GATE: A Simulation Code for Analysis of Gas-Turbine Power Plants

THE AMERICAN SOCIETY OF MECHANICAL ENGINEERS 345 E. 47 St., New York, N.Y. 10017 The Society shall not be responsible for statements or opinions advanced in papers or in discussion at meetings of the Society or of its Divisions or Sections, or printed in its publications. M Discussion is printed only if the paper is published in an ASME Journal. Papers are available ` from ASME for fifteen months after the meeting. Printed in USA. Copyright 1989 by ASME 89-GT-39 0 GATE: A Simulation Code for Analysis of Gas-Turbine Power Plants M. R. ERBES R. R. GAY ENTER Software, Inc. Menlo Park, CA and A. COHN Electric Power Research Institute Palo Alto, CA ABSTRACT The GATE (GAs Turbine Evaluation) code has been developed to evaluate the design and off-design performance of existing and advanced gas-turbine-based systems for power plant applications. By combining an intuitive, graphical user interface with detailed analytical models for the thermodynamic, heat-transfer and fluid-mechanical processes within gas-turbine-based power plants, GATE can be used by novices as well as experts for complex design and simulation studies. It can model a variety of gas turbine configurations and cooling technologies, and users can also interactively design and analyze an associated steam bottoming cycle. The basic formulations used in GATE are presented here, along with sample cases demonstrating the power and flexibility of the code. NOMENCLATURE Greek: C K P r T W X Y constant turbine flow constant of proportionality pressure stage pressure ratio temperature mass-flow rate cooling flow fraction of compressor inlet air flow cooling flow fraction of total local gas flow A efficiency penalty factor for cooling q (D stage isentropic efficiency cooling effectiveness O pattern factor yf turbine loading parameter Subscripts: 1 inlet to stage or section 2 exit from stage or section 3 compressor exit statepoint 4 turbine inlet statepoint b bucket c cooling flows des design conditions dsch compressor discharge conditions g gas flows max maximum mean mean value met n metal nozzle off off-design conditions w wheelspace INTRODUCTION Analyzing the performance of modern gas-turbine engines incorporating advanced cooling technologies is a complex task. Since the cooling flows are a large fraction of the inlet air flow in these machines, it is necessary to accurately model these flows in order to predict the performance of real engines. The design-point cooling flows are a complex function of turbine operating and design parameters such as compressor efficiencies, bleed pressures, turbine gas temperatures, and allowable metal temperatures. A large amount of detailed gas turbine design information is therefore required to accurately model the performance of modern gas turbines. Gas turbine manufacturers generally develop and use proprietary computer models ("cycle decks") to predict performance under design and offdesign conditions. These cycle decks are complex programs which are "tuned" to predict the performance of specific machines: they are not intended to be general-purpose gas-turbine design tools and manufacturers usually do not make them available for use by others. The GATE code was developed to deal with these issues and allow engineers to analyze the performance of a wide range of gas turbine designs quickly, easily, and accurately. It was developed under Presented at the Gas Turbine and Aeroengine Congress and Exposition June 4-8, 1989 Toronto, Ontario, Canada

sponsorship of the Electric Power Research Institute (EPRI) and is available from ENTER Software, Menlo Park, CA. It is a detailed, steady-state gas-turbine/combined-cycle performance evaluation code which runs on IBM-PC compatible computers. For the user interface functions, GATE uses the EASE+ graphical data-base management system from Expert-Ease Systems. EASE+ uses hierarchical menus, forms, context-sensitive help, range-checking and processing on data entry, and a variety of icon graphics and plotting tools to allow users to easily enter input data and graphically view the results. GATE is suitable for a wide variety of tasks, such as: analyzing the overall performance of a proposed gas-turbine-based power plant, checking the claims made by vendors about the performance of particular machines, simulating the performance of existing systems at off-normal operating conditions, or predicting the effect of proposed changes or enhancements to existing gas turbine systems. GATE can be used to analyze gas turbines with steam injection, reheat combustors, or compressor intercooling in simple-cycle configurations, combined-cycle plants, or steam-injected cycles. MODEL FORMULATION In order to use GATE to evaluate the performance of a particular gas-turbine-based power system, the user must first input enough detailed information to specify the engine design and operating point. GATE then performs a detailed design-point analysis, and automatically saves the results for use in subsequent off-design performance simulation. A compressor-mapping routine is used to evaluate the performance of the compressor sections. The expansion through the turbine is evaluated on a stage-by-stage basis, and the required cooling flows are determined using semi-empirical cooling-effectiveness relationships (Cohn and Waters, 1982 and Waters, 1983). GATE computes and outputs key plant performance characteristics, including: net power produced, overall heat rate and efficiency, statepoint temperatures and pressures, and component efficiencies and cooling flows. GATE can also be used to simulate the performance of a given gas turbine design over a wide range of off-design operating