Direct Off-design Performance Prediction of Industrial Gas Turbine Engine

Leonid MorozValentyn BarannikMaksym BurlakaAbdul Nassar

（1.SoftInWay Inc.,1500 District Ave,Burlington,MA,USA,Email:l.moroz@softinway.com；2.Soft In Way Turbomachinery Solutions Pvt.Ltd,Bangalore,KA,India）

Abstract：The modern gas turbine engine has been used in current power generation industry for almost half a century.Gas turbines are designed to operate with the best efficiency during normal operating conditions and at specific operating points.However,the real world is non-optimal and the engine may have to operate at off-design conditions due to load requirements,different ambient temperatures,fuel types,relative humidity and driven equipment speed.Also more and more base-load gas turbines have to work today on partial load too,which can affect the hot gas path condition and life expectancy.At these off-design conditions,gas turbine's efficiency and life deterioration rate might significantly deviate from the design specifications. During a gas turbine's life, power generation providers might need to perform several overhauls or upgrades for their engines.Thus,the off-design performance after the overhaul also might be changed.Prediction of gas turbine's off-design performance is essential to economical operation of power generation equipment.In this paper,an integrated system for complex design and off-design performance prediction(AxSTREAM?Platform)is presented.It allows to predict gas turbine engine's design and off-design performance almost automatically.Each component's performance such as turbine,compressor,combustor and entire secondary flow(cooling)system is directly and simultaneously calculated for every off-design performance request,making possible to build an off-design performance map including cooling system.The example of off-design performance estimation of industrial gas turbine engine is presented.The presented approach provides wide capabilities for optimization of operation modes of industrial gas turbine engines and other complex turbomachinery systems for every specific operation conditions(environment,grid demands and other factors).

Keywords：Virtual Gas Turbine Unit,Gas Turbine Unit,Off-design Mode,Power Control,Cooling System Control,Efficiency Improvement

Nomenclature

Symbols:

GTUgas turbine unit

VGTUvirtual gas turbine unit

MFRmass flow rate

AEFair excess factor

LHVlower heating value

Gmass flow rate

ppressure

Ttemperature

effefficiency

Npower

Dshaft diameter(for pump work calculation)

Cuflow tangential velocity(for pump work calculation)

Indexes

in inlet parameters

out outlet parameters

extr extraction

init initial

turb turbine

comp compressor

comb combustor

sp specified

i number of induction of cooling MFR to the turbine

j number of extraction in the compressor flow path

curr value on current iteration

prev values on the previous iteration

1 Introduction

Gas turbines are widely used all over the world.The same GTU frame could be installed in arctic or desert regions providing significantly different environmental conditions.But even little change of boundary conditions causes a significant influence on integral characteristics and reliability of the engine.It is well known,that ambient temperature elevation leads to unit efficiency and power deterioration and vice versa when the turbine inlet temperature is fixed[1-2].

There are two main reasons of GTU off-design operation:

·Environment-induced off-design

·Grid demands or driven device induced off-design

Environment-induced off-design is not desired in terms of driven equipment.It is usually a subject to mitigation utilizing the approaches that help to save fixed GTU power with ambient temperature rise[3]:

·Chiller application at compressor inlet;

·Water evaporation at compressor inlet;

·Humid air/steam injection to combustor chamber.

The off-design induced by grid demands or driven device is a requirement which could be achieved in several ways[4]:

·Turbine inlet temperature reduction.The reduction of temperature is carried out by the decline of injected fuel MFR in the combustor.The compressor inlet MFR,in this case,is almost unaltered and equal to the MFR at design mode;

·Compressor inlet air mass flow rate reduction.This approach is carried out reducing single inlet guide vane(IGV)angle or IGV together with several guide vanes;

·Combination of turbine inlet temperature decreasing and compressor inlet air MFR methods.The combined method is used when the power control by inlet compressor area does not allow needed power decreasing;

·Throttling of air flow at compressor inlet;

·Bleeding of air flow downstream the compressor;

·Bypassing of compressed air from the last stages of the compressor to first stages.

Despite the reason caused GTU off-design operation,the performance deterioration has to be estimated.One of the crucial aspects in off-design performance estimation is a determination of joint operation point of turbine and compressor to check if GTU would reliably operate avoiding any excessive temperatures of turbine blades and surge zones of the compressor and produce a certain amount of power.

