Performance analysis of combined humidified gas turbine ...

Performance analysis of combined humidified gas turbinepower generation and multi-effect thermal vapor compression

desalination systemsPart 2: The evaporative gas turbine based system and some discussions

Yongqing Wanga*, Noam Liorb

aCollege of Mechanical Engineering, Jimei University, Xiamen, 361021, PR ChinaTel. +86 (137) 7992 6575; Fax: +86 (59) 2618 3503; email: [email protected]

bDepartment of Mechanical Engineering and Applied Mechanics, University of Pennsylvania,Philadelphia, PA 19104-6315, USA

Received 4 February 2006; Accepted 16 June 2006

Abstract

This is Part 2 of the paper “Performance analysis of combined humidified gas turbine power generation and multi-effect thermal vapor compression desalination systems — Part 1: The desalination unit and its combination with asteam-injected gas turbine power system”. A combined power and water system based on the evaporative gas turbine(EvGT) is studied, and major features such as the fuel saving, power-to-water ratio, energy and exergy utilization,and approaches to performance improvement, are presented and discussed in comparison with STIG- and EvGT-based systems, to further reveal the characteristics of these two types of combined systems. Some of the main resultsof the paper are: the fuel consumption of water production in STIG-based combined system is, based on reference-cycle method, about 45% of a water-only unit, and that in an EvGT-based system, it is 31–54%; compared with theindividual power-only and water-only units, the fuel savings of the two combined systems are 12%–28% and10%–21%, respectively; a water production gain of more than 15% can be obtained by using a direct-contact gas-saline water heat exchanger to recover the stack heat; and the combined system are more flexible in its power-to-water ratio than currently used dual-purpose systems. Further studies on aspects such as operation, hardware cost,control complexity, and environmental impact, are needed to determine which configuration is more favorable inpractice.

Keywords: Integrated power and water production system; Water desalination; Humidified gas turbine cycles;Multi-effect thermal vapor compression desalination; Energy and exergy analysis

*Corresponding author.

Desalination 207 (2007) 243–256

0011-9164/07/$– See front matter © 2007 Published by Elsevier B.V.doi:10.1016/j.desal.2006.06.013

http://dx.doi.org/10.1016/j.desal.2006.06.013

Y. Wang, N. Lior / Desalination 207 (2007) 243–256244

1. Introduction

The objective of this two-part paper is to studythe energy, exergy, and water production perfor-mance of integrated power and water desalinationsystems that employ humidified gas turbines(HGT), of which the steam-injected gas turbine(STIG) and humid air turbine (HAT) or evapo-rative gas turbine (EvGT) systems are the mostrepresentative. Following the analysis of thermaldesalination unit and STIG-based combinedsystem in Part 1 [1], this is Part 2 of the paper,focusing on EvGT-based combined system andperformance discussion and comparison of thetwo combined systems. The calculation con-ditions and assumptions, as well as the perfor-mance criteria described in Part 1 are the same inthis part 2.

2. EvGT-based power and water system

2.1. System configuration of EvGT-based system

In an EvGT-based system, as shown in Fig. 1,part of the compression-generated heat of thecompressed air is recovered in heat exchangersAC1 and AC2 to heat water for air humidi-fication. The humidified air is then heated by theturbine exhaust in a high temperature regenerator(R). The exhaust gas from the regenerator is usedto produce steam in a heat recovery steamgenerator (HRSG) for desalination. A systemconfiguration that directly uses the exhaust gas toheat saline water was not employed, to avoidpotential corrosion and scaling problems causedby the higher temperature saline water that wouldbe generated.

As discussed in part 1 of this paper and furtherelaborated in this paper, the EvGT cycle promisessomewhat higher efficiency than the STIG cyclein a certain range of operating conditions, herefor pressure ratios somewhat below 20, but atwhich it produces less water for the same powerproduction when operated as a dual-purpose

system. At higher pressure ratios its efficiency issomewhat lower, but the water production higher.Beyond efficiencies and power-water productionrates, the practical decision on system preferencewould be based to important extent on the systemcost and complexity, and the EvGT requires(Fig. 1) 5 more components than the STIG: twoair-water aftercoolers (AC and AC2), an air-waterhumidifier (H), a water pump, and a regenerator(R). While we have not made a cost comparison,it is rather obvious that the EvGT system wouldbe somewhat more complex and expensive thanthe STIG.

In a conventional EvGT cycle, both the com-pressed air stream and the high-temperatureregenerator exhaust stream are used to heat thehumidification water and provide energy forhumidification, while in this study, the latter isused to produce saturated steam as a heat sourceand vapor compression driver for multi-effectthermal vapor compression (METVC) desalina-tion. Obviously, compared with a conventionalEvGT cycle, the power output W, the energyefficiency of power generation ηt (W as a fractionof the fuel energy Qf consumed in the combustor),and the exergy efficiency of power generation ge(W as a fraction of fuel exergy Ef), of the EvGTcycle in the combined system will decrease, as anexpense of water production.