conditions. The user need supply only the following data for an off-design run: the ambient air conditions at the compressor inlet, the inlet guide vane settings, the operating speed of the power turbine (which is usually the same as the design speed), and either the desired turbine inlet temperature or the required power level as percentage of the designpoint power level. GATE then calculates the compressor inlet flow rate and efficiencies, cooling flow rates, turbine stage efficiencies, and overall engine performance. A gas turbine in GATE consists of up to four spools. A spool comprises one or more sets of turbine nozzles and buckets (stages) attached to a single shaft, and rotating at a given speed. Each compressor spool is mechanically attached by a shaft to a turbine spool; the final turbine spool may optionally be designated as a power turbine free from a compressor spool. Each turbine spool has up to five stages, with a limit of ten total stages. Intercooling can be specified for each spool of the compressor. The heat sink for intercooling may be either a surface type heat exchanger, or evaporative cooling from a water source. Compressor bleed points for turbine cooling can be specified at any pressure ratio in each compressor spool, with a limit of fifteen total bleed points. Bleed flows can be chilled using a surface-type heat exchanger before ducting the air to the turbine. Alternatively, steam from an external source or from a steam bottoming cycle can be used to cool the gas turbine. The compressor-mapping routines (CMAP) supplied with GATE are used to predict compressor performance over a wide range of offdesign operating conditions. CMAP reads the necessary design-point information from the GATE data files, generates the compressor map parameters, and writes the results to a file. GATE accesses this map file during off-design analyses to determine the compressor operating parameters. GATE users can use CMAP to automatically generate compressor maps for particular gas turbine engines. These maps are generated using user-input design-point data and one of two empirical data representations supplied with GATE for typical compressor designs. The map generated for a particular engine can be reviewed and edited by the user. Thus, if more specific performance data is available for a compressor, the GATE user may enter that data manually to generate a more accurate map. CMAP is based on the data and methodologies presented in Johnsen and Bullock (1965) and Bettner and Sehra (1986), and is presented and discussed in detail in Gay (1988). The fuel for the gas turbine is specified by giving the lower heating value of the fuel and the mass or volume percentages of the major constituents. GATE uses stoichiometric calculations to determine the composition of the combustor exhaust gas. GATE calculates the thermodynamic properties of air, steam, and gas mixtures using a set of property routines which use polynomial fits based on data presented in Keenan and Kaye (1948), Keenan and Keyes (1969), and Yaws (1976). Included in the GATE code is a set of routines which can analyze an arbitrary steam bottoming cycle for any given gas turbine engine. The user designs the steam cycle interactively by selecting components from a library and graphically specifying their connections. GATE provides logic checks to ensure that the connections are made properly. The user then inputs data to specify the design conditions for the steam cycle components. The GATE steam-cycle analysis code then determines the performance of the complete combinedcycle power plant. Case studies are easily performed, because data for a particular plant configuration is stored separately from the data specifying the model layout and connections. The user can use the same model for different gas turbine engines, or the models can be copied and modified as needed for each case. Design-Point Gas Turbine Analysis A design-point gas turbine analysis differs from an off-design analysis in that, for the design point, the GATE user must input a large quantity of detailed information which specifies the operating point. A library of input data sets for common industrial gas turbines has been prepared and is distributed with GATE. Work is proceeding on developing models for a wide range of engines. The designpoint input data sets include: air inlet mass flow rate, pressure ratios for each compressor spool, specifications for the bleed and intercooling points, fuel composition and heating value, turbine inlet temperature, turbine stage geometries, maximum metal temperatures, and parameters to specify the cooling technology. The user also can modify parameters which control the calculation of mass flow leakages and parasitic power losses. When the design-point inputs has been correctly specified and GATE is executed, the code calculates the operating performance of the power system, including fuel consumption, net power, cooling flow rates, thermodynamic efficiency