The overall approach to search joint operation point is a utilization of the turbine and compressor maps(graphical method)[5-6].Maps utilization method is pretty simple and in combination with thermodynamic simulation allow calculating of GTU performance in the shortest time.However,the cooling system presence contributes the necessity of determining of additional factors(parameters),which leads to some simplifications and,as a consequence,may lead to inaccuracies,when maps are used.Some of these simplifications are determination of cooling air mass flow rate(MFR)as a percent from compressor inlet air MFR and simplified GTU components geometry consideration.

There are many papers devoted to an accurate off-design performance calculation without maps utilization[7-8].However,their utilization is time-consuming.

Therefore the methods that allow accounting for advantages of iteration maps method(relatively short time of calculation)with the 1D calculation of compressor,turbine and cooling system are of interest.The automation of off-design GTU parameters search process makes possible excluding errors related with transferring of large amounts of data for multiple variables.

Simulation of cooled GTU requires a utilization of various 0D,1D,2D and 3D models for calculation of GTU components(compressor,turbine,combustor,cooling system etc.),presence of efficient data transfer between the models and ability to incorporate custom models,and use logical operations and perform optimizations.The complex of the mentioned tools,methods,models,scripts connected in logical sequence is essentially a Virtual Gas Turbine Unit(VGTU).The development and validation of such a VGTU is presented in this paper.

2 The Description of GTU Prototype for VGTU

The 166 MW single shaft power generation stationary GTU was selected as a prototype,the VGTU will be developed for.

In the scope of this study,it was decided to limit the number of components and systems to be included in VGTU to compressor,combustor,turbine and cooling system(Figure 1).The thermo-structural analysis was not considered in this paper.However,it is planned to include it and expand the number of considered systems in future studies.

The detailed description of the prototype unit is given below.

Fig.1 Gas turbine unit scheme

2.1 GTU Scheme Description

In Table 1,the main characteristics of the investigated gas turbine at design mode are given.

Tab.1 GTU design characteristics

The compressor is 17 stages machine with IGV and three extractions to cooling system:after nozzle of the 11thstage,after rotor of 16thstage and at compressor outlet.The power of compressor at design mode is 137 MW.

Compressor flow path was designed and calculated with task formulation of finding inlet MFR for given outlet pressure utilizing streamline solver[9].As a result of compressor calculation the outlet MFR,outlet total temperature of the compressed air and boundary conditions at extractions were determined.

Fuel MFR was determined by thermodynamic calculation of combustor.Based on energy balance equations(1-4)the fuel MFR was received.The combustor outlet temperature is equal to 1550 K.Methane was used as a fuel.

where

Iturb_inturbine inlet enthalpy;

pturb_inturbine inlet pressure;

Tturb_inturbine inlet temperature;

AEFair excess factor.

where

Gfuelfuel MFR;

Gcomb_outcombustor outlet MFR;

Gcomp_outcompressor outlet MFR;

Icomp_outcompressor outlet enthalpy;

LHVlower heating value.

I0stoichiometric air-fuel ratio.

The combustor outlet MFR in the first iteration was set as

where

The initial guess for air excess factor was arbitrarily selected.The pressure drop at combustor was equal 0.05 for all calculations.

Turbine is a three-stages axial machine with 14 cooling inductions and cooled duct at the turbine inlet.The turbine flow path was created and calculated utilizing streamline solver with the same task formulation as a compressor.

The expansion process of the cooled turbine in the h-s diagram is presented in Figure 2.

Fig.2 Expansion process of turbine in h-s diagram

The dotted lines show the process without cooling.The solid lines represent a real expansion process.Blue lines represent expansion processes in stationary nozzles.Red lines represent expansion processes in rotating blades.

2.2 Cooling System Description

As was mentioned above the cooling system is presented by three extractions in compressor part,which then divided on 14 cooling flows inducted to turbine flow path.

Schematic ally the cooling system is presented in Figure3.

The first extraction(after 11 stages)separates on three flows that are inducted to the second rotor internal,third stator and third rotor internal endwalls,Equivalent hydraulic network is presented in Figure 4.

Second extraction(after rotor of 16thstage)is separated on two cooling flows that cool the rotor blades of the first and second stages and inducted to turbine flow path through holes at the blade tip.

Finally,the cooling extraction at compressor outlet is separated on 9 flows:external and internal endwalls of the duct;leading,trailing edges,internal and external endwalls of the first stage stator;leading edge and internal endwall of the second stage stator and internal endwall of the first stage rotor.