In an EvGT-based combined system, themaximal pressure and thereby the maximumtemperature of the steam produced for desali-nation is up to the state of the exhaust gas and thedesign of the HRSG. Steam of different pressures,lower than the maximum value, can be obtainedfrom the HRSG in an EvGT-based system, whichis unlike the STIG-based system where the pres-sure of the motive steam is the same as that of thesteam injected into the combustor. Our calcu-lations show that the highest pressure can help geta maximal water production. For a specifiedEvGT cycle, the parameters of the gas turbineexhaust gas are fixed. Although the heat energyrecovered in the HRSG by the water/steam with

Y. Wang, N. Lior / Desalination 207 (2007) 243–256 245

Fig. 1. EvGT-based power and water combined system. AC1, AC2, aftercoolers; C, compressor; CC, combustor; FC, fuelcompressor; G, generator; H, humidifier; HRSG, heat recovery steam generator; P, pump; R, regenerator; T, turbine; TDC,thermal desalination unit.

different pressure pm is the same, a higher pmyields a more exergy-efficient heat transferprocess. As shown in Fig. 2, for instance, thetemperature difference between the gas and thewater/steam is obviously smaller for pm = 1.5MPa than for pm = 0.5 MPa, resulting in higherexergy recovery by the motive steam. Thisimproved exergy utilization in the HRSG isenough to make up the increased exergyconsumption in the METVC unit under high pm(Fig. 5 in Part 1), thus resulting in an increasedwater production. The following analysis of theEvGT-based system is based on the maximal pm(Tm) allowed in the HRSG. Calculations showthat these maximum pm are higher than 0.3 MPa

Fig. 2. T-Q diagram in the HRSG in the EvGT-basedcombined system for different pm.


within the parameters range studied. Based on theperformance analysis on METVC and MEE unitsperformed in Part 1, a 6-effect METVC unit isalso chosen for the EvGT-based system, as thatfor the STIG-based system.

The pressure ratio β of the compressor, theturbine inlet temperature (the “firing” tempera-ture) TIT, and the humidification rate xh are themost important parameters influencing the per-formance of the EvGT cycle, and of the combinedsystem.

2.2. The influence of humidification rate xh

Just as the steam injection rate xj in the STIG-based combined system is a key parameter whichdetermines the energy distribution between powergeneration and water production, the humidi-fication rate xh, which is the ratio of the watermass evaporated in the humidifier and the airmass through the compressor,

(1)h h ax m m=

is such a key parameter in the EvGT-basedsystems.

Fig. 3 shows the influence of xh on the powerand water production, and Fig. 4 shows its influ-

Fig. 3. Normalized power and water production of theEvGT-based system as a function of xh.

ence on the energy and exergy utilization of theEvGT-based system. Clearly, increasing xh causesthe power output to increase, and the waterproduction to decrease. For a specified pressureratio β, the temperature of the compressed airfrom the compressor is fixed. More heat will beused to evaporate water in the humidifier if ahigher xh is wanted, leading to a lower tempera-ture of the moist air at the outlet of the humi-difier, more energy recovery in the regenerator,and less energy available for desalination (Fig. 4),and the result is higher production of power butlower of fresh water (Fig. 3).

2.3. The influence of pressure ratio β and firingtemperature TIT of the cycle

Fig. 5 shows the exergy efficiency ge andthermal efficiency ηt of power generation, as wellas the exergy recovery rate for desalination ξe,D,which is the exergy recovered by the motivesteam as a fraction of fuel exergy Ef, and energyrecovery rate for desalination ξt,D, which is theenergy recovered by the motive steam as a

Fig. 4. Exergy and energy utilization of the EvGT-basedcombined system for different xh.


Fig. 5. Influence of β and TIT on ge, ηt ,ξe,D and ξt,D in the EvGT-based combined system. (a) Variations of ge and ξe,D withβ and TIT. (b) Variations of ηt and ξt,D with β and TIT.

Fig. 6. Normalized power and water production of theEvGT-based combined system.

fraction of fuel energy Qf, for different β (from 10to 30), TIT (1100oC and 1300oC) and xh (0.05 and0.1) of an EvGT-based system. We can see thepercentage of fuel energy/exergy converted intopower and consumed by desalination. Althoughξe,D and ξt,D only represent the thermal energy/exergy provided for desalination, they will deter-

mine water production, as shown below, sincethat pumping work is only a small fraction of thedesalination energy consumption. Fig. 6 showsthe normalized power and water pro-duction, andFig. 7 the exergy and energy utiliza-tion fordifferent β.

