gas analysis is: % volume CO2 = 12, HO = 12, Np = 70, 02 = lube size is 2 x .120 and tubes are of carbon and low alloy steel, SA GA and SA 213 T11 materials. Tube pitch is 4 in., square. Assume Bing factors of 001 sq ft h F/Btu on each side and a 1% heat loss.180 ‘Waste Heat Boiler Deskbook, ‘Assume also that the superheater is of convective type and the configuration is counter flow. Sotution: ‘The enthalpy of steam at exit from steam tables = 1397.5 Btu/Lb and that at inlet is 1202.6. Hence the duty = 30000 x (1397.5 ~ 1202.6) = 5847000 Btu/h. The gas temperature drop = 5847000/(150000 X99 x 2851) = 138°F. Hence the exit gas temperature = 862°F, The average gas temperature = 931°F; the average steam temperature is 637°F; hence the gas film temperature = 784°F. Note that the steam inlet temperature is 492°F. From Appendix E, the gas properties at 7RA°E are: Cp = 2795, j= 07566, k = 02826. ‘Assume that there are 18 tubes/row, and the effective length is oft. Use 9 streams for steam flow. Hence w = 3333 Ib/h, G = 150000 x 12/118 x 10x (4-2) = 5000; Re = 5000 2/12/.07566 = 11015. Using Grimsons correlation, Nu = 229 x 11015692 = $2.1 = hex 2/(32-x..02826); Hence, he = 13.90. hn using the equations (3-17) and {G-18)can he shown to be 15. hr using a C factor of 385 = 2.44 x 3883 X335/1.73818 = 198, 1/U = 1/139 + 15) + 001 x 2/1738 + 2 In (2/1.738)/24/20 + 001 + 2/(1.738 x 198) = 074; hence U = 13.5 Btu/sq fhe [Log-mean temperature difference for counter flow arrangement {s: (1000 783) - (862 - 492) /In (370/217) = 287°F. Hence surface area 5 required = 5847000 /(13.5 x 287) = 1509 sq ft = 3.14 x2 18 x 10% 1Na/2; hence Na= 16 rows. In order to compute the pressure crop inside the tubes, the effective length should be known. Since there are 9 streams and 18 sections and 16 rows, Lg = 10% 2x 16 x32 2.5 x 1.738 = 459 ft, using 2.5xdj for the effective length of each tube bend. Using Equation(3- 40), AP; = 3.36 x.02 459 x 3.332 x .935/1.7385 = 20.3 psi where .935 is the specific volume of steam at the average temperature. ‘Gas pressure drop may be obtained from the charts or by using Equations (3-19) to (3-21). Gas density at the average temperature of 931°F. P= 0285 Lb/eu ft; f= 1015-5 x (04 + 08 x2) = 05. AP = 93 10-10 50002 x 16 x.05/.0285 = 66 in. we. 7"Tube Boilers 181 jerror as discussed in Chapter 4 and is illustrated by the swing example. _ Determine the performance of the superhester designed above the gas flow is 125,000 pph and inlet temperature is 950°F. E The procedure involves tial and error. The steps are: 1. Assume an exit temperature for steamy compute assumed duty = 750°F; the enthalpy of steam 378.5. Assumed duty = Qu = 25000 x (1378.5 — 1202.6) = 4397500 iy. . ©. The gas temperature drop = 4397500/125000/99 /-283 = 125°F. mce the exit gas temperature = 825°F. Log-mean temperature = (125000/150000)$92 x 13.92 = 12.4; in = Beu/h. Since the transferred duty is higher than that assumed, the eam exit temperature will be higher. It can be shown that the duty lances at 4570000 Btu/th; the exit steam temperature = 761°F and METHOD OF ERFORMANCE CALCULATIONS ‘Another method of predicting the performance of superheaters €conomizers is by using the NTU (Number of Transfer Units)18 ‘Waste Heat Boller Deskbook method. This procedure avoids iteration in the sense that once U is determined, the duty can be arrived at without assuming the exit temperature of either fluid. It should be noted however that in order to estimate U, one should have a feel for the gas/steam temperature profiles. The duty Qe = Cmnin (Ti~ tsi) (3-41) exp NTU -C)) Where 1—C exp[ NTU (1-C)] (3-42) for counter flow arrangement is given by: NTU = US/Cmin (3-43) and C = (WCP) minl(WCP max (3-44) Example 18: Solve example 6 using the NTU method. Solution: (WCpmix = 25000 x 68 = 17000. (The specific heat of steam in the temperature band is estimated as .68 Btu/Lb F). Let us assume that U has been computed; U = 12.12 Btu/sq ft h F. NTU = 12.12 x 1507/1700 = 1064. C= 17000/35071 = 485. 1064x4854 4) — Cer 126A) e=[l-e 86 thence Qt = 586 x 17000 x (950 ~ 491) = 4570000 Btu/h. The gas and steam exit temperatures may be computed as before. METAL TEMPERATURE CALCULATIONS Metal temperature calculations are important for superheaters as they form the basis for selection of materials and thickness of tubes. The allowable stress is computed at the mid-wall temperature13 thickness as discussed in Appendix F. ‘The outer wall temperature determines the material to be used ‘oxidation considerations. Table 3-7 shows the maximum outer BP temperatures acceptable for some commonly used superheater Fr erials. Table 3-8 (pg 184) gives the thermal conductivity data to B ised in computing the wall temperatures. ‘Table 3-7. Ferrous Tube Material Specification and ‘Maximum Allowable Metal Temperatures Maximum Allowable f_ Nateriat, ASME Specification Temperature, *F ‘Saaz 980 A210 gr At 950 SAzI0gFC so SA-178 gr A 900 50-170 gC 950 '$A-200 gr Tae 975 sais grT2 1.025 A219 ge TH 1.050 SA213 gr T22 135 SA-213 pr TO 11200 $A-213 gr TP 204 HL 400 5A210 gr TP 321 H 1400 - There are several aspects to be looked into before performing etal temperature calculations. While the overall heat transfer elt can be seen that a central inlet and outlet arrangement izes the non-uniformity in flow, while an end inlet and exit } maximize flow non-uniformity. Multiple entry and exit help ppimize flow non-uniformity. A low pressure drop through the theater is also not preferred as at low loads, the maldis-* ‘Waste Heat Boiler Deskbook 188 «oe we & oF Tm Ae ca woe wa vee IN Kor ID Hse ua wou a ae uo ee ee e @ ee « e @ on aK "OH sais son 20D «we 8 « @ w @ e33 (AI ON eee 8 & & «0 Pais oan West Gort oocr ott @0ir aio! 005 00d oo: cod 00s oo» owe 000 iy hava sqeiap30 ArsnoMpLOD Te EAI185 oy sseydeoriu # Jo oBreypsp pur Addn yo pouiaut ayi uo Sipuedop zoproy eyo yiBua| ayy Suojeuopnaysnp omssoad anes “-g amy w pa a OE ain har”186 ‘Waste Heat Boller Deskbook of each stream, the flow can vary, like the current flowing through several different resistances arranged in parallel. In addition, one has to be careful atout the non-uniformity in ‘gas flow across the superheater. If the bundle is located in a turning section or a bend there is likely to be a large variation in the velocity profile across the duct cross-section, resulting in high local heat fluxes in some portions. In order to compute the tube wall temperature," the local ‘maximum heat flux at the gas temperature zone desired should be obtained. ae = 12QSF Dl) + (Tet (he h) + fills + Bla + dali + de Iida) 24Ke) (3-45) ‘The second term is obvious; it is the product of overall heat transfer coefficient and the difference between the gas and steam temperature. In this calculation, the local maximum convective and non-luminous coefficients should be used. While estimating fi, the actual flow through the tube in question should be used after accounting for mal-distribution due to header arrangement and different tube lengths of a parallel muti-stream assembly. ‘The first term involves external direct radiation Q,, which can bbe due to presence of flame in the furnace or radiation from a cavity upstream of the superheater. The factor ? accounts for the effect of tube pitch and the fraction absorbed by each row of tubes in a bundle. EXTERNAL RADIATION Q; If the superheater is located at the exit of a furnace or a cavity and is capable of receiving radiation from the flames or hot gases, Qr may be estimated as follows. Q, = 17Bx10~ "xe gx OH Ty + 460)" ite + 460)'] XA (3-46) g is determined as discussed earlier; the beamlength used ‘would be that of the cavity or space ahead of the superheater bundle. A is the opening to the tube bundle in sq ft.be Bollers 187 the fraction F absorbed by each row depends on the shielding of the tube rows and is affected by the pitch and diameter of es. The first row facing the radiation would naturally receive jum energy; the factor F decreases as we oss tube rows. SUTION OF RADIATION TO TUBES. External radiation may be estimated by the following formula: AIR. ds sin (dls) + {(s4d)— 1) — sep G-47¥ [be term sinc? (dis in radians.) bThe second row: would absorb af7-2) of the energy and the third Fala] and so on. can be seen that significant amount of extemal radiation gets bed in the first few rows particularly if the si/d ratio is small. it can be seen that the radiation to the first row can be sized. by using. large s;[d ratio. This is one of the reasons for ‘a wide pitch in radiant sections. The energy gets distributed in even manner if syd is large. aple 18: . supetheater is located beyond a cavity of dimensions 3 ft Se-10 ft high and 5 ft deep, with the opening of 50 sq ft to the theater. Gas temperature is 1600°F and % volume CO2 and HO each. If the average surface temperature cf the superheater is 700°R, determine the external radiation to the bundle. lf the tube size is 2 in. OD and the pitch is 8 in. wide, inline, “mine the energy to each row if there are 6 rows deep. What ‘We have to obtain eg and Q, first. The beam length of the 3.4 x volume/surface = 1.7/(1/10 + 1/3 + 1/5) =2.68 ft = 817 tial pressures of COz and HzO = .12. Average gas temperature x Lid x 24/(24 x 198. Qr= 1173 x 9x 198 190 = 303. eg= 9x (e308 x 817) £20.64 11.6%) x 50 =250,000 Btu/h.188 Waste Host Boiler Deskbook Using a pitch of 8 in,, s1/d = 4; substituting in (3-46) a =3.14/2/4 ~ 25x (sin 1(.25) + 155-4] = 361. row 2 would receive: (1-361) x .361 = 231 row 3=1~[1-(361 +.231)x 361] = 147 row 4=1~[1-4361 + 231 +.147)x 361] = 094 Hence the first row receives 250000 .361 = 90250 Bru/h and so on. Ifsy/d = 6,a = 248, row 2 = 186, row 3= .140, row 4=.106 and so oon. Note that the first row receives less energy now and hence the heat flux will be lower. It can also be seea that a substantial amount of energy gets absorbed by a fewer rows if syd is small. As a result, the heat flux will be higher when the sy/d ratio is smaller. Heat flux due to external radiation in row 1 with sy/d = 4 will be, assuming a height of 10 ft and 7 tubes /row, q= 90250 12 /2x 7x 10) = 7735 Btu/sq ft h. TTo this must be added the heat flux due to convection and non- luminous coefficients in order to arrive at the total heat flux, which will be used in the estimation of tube wall temperatures. 4max is computed by using the local maximum convective and non-luminous heat transfer coefficients. Then, the temperature drop across each resistance is estimated in order to obtain the tube wall temperature, For example in order to compute the maximum outer wall temperature in example 6, the gas inlet conditions should be used swith the gas temperature of 1000°F and steam temperature of 783°F. Example 20: Determine the maximum outer wall temperature for the superheater in design case. Neglect extemal radiation. Solution: At the gas inlet zone the gas temperature = 1000°F and steam temperature = 783°F. he = 14.11 and hg = 1.935. Using a 20% non— uniformity in he , the local U may be computed. ke = 1.2% 14.1 16.95. 1/U = 1/(16.95 + 1.935) + 001 x 2/1.738 + 001 + .000585 + 2/1.738/198. Hence U = 1624; q = 16.24 x (1000-783) = 3524 Btu/sq ft h. Temperature drop across gas film = 3524/(1695 + 1.995) = 187°F; drop across the fouling layer = 3524 x 001 = 3.5°F. Hence the outer ‘wall = 1000-187-3.5 = 810°F; since the drop across the tube wall isPie Boilers 189 2°F, (3524 x .000585), the above value may be used for I selection, te that if external radiation had been present, the tube wall ure will increase. FIN PARALLEL STREAMS Ein‘order to determine the tube wall temperatures, the actual jow through each tube and not the average should be used. If fective length of each stream or parallel pass is the same, then low through each tube will be the same; however due to the construction and configuration used, the parallel passes may ifferent effective lengths, which can result in different flow ighthe parallel passes. The tubes with the lowest flow should By oked at carefully while determining the wall temperatures. fective length of one is 50 ft and the other, 60 ft. IF the total fin the superheater is 300,000 pph, determine the flow in each ‘There are a total of 60 parallel tubes handling 300,000 pph of ty for the estimation of hi, a value of 300,000/60 = 5000 pph low in each pass is required. Let us assume that the non— mity is due to the difference in the effective length of the pres Let us assume that 30 tubes have Le = 50 and the other 30 Le 60 ft. Since the pressure drop between the headers is the same, AP= e; where J is a constant. Let wa and wp be the flow in the pass an effective length of 50 and 60 ft respectively. Then, wa? x 50 =190 \Weste Heat Boiler Deskbook = 10,000. Hence wp = 4722 pph and wa = 5278 ph. Thus the tubes with an effective length of 60 ft have a flow of 4722 pph through ‘them, while 5278 pph flows through the shorter tubes. “The effect of variations in specific volume was neglected in the above analysis; another iteration could be made by computing the pressure drops using the actual conditions in each pass. MINIMIZING SUPERHEATER WALL TEMPERATURES Superheater tube wall temperatures could be reduced by locating them in’a cooler gas zone; however this would reduce the log-mean temperature and thus increase the surface area requirements. Parallel flow