Fig.3 Cooling system scheme

Fig.4 Cooling system equivalent hydraulic network

3 The Description of the VGTU

In this study VGTU is used to simulate the off-design operation of the considered GTU porotype.Turbine inlet temperature was kept constant for each off-design mode.

The scheme of calculation process is presented in Figure 5.

Green circle block is self-explanatory.Orange rectangles represent custom scripts.The scripts were added in order to perform additional calculations not available among off-the-shelf tools.Yellow blocks represent some available computational tool.In particular,for the compressor,it was streamline solver,as well as for turbine.For cooling system,it was hydraulic network 1D numerical solver.Pink diamonds represent conditional statements for process control according to a predefined condition.These blocks allow implementation of loops required to converge required parameter,for example,MFR for each cooling flow.The conditional block could be also used for implementation of alternative paths of the calculation process.Red octahedron is self-explanatory.

The top diamond on the diagram represents condition of equality of cooling flows MFRs.If balances of cooling flows mass flow rates extracted from compressor and inducted to turbine do not correspond,the reassignment of cooling flows MFRs is performed.The equality of turbine inlet MFR and combustor outlet MFR is done by bottom diamond.In this case,the difference in mentioned above MFRs leads to reassigning of turbine inlet pressure.

All described above calculations were performed automatically.Thus,eventually,the capability to simulate off-design performance for any compressor inlet BCs as in real test facility was achieved.

Input/output parameters for all blocks are presented in Table 2.

Fig.5 VGTU operation flowchart

Tab.2 Inputs/Outputs

Indexesifor given GTU is varied from 1 to 14(number of cooling flows),andjfrom 1 to 3(number of extractions).The data transfer for all presented in the table parameters for all cooling flows was automated.

4 Validation of VGTU

The validation of the developed VGTU was performed comparing the estimated performance data with real GTU test data[10].The comparison was performed for the off-design caused by variation of ambient temperature from+5℃to+30℃.The ambient pressure and turbine inlet temperature for each off-design mode were kept constant and equal to the ones at design conditions.

The pressure losses in inlet and exhaust ducts were not taken into account.

The VGTU performance obtained for different values of ambient temperature is presented in Figure 6.The data in the Figure 6,Figure 7 and Figure 8 are presented in relative values that were determined as a ratio of parameter value at the current mode to a parameter value at the design mode.

As we can see from Figure 6 the degradation of unit efficiency is significant when ambient temperature increases and equals to 0.11%in relative values and vice versa respectively.

The similar trend is observed for GTU power.The dependency of GTU power vs ambient temperature is presented in Figure 7.

The data for power correction factor obtained utilizing the VGTU and the GTU test data from[10]are presented in Figure 8.Power correction factor is a ratio of power at design mode to power at off-design mode.It is clearly seen that the power correction factor obtained on VGTU is in good agreement with the power correction factor variation from[10].This allows concluding that VGTU performance data results are plausible and that VGTU can be used for GTU off design performance data gathering.

Fig.6 GTU efficiency vs ambient temperature

Fig.7 GTU power vs ambient temperature

Fig.8 Power correction factor vs ambient temperature

5 GTU Performance Augmentation by Cooling Air MFR Control

It is well known that GTU part-load control by turbine inlet temperature envisages reduction of the turbine inlet temperature to obtain part-load mode of GTU.Authors of this paper performed simulation of part-load mode of VGTU(these results are not presented here,but they will be published in ASME Turbo Expo 2018 paper)and determined that the tem-perature of hot gas can become even less than blades material allowable temperature,but cooling mass flow rate was almost the same as at design mode.In other words the cooling flow is simply wasted at deep part load modes.It is obvious to assess possibility of cooling mass flow rate control in order to adjust it according to the temperature of hot gas and improve efficiency of GTU at deep off-design modes.

The proposed VGTU was utilized to perform the assessment.For this,VGTU cooling system was modified by addition of control valve to the cooling channel going to first nozzle vane.First vane required the most significant amount of cooling.Thus the effect from reduction of cooling flow to first nozzle has to be pretty significant.

6 Determining the Quantity of Cooling Air MFR

There are dependencies in the literature sources that allow approximately calculate the required cooling air MFR for given type of cooling and temperature of main flow.

In the presented paper the cooling air MFR was determined from[11](Figure 9).