U and heat rate, exhaust temperature and gas properties, and thermodynamic statepoints throughout the system. GATE automatically saves the necessary parameters for the off-design performance simulation. GATE can analyze the effects of bleed flows and/or intercooling at each segment of the compressor. At each intercooling point the gas flow is extracted from the compressor, cooled in an external heat exchanger, and injected back into the compressor. Bleed flows for turbine cooling may optionally pass through an external heat exchangers before introduction into the turbine. The compressor map generated during a design-point run is automatically stored in a map file for each engine, which is then accessed by GATE as input for off-design analyses. A turbine map is not used in GATE; instead, a detailed stage-bystage analysis of the gas turbine expansion process is performed. Each stage consists of a nozzle (stator) followed by a bucket (rotor). Before each stage steam injection can be specified, or a reheat combustor may be used. If present, the reheat combustor heats the turbine gas to a user-specified temperature. In each case, the appropriate mass and energy flows are mixed with the main gas stream before the flows enter the nozzle. Before the expansion process through each rotor is analyzed, the nozzle cooling flows are mixed with the main gas stream. For a design analysis, the cooling flow rates to each blade are predicted from a semi-empirical cooling-effectiveness relationship (Cohn and Waters, 1982 and Waters, 1983): and, W (D ) 1.25 = C I (f) Wg 4)max"`" J (Tg,max - Tmet) (Tg,max - Td (2) A set of parameters is used to correlate tmax as a function of blade geometry and the specified cooling technology: this parameter set can be modified or input directly by the user. The GATE user also inputs the metal temperatures allowed on the blades and the correlation coefficient C used in equation (1). Default values for Care supplied with GATE for several typical cooling technologies such as convective cooling, film cooling, and transpiration cooling. The maximum gas temperature is the highest gas temperature in contact with the blade: it is normally higher than the bulk gas temperature around the blade. GATE calculates the maximum gas temperature using a user-input value for the pattern factor, O, defined as: _ (Tg,max - Tmear) (Tmean - Tdsch) (3) In addition to calculating the blade cooling requirements, GATE also allows for leakage of coolant flow both at the nozzle and at the bucket. The ratio of the coolant flow leakage to the main gas flow is input directly by the GATE user; default values for each cooling technology are written into GATE. Once the nozzle cooling flows are determined, the properties of the mixture of the cooling flow with the main gas flow are calculated to determine the initial conditions for the expansion across the bucket. The isentropic expansion across the bucket is modeled by dividing the total pressure drop into a number of equal increments (specified by the user); an isentropic expansion is then calculated across each step. The gas properties are recalculated at each step to allow for variations during the expansion. The actual work across the expansion is calculated as the isentropic work multiplied by the stage efficiency. The turbine stage efficiency is computed using the following formula: 11 = 1 max - (yr -0. 5)2- Xn A n - XbAb - X.A w (4) The user must input the maximum stage efficiency, fl max and the cooling flow penalty factors, A. The turbine loading parameter, Ni, is defined as the actual work produced in the stage divided by the onehalf the square of the rotor speed. Since the stage efficiency depends upon the work produced, iterative calculations are needed to determine the actual stage efficiency. After the expansion calculations are completed, GATE computes the cooling and leakage flows for the bucket using the same procedures described above for the nozzle. The exit conditions for the turbine stage are calculated by mixing these mass and energy flows with the main gas stream. Off-Design Gas Turbine Analysis GATE can simulate the performance of power systems over a wide range of conditions. The off-design analysis procedures automatically determine the inlet air flow, compressor pressure ratios, compressor efficiencies, turbine inlet temperature, and turbine stage efficiencies as functions of operating conditions. In order to predict the off-design operating characteristics of a gas-turbine power system, three sets of input information are required. First, the user must set up and execute a design-point analysis of the engine. GATE will automatically save the results of this calculation in a reference file to be accessed during the off-design calculations. Second, the user must generate a map of the compressor performance using the CMAP, the compressor-mapping program provided with GATE. The third set of input information required specifies the off-design operating point of the gas turbine using the following parameters: One of the following: the turbine inlet temperature, the first stage rotor inlet temperature (sometimes called the firing temperature), the turbine exhaust temperature as a function of compressor outlet pressure, or the desired power level as a percentage of the design-point power. The ambient air conditions at inlet to the compressor: these consist of temperature, pressure and relative humidity. The compressor inlet guide vanes settings. The operating speed of the power turbine specified as a percentage of the design-point operating speed (usually 100%). The above set of input parameters provide a convenient format for running a variety of simulation studies, but misleading results can be obtained if the user does not take into account the effects of certain input combinations. For example, the user can enter either the firing temperature or the desired power level. Gas turbine engines are typically designed to operate at a given firing temperature. In order to determine the performance of the engine over a range of ambient