arrangement could be used versus counter flow. The high steam temperature section will then be at the cooler gas region, resulting in lower wall temperatures. The tube side coefficient could be increased by using a higher flow per tube; this increases the steam pressure drop and also improves the flow distribution at low loads, although this may incur an operating penalty in the form of additional pump power consumption; smaller diameter tubes minimize the non-uniformity around the tube periphery and hence rediice the wall temperatures. Wherever radiant superheaters are used, it may help to use a few rows on convection screen ahead of them to absorb the direct radiant energy from cavity or furnace. Header arrangements should be such that the mal-distribution is minimized. A central inlet and ‘exit is preferred to end connections. Maltiple inlets and exits also minimize the mal-distribution. Whenever the steam temperature has to be controlled, it is desirable to split up the superheater into two segments and use a desuperheater spray or some other control mechanism between the stages. This ensures that the tubes are not overheated. One has to be more careful with finned superheaters as the tube wall and fin tip temperatures can increase significantly if a high fin density is used. The procedure for computation of wall and fin tip temperatures for finned tubes is outlined in Appendix A. ‘Steam purity is particularly important. If the drum internals are not properly designed, wet steam can be generated and solids can 7 aFe ube Bolles m1 sposited inside the tubes; as discussed in chapter 2, the tube B jiemperature can increase as-a result of any deposit or scale ition. Hence good drum internals are required if steam fer perature is high or if the superheater is located in a high gas PEperature' zone. Good steam purity also minimizes deposit sition in steam or gas turbine blades. TEMPERATURE CONTROL E Sein utility or industrial boilers the steam temperature may to be controlled in waste heat boilers. Several options are able for doing this such as the cooling of a controlled amount of iin the drum, bypassing some steam flow around the rlieater so that its duty is reduced and the use of spray water for estage desuperheating Spray attemperation or desupetheating is widely used in waste oilers. One has to ensure however that the water that is being gi {6d into the steam has a good purity. Demineralized water ot ate is preferred. A rule of thumb is that the total solids in ‘water nsed for injection should be mecrly the same as the [) dissolved solids expected in the steam ‘Computation of spray water is important from heat and mass ince considerations. The flow through the superheater ahead of emperator will be lower by the amount of spray quantity as ithe flow in the economizer and hence affects the gas and steam Br Perature profiles. Presented below is the method of computation ePiay quantity and a BASIC program that permits rapid lation of the same. ERMINE SPRAY WATER DESUPERHEAT STEAM {hn steam plants, steam is often desuperheated in spray type g iPerators to control the final steam conditions Figure 3-16. With BPelp of this program written BASIC, one can determine several tant variables such as:‘Waste Hest Boiler Deskbook 192 savas > soe onto insets, wea 305 SSI 4 ant 2 z 4 oy, eo nountienry -q 9 ¥91-E LEEe Ballers wa ied value. din temperature obtained by spraying a given quantity of peratures. je’equation for energy balance around a desuperheater is: ce With = (Wes Wille (3-48) £100,000 Ib per hr of superheated steam at 1500 psia and 900°F is, Seduced 0 775°F in a spray type aitempento. Deeine the (inode, the screen asks for steam pressure (100), temperature Jpsteam flow (100,000), and spray water temperature (300), Rare inputted as shown in Figure 3-17, iEhe computer prints out the saturation temperature (97°F at ipsia pressure) and the maximum quantity of spray water to e’steam temperature cannot get lower than the saturation rature in normal spray operations. Do not inject more than the iim spray water quantity, which is 28,889 Ib per hr. The screen asks for the spray water flow or the final steam ature, which is inputted as shown. The final results are aout. For given conditions, spray water required is 7743 Ib per ithalpies of steam and water are also shown in Figure 3-17.198 Waste Heat Boiler Deskbook Figure 3.17, Program Listing with Resuls [Hesting, Piping, Air-Conditioning) 2000 Te203.26(0-92) (1.88707 /24.696 Joao wasnoenot fara aaoaoere (2642-527 Gozo oleaeqov2tz0i/3/ saneonen sayeisazeacon [efe eied.s4-nezecon/rerisaenent-06/m) "204601 Be Riceetisteeasisynone195, SO gett arene) sppaaoree eS) cen) (n) eesey-AeE-)) * ae c/a) oor shaar WAGER TLOH,7DNAL STERN PEnP(2NPOT 2580 FOR RIOR) =" 78,2 Be ep v0 coro 200, GE wocon nasa AGorrDie STEAK ES WICHECK GPAAE WATER QURNTITI“:COTO 320 Sector recwon (Raa) oS (22-27) 3r mach tare tael3e(cI-Feas) Siic-Somrrecreecniies)vniey racks nam sonteatonsscomo 280 etnes saraires TT mecaeespaaer® *Past Waste Heat Boiler Deskbook: Example 23: 50,000 Ib per hr of steam at 800 psia and 750°F is to be sprayed ‘with 2000 Ib per hr of feedwater at 250°F. Determine the final steam conditions. Solution Key in the data listed in Figure 3-17. Final steam temperature is 678°F. The maximum spray water to saturate the steam is 8669 Ib per hr, Input data and results are shown in Figure 3-17. DESIGN OF ECONOMIZERS ‘The design procedure for economizers is similar to that for supetheaters and hence will not be elabcrated. The arrangement can be as in Figure 3-14. Steaming at part loads is a concern in gas turbine applications. Methods of overcoming this concern are outlined in Chapter 1. The tubes can be bare or finned. Low temperature corrosion is a concer and Appendix B discusses this topic in detail. ‘One may use a heat exchanger to preheat the feed water in case the wall temperature is low enough to cause condensation of acid vapors on the tubes. It should be bome in mind that the metal temperature is decided by the feed water temperature and not by the ‘gas temperature as the tube side coefficient is several times higher than the gas side coefficient. Typical water velocity is 2 to 6 ft/s. | ‘Tube sizes range from 1.5 1035 in. Depending on the type of exit duct work, the economizer could j be arranged in several ways. Figure 3-18 shows a few possibilities. In Figure 3-18a, the economizer sits on top of the evaporator. The water flows downward. One has to make sure that economizer steaming | does not arise under all operating conditions. The stack can be mounted on top of the economizer if it is not too tall or the wind velocity is not severe. ' In Figure 3-18b the economizer is arranged in two gas passes ' with the hotter water having an upflow direction, which helps as ‘ far as steaming is concerned, as the bubbles of steam take a natural path upwards. The gas exit is at the same level as the boiler. ee ———I oro" oo | _Levaporator: | econonmser Jo 0.0 0, 2 Figuri3-18a-c Arrangement of Eeonomizes (3-<) sented below is the printout from a program developed by ithor, Figure 3-19: It shows an HRSG with a radiant section, State superheater with attemperator, an evaporator with I types of fin arrangement and an economizer. The program188 Waste Heat Boiler Deskbook ee “Ripe 184A Lage Exons Col [ABO Inds}20 ‘Waste Heat Boller Deskbook Figure 219, Pintot of Results From HRSG Program. EERE EEEEe Bolles 201 Halically arrives at the firing temperature required to generate ed final steam quantity ifthe fuel analysis is specified. he mechanical configuration is inputted along with gas inlet Pr fons entering the burner system (if used). The steam and gas Ee pressure drops, tube wall temperatures, spray water, firing GE jature; gas analysis and properties are all automatically led in seconds; provision exists to supetheat steam from other ‘orto withdraw a certain quantity of saturated steam from the eat boiler for process purposes and superheat the balance V. Ganapathy, “Charts can help give quick engineering Bp ates of gas pressure drop in tube banks,” Oi and Gas Journal, ph 1, 1982 V. Ganapathy, "Charts simplify estimation of non-luminous ransfer coefficients,” Hydrocarbon processing . V. Ganapathy, "Basic programs for steam plant engineers,” Bel Dekker, New York, 1984 V Ganapathy, “Simplified approach to designing heat V Ganapathy, "Determine spray water to desuperheat ” Heating, Piping, Air-conditioning, December 1987 7. V. Ganapathy, "Steam Plant calculations manual, Marcel i. fon factor for pressure drop; subscript g stands for gas.20 ‘Waste Heat Boiler Deskbook {f-fouling factor, sq th F/Btu — subscripts i and o refer to inside and outside tubes. F—fraction of energy absorbed by direct radiation FF, F3 factors used in equations (3-25), (3-35) G ~ gas mass velocity, Lb/sq th H— enthalpy, Btu/Ib 4 denotes difference in enthalpy. Subscript s stands for steam h— heat transfer coefficient, Btu/sq ft kF; subscripts c-convective, i~ inside tubes, x—non-luminous, o-outside HL heat loss factor; 1% heat loss means Al =.99 k~ gas thermal conductivity, Btu/ft F K -factor used in equations (3-16) , (3-18) Km tube metal thermal conductivity, Btu/ft h F L=length, ft or m; subscript e stands for effective length, ft. Na, Ni number of tube rows deep Nw—number of rows wide NTU-Number of Transfer Units ‘Nu—Nusselt Number Po, Pey~ partial pressure of carbon dioxide and water vapor Pr—Prandtl Number AP pressure drup; subseripts u or g stand for gas,in we and subscript for tube side, psi 4~ heat flux, Btu/sq ft h; subscripts i and o stand for inside and outside Q—duty, Bru/h; subscripts and t stand for assumed and transferred: c mand r for convective, non-lumirous and radiant Re~ Reynolds Number S~ surface area, sq fis, selongitudinal and transverse pitch, in AT ~log-mean temperature difference, F Ty, Tz temperature entering and leaving, for gas Tz— average gas temperature, F or K f,~saturation steam temperature, F U- overall heat transfer coefficient, Bu/sq ft h F; subscripts and o~ stand for inside and outside tubes. v,V ~ specific volume, cu ft/Lb and velocity, ft/s Wo, We ~ gas flow, pph Wi Ws, Wy~ Flow inside tubes, pph, inside, steam and feed water € ~ effectiveness of exchanger, Equation (3-42) 2 (iBly-used in various applications as discussed earlier in Chapters d-3. Methods of performing their design: and performance lations were elaborated in Chapter 3. The mechanical Gp guration of the WHB had to be known in these calculations and the duty and heat balance. x However when consultants and system engineers are in the tages of developing a project they need some basic information HRSG performance and capabilities at different plant loads peePerating points. For example in a gas turbine cogeneration ge if the HRSG performance can be simulated at different pent conditions and loads, they can study the complete system mance knowing the steam generation, auxiliary fuel iption, system efficiency etc., and use the informanon to select auxiliaries such as steam turbine, condenser and deaerator or pate the overall plant economics. They may obtain this information from HRSG suppliers; er, during the early stages of a project, a lot of time can be by consultants if they depend on HRSG suppliers for this tion for the simple reason that too many alternatives or gas gies may be involved and HRSG suppliers may take a long time Bsign a HRSG for each alternative and then provide the205, ‘Waste Heat Boiler Deskbook performance data for different operating conditions; also, they may not respond as quickly as the consultants would like to, particularly if itis a study and the gas inlet conditions are not firm. "Also it would be in the interests of consultants to be able to simulate the HRSG thermal performance and not be tied to any particular HIRSG design or supplier. Since they are only interested in the thermal performance, the mechanical details are not important at this stage. Once the study is completed and a particular gas turbine or heat source is selected, then the consultants can approach different HRSG suppliers for proposals, knowing what kind of ‘performance and steam production they can expect for their system; this also makes them knowledgeable and the process of evaluation of HIRSG bids can be quickened. ‘Another reason for consultants to perform this simulation theinselves is that they can optimize the steain system (whether to maximize HP or IP steam, where to take off the steam for deaeration, how much LP steam should be generated etc.) as they are more familiar with cost of steam, fuel and the utilities and the steam needs of the customer. Is there a method of evaluating HRSG performance without knowing the mechanical design details such as tube size, length, fin density etc? Can one obtain information on steam generation capabilities and temperature profiles without doing a mechanical design? The answer is YES and this chapter will elaborate the method, which can be effectively used by plant engineers, consultants and even HRSG designers to simulate and optimize a HESG, be it a single pressure system or a complex multi-pressure unit and integrate it into the steam plant. DESIGN AND PERFORMANCE CALCULATIONS ‘There are two basic types of calculations performed while selecting a HRSG; one is the “design” calculation; in this mode, the basic configuration or the disposition of various surfaces is arrived at including the gas and steam temperature profiles and the steam production. Once a unit is designed, it means that the surface areas of medBe piign and PerformanceStmulation 207 rious components such as the superheater, evaporator and fizer and the configuration are frozen. ‘0 important variables namely the PINCH point and ACH Point determine the complete design, the design fe sature profiles and the steam production, as explained below. {ihe "performance calculation” tells the user. how the HRSG been designed performs at different other gas flows, inlet atures, or steam parameters. This is a complex iterative lure and will be discussed later. There is only one design point Bf ere are several off-design performance points. gure 4-1 shows the temperature profile for a simple HRSG for le pressure case with a superheater, evapcrator and econo- The desired superheated steam pressure and temperature are ‘as also the feed water temperature and blow down. The gas Eonditions such as the gas flow, inlet temperature and the gas is are also known. It 1s desired to obtain the gas and steam ature profiles and the steam flow in thé design mode. {Once the pinch and approach points, which are defined in the Him, are chosen, the complete gas and steam temperature pp 28d the steam flow fall in place. In other words, if the pinch FPproach points are selected the design may. be said to be lete as shown below. saturation temperature fis known or may be obtained from dam tables as the evaporator pressure is known. (It can be $y" to be a few psi above the superheated steam pressure). Since fech and approach points are assumed, Ty3 and fy? can be Bp 84. (Tg3 = ts + pinch point and toa = ts ~approach point) 12 = duty of superheater plus evsporator = Wea xI(hsp~hyo) + BPP = wg *Cp *(Tg1-Ty5) x hag 4) 'm the above, Qz2 can be computed as T1,Ty3 and Wg are Cp Is obtained from the gas analysis. iy is the heat loss208 ‘Waste Heat Boiler Deskbook factor—typically ranging from .98 to 99 depending on the size of the HIRSG and insulation or casing design. Example 8 of Chapter 2 shows how one can compute this loss. ‘Wea, the design steam flow rate may-be solved for; saturated steam enthalpy fis1, superheated steam enthalpy hg2 and enthalpy of feed water fiz2 are obtained from steam tables. The superheater duty = Q1 = tosd» (hag han) = Wy XC p (Tega Ign) hy (4-2) Hence Tgz may be obtained. Since Qx2 = Q1 + Oz, Q2 the evaporator duty may be obtained; also, (Qp = Wg xCp x (Tg2 — Tya) * hag (£3) ‘The economizer duty is then = Q3 = Waa x (1+ bd) x (ha hot) = We x Gp X (Tg ~ Tye) HIF (44) Hence one can solve for Tyg from the above equation. Thus the complete gas and steam temperature profiles and the steam flow are arrived at. (hg is the enthalpy of saturated water.) 1 ie Ne 2 p+ Finch Point = Tg3 = ts |— Approach Point nn Evaporator Economizer +3 Superhester 1 t Flgure 41, HRSG Temperature Profiles delot of confusion exists among engineers regarding the above Griportant variables. There are several important facts that febbe kept in mind before selecting them. Pinch and Approach points cannot be’ arbitrarily selected. If >a temperature cross situation can arise as discussed below GD Wy Cp x Tyr Tgs) = Wea ® a-ha) (a5) Bots = Wy x px (Tgr-Tge )= Wat X heaton) 6 Bix the above equations, blow down and heat oss were neglected. ing variations in Cp, we have, dividing (£5) by (4-6) © (Tyr Tgp) Tg1-T a) = (ha Hey) 2 ~hiyt) =X (47) FFX is'a constant for the steam parameters ir consideration. Two (4-9) (4-10) “Let steam pressure = 585 psig, steam temperature = 70°F, Feed tr temperature = 250°F. Let approach point = 20°F. ts = 488°F; ‘a2 = 468°F. From steam tables, hg2 = 1351.8, gt = 219.5, hw10 ‘Waste Heat Boiler Deskbook 450.7. then X = .796. From (4-10) Tgie = (488-796 x 250) /(1-796) = 1416F Above 1416°F, the feed water temperature governs the profile To illustrate this, let us assume that we selected a pinch point of 30°F when Tg7 = 1600°F. The Tg3 = 488 + 30 = 518°F. Then from (4-7), (1600-518) /(1600~Tg4) = .796 oF Tyg = 240°F, which is below 250°F, resulting in a temperature cross situation, which is not practical. Let us say we selected 280°F for Tye, then: (1600-T'g3)/(1600- 280) = .796 or Tg3 = 550°F, a pinch point of 62°F. This is a feasible value. With a higher approach point, Tige increases and the pinch point would not be governing the temperature profile. Let us select = 400°F in the above case. Then X'= 862. Hence Tgyc = 1978°F. 1600°F, the pinch point can now be lower. Try 30°F. Then. (1600-518) /(1600-Tgq) = 862 or Tyg = 345°F, which is higher than Ty and hence thermodynamically feasible. However whether it is feasible from a mechanical size consideration has to be seen. Thus temperature cross situations can easily be handled by increasing the pinch or approach points or both. However the point that has been brought out is that they ‘cannot be arbitrarily selected, and the inlet'gas temperature plays a role. Hence one has to be careful in selecting pinch and approach points in the case of fired units or in HRSGs where the gas inlet temperature is high above 1000°F. It is suggested that pinch and approach points not be selected in the fired mode (in the case of gas turbine HRSGs) for the above reason. Note that the gas flow does not matter 2. Pinch point, has to be selected such that the HRSG ‘evaporator is of reasonable size or one that can be built and shipped. Ifa very low value is selected, then the surface area required may be too much and the HRSG will be cost prohibitive. If a very high value is used, then one may not be able to recover adequate energy from the gas stream. See Chapter 3 for an example on optimum pinch point selection. Choice of Pinch and approach points are also affected by the type of surface used, whether bare or finned. With bare tubes, it is impractical to obtain a pinch point of less than 100 to 150°F. The number of tube rows would be too many and hence the gas pressuregn nd Performance Simation am would be very high. Appendix A compares two designs with and finned tubes for the same pinch and approach points. ‘The following table may be used as a guide for selecting pinch approach points. The author suggests that for gas turbine , these values be selected in the unfired mode even though the yy be operating in the fired mode. Table 4-1. ‘Suggested Pinch and Approach Points Pinch Point, F Approach Point, ‘type bare finned friet ges temp, F F1200-1800 130-150 30-60 40-70 5750-1200 80-130 10-30 10-40 _ or a gag turbine application, pinch and approach points of 15 E20°F are reasonable in the unfired mode. For a dirty gas [plication such as MSW, with a gas inlet temperature of 1500 to LW", a pinch point of 130 to 150°F and approach of 40 to 70°F may sonable. . 3. Pinch and Approach points should be selected in the unfired fice for gas turbine applications, even though the unit may be gpating in the fired mode all the time. The following are the ja: A temperature cross situation can result as discussed above. sultants who are not familiar with HRSG performance fe times call out the design pinch point or approach point without izing the implications of the gas inlet temperature or the B.ticality of the value chosen. j.b. For example if a pinch point of 20°F is selected with a firing BPerature of 1600°F, the boiler size would be huge and the cost and § Pressure drop would be unreasonable. [tis difficult to visualize a BASS size in the fired mode; having designed several units the Bio recommends that the pinch and approach points be selected fe °° unfired mode and the performance evaluated in the fired His: ®- Through experience, it can be said that pinch and approach gp oF 10 to 20°F are feasible for unfired gas turbine units.ae ‘Waste Heat Boiler Deskbook ¢. Steaming in the economizer is a concern with gas turbine units. This occurs during cold ambient unfired conditions when the gas flow is higher and the inlet gas temperature lower than normal conditions. The reason for this will be given later with an example. Hence if one selects the pinch and approach to avoid steaming at the unfired cold ambient conditions, steaming can be avoided at any other operating condition. Note that as the gas inlet temperature increases, the pinch and approach points will increase and vice versa. d. If the selection of steam temperature is done in the fired ‘mode, it will not be achieved in the unfired mode due to the lower as inlet temperature. In units where the steam temperature has to be controlled over a wide load range, the steam temperature has to be achieved in the unfired as well as in the fired modes. The excess surface area of the superheater or the amount of spray water required for steam temperature control cannot be visualized if the temperature profiles are selected in the fired mode. Also, several performance checks have to be made to ensure that the desired steam temperature range is being achieved at different unfired and fired conditions. On the other hand if the steam temperature is selected in the unfired mode, it can be certainly achieved in the fired mode and through some steam temperature control methods, we can obtain the desired value at any other higher gas inlet temperature. 4. Once the pinch and approach points are sclected and the design temperature profile is arrived at, the design is nearly complete. If the HRSG operates under a different mass flow, inlet temperature, gas analysis or steam pressure, the HRSG performance or the temperature profile and the steam flow would change. This information is obtained through "performance" calcula- tions, which is discussed later. The pinch end approach points would change with different case of inlet conditions. They fall in place and have to be evaluated using complex iterative procedure. The important thing to keep in mind is that pinch and approach points are not constants but vary with variations in gas flow, inlet temperature, gas analysis, steam pressure and feed water tempera- ture. Hence consultants and engineers should not use these terms loosely but associate them with a particular gas inlet condition andsign and Performance Simulation 23 E parameters. We will now illustrate the design and per- ice calculation procedure with examples. AND CALCULATIONS 1e Superheater and economizer are assumed to be in jow arrangement, which is the widely used configuration. PA gas turbine HIRSG is to be designed for the parameters shown ible 4-2. Determine the gas/steam profiles and the steam flow. : .0 in. WC. Let superheater pressure drop ‘The drum pressure = 450 + 7 = 457 psig. The saturation qiperature is 460°F. Gas temperature Teaving the evaporator = 460 40°F. Compute the gas properties for the given analysis. The data Shown in Table 4-3. Using an instantaneous specitic heat of 0267 he range 900 to 480°F and a heat loss factor of 0.99, the duty in [re superheater and evaporator is: Q1 + Q2 = 150,000 (0.267) (0.99) 480) = 16.65 x 108 Btu/h = Wea{(1,330-8-431.2) + 0.02442.3- J2)] = 899.8 Was. Where 1,330.8 = enthalpy of superheated steam G0 psig, 650°F, 442.3 = enthalpy of saturated water at drum E sure, 431.2 = enthalpy of water entering the evaporator at 450°F. Bis the blow down factor. From the above, Ws = 18,510 pph. Superheater duty, Qr = Hy>10(1,330.8-1,204.4) = 2.34 x 10° Btu/h, where 1,2044 is the geclPy of saturated steam. Gas temperature drop in the heater = 234 x 106/(150,000) (0.273) (0.89) = 58°F Hence, gas B perature to evaporator = 900-58 = 842°F. Q7 = Evaporator duty = 14.31 10° Btu/h. Economizer duty = 18,510 (1.02) (431.2- .19 x 108 Btu/h, where 209.6 is the enthalpy of feed waterau ‘Waste Heat Boiler Deskbook Table 42, ‘Data for "Design" and "Performance" Calculations Design Pot = Port Pert 1. Case no. 1 2 3 4 2. Gas flow, pph 150,000 165,000 165,000 165,000 3. Exhaust tomp., °F 900840 aH 4. %yvol CO2 3 3 3 3 20 7 7 7 7 Ne ~% OB 75 1s 2 1 15 18 15 5. Steam press, pig 450,«S450—« 4500 6. Steam tomp., °F 650 2 2 @s0 7. Food watertomp.,"F 240 240240 8. Blow down, % 2 2 2 u 9. Process steam, poh — = 2500 : 10. Hoatloss + margin, 4 1 1 . 14. SH press, drop, psi 7 2 2 20 2 2 13. Approach point, °F 10 2 2 44. Steamiow, poh 2 26,000 26,000 15. Ambient temp, 80 50 50 Note that steam is required at a controlled temperature of 650°F in ‘case 4. In cases 2 and 3 it is uncontrolled. Also, in case 4, 2.500 ph of saturated steam is taken off the drum and the balance of 26,000 pph is to be superheated to 650°F. The steam exi pressure is 300 psig in case 4. It will be seen later that cases 3 and 4 are fired and casos 1 and 2 are unfired, Table #3. Part a—Gas Properties—Unfired Gas (5 vol CO =3, HO =7, N2=75, 02=15) Temp, F Cp # 900 02738 0.083 0.0308 65002858 0.0724 0.0281 40002584 0.0612 o.0at8 (il215 45, HzO = 7.87, No = 74.85, 02 = 1401) & # k 1,050 0.2800. 0.0887 0.0330 700 02689 0.0743: 0.0287 $50 02583 0.0586 . 