Fig.9 Efficiency of different types of air cooling

On the Figure 9:

1-blade with radial channels;

2-blade with semi-closed cooling system;

3-deflector blade;

4n-nozzle blade film cooling;

4r-rotor blade film cooling.

where

relative cooling air MFR

Gcoocooling air MFR;

GgasMFR of main flow.

where

T*gastotal temperature of main flow;

Tblmaximal allowable temperature of blade without cooling;

T*cooltotal temperature of cooling flow.

Since the blade cooling type for presented GTU is film,the necessary cooling MFR was determined by the“4n”line.

7 Results of Cooling Air MFR Adjustment

The effect of cooling air MFR control on VGTU performance for different power levels is presented in comparison with performance without any cooling system control.

The dependency of VGTU performance for different power level is presented in Figure 10.

Fig.10 Relative GTU efficiency vs GTU power level

As we can see from Figure 10 the efficiency of VGTU with controlled cooling system is higher than one of VGTU with uncontrolled cooling system.The efficiency increment at 40%of power is about 3.5%in relative values,at the 70%it is about 1%.It should be noticed that cooling air MFR was controlled for first nozzle only.It can be assumed that the cooling air MFR controlling at another flow path elements allows to get even higher efficiency increments.The VGTU efficiency curve bend between 60%and 50%part load modes was caused by full closure of the first nozzle cooling channel.The rest of flow path elements have no control in this study,but VGTU allows controlling the other cooling system branches either.

The distribution of cooling air MFR at different VGTU power levels is presented in Figure 11.

It is clearly seen from Figure 11 that the decrement of the cooling air MFR is much more significant for the case of controlled cooling system.The turbine inlet temperature is less than maximal allowable temperature of flow path metal in the range of 50%～60%of power.

Fig.11 Relative cooling MFR vs GTU power level

It should be noticed that presented methodology of cooling system calculation allows to define the flow directions inside the blade,which is possible at some deep part-load modes.This assessment was not performed in this study.It is planned to include such an analysis and perform cooling system control simulation at VGTU part-load modes taking into account possible penetration of hot gas inside the blade and its influence on blade temperature and adjust cooling flow rate accordingly.

The dependency of relative turbine inlet temperature from power level is presented on Figure 12.

Fig.12 Relative turbine outlet temperature vs GTU power level

As we can see the difference in turbine inlet temperature in the range 100%～60%is not significant for controlled and uncontrolled cooling system types.The difference in 3%in relative values is observed when the main flow temperature is less than maximal allowable metal temperature and the cooling air MFR for stationary elements is equal to zero.

8 Approximately Ecomomics Evaluation

Preliminary economics analysis was performed assuming that the GTU works at 40%part-load mode from 20%to 50%of time in a year.Thus,if the fuel price is equal to$0.17 per kg of the fuel with LHV equal 50 kJ/kg,the savings may be from$154,000 to$387,000 per year.

9 Conclusions

The developed VGTU allows simulating the behavior of real GTU including off-design and part-load modes,automatically calculating compressor,turbine and entire cooling system performance,and their matching.

The validation of proposed VGTU with the test data for the case of different ambient temperature values was done.The validation showed good agreement of the VGTU performance data with the real test data.

The assessment of possibility of cooling mass flow rate control and its influence on turbine performance was performed.The efficiency increment at 40%of power is about 3.5%in relative values,at the 70%it is about 1%.It should be noticed that cooling air MFR was controlled for first nozzle only.It can be assumed that the cooling air MFR controlling at another flow path elements allows getting even higher efficiency increments.

Preliminary economics analysis showed that the savings are in range from$154 000 to$387 000 per year.The final value of savings depends on part load mode power level and operation time percentage per year.

10 Off-the-Shelf Software Tools Utilized in the Study

AxSTREAM?turbomachinery design,analysis and optimization tool[12-13]was integrated into VGTU for simulation of compressor and turbine.

AxSTREAM NET?1D hydraulic networks analysis tool[14]was integrated into VGTU for cooling system simulation.

AxSTREAM ION?[15]system engineering infrastructure for the design of engineering systems was utilized for the development of VGTU,including operation flowchart design,integration of the off-the-shelf and custom software tools,and execution.

Acknowledgments

We wish to thank the many people from SoftInWay Inc.team who generously contributed their time and effort in the preparation of this work.The strength and utility of the material presented here are only as good as the inputs.Their insightful contributions are greatly appreciated.