conditions, the user would therefore normally input the design value of the firing temperature, and run the off-design analysis over the desired combinations of temperature, pressure, and relative humidity. However, the user can also enter the required power level instead of the firing temperature, which can lead to unrealistic performance predictions. If the user specifies a desired power level, GATE iterates on the firing temperature until that power level is achieved. There is no protection written into GATE to keep the firing temperature below the maximum specified by the manufacturer. For instance, if the GATE user specifies an off-design run at full power and at ambient conditions of 100 F (311 K) and 90% relative humidity, GATE will raise the firing temperature to generate the specified power. However, the resulting firing temperature will be significantly above the design value. In an actual gas turbine engine, the control system senses the turbine exhaust temperature (which tends to move proportionally with the firing temperature), and cuts down the fuel flow if it exceeds the maximum allowable level. The off-design methodology in GATE predicts the flow rates and pressure levels in the gas turbine expander, and then adjusts the compressor operating point (using compressor map calculations) to match the required conditions in the expander. The basic assumption in this approach is that the flow rate through a turbine stage varies directly with the pressure and inversely with the square root of temperature (Spencer et al., 1974 and Salisbury, 1950). In addition, if the flow is not choked, the pressure ratio across the stage influences the flow rate as indicated below (Erbes, 1986): where, W = K (P 1 /JTI) (5) r = P2 /P 1 (6) During the design-point analysis, GATE computes and saves the values of the turbine flow coefficient (K) at the entrances to the first and second turbine spools and at the diffuser entrance. During the offdesign analysis, GATE iterates on the compressor operating point(s) until the turbine flow constants match the design-point values. The local turbine cooling flow ratios for the off-design analysis are calculated from the design-point cooling flow ratios by adjusting for pressure and temperature. In GATE, it is assumed that the cooling flow ratio, Y, expressed as a function of the total local gas flow, varies as the square root of turbine inlet temperature divided by the turbine inlet pressure: Yc = We /W9TT /P4 - (7) This relation is derived by noting that the mass flows rates of the bleed extractions are proportional to the square root of the pressure difference between the compressor and the turbine: WeP3 - P4 (8) and, since this pressure difference is assumed to be proportional to the compressor discharge pressure: We o ) P4 (10) Now, by noting that for the choked flow at the turbine inlet, the following relationship holds: Wg P4/ 4 (11) The relationship shown in equation 7 is therefore derived by combining equations 10 and 11. Equation 7 leads directly to the following relation used in GATE to determine the off-design values for the cooling flows: Yc,off Yc,des FT4 - / P4)des (T/ P4)off (12) If there is only a single compressor spool, GATE need calculate only a single compressor operating point. The compressor speed is input by the user for the off-design analysis, and GATE iterates on the compressor pressure ratio until the turbine flow constant at the entrance to the first expander spool equals the design-point flow constant. The compressor map is then used to determine the inlet mass flow and the compressor efficiency. If an off-design GATE run is made for a specified power level (or turbine exhaust temperature), a second iterative variable must be added to the calculations: the turbine inlet temperature is adjusted until the computed results match the desired power level (or exhaust temperature). Otherwise, the turbine inlet temperature (or the first stage rotor inlet temperature) is held constant at the value input by the user. For multiple-spool engines, the iterative calculations become much more complex. For example, for a three-spool aero-derivative engine, there are seven key unknowns and seven equations used to solve the problem. The seven unknowns are the efficiencies, pressure ratios, and operating speeds of the two compressor spools, and the compressor inlet mass flow. The equations used by GATE to find their values are compressor-map functions for flow and efficiency for each spool (four equations), and the mass-flow functions for the first two turbine spools and the diffuser entrance (three equations). The solution algorithms for all of the possible gas turbine configurations which can be modeled by GATE are built into the analysis code: the user does not have to set up these equations for each GATE run. The user can control the relaxation parameters and convergence tolerances used by GATE. Steam Cycle Analysis A steam cycle analysis (CYCLE) code is included with GATE which allows the user to define and analyze an arbitrary steam bottoming cycle. This code predicts the performance of a complete combined-cycle plant by combining the results of the GATE gas turbine model with the results from the CYCLE code. The user creates a model by drawing the desired cycle configuration on the screen, selecting the necessary components from a graphical menu of icon representations. A list of the steam cycle components currently included with the GATE/CYCLE code is presented in Table 1. (P3 - P4) ` f'3 (9) by combining equations 8 and 9, it follows that the cooling mass flow is proportional to the square root of the turbine inlet pressure:

A Table 1: Steam Cycle Components Modeled Components Modeled Gas Turbine Steam Turbine Superheater Evaporator Economizer Deaerator Condenser Pump Pipe Valve Mixer Splitter Sink Source Makeup Exhaust Stack Table 2: Heat Exchanger Design Control Parameters COMPONENT METHOD SPECIFIED PARAMETER Superheater 1 Heat Transfer Surface Area 2 Effectiveness 3 Outlet Steam Temperature 4 Approach Temperature Difference 5 Outlet Steam Enthalpy Evaporator 1 Heat Transfer Surface Area 2 Effectiveness 3 Steam Production 4 Pinch Temperature Difference Economizer 1 Heat Transfer Surface Area 2 Effectiveness 3 Outlet Water Temperature 4 Economizer Exit Subcooling 5 Bypass Economizer After the component icons have been positioned on the cycle diagram, the user draws the connections between the components using a mouse or the cursor keys. The graphical tools provided with the EASE+ interface can be used to automatically route the connecting streams on the diagram, or they can be positioned manually by the user. Logic is provided with the connection procedures to ensure that all of the ports on each component are properly connected. For example, the user is not allowed to connect a gas turbine exhaust gas to the inlet water port of an economizer: gas stream outlet ports can only be connected to gas stream inlet ports. To prepare for the cycle calculations, the user must enter enough data to specify the design conditions for each component. Only a few input parameters are required for each component, and default values are supplied if the user fails to enter a necessary value. The data is entered through forms which are accessed from the cycle diagram by positioning the cursor on the desired component and hitting a single key. These forms contain data input fields which the user can modify, and also display the results from the model analysis. Context-sensitive help messages are available to guide the user through the selection of the input parameters. The data form for the superheater is shown in Figure 1. The parameters set by the user on the input forms control the calculational methods used in the design and off-design analysis of the components in the steam cycle. A number of different choices are available for each component: a listing of the control options for sections of the heat-recovery steam generator (HRSG) is shown in Table 2. The heat transfer analysis calculations follow the effectiveness-ntu (Number of Transfer Units) methodology outlined in Kays and London (1984). For design-point analyses (methods 2-5 in Table 2), the user-supplied data and the inlet stream conditions passed from the system control program are used to determine the performance of the heat exchanger and to calculate the heat transfer surface area. For off-design analyses, the user should select the fixed heat-transfer surface-area option (method=l in Table 2) for each section of the HRSG. The component analysis procedure then calculates the heat exchanger effectiveness and the outlet stream conditions from the specified area, which is saved from the design run or input by the user. The steam cycle drawn by the user can be of arbitrary complexity. A representation of a simple single-pressure HRSG can be created with only a few components, or a model can be developed for a complex, three-pressure level HRSG with a reheat steam turbine, as shown in Figure 2. To analyze the performance of the steam cycle, the CYCLE code first reads the connection and component data from the EASE+ graphical data base. The data for the gas turbine exhaust gas is then read in from the results file produced by GATE for the specified run. Next, the CYCLE code analyzes the connection data to determine the order of calculation for the components, a procedure known as flowsheet decomposition. This is a complex procedure which must ensure that each component is included in the calculations and that the order selected will allow the iterative calculations to converge. The CYCLE code proceeds to analyze the performance of the cycle by calling the appropriate model routines in the order determined by the flowsheet decomposition procedure. After execution of each component model, the output data from that component is passed to all connected components. One system iteration is completed when all of the component models in the steam cycle being analyzed have been executed. At the end of each system iteration, the CYCLE code uses a number of different criteria to determine if the model calculations have converged. First, the calculated output variables from each of the components must match the values from the previous system iteration within an error tolerance selected by the user. Second, there must be a mass and energy balance around each of the components in the model and around the system as a whole. Finally, the data passed between components must match: that is, if the outlet port from one component is connected to the inlet port of a second component, then the data stored in the program data structures for those two component ports must be equal (within the user-specified error tolerances). A typical GATE/CYCLE run will require from two to ten system iterations to converge completely, depending upon the complexity of the steam cycle being modeled, the convergence tolerances selected by the user, and the accuracy of the initial guesses used to initiate the calculations.