0208 Units: Gy—Beu/Ib°R, w-lb/fth, k-Bta/ft h°F Anterpolate for gas properties at intermediate temperatures) 19 x 108/(150,000) 109°F- The gas specific heat at the average gas temperature ‘economizer, obtained from Table 4-3 by interpolation, is 0.26. ‘the exit gas temperature = 480-109 = 371°F. The temperature {is shown in Figure 4-2. Using a similar approach, the temper- Profiles for any other pinch-or approach points can be obtained. (sows = T5050 al. COy = 3,140 = 7, Ny = 7,09 = 15.504 =0 Qusre | wate. | uy [Pal Pew [Pam] Pec age iron | ui] | em || 5 |26 Waste Heat Boiler Deskbook ‘To proceed with the performance calculations for case 2 shown in Table 4-2, a few parameters should first be computed, as discussed in insert B. These parameters help relate the heat transfer coefficients in the "design" mode to those in “performance.” For the superheater: Ky = Qz/(4Ty) (W,?5) Fg) where AT = log-mean temperature difference = [(842-£60)-(900-650)}/In{(642— 460) /(900-650)] = 311°F, Fy = Cp-33,9-57), 5? = 0.135, using a Cp 20.273, k = 0.029,and jt = 0.0826. Hence Kz = 2,340,000/ 150,000°-45 /311/0.135 = 24.10. Similarly for the evaporator Kz = 387.6 and K3 = 2184 for the economizer. Ky, Kz and K3 will be used to ‘compute (UIS)p, the prodiict of U and $ in the performance modes a5 discussed in insert B. PERFORMANCE CALCULATIONS Let us see how the unit performs when the conditions are as shown in case 2, Table 4-2. The gas flow is 165,000 pph at &40°F. The gas analysis, feed water temperature and steam pressure remain the same as earlier. The performance of the HRSG is arrived at through an iterative process described in Inserts A and B. ‘Trial 1. As a first approximation, assume that the steam flow is proportional to the gas flow and temperature drop. W; = 18,510 (165,000/150,000) (840-371) /(900-371) = 18,050 ph. Superheater Performance. “Let fs, the steam exit temperature = 640°F. Then, from steam tables, the enthalpy = 1,325 Btu/Ib. The assumed duty = 18,050 (1,325-1,204.4) = 2.177 x 106 Btu/h. Gas temperature drop = 2,177,000/ (165,000) (0.99) (0.271) = 49°F. Hence,gas temperature leaving the superheater = 840-49 = 791°F. Compute the transferred duty, Ore, using Equation (4-12) in insert B. Fy = 0.135, We = 165,000, Kz 2241, Wsq = 18510, Ws = 18,050. Hence (US)p = 1165,0000-65(0,15) (24.1) (18,050/18,510/015 = 7,974. AT = log-mean temperature difference = {(840-640)-(791- 460)} /In{(@40-640)/(791-460)} = 260°F. Hence, Qe = 7,974 (260) = 2,074,000 Btu/h. This is close to the assumed value. Ifit were not, we Oe“Fo ign and Performance Simulation a7 ompute Fe at the average gas temperature in the evaporator. £0,129, K2 = 387.6. Then, (US)p = 165,006 (0.129) (387.6) = 23, Using Equation (4-18), [(791-460)/(Tg3-460)} 23/163,000/0.99/0.266) = 17.00. Hence Tz3 ~ 480°F: Q) = zer Performance. Let the water temperature leaving the economizer be 450°F. [> = 431.2 from steam tables. Assumed duty Q3a =. 1.02 (18,050) H2-209.6) = 4.08 x 10° Beu/h. The gas temperature drop = £50,000 /165,000/0.99/0.26 = 96°F, exit gas temperature EDT. Fy = 120, Ks = 2184, Hence (US)y = 218.4 (165,000) (0.120) 4/535. Transferred duty = Q3¢ = 64,535(72.7) Fre 72,7 i the log-mean temperature difference. Since the transforred duty ie more than the assumed, let us fort the calculations with say tw = 457°F. Q3_ = 18,050 (1.02) 209.6) = 4,230,000 Btu/h. The exit gas temperature = 381. AT = en Q3a, = 64,535 (65) = 4,190,000 Btu/h. Since this is closer to Ep) let us continue. The total transferred duty = Q1+ + Ore + Os: = +1352 44.19 = 19.78 MMBtu/h. The corrected steam flow, Wee = 8 x 108/[1,325-209.6 + 0.02(442-209.6)] = 17,660 pph, per Equation §22). Since this is not close to the assumed value of 18,050 pph, fer tral is warranted, Try Ws = 17,770 pph .69 x 108 Bru/h, 2 | | Let the revised steam flow = 17,700 pph. Follow a similar dure as before. G Let 52 = 640°F. Og = 17,700 (1,325-1,204.0) = 2.134 MMBtu/h. b emperature drop = 2134,000/(165,000) (0.99) (0.271) = 48°F. T g2 0-48 = 792°F, AT = 260°F. Fy = 0.195. Kj = 241. Then, (US)p=218 Waste Heat Boiler Deskbook 165,0009-55(0,135) (24.1) (17,700/18,510)0-15 = 7,957. Q1s = 7957(260) 2.07 MMBtu/h. SinceQz is less than Qua, try a lower steam tem- perature, say 635°F. Then Qza = 17,700(1,332-1,204.4) = 2.081 ‘MMBtu/h. Gas Temperature drop = 47°F. Tgo = 840-47 = 79°F. AT: 264°F. Hence , Q7¢ = 7,957(264) = 2.1 MMBtu/h. This is close enough. Contin, Evaporator Performance. Solve for Ty3 as before. (793-460) /(Ty3~460)] = 17.00; hence Ty =480°R. Q2 = 165,000 (0.99) (0.266) (793-480) = 13.6 MMBtu/h. (The factor 17 computed from Trial 4-1 is unchanged.) Economizer Performance. Let tazg = 455; hag = 436.8; Qsa=17,700(0.02) (436.8-209.6) = 4.1 MMBtu/h. Gas temperature drop = 96°F. Tyg = 480-96 = 384°F. AT 68°F. Using the same (US)p as before, Q3¢ = 64,535(68) = 4.36 MMBtu/h. Since the variation between Q3q and Q5 is large, try 172 58°F. Then, Qsa = 4.14 MMBtu/h. Tye = 383°F. AT = 64.6°F. Hence, Q3t = 64.6(64,535) = 4.16 MMBtu/h. This is quite close, The total transferred duty = 2.1 + 13.6 + 4.16 = 19.86 MMBtu/h. The corrected steam flow, Wse = 19.86/[(1,322-209.6) + 0.02(442-209.6)] = 17,770 pphSince this is close to the assumed value of 17,700, let us stop here. The final temperature profile is shown in Figure 4-3. The gas pressure drop, using Equation (4-25) = 6(165,000/150,000)210.5(840 + 383) +460)/[0.5(900 + 371) + 460)] =7.1 in. WC. Performance Check—Fired Case. Let us check the performance for case 3 shown in Table 4-2, where it is desired to make 26,000 pph of steam. The steam temperature is uncontrolled. It is obvious that with same inlet gas conditions as in the earlier case, we need additional fuel input to the HRSG to generate 26,000 pph. The procedure is similar to the earlier one. However, additional steps are necessary to iterate for the firing temperature, as discussed in Insert A. The method of computing the fuel input, firing temperature and gas analysis is discussed elsewhere. Let tis only check the final results which are shown in Figure 4-4.1 eo { 0 “st my zr 5° | manne petomane eens ‘Ga ph = 100 wo OW MO =T. Nn TE, Om 16, SO Tasers. | wren] Say [Pom Few [Pam] Pcs [ogc] ety | Sees | ant ao | poe | 8g eae Figure 43. Performance Case? Results 3 este—putomse ne tins {Ss woe = 1800, el Oy = 3, MD = 7. Ne = 7, Ok = 1, SO 8 TT] comm | gman, | aay Pe] Rr [Pam [Pe [oes] Lf eee | to | wy oom |" | |“ er] sce | to|e | ar] as | See atoms |teno| a7 | swan SG = 85. OO, BNO FR | at paceman rte tneune 56,2 an 2 3p = 2, Semin emma ti we sons Figure 44, Performance Case 3 Resuls20 ‘Waste Heat Boiler Deskbook Superheater Performance. Table 4-3b shows the gas properties for the gas analysis after combustion. From the printout, Figure 4-4, itis seen that the HRSG ‘gas inlet temperature is 1,034°F and the burner fuel input is 9.29 MMBtu/h (LHV basis). We = 165430; W, = 26,000; typ = 677°F. Fy at the average ges temperature is 0.142. The saturation temperature is 462°F, at the corrected drum pressure of 463 psig. Q3a = 26,000(1,346.0-1,204.3) = 3.69 MMBtu/h. Gas temperature drop = 3,690,000/(165,430) (0.99) (0.278) = 81°F. Exit gas temperature, Tg = 1,034-81 = 953°F; AT = 420°R. Ky = 24-1. (US)y = 165,430°-45(0.142) (24.1) (26,000/18,5109°15 = 8,840. Then, Qty = 4208,840) = 3.71 MMBtu/h. Evaporator Performance. Fg = 0.135; K = 387.6; hence (US)p =165,430°6(0.135) (387.6) = 129437. Using Equation (4-18), [(953-462)/(Tg3-462)} = £128,437 /165,480/0.99/0.27) = 18.67. Hence, Ty3 = 489°F. Q2 = 1165 430(0.99) (0.27) (955-489) = 20.52 MMBtu/h. Economizer Performance. tg = 435; hug =414.45; Qzq = 26,000(1.02) (414.45~-209.6) = 5.43 MMBtu/h. Gas temperature drop = 128°; Tog = 489-128 = 361°F. AT 89°F. Ky = 718.4; Fg = 0120; (US)p = 165430055 (218.4) (0.120) = 65,000. Hence Qs¢ = 83(65,000) = 54 MMBtu/h. Total energy transferred = 3.71 + 20.52 +5.4 = 29.63 MMBtu/h. Wee = 29.63 X 108 /1(1,346.7-209.6) + 0.02 (442.6-209.6)} = 25,970 pph. The gas pressure drop could be corrected as before. Tris gives an idea of the complexity of performance calculations if fuel firing is involved. Several iterations of performance calculations would be required before the correct firing temperature is arrived at. Also, if the steam temperature has to be controlled, the superheater has to be split up into two stages with’a spray desuperheater in between. The method of computing the spray water for steam temperature control is discussed in Chapter 3. In such an HRSG, more iterations are involved before the spray water flow and the final temperature profiles are arrived at. Without a computer it would be extremely tedious and timeSESE TESS can anemaenanes SSI [oe | ital oe | po || SF | SoS Se S| se |5 S| <,| » Bea. | f BARBS ‘anf be HSS = ES, C0, 2 0 = or, Roe on Ova 004 0420 Pe ee eon 52 ane = 2,3 papa = 2 routs hss metus h nnn chs Figure 45, Peeformance Case-4 Results Note that gas inlet temperature is 1,067°F. The spray quantity iperheater while actually building the HIRSG. Also, note the higher steam pressure drop in the superheater i to the lower steam pressure. The economizer flow includes the 500 pph saturated steam taken off the drum.mm Waste Heat Boller Deskbook A note of caution on U, § and U times S values. Note that US values could be computed for each surface from its Q and AT data. For instance in the “design” case, for the supecheater, US =234,000/311 = 7,524. These would naturally change depending upon the gas flow, analysis and temperature profile. Hence, these values should be interpreted with caution. ‘After arriving at the LIS values, some engineers try to split up the U an $ values and compare alternate designs based on $ values alone. This can lead to very misleading conclusions and the author strongly.recommends against it. particularly if extended surfaces are used. With finned tubes, the gas side heat transfer coefficient and fin efficiency are affected by variables such as fin density, height, thickness and fin or tube material. By using tubes with high fin density, say six, one could show more surface in the HRSG, but due to the lower U associated with it, it does not mean that the energy transferred is more compared to a design which has a lower fin density, say two to four, and hence, lower S. Lower fin density should be used winenever possible to incréase U and minimize gas pressure drop and fin and tube wall temperatures. This is more important in surfaces with low tube side heat transfer coefficients such as superheaters. One could show that S can be 100 to 200% more by using six fins/in, compared to two, but due to the higher U, the duty can be the same or even more. ‘The author has performed studies on optimization of finned tubes and advises engineers against comparing and selecting HRSGs simply because the surface area, S, is more compared to another design which used lower fin density. Unless the engineer is familiar with all aspects of heat transfer with extended surfaces and the impact of each variable on U, comparisons of § alone can be misleading and should be avoided. Limitations and Software. ‘The approach discussed has a limitation. It cannot be used in HIRSGs which have a radiant section. However, the author is of the view that 80 to 90% of HRSGs fall under the category discussed in Figure 4-6, and hence the methodology discussed can be applied to a wide variety of HIRSGs used in the industry.an re ‘a Diese { so TTT TAT fw " Lu VW U U Vy Pay), s.sHsEinr Bo teat. EAP oat xn nt = i go 6. EVAP [pele 4. EVAP + ECO noc 4 ~ Steamin 4 arn 2. SH jroo 6 1. £00 fete 5) Figure46. Various modules can be combined to represent ‘muhiple pressure and complex HRSG configurations. (Hydrocarbon Procesing] t While! the method of predicting petformance using U values sed on actual tube geometry, fin configuration, etc,, gives accurate esults, this methodology has been checked against several designs ind operating results. For the purposes of engineering analysis, trend projections, evaluation of alternate designs and for studying the Erlfect of different gas/steam parameters on performance, this ppproach is very effective and hence a powerful tool,228 Waste Heat Boller Deskbook Considering the complexity of the calculations and iterative nature of the procedure, particularly if multipressure HRSGs are involved, program COGEN has been developed by the author for HIRSG design and performance evaluation. For more information on the software and its availability, contact the author at P.O. Box 673, Kbilene, Texas 79604, USA. INSERT A ‘Performance calculation procedure ‘The procedure is discussed for a single pressure HRSG. Figure 4- 6 shows the various configurations of HRSGs considered. The first case is quite involved. The methodology for this case will be discussed: The gas flow, gas inlet temperature and analysis, steam pressure and feed water temperature are assumed to be Known. The design calculations, which are the basis of establishing an initial design, are assumed to be done and the results available, along with Ky, Ko Ke factors. 1. Assume the steam flow. A good estimate is obtained by using @ ratio of the "performance" to “design” gas flows and temperature drop. 2. Solve the superheater performance. This is an iterative process. See insert B, see Equations (4-17 }to 4-15). Ifthe transferred and assumed duty are not equal, repeat with another steam temperature or else continue. 3. Solve the evaporator performance. Obtain the duty and exit gas temperature using Equations (4-16 Jto (4-19) . 