FORM:SPHT 11 1112/17/88 15:27 SUPERHEATER: Streams and Control Parameters Component ID: SPHT Description: Superheater Flag for Method.. (FLAGS[0]): 1 Input value.. (VALUES[0]) : 35000.00 Method = 1: Surface Area (ft2).. (VALUES[4]): 35000.00 2: Effectiveness....... (EFF): 0.93 3: Steam Outlet Temp (F)...(T[3]): 952.44 4: Temp Approach (F)...(DELT): 42.06 5: Steam Outlet H (Btu/lb)..(H[3]): 1473.98 Debug output flag.. (DBGFLG): 0 External toler. Max internal iters. (ITRINT): 30 Internal toler. (TOLEXT): 0.00 (TOLINT) : 0.00 Flowrate (W) Temp. (T) Pres. (P) (lbm/hr) (F) (PSIA) 0. Gas Inlet: 2.25e+006 994.50 15.17 2. Steam Inlet: 2.66e+005 565.95 1188.00 Enth. (H) Quality (X) (Btu/lbm) 243.33 1183.92 1.00 1. Gas Outlet: 2.25e+006 868.77 15.02 209.01 3. Steam Outlet: 2.66e+005 952.44 1128.60 1473.98 1.00 Fl Help F2 Range F8 Skp-Frm F9 Done F10 Nxt-Frm ESC Quit Figure 1. GATE/CYCLE Superheater Input Data Form Figure 2. Model Diagram for GATE/CYCLE Combined-Cycle Configuration The GATE code is written to run under the EASE+ program shell, which is currently available on personal computers running the DOS operating system and on VAX computers running VMS. On the PC, each GATE/CYCLE model is limited to approximately 100 components because of the memory limits in DOS. The execution time required varies with the speed of the machine, the complexity of the model, and the accuracy of the initial guesses used. On a 80386- based 20 MHz PC, the minimum analysis time for a steam cycle model with 25 components (as shown in Figure 2) is about 10 seconds. This corresponds to two system iterations, which is the minimum number required for system convergence. RESULTS AND DISCUSSION To illustrate the flexibility and power of the GATE code, the steps required to predict the performance of a combined-cycle plant as a function of ambient temperature and load are outlined below. The gas turbine selected for these sample test cases was the GE MS7001E engine. The first step is to develop and test a design-point model for the GE 7E engine. This requires data on engine design and performance at the design operating point which may be difficult to extract from 6

0 published literature. Fortunately, however, a design-point model of Table 3a: Summary of Results for Sample GATE/CYCLE the GE 7E engine has already been developed and tested and is Runs at 59 F. supplied with GATE. The user simply chooses this model from a menu of engines and executes the design-point analysis. PARAMETER MAX 24.7 40% POWER IGV POWER The design-point model for the 7E assumes that the turbine is fired by natural gas. At ambient conditions of 59 F (288 K) and GT Inlet Air (Ibm/s) 609 495 499 14.7 psia (101.4 kpa), GATE predicts that the gas turbine will Gas Turbine Power (MW) 76.7 59.9 30.5 generate 76.7 MW net electric power. The inlet air flow of 6091bm/s GT Efficiency (LHV, %) 32.0 29.6 25.0 (276 kg/s) suffers a 3.5 in H 2O (870 Pa) pressure loss to 14.6 psia GT Exhaust Temp (F) 992 1060 722 (100.7 kpa) at the inlet to the compressor, then is compressed to HP Steam (Ibm/hr) 206,400 194,700 38,900 172.4 psia (1189 kpa), for an overall pressure ratio of 11.8. Cooling IP Steam (Ibm/hr) 55,700 42,700 50,700 air extracted from the compressor and injected into the turbine totals LP Steam (Ibm/hr) 39,000 33,300 55,900 11.9% of the inlet air flow. Natural gas is consumed at the rate of HRSG Stack Temp (F) 344 323 381 10.6 lbm/s (4.81 kg/s); the combustor exit temperature is 2056 F Steam Turbine Power (MW) 38.4 36.5 12.1 (1398 K) and the inlet temperature to the first rotor is 1985 F (1358 Net Power (MW) 112.3 93.8 41.6 K). This exhaust gas exits the turbine at a temperature of 992 F (806 Heat Rate (LHV, Btu/kW-hr) 7280 7360 9997 K) and a flow rate of 6191bm/s (281 kg/s). Cycle y ent Efficiency y ( LHV, %) ) 46.8 46.4 34.1 After the design-point gas turbine model was completed, a model was constructed for the steam bottoming cycle using the EASE+ graphical interface tools described above. For these sample cases, a Table 3b: Summary of Results for Sample GATE/CYCLE three-pressure-level HRSG with a reheat steam turbine was used, as Runs at 0 F. shown in Figure 2. This HRSG consists of three evaporator sections, two superheat sections, and two economizing sections. The low- PARAMETER MAX 24.7 40% POWER IGV POWER pressure evaporator supplies the steam necessary for deaeration of the boiler feedwater. The high-pressure system is used to generate GT Inlet Air (Ibm/s) 697 556 562 superheated steam which flows through the high-pressure section of Gas Turbine Power (MW) 96.2 74.5 30.3 the steam turbine. Additional steam is generated with the GT Efficiency (LHV, %) 33.5 31.4 25.0 intermediate-pressure economizer and evaporator, which is mixed GT Exhaust Temp (F). 962 1022 596 with the steam exiting from the HP turbine section, reheated in the HP Steam (Ibm/hr) 218,400 202,600 55,600 intermediate-pressure superheater, and sent through the reheat IP Steam (Ibm/hr) 64,800 49,900 43,400 section of the steam turbine. Operating pressures of 1500 and 500 LP Steam (Ibm/hr) 42,700 36,500 32,700 psia (10.3 and 3.4 MPa) were selected for the main steam drums, and HRSG Stack Temp (F) 358 335 325 the LP drum was set to operate at 25 psia (172 kpa). An isentropic Steam Turbine Power (MW) 40.6 37.8 9.0 efficiency of 85% was assumed for each section of the steam turbine, Net Power (MW) 133.5 109.6 30.9 the pump efficiencies were assumed to be 80%, the generator Heat Rate (LHV, Btu/kW-hr) 7350 7390 10,760 efficiencies were assumed to be 98%, and the balance-of-plant (BOP) Cycle Efficiency (LHV, %) 46.4 46.2 31.7 power consumption was assumed to be 1.5% of the shaft power production. Design control parameters were then selected for each section of the HRSG. For the HP and IP evaporators, a pinch-point temperature Table 3c: Summary of Results for Sample GATE/CYCLE difference of 30 F (16.7 K) was used to determine the design Runs at 120 F. operating point, and a fixed steam generation rate of 39,000 lbm/hr PARAMETER MAX 24.7 40% (4.91 kg/s) was specified for the LP evaporator that would exactly POWER IGV POWER match the amount of steam required for complete deaeration of the boiler feedwater. The design parameters for both superheaters were GT Inlet Air (Ibm/s) 530 438 441 set so that the exit steam temperature was 30 F (16.7 K) less than the Gas Turbine Power (MW) 60.1 45.3 30.3 inlet gas temperature, and the economizers were designed so that the GT Efficiency (LHV, %) 29.8 27.1 24.5 exit boiler feedwater was 20 'F (11.1 K) below the boiling GT Exhaust Temp (F). 