4, Solve economizer performance using Equations (4 - 20) to (4-23). ‘This is again an iterative procedure. Calculate the total transferred duty. 5. The steam flow is then corrected based on the total transferred duty and enthalpy rise, Equation. (4-24). If this is close to the assumed steam flow in step 1, continue or else repeat steps 1 t0 5. 6. If the final steam temperature is greater than that desired, the steam flow is corrected for the desired steam temperature. Se ha wi the the iADesign and Performance Simiuation 5 If the desired steam flow is zero (unfired mode) or less than the ted flow, proceed to step 11. If the desired steam flow is larger than the corrected flow, SIE the final’ steam temperature is higher than desired, ulate the interstage spray quantity based or. split superheater Another round of fine tuning is done to check the temperature. Tk can be’ easily seen that a lot of iterative'catculations are sved. For each round, the gas and steam’properties have to be rputed based on the gas analysis and temperature. If there is INSERT B Equations used in performance caloulations Ota = Wotha-tat) = Wel Cp (fT yiTen) (1D jere t:2 + exit steam temperature and hs2, the enthalpy. Compute ‘exit gad temperature, Tyo, from the above H-The transferred duty is then: (4-120) = AT = log-mean temperature difference BAT = FTgr-tod-Tgz-tep) nl Tg rte) / ga. ts) (4-128) gming counter flow configuration, which is widely used. (US)p is g2iptoduct of § and U in performance mode and is obtained from the226 ‘Waste Heat Boiler Desktook (US) value in the design case by adjusting as follows for the gas properties and flow. (US)p = We?55 FeKy(Wol Weg)025 (413) Kz is obtained from Qy, AT, Wy and Fg values in design case: Ky = Qu/lAT(Ws?5) (Fy) ) (4-14) Fg = (Cp 330.67) :032) (4-15) If for the assumed steam temperature Qia and Qre do not come close (say within 0.5%), another iteration is warranted. All of the above steps are repeated until Qzg and Qy4 match. Evaporator Performance. From energy balance, Q2 = WelCp) (Hf) (Ty2-Tg3) = (US) AT (4-16) where AT ~ [gg i) (Tga —to)fln [CT 32-19) 99-9)] = (Tgp -Te3iftn [CT g2-t5)|(Tg3-t5)) (17) From Equations (4-16) and (4-17) after simplification, we have: gp-ts)/(Fg3-ts)] = e(US)pIW, C, yp (4-18) where: (US)p = W425 Fy Ky (£19) 2s computed as in Equation (4-14) from the design conditions. Fg is computed for the performance conditions. Ts is solved from Equation (4-18) without iteration. Qp, the duty, can be obtained from Equation (4-16) Economizer Performance. ‘Assume fap, the water exit temperature. Then, O30 = We (haty-heoy) (1 + bd) = Wy Cp (Tqs-Tya@hlf- (4-20) It rey cor ap Poi tem dec; cor surf ‘ner, evapDesign and Performance Simulation a Obtain Tye and then the AT, assuming counter flow conditions (421) (422) (423) Wee =(Qre + Qre + Qspilths hwy) + bdlhf-kwp] (4-24) Wee is not close to the assumed flow, Ws, the calculations are peated starting with the superheater. The gas pressure drop is Forrected for performance conditions: BoP = (AP)a (Wel Wed)” [(Tavg + 460)|(Tavga + 460)! (4.25) From the above calculations we can draw some conclusions: © 1, Ina HRSG with an economizer, the pinch and approach points decrease with a decrease in inlet gas temperature and vice versa; for a gas turbine HRSG, this means that as the ambient smperature decreases (case 2), the pinch and approach points Piecrease. The reason is that the energy transferring ability of the momizer has not reduced (namely U x'S), as itis a function of the "irface area and overall heat transfer coefficient, which has g hcreased due to the larger mass flow. With a reduced steam flow, due to the reduced duty of the faporator, the economizer can now bring the water closer to2 ‘Waste Heat Boiler Deskbook saturation temperature than before. Hence the caution that steaming is likely in the cold unfired ambient gas turbine exhaust conditions. Tt can also be seen from cases 3 and 4 that as the firing temperature increases, more steam is generated and the pinch and approach point increase. 2. The other point to keep in mind is that dué to the lower water flow, the exit gas temperature from the economizer will be hhigher in case 2 versus case 1. So the HRSG efficiency is affected. It will be also be seen later that as the inlet gas temperature increases, more steam is generated and the stack temperature reduces. These trends are important and specification writers and consultants should ‘be aware of them. 3. The superheater duty is not obvious in the fired mode as seen from case 4. It is higher compared to the situation where the steam is simply raised from saturation to 650°F due to the spraying of injection water for temperature control. Hence the suggestion that temperature profiles should be selected in the unfired mode and the performance checked in the fired mode. SOFTWARE FOR HRSG SIMULATION — COGEN Predicting the off-design performance of a HRSG is a complex procedure involving several iterative loops even for a single pressure HIRSG as seen from above. If more than one pressure level is involved | or if the fuel input has to be arrived at automatically or if a common economizer feeds several pressure levels, the HRSG performance cannot be evaluated manually unless one is prepared to spend several hours on calculations. The author hyas developed a software, "COGEN", which is available from him. This may be used to arrive at a design and then predict its performance in a matter of minutes for complex HRSGs. Several alternates may be studied and the system optimized. These features will be explained through examples. Figure 4-6 shows the various configurations of HRSG that can be combined to form a complex single or multi-pressure HRSG for simulation. ee‘Design and Performance Simulation co With this software, consultants, plant engineers and sneration system analysts can predict the HRSG performance ‘are computed; then, for any other inlet gas flow, temperature, fim pressure or feed water temperature, the revised gas/steam iperature profiles and steam flow are arrived at based on a quick rerging iterative logic developed by the author; this is the sunfired: mode, a gas or oil'fired’ burner kicks’in-if there is jentoxygen in the gas stream and automatically computes the Slinput and firing temperature and revised temperature profiles. Different configurations as seen in Figure 4-6 can be handled by Gis program; by combining the modules, multi-pressure units can be Bralvated in ONE run. In order to use this program one need not know e:size, pitch or fin parameters, etc., need not be known. In short, this is a powerfll tool for consultants, plant engineers End cogeneration system designers, who would only like to simulate fije HRSG performance under different conditions but do not want to St involved in detailed thermal or mechanical design aspects, [hich is best done by HRSG suppliers. cial Features of the Program. 1. The software arrives at a HRSG "DESIGN" based on plant rating data or initially suggested pinch and approach points or then evaluates its "PERFORMANCE" under different inlet gas Fonditions and steam parameters in unfired or fred conditions. Single smulti-pressure HRSGs can be evaluated in ONE run. ¢.- There is no limitation on the number of modules. Up to six PPessure levels can be easily analyzed. The HRSG configuration is Eilected by combining the various modules shown in Figure 4-6. mplex HIRSG systems can thus be modelled. Each module can erate; at its own pressure, feed water or steam conditions. Evpplemental firing ean be introduced ahead of the HRSG or en modules.20 ‘Waste Heat Boiler Deskbook 2. One need not actually design a HRSG in order to evaluate its off-design performance. Tube configuration, layout, surface area, boiler size ete. need not be known. Heat transfer coefficient U is not computed; instead the product of U and surface area S is computed and then corrected for different gas flows, analysis and temperature. 3. Unfired or supplementary fired HRSGs can be analyzed. Gaseous (up to 14 constituents) as well as oil fuels can be used in the burner system. The program automatically ‘arrives at the firing temperature, burner fuel input if the required steam flow in any evaporator is more than that obtained in unfired mode, Source of oxygen can come from the exhaust gas or from atmospheric air (fresh air fired case); augmenting air can also be added if there is insufficient oxygen in the exhaust gas; this is important in HRSGs for Steam Injected Gas Turbines, which have ow levels of oxygen and high water vapor content. 4. The effect of gas analysis is considered; (Ux 5) values and specific heat values are corrected for different gas analysis and temperature making the evaluation realistic. 5. Ifthe final steam temperature is controlled and if for given gas inlet conditions, the steam temperature exceeds the desired value, an interstage desuperheater is simulated and spray water flow required to achieve the final temperature is computed along with the gas and steam temperature profiles. This can be done for any module in multi-pressure units. 6. The extent of “Steaming” in any economizer can be computed, if it does occur under different gas end steam conditions. The "DESIGN" can then be easily revised if desired. 7. The HRSG is modelled such that some saturated steam from the evaporator can be taken off for process heating or other applications if required and the balance superheated. Saturated steam from other boilers can also be introduced into the superheater thus increasing its flow. 8. The common economizer concep: can be used to supply feed water to any number of modules; the program arrives at the water flow and exit water temperature through several rounds of complex iterations. 9. One can arrive at a design and fine tune it based on its off- design performance within minutes.AA gas turbine HRSG generates saturated steam at 200 psig using water at 240°F; blow down is 3%. The gas flow is 150,000 pph JF design the HRSG and study the effect of fuel input on ance while generating 40000 and 60.000 pphof steam. Fel bol CO2 =3, HO =7,Nz =75, and 07 =15- ‘The data and results are shown in Example 1-5, Chapter T and system efficiency is computed in the unfired and fired modes. It shown that the system efficiency as estimated by ASME PTC increases with additional fuel input. * CCOGEN may be used to improve the efficiency of HIRSG systems o ways: a. Use add-on surfaces such as condensate heater, deaerator coil [heat exchanger to lower the stack gas temperature. Figure 4-8 With a condensate heater, the efficiency is improved through Sluced demand of deaeration steam. "'b, Rearrange location or disposition of heating surfaces to Brive at a lower stack temperature with the steam production ‘ining unchanged in the fired mode or more steam generation in Biited mode.22 ‘Waste Heat Boiler Desktook Ges Hlowpate 180000" cts Mattes 3 gee tao so mso~ soo gat temp wet/ot duty prose flow petm plac appech Estoutce sprouse ase bya Baigy gg ye" ME sb tonp-ts 60 rel tanta 60 heat Looe 2 gat tenp to msc 900 G8 Hlowppbw 0000 val coze'9) tor? 2's TE G2 © 26 come 0 gas temp wet/ata duty press flow prea plach appcch Gout nate omen pet poe ee AE » Berataren te 249 acy alee ~ 0 ‘mb temp-tm 60 cet nants 60 heat ton Flowppoe 1800008 vol coam'3) tor? a2'e'75 Ge = 5 Some 0 gaa temp vet/atm duty pr Goeth akouter aa bin peig ‘pon ee OP ovepeayew tem 949 bebe ib =a Figure 47, Printout of results of eample 2 from COGEN software. « e st 3 teDesign and Performance Simulation 2 Figure 46 HRSG with condensate heater and glyeot heater. (ABCO Industri] faproving HRSG Efficiency Using Secondary Surfaces E Figures 48, 4-10, 4-11, 4-12 show the concept of adding a ondensate heater, a heat exchanger or a deaerator to improve the ficiency. The system consists of a HRSG generating superheated sm for producing electrical power via a steam turbine condenser ystem. The steam for deaeration is taken from an extraction point in ie turbine as it is more economical to do so than take it off the . The reason for this will be given later. : ' jondensate heater. This is basically an economizer used for preheating the mixture condensate returns and make up before it enters the deaerator, sBure 4-10. A 20°F approach is used; that is if the deaeration Smperature is 240°F, the design condensate heater exit temperature 20°F. The following aspects of this option should be considered, 1. The log-mean temperature différence between the gas and ater is higher than in the deaerator option described below; as a ult the surface area requirements will be lower.‘Waste Heat Boller" Deskbook [eamog] “aeays worex=e -op Suppnpar syesuapuod yeaypxd 01 7 HOR “do uy sozjurot099 =) Sy 30 wreomsuMOp pappe s1 s9yOy aESUaPUOD “OTP AMBIE tame sunt fq paprosd hquo Araa0307 e041 sey ruondo quamaSuesresyseg-6-p andr Teneo = vara] aT =a ai | 5 leosratoers (cram aR wrcrone a TC sarees 207 mrssF iS Design and Performance Simulation WaEToROOE rarer) a ‘M26 Wiaste Heat Boiler Deskbook 2. Due to the low water temperature at the inlet, the tube wall temperature will be lower and hence condensation of water vapor can occur resulting in corrosion. It is suggested that the tube wall temperature be above the water vapor dew point of the exhaust gas If there is steam injection in the gas turbine, the water vapor dew point is higher and hence suitable precautions have to be taken such as better choice of tube materials. If the exhaust gases contain sulfuric acid vapor or sulfur dioxide, then this option is not recommended. 