1041 1080 873 temperature. HP Steam (Ibm/hr) 200,200 180,700 110,400 IP Steam (Ibm/hr) 46,500 36,900 44,200 After the design parameters were input for the selected steam LP Steam (Ibm/hr) 35,000 30,600 32,900 cycle configuration, the GATE/CYCLE analysis code was executed. HRSG Stack Temp (F) 330 313 319 The results from this run are presented in the first column of Table 3. Steam Turbine Power (MW) 37.4 34.0 21.6 The calculated net combined-cycle power production was 112.3 Net Power (MW) 94.9 77.1 50.5 MW, with the plant operating at an efficiency of 46.8% (LHV) or a Heat Rate (LHV, Btu/kW-hr) 7230 7410 8370 heat rate of 7280 Btu/kW-hr. As mentioned above, the GE 7E gas Cycle Efficiency (LHV, %) 47.2 46.1 40.8 turbine produced 76.7 MW at these conditions: the steam turbine produced 38.4 MW. Also shown in Table 3 are the steam productions levels in the evaporators and the gas turbine and HRSG exhaust gas temperatures. Although not shown in Table 3, the

0 CYCLE code also calculated all of the state points for the cycle and a number of key design parameters for each of the heat exchanger sections, such as effectiveness and heat transfer surface area. With the design-point calculations for the gas-turbine combinedcycle plant completed, it was a simple matter of changing a few key parameters in the data entry forms for the major components to predict the performance of the combined-cycle plant under varying loads and ambient temperatures. The gas-turbine compressor mass flows calculated by GATE for three temperature levels - 0, 59 and 120 F (255, 288 and 322 K) - are shown in Figure 3. The predicted off-design values for inlet air flow and turbine power production are within 2% of the performance data supplied by the manufacturer. Additional results from these GATE runs are shown in Tables 3a-3c. Three GATE runs were executed to produce the curve at 59 F (288 K) in Figure 3. The first run was the design-point case, which is the highest power point on the 59 'F (288 K)curve. For the offdesign runs, it was assumed that the gas turbine power was reduced using inlet guide vane (IGV) modulation and exhaust-gas temperature control: this simulates the functioning of a GE SPEEDTRONIC control system. With GATE, the user must manually input the IGV position; with the GE control system, however, the IGV positioning is automatic. To reduce power, the inlet guide vanes are closed to reduce the compressor inlet air flow; the fuel flow to the engine is then adjusted to maintain the exhaust temperatures at the level specified by the trim curve of exhaust temperature vs. compressor discharge pressure shown in Table 4. The second point at 59 F (288 K) was an off-design GATE run with the inlet guide vanes at 24.7 deg (fully closed). This is the most restrictive setting of the guide vanes for this engine type. Table 4: Trim Curve Used in Off-Design Analysis of GE 7E Engine. COMPR DISCHARGE TURBINE EXHAUST PRESSURE (psia) TEMPERATURE (deg F) 15.0 1080. 130.6 1080. 150.5 1040. 191.5 955. With the compressor guide vanes fully closed, the above procedure can no longer be used to reduce power; instead, the inlet guide vanes are maintained at their fully closed position and the fuel flow (and thus the turbine inlet temperature) is reduced to lower the power production. With GATE, the user models this process by simply setting the appropriate off-design calculation flag and inputting the desired power level as a percent of full power. With these inputs, GATE automatically adjusts the turbine inlet temperature until the desired power level is reached. A 40% power level was specified for the final point on the 59 F (288 K) temperature curve in Figure 3. Similar GATE runs were made at 120 'F (322 K) and 0 F (255 K): the calculated compressor mass flows for these cases are also shown in Figure 3, and the overall performance results are presented in Tables 3a-3c. These two curves were generated by copying the three cases at 59 F (288 K) discussed above and changing a single parameter (ambient temperature) on the off-design input forms. The lowest power point on each curve therefore represents a gas turbine operating with the guide vanes set to 24.7 deg, and a forty percent 700 O o0f 650 O... * 59 F t^ a120f 600 0 I L 550 a 500 0...^ ld. 450 40 60 80 100 Generator Output (MW) Figure 3. Compressor Inlet Air Flow vs. Generator Output Power power setting. The second point on each curve was obtained with this same guide vane setting and with the control system trim curve (Table 4) used to determine the required fuel flow. Since the fullpower points on the off-design temperature curves, at 0 and 120 F (255 and 322 K), are not at design conditions, GATE was executed in the off-design mode for these two points using the trim curve and a 0 degree inlet guide vane setting. To predict the off-design performance of the steam bottoming cycle, the data sets for the 59 F (288 K) full-power design point were copied to off-design cases for each of the above GATE gas turbine runs The input parameters were then adjusted as needed for off-design calculations. The major change necessary was to set the parameter controlling the calculational mode for each of the HRSG sections to use the heat transfer surface areas calculated during the design-point run. Since the CYCLE code writes all calculated values into EASE+ data fields, the surface area for each heat exchanger section was already stored in the input forms and did not need to be entered again. For these off-design runs, it was assumed that the steam turbine and pump efficiencies and overall heat transfer coefficients did not change from their design values, although simple off-design correlations are incorporated into the CYCLE code. After changing the heat