3, The gas pressure drop will also be lower due to the lower surface area requirement. Table 44. Results of Study of Deaeration ‘Steam and Electric Output* ———————— Option yo @ ® @ Base Cond Heat Lip Design reference case hr exch evap Exhaust gas, 108 Ibyhr 550 550 550 550 Gas-inlet temperature, F 975 975 975 B75 ‘Stack-gas temperature, F 374 310 323 297 ‘Steam to turbine, 103 tbyhr ** go 80 «68080 Steam to deaerator (from turbine), vor 10,250 1730 3400 0 Feedwater temperature, F 240 240 151 «240 Mixture temperature (0 deaerator),F 107 220 200 107 Electric power, KW 6528 6830 6770 6890 * Assumptions: Natural-gas fired gas turbine; gas analysis, % vol: C02 =3.5, Hz0 = 10, Nz = 74, O2 = 135; pinch point = 20°F; approach point = 20°F, blowdown = 2%; deaerator pressure = 10 psig; makeup temperature = 60°F; condenser pressure = 25 in. Hg abs (1.22 psia); heat loss in HRSG = 1%. "Conditions: 620 psig, 650°F. a t t t te ar o fo in sb su to ab an oF6 Design and Performance Simulation 27 Figure 4-11 shows a heat exchanger used to improve the Glliciency of a HIRSG system. A water to water exchanger preheats Bie make up using the feed pump discharge before it enters the economizer. The make up water temperature increases thereby Ireducing the deaerating steam requirements. Since the feed water to economizer is lower, the stack gas temperature reduces. 1. If the gas stream contains sulfuric acid, then this scheme is [suitable due to the low tube wall temperature at the economizer Holet. Better tube materials may have to be.used if water vapor dew Fpoint is high. 2. This scheme is attractive ifthe make up water temperature tow say 40 to 60°F. Then, the feed water-can be cooled further in fe exchanger. The size of the exchanger will also be smaller due to Give larger log-mean temperature difference. Also, if the make up Hemperature is very high, on the order of 150 to 180°F, the exchanger Bay not be feasible due to the resulting low log-mean temperature Find possible temperature cross conditions inthe exchanger. Ef 3. Due to the lower feed water temperature, the size of the ieconomizer will increase. This cost has to be evaluated along with he cost of the exchanger, piping and associated values Deaerator coil. Figure 4-12 shows the deaerator coil used to recover additional from the gas stream. The bundle generates low pressure steam por deaeration and is essentially another evaporator. This may be an p itegral deaerator or simply an evaporator generating low pressure Pteam, which is taken to a deaerator. 1. The surface area required will be more due to the low log- piiean temperature difference compared to the condensate heater. 2. This scheme is suitable even if the gas stream contains furic acid as the evaporator pressure can be raised to a value close EP or above the acid dew point. The tubes will be at least 5 to 10°F tbove the'saturation temperature. 3. This system is more expensive than the others due to the size Bi the use of drums and associated trim and controls. {4 The gas pressure drop will also be higher than the other potions, due to the size.28 Waste Heat Boiler Deskbook USE OF COGEN TO ANALYZETHEOPTIONS —- Example 3: ‘A HRSG generates about 80,000 pph of superheated steam at 620 psig and 650°F from 550,000 pph of gas turbine gases at 975°F; the steam is expanded in a steam turbine to generate power, with deaeration steam taken from an extraction point in the steam turbine. Study the options assuring that the gas stream contains no sulfur dioxide. Solution: Figure 4-13 shows the results from COGEN in the design mode for all of the options. A 20°F pinch and approach were used. The steam required for deseration was estimated for each case through heat balance as described below. The power output from the steam turbine was obtained through computation of turbine steam rates. A Basic program for this is given later. ‘Table 4-4 shows the summary of results. Since the deaerator option generates the maximum steam output to match deaeration requirements, the electrical power is the highest. One may also predict the HRSG performance at any other condition sing COGEN and sae how the system behaves. DEAERATION STEAM CALCULATIONS Estimating steam quantity for deaeration is an important aspect of plant mass and energy balance calculations. Steam for deaeration should preferably be taken from an extraction point in the steam turbine if available. This results in a better system efficiency compared to the case where the steam is taken from the HRSG exit. Figure 4-14 shows the two schemes. COMPUTING DEAERATION STEAM AND ELECTRIC OUTPUT To illustrate the procedure, consider two sources for deaeration steam for the base case (Option 1), Example 3. Assume, frst, that thegas Seog int/atm duty prone flow pata plech appech Grover ascents amt b/m peg) ‘poh Me P amp-ta 60 rel'nuarte-0\ best: toute i gue tong to moon 975 pflowppbm 550000 ¥ tol cof 3-5. Moe 30" M2 = Ta Ga 10.2 soze° gus tone wat/atn auty. prem flow. pata posh appech Exec Grout web/e psig poe her E Ds meas eso 9.31 620 acoas 100.0 ° vat/ats duty pewes flow pata piach appech soreuier wr b/n pets “pn NT PSE 5 2 goat ae sooes rove @ GPF flowgete 550000" 8 vol coz 3.8 mio 102 e902 13. soe O gus ewe was/at ety pew Eso? cour wwebya bens240 ‘Waste Heat Boiler Deckbook DDeseraton steam! (8785 te) ae 107 Is, eopsigso soooommr Feacsion 2408 *et.700 nr Condensate: 1086 Figure 4:4. Two alternatives for deseration stam take off location; steam from. HRSG (top) and from steam turbine (baton). Poser)Pisc Design and Performance Simulation 2H jeam is taken from the HRSG exit (Figure 4-14 left) rather than ffom the turbine. Makeup temperature is 60°F and condensate enters mixing tank at 108°F (2.5-in. Fig condenser pressure). Neglecting ved and vented steam, a mass balance around the mixing tank energy balance is then performed around the deaerator, giving: ,700) (208) = (1700) (28) +(80,000 ~ X) (76) + 1319X, where 1319 :/ib'is the enthalpy of steam at 620 psig, and 6£0°F, 28 Btu/‘b the .ke-up enthalpy, 76 Btu/Ib the enthalpy of condensate returns, and O8:Btu/Ib that of feedwater after deaeration. Solving this equation, = 8741 Ib/hr steam: using 8785 Ib/hr allows for losses. To obtain the snsion from 620 psig and 650°F to 1.22 psia (25 in. Hg) EA program developed for this purpose gives ASR [iyewh: Allowing for 4% mechanical losses, P = 0.96 (80,000 ~ 8785) 14 = 6137 KW. Next, consider deaeration steam taken from the turbine ‘Thus, greater electric output is obtained by taking steam from be steam turbine, and so this is the preferred procedure for all other Engineers involved in cogeneration projects and power plant E!ies often need to calculate the steam properties during expansion G9 steam turbine to evaluate the theoretical and actual steam rates‘Waste Heat Boiler Deskbook [Bigssxo01g vso.phy] synsax woys ze 0 0} wezSerp 1) QSt-p amnShr ‘autqmy v uy urea}s Jo woys suedxg :@39)) est angTSR’ = 5,413/(hy —h25} (4-26) snp = hy ~ hy — hag E100) 77) 3 413)(hy — hz) (4-28) hexhy (1-0 hf (4.29) where the saturated vapor and liquid properties are obtained the dryness fraction, x, is S15y #(l~ asf (4-30) ‘The logic used in the program is shown in Figure 4150. eAimple 4: | E Two examples illustrate the use of the program. Example 4: bine to 150 psia with an expansion efficiency of 80% E Calculate the steam Properties at inlet and exit as also the ‘Key in the program in Figure 4-16a, In the RUN mode the screen for Phe inlet nevsstire and temoerature. If saturated, Kev 0 for2a ‘Waste Heat Boler Deskbook So tor ances neaceazaysmucr(ar cams vor Pon Hoary ea, 20 FRO" Ho ieanany/(er-sojaaerearcacayes Hop iecaremsay/(areigy anes eaeranrearcoro 290 feo rans aecnsayn.ovmcesne.cse io sesso frrcacaoene+-26t)-62/2)40e1.osiaomanenea/s/t 3a Sie aeztaoi ay seotessinh. anegutan autem achat eVr2eLen- {20 For ea 50 seaagryenczy poets /2se(eP+40(2)9109(P) RIE) cPOROE(L)*P-3CCE 29 rena sxarman gs eangeyssiecT) SPAN) igure 4162. Listing of program for expansion of stam [Hydrocarbon Processing] PeedGriple S: Saturated steam ac 1,000 psia is expanded to I psia in a.turbis ‘This example shows that by rearranging the heating surfaces p may lower the stack temperature while generating the same ating eam at 625 and 205 psig. There ae two options, shown in ve 4-17'and 4-18. In gure 4.17, the HP stage is followed by the as temperature. [Hlence, we need a lower inlet gas temperature in Figure 4-18 in to generate the same quantity of HP and LP steam or we can fate more steam (HP or LP) with the same gas inlet conditions.246 Waste Heat Boiler Deskbook Figure 4-166 Data and results for examples Example 4: ‘STEAM PROPERTIES AFTER EXPANSION — BY V. GANAPATHY INLET PRESS-PSIA, TEMP-F (IF SATURATED INPUT 0 FOR, TEMP) = ? 650.750 : EXIT PRESS-PSIA, EXPN EFF ~% =? 150,80 INLET PRESS — PSIA - 650 TEMP ~ F = 750 ENTH = 1977.067 ENTRPY = 1.601213, EXIT PRESS = 150 TEMP = 458.7688 ENTH = 1250.208, SAT TEMP = 958.5408 [ASR = 28,9075 TSR = 21.523 EXPh = BOEXITOLTY = 1 Example 5: ‘ ‘STEAM PROPERTIES AFTER EXPANSION BY V. GANAPATHY j INLET PRESS-PSIA, TEMP-F (IF SATURATED INPUT 0 FOR, TEMP) =? 1000.0 EXIT PRESS-PSIA, EXPN EFF -% =? 1,100 INLET PRESS-PSIA = 1000 TEMP ~F = 547.2928 ENTH = 1198,088 ENTRPH = 1.392013, EXIT PRESS = 1 TEMP = 101.7387 ENTH = 776.87 SAT TEMP = 101.7387 ASR = 8201008 TSF EXIT QLTY = .682389 8.201008, EXPN EFF -% = 100 Soluti ‘The results are shown in Figure 419, output from COGEN. Since the burner duty is small and the firing temperature less than 90°F, only the design modes are analyzed to show how the rearrangement of surfaces is helpful. The gas analysis was assumed to be close to the inlet gas analysis. The saving in fuel input is about 12 MMBtu/h for the improved option; assuming that the HRSG operates in this mode for 7000 hours/ year, the annual savings is $210,000 at $2.5/million Bru,Design and Performance Simulation 247 ;Alternatively we could have generated more HP or LP steam Biche configuration in Figure 4-18, if the gas inlet temperature [odie were the same as in Fivure 4-17248 ‘Waste Hest Boller Deskbook ano tang-fe 60 rel mumts 0. neat losecte 1 gaa temp to Mou 200 Se EELS aT ae aa aes cae ae soax 0 us tenp wat /otm uty. pene flow pet pinch apgrch Glace Gace webaper mm ee ‘gus tex wab/ata_ duty eens flow pate pinch apgech ELSE DMI bie pig om NP pnp 662 420395 390 86.76 205 202786 100.0 20 14 igure 419, Results from COGEN for example 6 for the? options shown in Fig 4-17 and 18, Table 4-5. Data for HRSG design Exhaust gas flow, pph = 121,000 Exhaust temperature, F = 770 HHP steam flow, poh » 75-77,000 HP steam pressure, psig = 635 HP stoam temperature, F = 600 LP steam flow, pph = $5—100,000 LP steam pressure, psig = 205 LP stoam temperature, F = 390 (sat) Food water temperature, F = 230, Exhaust gas analysis = % volume C02 = 2.6, H,O =21, Np= 63, Op = 136 i jFE ¢ Design and Performance Simulation us SESULTS .. . DESIGN case unrtReD to, Fe} Mate 0 heat leas-t. 2 gus temp to HG 1400 poe Is0060 vol Cote? Hae te “He's'7s Gos 'Ssee. 0 timp wat/sta ty press flow patm ech appreh fot Teor oth eg) ER moo AESULTS... PERORMINCE CASE... UNFTAED pita 6, faT hota 60 heat lse-te 2 gas tmp to HESGe 1200 hm 16ND Evol Caan FHSS 12 Hee DE GSE Shae is tem watisin duty, past flow ptm pach appech a oe SH Me au ane as anes 100.0 190k we mone Ses es 88 : sae ae 's example illustrates how COGEN may be used to simulate gen of an operating HRSG from field data and then predict its gr™ance at a different condition.250 ‘Waste Heat Boiler Deskbook USING FIELD DATA TO PREDICT HRSG PERFORMANCE Example 7: ‘Table 46, Data For Design and Performance “This is an example where the design is simulated from plant opérating Gata, Table below shows the operating cata for a HRSG for incinerator exhaust gases. By playing with pinch and approach points, a design is artived at matching the operating data, Then lis periormance is predicted for case 2 conditions. The HRSG consists of an evaporator and economizer, module 4 ‘and generates saturated stoam. 1.Case 1 2 Gas flow, pph 150,000 165,000 3: Gas inlet temp, F "1400 1300 44. Steam pres, psig 400 218 5. Feed water temp, F 250 240 6. Steam flow, pph 42,400 2 7. Exit gas temp, F ‘400 2 8. Blow down % 3 3 [Gas vol %-CO2=7, H2O = 12, N2= 75, O2= 6 Heat loss % = 2] sults from COGEN are shown in Figure 4-20.” The results of COGEN may also be used to check if there is fouling of heat transfer surfaces. The results from COGEN are based oon the fact that there is no appreciable fouling. For example in the above case 2 if the stack gas temperature were 420°F versus 384°F predicted, then one can assume that there is some fouling in the HRSG. If the gas temperature at the economizer inlet were measured and found to be much higher than 544°F then the fouling could be in the evaporator. MULTI-PRESSURE HRSG DESIGN AND PERFORMANCE ‘Asa final example, the design and performance of a multi pressure HRSG with supplementary firing will be simulated. This ‘example shows how one can simulate the behavior of complex andMode + eva Modale2 fear]. Modate 3 fear] [Eco] Modste + Zc0] Modle S| Modole2, | Module2 Module 5. Mod a San wae OME EE Modate 6 ‘tio” Fee Y Stein 3,'&, anda Sactec Ce igure ¢21a. HRSG configumation for problem. (S<-Teck publishers) This is a multipressure HRSG. A common sconomizer provides [fd water to both HP and IP evaporators. The configuration is Biained by combining four modules, namely 22,5, and 3. Both HP [id IP steam are superheated, while LP steam is saturated. The ign case assumes that 375°F water is obtained from the common Since its flow is unknown to begin with, inputting of zero for the Ester flow will enable it to be computed automatically. In the Fietormance case, the desired HP stcam is obtained by firing the duct ESsrner, which automatically arrives at the firing temperature and ‘input. Note that augmenting air is also used for the burner. A EEsuperheater is also simulated. The program also automatically es at the flow through the common economizer and its water temperature, which is the inlet feed water temperature to HP [E:IP steam systems. All these complex iterations are completed fat5in 30 to 40 seconds with a 386 processor. The design case takes22 ‘Waste Heat Boiler Deskbook RESULTS... DESTOR CASE rae ‘and tenp-f060 rel bute heat Joss-te 1 gas temp to HiSG = 900 es Hlowppbe 1218000" vol cO2e'2.6 #20 = 22 N2= 6302 = 12.4 S020 ms tony vat/stn duty press flow ptm plnch aprch owt? Tneoub-F Worm psig) ph pee st 900 977492 G00 ge ee woo Bie Or sae Sse a Wid 10 nsw ine ae ia Ss 3 as 1s ee eo #4 m6 240 5 0 wns 0.0 can.eco OW 36 255 190 2D 7.08 MOIS OTF RESULTS... PERFORMANCE CASE... FIRED anb temp-fo 60 rel hunt = €0 beat ore-Z= 1 gas temp 20 HRSG « SO0 sat Florpphe. 1215000 vol CO2 = 2.6 WO) 22 N2= GI O2= 12.4 S02 od gat tem vatiste duty press flow pst pinch appre Grout Toute itm pela pph TB re eum 900 1003 00 as.7 «0 zt a SH 100s ‘S71 458690 M-st_ gop 124936 100.0 desu "S51 Gs) Seo sa 'O.co gn rae ® EP $71 6 366 453 OPIS 623126986 100.0 105127 EVA 658 417 366395 84.72 218 98382 100.0 2328 Feo 417352240 355 29.41 650 28575 0.0 con.eco fp 352° 255 190 240 26.51 1026813 100.015 25 flog ater SG = 1262984 ol cad» 2.9 Nw 22.10 92 = GIT 21.96 ‘fuel GAS: analysis. volume 2 emne +2 3. propane = 1 e-bturey ft = 942 Teobealin "> 2taae aug air-pph 25000 NOTE: TEMPERATURE PROFILES ARE NOF GENERATED BY SOFTHARE, fuel toput 48 on UW bast igure 421b. Printout of remus from COGEN. [Sck-Teck publishers)Design and Peformance Simulation 23 900 190 ae a irene ea ea con ot eae BaP * com CO BAP Case no. 1 2 . Remarks design performance - Heat source gasturbine gas turbine Gas flow, pph 1215,000 —"1215,000 . Gas temp to HRSG, F 500 900 HP steam pressure, psig «0 600 IP steam press, psig 205 205 - LP steam press, psig 10 10Waste Heat Boller Deskbook 9. Feed water temp, F 240 240 410. Feed water to HP, F 375 2 11. Feed water to IP, F 395 2 12. HP pinch point, F 150 2 13. IP pinch point, F. 2 2 14. LP pinch point, F 5 2 15. HP steam flow, pph 2 125,000 16. HP steam temp, F 0 0 17. IP steam flow, pph ? 2 18. IP steam temp, F 500 2 19. HPSH press drop, psi 15 2 20. IP SH press drop, psi 10 z 21. Feed water temp to LP, F 0 390 22. LP steam flow, pph 2 2 [fuel used in case 2N. gas:% vol Cl = 97, C2 = 2, C3 = 1, ‘augmenting air flow = 25,000 phy; exhaust gas % vol CO2 = 2.6, HzO 52, Np = 63, O2 = 12.4; heat loss = 1% and blow down = 2% each pressure level] REFERENCES 1. V. Ganapathy, "Simplify heat recovery steam generator evaluation’, Hydrocarbon processing, March 1990, p 77 2. V. Ganapathy, "Analyze options for deaeration steam", Power, September 1989, p 35 3. V. Ganapathy, "HRSG temperature profiles guide energy recovery" Power, September 1988 4. V. Ganapathy, "Win more energy from hot gases", Chemical Engineering, March 1990, p 102 5. V. Ganapathy, "Program computes turbine steam rates and properties", Hydrocarbon Processing, Nov 1988, p 105 6. V. Ganapathy, "COGEN”, software published by Sci Tech, New Jersey, 1990NOMENCLATURE [ji-—Blow down fraction; if blow down = 2%, then bd = 0.02. Br p—Gas specific heat, Btu/Ib°P. E;—A factor accounting for gas properties, defined in Equation 45. —Heat loss factor; if heat loss = 2%, then hlf = 0.98. 2 s1 —Enthalpy of superheated steam and inlet steam, Btu/Ib. Bi; hewz —Enthalpy of water at eco inlet and exit, Btu/Tb. as thermal conductivity, Beu/ft KF. ‘Kp, Ka —Factors obtained from design conditions, Equation 4-14 02, Q3,—Energy absorbed in superheater, evaporator and economizer, Btu/h; subscripta = assumed and t'= transferred. 2 Tg2,T 3. Tot —Gas temperature distribution, °F. Surface area, sq ft. "—Log-mean temperature difference, °F. Taaga —Average gas temperature in HRSG in performance and design modes. tw —Water temperature at inlet and exit of economizer, °F. sf 152 Saturated and superheated steam temperature, °F. Overall heat transfer coefficient, Btu/sq ft h°F. 'SJp Product of U and S in performance mode. Eo, Wed —Gas flow in performance and detign modes, pph. By, Wea —Steam flow in performance and design modes. {2 ap —Gas pressure drop in design and performance, in. WC. [Gas viscosity, lb/ft h.highlight some of the areas which should be emphasized or rated so as to result in a boiler that will be functional and cost git APPLICATION OR SYSTEM DESCRIPTION ~The application or the system of which the HRSG is a part yuld be described completely. Often I come across specifications Hit simply bay waste heat boiler for flue gas heat recovery without Bentioning the process or the source of the gas. Depending upon the process the gas stream may have peculiar acteristics which can be inferred if the application is stated. instance if the flue gas is from a municipal solid waste Hcinerator,|we know that it can be dirty and hence precautions Brould be taken to have cleaning lanes, soot blowers and avoid If the flue gas were from a fume incinerator then it is much inet and hence extended surfaces may be used to obtain a compact ign. If the effluent is reformed gas from a hydrogen plant, one can Fe that provision is made for exit gas temperature control and design the boiler to handle the high gas pressure. Extreme258 Waste Heat Boller Deskbook ‘operating conditions or upset conditions should also be described so that the design can accommodate these conditions. Mention should also be made as to how and where the steam that is generated is being utilized so that steam purity could be given consideration. If steam is used in a steam turbine drive, depending on the pressure, the purity could be demanding. Proper drum internals, Figure 5-1, should be used to give the desired purity. In the Cheng cycle discussed in Chapter 1, where the superheated steam is injected into the gas turbine, a combiration of internal and external steam separation devices were used to achieve a steam purity of less than 50 parts per billion solids. In high pressure boilers, a combination of cyclones, and chevrons may be used to obtain a steam purity of less than 100 ppb solids. If steam is used in a steam turbine or injected into a gas turbine for NOx [purposes or for increasing its power output, the specification writer should obtain information on steam purity requirements from the suppliers of these equipment so that the boiler drum internals can be properly designed. If the steam is used for process heating or cooling and is ata low pressure, a demister pad which gives 1 to 3 ppm solids may be adequate. A flow diagram showing the flow path of water and steam in the system is helpful, particularly in HRSGs for multi-pressure units. In some boilers saturated steam may be taken off the drum and the balance superheated. In some systems, external superheated steam from another boiler or source mey be mixed with the saturated steam from the waste heat boiler and the total may be superheated in the waste heat boiler. In order to understand and properly design the various heat transfer surfaces such a flow scheme would indeed be helpful and avoid confusion later. The take off point for deaeration steam is also an important aspect of any HRSG system; as discussed in Chapter 4 if steam is taken off the steam turbine at a suitable extraction point, the plant efficiency is better than if it had been taken off from the boiler outlet, This is because the steam that is drawn off the turbine has performed some work in the process of expansion. The scheme or flow diagram should suggest the source of deaeration steam or any other steam for process.catying Waste Hat Boers“ 29 Echevron’ 2eyctones ‘downconer arises Steed water pipe estear FigureS1. Arangementof drum intemals: 2. SPACE AND LAYOUT GUIDELINES <~ Waste heat boilers may be located inside a building, where the ;pace is limited or could be located outside but within a plant as part Gof the expansion scheme or there could be no restrictions as in a new E project; anyhow space limitations and layout restrictions if any should be outlined in the specifications. * ‘The boiler configuration could be developed based on the space. ailability. As seen in Chapter 1, for unfired gas turbine exhaust a iwo-pass could be used instead of a single gas pass design to minimize or space. If height is a concern, a short but wide boiler could be igned. There are cost implications to these special designs and the specification should state any requirements in the early tages of the project. As discussed in Chapters 1 to 3, several ifferent boiler configurations could be developed for the same gas steam parameters.20 ‘Waste Heat Boiler Deskbook 3. GAS PARAMETERS All of the data pertaining to the waste gas should be stated clearly. a. Gas flow in mass units NOT in volumetric units. One of the common errors made by specification writers is to state the gas flow in cubic feet per minute. This should be avoided. ‘The gas mass flow to the boiler should always be given in pounds per hour; one can confuse cfm with acfm or scim and come up with a mass flow that is different from the actual flow due to an incorrect estimation of gas density which is dependent on the gas molecular ‘weight, temperature and pressure. ‘Thus several variables are involved and a wrong assumption of any one of them could result in a different density and hence mass flow. The author is familiar with a case where a specification ‘mentioned that the gas flow was 5000 cfm and did not say whether it ‘was acfm or scfm. Two different boiler designers came up with designs in which the mass flows were apart by a factor of 4; one designer had assumed that the flow was in scfm and the other thought that the flow was in acfm. A lot of time was wasted in the process of explaining the basis of design to the consultant, which could have been avoided. If the gas is exhaust from a gas turbine, it is prudent to state the {gas flow data at different gas turbine loads and ambient conditions and also specify the design point so that comparison or proper evaluation of boiler performance could be made. As mentioned in Chapter 1, the mass flow and exhaust gas temperature vary significantly with ambient conditions. If a bypass damper is used the leakage flow through the damper should be accounted for. In some wood based cogeneration projects, a portion of the hot gases are taken off for drying the wood products and the balance goes to the HRSG. Hence the gas system should be described adequately so that the HRSG mass flow can be properly evaluated. Gas temperature to the boiler should be stated along with the gas flow. If the duct work between the waste gas source and the boiler is significant, the heat loss in this should be accounted for.Eiying Waste Heat Bolles 2 i ana. P'This i another important aspect ‘which is neglected by B itication writers. Depending upon the composition, the gas B operties such as specific heat, viscosity and thermal conductivity ‘vary; Appendix E shows how to compute the properties. If the {properties are different naturally the’ heat transfer coefficients hnence the duty and gas temperature profiles will be different as ssed inf Chapter 4. i ‘The difference in duty could be as high as 10% if the moisture hhtent is significantly different. Hence-if a steam: injected: gas ine’ exhaust boiler is being specified it is advisable to state the flow, exhaust temperature and analysis pertaining to the snode ‘operation of the turbine, whether steam injected or dry. in the Cheng cycle, discussed in Chapter 1, the gas properties Pary significantly between the dry mode (7% volume of water vapor) Brad the injected ‘mode (25% volume of water vapor). Appendi< E ows how the gas properties differ. As the size of the boiler gets ger or the steam production increases, even ato 3% difference in butput can affect the project economics. ‘Also, depending upon the process, the gas stream could have [aces or small fraction of Hel, Clz, SO2,H2S, or Sg, It is in the fiterest of, the specification writer to mention this as the design. Gyyould have to consider methods of minimizing low or high Fmperature corrosion concerns as discussed in Appendix B, even jough the heat transfer aspects or sizing may not be affected. ‘in hydrogen plants for example the’ data on partial pressure of rarogen is used in the selection of boiler tube material as discussed Chapter 1. Hence gas analysis is relevant. Gas velocity and. pressure drop through the boiler are also affected by gas density P vhich in tin depends on the gas analysis used. Evenvif en analysis is not readily available, the specification Should describe the process of obtaining the flue gas-eg. natural gas Kombustion at 20% excess air-and the boiler designer can perform ombustion calculations to arrive at the flue gas analysis and flow. Gas pressure and pressure drop. 