exchanger control parameters, the code was executed for each gas turbine case. The results for these nine runs are presented in Tables 3a, 3b and 3c for the three temperature levels. The points labelled "40% POWER" in these three curves represent the lowest power points with the reduced fuel flow and turbine inlet temperatures; the points labelled "24.7 IGV" are those points where the inlet guide vanes are set to 24.7 degrees and the trim curve is used to determine the fuel flow; and the points labelled "MAX POWER" represent the maximum power points at each temperature. These results show that the best cycle efficiency (47.2%, LHV) is at the full-load point at 120 F (322 K). At these conditions, the cycle performance is better than at the design point of 59 F (288 K) because, although the simple-cycle gas turbine efficiency has dropped from 32.0% (LHV) at the design point to 29.8% due to the ambient temperature change, the gas turbine exhaust temperature has increased from 992 F (806 K) at the design point to 1041 F (834 K), enabling the steam cycle to generate more power per pound of gas

0 turbine exhaust. When the ambient temperature is changed from 59 F (288 K) to 120 F (322 K), the gas turbine power drops nearly 22%, from 76.7 MW to 60.1 MW, but the steam cycle power production drops only about 3%, from 38.4 MW to 37.4 MW. The largest power production (133.5 MW) is at the full-power point at 0 F (255 K). The power increases at this lower temperature because, as compared with the design point, the gas turbine passes considerably more mass flow and therefore generates more power. Also, because the heat transfer surface area was fixed in the steam cycle system analysis, the larger gas turbine mass flow causes more steam to be generated in the evaporators, which increases the power produced in the steam turbine. The gas turbine efficiency for this case (33.5%) is actually higher than at the design point, but the overall cycle efficiency drops slightly, from 46.8% at the design point to 46.4% at 0 'F (255 K), due to the lower gas turbine exhaust temperature of 962 'F (790 K), compared with 992 'F (806 K) at the design point, and the corresponding decrease in the steam cycle efficiency. Because of the low gas turbine exhaust temperatures for the 40% power points at 59 F (288 K) and 0 F (255 K), small adjustments to the model were necessary. If the CYCLE model were run with no modifications to the input data, boiling would occur in both the HP and IP economizers since the steam production is so much lower than for the design case. It was therefore assumed that the steam cycle was configured to allow the boiler feedwater flow to bypass these economizers. To model this with the CYCLE code, the only change necessary was to set the flag in the economizer input form to indicate that the component was to be bypassed. In addition, for the 0 'F (255 K) case at 40% power the gas turbine exhaust temperature of 596 F (586 K) was so low that essentially no steam would be produced with the HRSG operating at the design pressures of 1500/500 psia (10.3/3.4 MPa). It was therefore assumed that sliding-pressure control would be used and the evaporator pressures were set to 450 and 150 psia 3.1/1.0 MPa). This involved setting the pressures in the two evaporator input data forms and in the corresponding pump input forms. These were the only changes necessary to model these extreme off-design performance runs with the GATE/CYCLE code. REFERENCES Bettner, J.L. and Sehra, A.K., 1986, High-Efficiency Axial Compressor, EPRI Report AP-4943. Cohn, A. and Waters, M., 1982, "The Effect of Alternative Turbine Cooling Schemes on the Performance of Utility Gas Turbine Power Plants", ASME paper 82-JPGC-GT-19. Erbes, M. R., 1986, Phased Construction of Integrated Coal Gasification Combined-Cycle Power Plants, Ph.D. Thesis, Stanford University, Stanford, California. Gay, R.R., 1988, The GATE Code: Thermodynamic Analysis Code for Gas-Turbine-Based Power Plants", presented at EPRI Conference on Heat Rate Improvement, May 10-12, Richmond, VA. Johnsen and Bullock, 1965, "Aerodynamic Design of Axial-Flow Compressors", NASA SP-36. Keenan, J.H. and Kaye, J., Gas Tables, 1948, John Wiley & Sons, New York. Keenan, J.H. and Keyes, F.G., 1969, Steam Tables, John Wiley & Sons, New York. Salisbury, J.K., 1950, Steam Turbines and Their Cycles, Robert E. Krieger Publishing Co., Huntington, N.Y. Spencer, R. C. et al., 1974, "A Method for Predicting the Performance of Steam Turbine-Generators", GER-2007C, General Electric Co. Waters, M., 1983, Gas Turbine Evaluation (GATE) Computer Program, EPRI report AP-2871-CCM. Yaws, C. L., 1976, "Correlation Constants for Chemical Compounds", Chemical Engineering, August 16, 1976, pp 79-87. CONCLUSIONS The GATE code is a powerful tool for the analysis of the designpoint and off-design performance of gas-turbine based power plants. It combines detailed analytical methods based on first-principle models with an intuitive, user-friendly graphical interface which enables GATE to be used by novices and experts alike. GATE can be used to model a broad range of gas turbines and associated power systems. The library of engine design-point models combined with the off-design analysis capability and the flexibility of the steamcycle analysis code make it possible for GATE users to rapidly and accurately predict the performance of gas turbine engines under a wide variety of design and operating conditions. Work is continuing on GATE under EPRI sponsorship. The planned scope of future work includes verifying and improving the aero-derivative gas turbine engine modeling capabilities of GATE and adding detailed modeling of the design and off-design performance of steam turbines.