'A fire tube boiler may have to selected if the gas pressure is igh; if it is marginally higher than atmospheric pressure and 4262 ‘Waste Heat Boiler Deskbook ‘water tube boiler is used, the casing would have to be strengthen or as discussed in Chapter 1, the boiler or superheater or economizer ‘may have to be located within a pressure vessel. If the gas pressure is very high, on the order of a few atmospheres, the gas properties would also be affected. See Reference 2 on computation of gas properties at elevated pressures. Appendix E shows how gas pressure affects heat transfer coefficients in flow inside or over tube bundles. Gas pressure also affects the density and hence the velocity and gas pressure drop. Gas pressure drop through the boiler is an operating expense. As ‘mentioned in Chapter 1 in the case of gas turbine exhaust each 4 in we of additional pressure drop results in nearly 1% decrease in the gas turbine power output. The specification may suggest a pressure drop but should give credit if a HRSG vendor has a lower gas pressure Grop than suggested or penalize the design ifit is higher. ‘The life cycle cost of this expense should be evaluated, though the initial cost of the design with a high gas pressure drop may look appealing. D. Nature of gas. ‘The nature of the gas stream should be stated, whether it has particulates, whether it a dirty gas and has potential fouling or slagging concerns. If an ash analysis is provided one can infer from it whether the ash has slagging characteristics; if salts of sodium and ‘potassium are present, then as discussed in Chapter 1, we know that the melting points of their salts could be low and can result in slagging problems at the front end of the boiler. Depending on the type of combustion or incineration device and the process, whether itis a fixed bed, moving bed or fluidized bed, the carryover of particulates could be different and hence has a bearing on the boiler design. The particulate concentration in sgrains/scf should be stated. Bare tubes have to be used in the case of a dirty gas, while extended surfaces could be used if itis clean, thus making the design compact. The author is familiar with cases where extended surfaces were used on boilers which should have had bare tubes and which hhad to be replaced completely within a few years of startup. BlE BOILER DUTY E* The duty or the energy to be recovered from the boiler should be statéd. This includes the steam and feed water parameters: fd the steam generation. This may not be difficult in cases such as ingle pressure boilers. However, if this is‘not clear, different ers can-come up with different combinations of evaporators:and The specification should alternatively state the criterion used ‘optimization; the cost of energy, steam and fuel should be G covided so that the boiler supplier can make his own economic ialuation. "Inthe case of multiple pressure boilers as discussed in Chapter EY one can have different combinations of HP, IP and LP steam; the Secification writer should analyze the HRSG temperature profiles Eising software such as COGEN and arrive at the temperature (profiles and duty of each section at different operation conditions. Some specifications require that the pinch point or approach have certain values. This is a poor way of specifying boiler Fluty. As discussed in Chapter 4, the pinch and approach points are a tion of the gas flow and inlet gas temperature and vary with rating conditions and load. There could be temperature cross uations and one cannot arbitrarily select these values. Some specifications call out the desired exit gas temperature fom the boiler; now this may or may not be achieved with the Elowest practical pinch and approach points as discussed in Chapter Hence one should be familiar with HRSG temperature profiles ind how they vary with gas inlet conditions before suggesting the Bj tuty or temperature profiles. Part load conditions if any should also be addressed in the Ppecifications. In the case of incinerator exhaust, the gas exit mperature would be lower at lower loads and concerns about low B emperature corrosion should be addressed.2 ‘Waste Heat Boiler Deskbook Some engineers are under the impression that low temperature corrosion in the economizer can be avoided if the boiler exit gas temperature is increased at lower loads ty using methods such as ‘bypassing a portion of the economizer. This is incorrect. As discussed in Appendix B, the tube wall temperature of the economizer is not affected by the gas temperature but by the feed water temperature. Hence by bypassing the economizer, we are not increasing the tube wall temperature of the economizer. If the objective is to ‘minimize corrosion in the ductwork between the economizer and the stack, then probably it could be achieved by this method. ‘The steam temperature varies with gas flow and inlet temperature conditions. It is desirable to know if the steam temperature could be allowed to float within a range or-a constant temperature is required at various gas flow conditions. Oversizing of the superheater may be necessary to achieve the desired steam temperature at lower loads. Provision should be made for steam temperature control at higher loads. In the case of gas turbine HRSGs it should be stated clearly if the same steam temperature is required in the unfired as well as in the fired modes. As shown in Chapter 4, the superheater should then be designed to generate the steam in unfired mode and the temperature would have to be controlled in the fired mode. If a lower steam temperature is acceptable in unfired modes, then this should also be stated as it will help reduce the size of the superheater. 5, AUXILIARY FUEL DATA In case supplementary fuel is needed, as in gas turbine applications, the fuel analysis and augmenting air requirements if any should be stated. The burner supplier can provide information on augmenting air requirements based on the gas turbine exhaust gas analysis, the firing temperature and the fuel used. If the bummer is required to bum a different fuel at a later date, this requirement must be stated in the specifications. Modifications that may be required to be done later to the boiler configuration could be reviewed early in the design and handled less expensively.28 EMISSION GUIDELINES Fn the case of gas turbine exhaust, where a SCR or catalyst is F a for CO or NO; contro, the incoming levels of these pollutants as fil as the desired outlet level should be stated. " ven if the SCR is likely to be added after several years, the should be designed with this requirement in mind. Space fould be provided for the SCR.In addition, the boiler cross section FEED: WATER ANALYSIS AND: BLOWDOWN Often the feed water analysis is omitted from the specifi- Ecjons. This is an important data and is helpful in determining, the down requirements. The boiler water quality depends on the steam pressure. Both ‘A (American Boiler Manufactures Association) and ASME have Fesued guidelines, Tables 5-1 and 52. Forde to calculate the blow down, the extent of make up water and Fondensate returns should also be known as illustrated in the fecample below. féxample 1: EA boiler generates 25,000 pph of saturated steam at 400 psig. Fron of the feed water is condensate returns ard 40% is make up ater, which has a'TDS of 250 ppm. The boiler drum is maintained £2000 ppm Determine the blow down, make up flow and condensate Solution: Let us define a few variables. See Figure 5-2. Let E= evaporation, 25000 pph R= condensate returns, pph M-= make up water flow, pph F = feed water flow, pph B= blow down, pph S = steam to process, ph = E-Ri ows sr gpie> pts git gO eOO™ a iret — a So0l> Lal ee i ee ee i ee a ee > oe colon: ESS rs Seem een ognes coos wos sete vrs © Gorey oorse vote worse Saqmousine owe “outs | “vores “ores ects “cootr cious 010 == sivos «ico «—«aus_—=«ceocr «sro « (noah see Glows oovor «somos savor «orogens foe now toposes 10 Hom ww TORETOST) UST=TOUT) WOOT=T08) OUETSZOET=TO | Menseo: “weoreso epeare irears uss fund woos uo uote och unsp [snsussuog 6261 SINSV Wo. peidepy) g sy Aven 2018 polsoBBng “1-5 O1geLsss0001d prone 5B Mo] SB 198 0¢ Ly si y0q evig “(g e104 pur Ajugerye Hos Ue 006) ea | 1uoo Hd-eyeudsoyd porsuipioco 10 ywens6u0o pounsse (phy winysseyod 10 wy Pm yo Wooo" : eluyeen ese Boljgeds Ateripinpuy oq isn (isd 008) 2g 12'9 MoI sH0}0q UA je) elfséeid 1045 sonjen osn (gWUWimig OOO'OSL<) gWAKA Z'ELY< BBxny TeEK [0) ing Waste Hoat Boilers Ig eges 401 S010N268 ‘Waste Heat Boller Deskbook Aa ‘Tables Watertube boilers Recommended Boller Water Limits ‘nnd Associated Steam 2a ‘At Sleady State Full Load Operation ‘Dram IABMA= 18821 Ove ane ge Srl Rare Peoure —TauDiched ‘Tes Akay Berane TaalOsshed Ses ro St Br Sr our— ppm 9uD Sa ner 0) pom em AK epeced vel oa yongsoo 140700 ® one sevaso —wtnsteo 120400 * 0230 even stoasto 1084500 : ease ous oso 48-200 3 ox-08 gaa “tem S018 2 oxas core 565 ENE 1 aras corsa 1 NOTES) 1 a reovas0 0 wa a pera a WA os 201-2900 s wa 008 peor ‘ue Tan ERS onandieome 08 Nm NR ons movant Be eral vate wtint age fete TDS nt ed wae Hh ala a hss, oie few ci ena ‘2a er whine ange ar dct opal ba ce ae TDS lor ma hr wave tee iyi bor ae worn sis bela 8. Deed by ar war amen. 2 Tyerhmear ae ce, ‘The total dissolved solids (TDS) in blow down water is 2000 ppm; TDS of make up water is 250 ppm; the steam or condensate returns is assuined to have 0 solids. Mass balance gives the following equation FeM+R (5-2) since the condensate returns and make up are in the ratio of 60 to 40%, R/M=15 63) _éiFigure 52: Scheme of blow down system: 250;M +0 x R = solids in feed water = 2000 B 6-4) orM=8B 65) also, R+ M = 25000-+ B 5-6) bstituting (5-3) and (55) in 5-6), 25M = 25000 + Bor 20 B = 25000 +B 6-7) & ce B = 1316 pph; M = 10526 pph; R = 15790 pph; F = 26316 pph case all of the feed water is make up, then R= 0. Then: F=M (5-8) 250 F = 2000 B 69)m Waste Heat Boiler Deskbook F=25000+B 5-10) hence B = 25000)7 = 3572 pph and F = 28572 ph. Note also that the deaeration steam requirements will change with the ratio of R/M . The more the condensate returns, less the deaeration steam. In addition to the feed water or make up water analysis, the specification should give the feed water temperature. It is usually the saturation temperature corresponding to the deaerator pressure if thermal deaeration is used. 8, SURFACE AREA AND FIN CONFIGURATION In applications such as gas turbine exhaust, extended surfaces are widely used. Depending upon the fuel used in the gas turbine and HRSG burner, up to 5 or 6 fins in. may be used for the evaporator and ‘economizer. Lower fin density should be used for surfaces with a low tube side coefficient such as superheaters or air heaters as discussed in Appendix A. ‘Engineers who are not familiar with heat transfer aspects of extended surfaces sometimes fall into the trap of assuming that more surface area means more duty. This can be wrong as shown in Appendix A. As surface area is determined ty a combination of factors, specifying the fin density to be used or purchasing a design based on surface area alone should be avoided. In applications where the gas stream could be slightly dirty, the maximum number of fins to be used from cleaning considerations may be specified. Surface area is also affected by the gas velocity used. One should also note that the surface area is sometimes specified on tube inner diameter basis in case of fire tube boilers. Finally, the basis for evaluation should be the overall operating and installed costs and not surface area.1g Waste Heat Bailes m COST DATA FOR ELECTRICITY, L AND STEAM Information on the cost of utilities such as electricity, fuel and jgner can make his own economic evaluation of the design. In a HRSG for example, there can be variations in gas pressure drop, ‘consumption and steam production among different designs. If the cost data and the period of operation in each mode (ed, fired low load, high load ete.) are known, a:study. may be, Field the lowest life cycle cost. ‘The consultant should also use a life cycle costing analysis to I investment alone. While selecting an option based on low cost: may look attractive in the short term, in the long term it xy proveto be uneconomical. Hence the cost of major utilities such steam, ful, electricity should be stated in the specifications. ‘One of the parameters that is often suggested in the par BE specification is the drum hold up time or the duration between ipty. The objective is to ensure that steam can be generated for a minutes even if the feed pump or the heat source is cut off ithout setting off the level trips. The drum sizing is usually based on steam release rates and the Figure 53 shows the formula used for computing the hold up as, function of the liquid level. The formula may also be used to ‘pute the duration between levels as shown in Example 2 below. imple 2: A oller generating 20000 pp of steam at 400 psig has a 42 in, is 2 in. below the center line and the low level cut off (LLCO) BS 4 in. below it, determine the duration available between NWL and .CO at the design steaming rate.