Figs. 5 and 6 reveal that ge, ηt have a similartrend as the power output w, consistent with thedefinition of ge and ηt; so do ξe,D, ξt,D and waterproduction mw. This behavior is the same as in theSTIG-based system.

Different from the STIG-based system, ahigher β lowers ηt, ge and w, as known in [2–4],while it raises ξt,D, ξe,D and mw in the EvGT-basedsystems. A higher β results in a higher air tem-perature at the compressor outlet, and then, for aspecified xh, a higher humid air temperature at thehumidifier outlet, leaving less heat and exergy forrecovery by the humid air in the regenerator, andmore available for desalination (Fig. 7), and theresult is lower power output and higher waterproduction.

A higher TIT is beneficial to the power gene-ration of the EvGT-based system, as determinedby thermodynamic principles [2–4]. Different


Fig. 7. Exergy and energy utilization of the EvGT-based combined system for different β.

from the STIG-based system in which a higherTIT also leads to a distinct increase in waterproduction 1 kg/s fuel consumed, in the EvGT-based system the TIT does not have such anobvious positive influence on mw, and even con-trarily, for higher β the mw of the EvGT-basedsystem with a higher TIT may equal or even belower than that with a lower TIT, as shown inFig. 6. This means that in the EvGT-basedsystem, for 1 kg/s fuel consumed, increase of theTIT will indeed increase the power output, butthe water production may increase a little, remainthe same, or even decrease. One can see fromFig. 8 that shows the exergy and energy utili-zation of an EvGT-based system for β = 25, xh =0.05 and TIT = 1300EC and 1100EC, that morefuel energy, and correspondingly almost the sameamount of fuel exergy, is recovered for desa-lination when TIT = 1100EC than at TIT =1300EC. This can also be seen clearly fromFig. 5. The reason is that when changing TITfrom 1300EC to 1100EC, the air mass that 1 kg/sfuel can heat in the combustor will, obviously,

increase, from 45.7 kg/s to 55.0 kg/s in thesample cases, leading to an increased mass flowof exhaust gas, from 49.0 kg/s to 58.8 kg/s,through the HRSG. The energy and exergy thatcan be recovered for desalination in the EvGT-based system depend mainly on the pressure,temperature and mass flow of the exhaust gasentering the HRSG, with the latter two being themain factors (since the pressure of the exhaustgas in the EvGT-based system is always close tothat of the surroundings). In the sample cases,based on 1 kg/s fuel consumed, the mass andtemperature of the exhaust gas are 49.0 kg/s and148EC respectively for TIT = 1300EC, and58.8 kg/s and 117EC for TIT = 1100EC. For TIT= 1100EC, the decreased exhaust temperaturecauses decreased specific enthalpy and exergy,while the increased exhaust mass raises the totalamount of energy/exergy that could be recovered,resulting in a higher ξt,D and an almost equal ξe,Dcompared with that when TIT = 1300EC. It is theexergy, not the energy, that determines the waterproduction, so almost the same amount of water


Fig. 8. Exergy and energy utilization of the EvGT-based combined system under different TIT.

is obtained in the two cases (Fig. 6). It is note-worthy that in both STIG- and EvGT-basedsystems, changing β, TIT or xj/xh for 1 kg/s fuelconsumed, change both the temperature and massof the exhaust gas for producing steam fordesalination, but at all the conditions except theabove-discussed ones, the temperature, not themass flow of the exhaust gas, is the key factorinfluencing the amount of the energy/exergyrecovery. Only the influence of the temperature istherefore mentioned in all the other conditionsexcept the above-discussed ones.

2.4. Energy recovery from the stack gas

A direct-contact gas-saline water heatexchanger is also considered in the EvGT-basedsystem for recovering the stack gas energy. Partof the saline water from the end condenser of theMETVC unit is fed to the direct-contact heatexchanger, heated to 63EC by the stack gas, andthen, after mixing with the saline water from theend condenser according to the mass flow andtemperature required by each effect, used as the

Fig. 9. Net water production of EvGT-based combinedsystem with and without stack heat recovery.

feed of a 6-effect METVC unit, in which nopreheaters are used because the feed saline waterhas been preheated by the stack gas. The netwater production with and without stack heatrecovery are both shown in Fig. 9, and the resultagain indicates a distinct positive effect of stackheat recovery.


3. Discussions on STIG- and EvGT-basedcombined systems

3.1. Energy saving of the STIG- and EvGT-basedcombined systems

Fig. 10 shows the optimal thermal efficienciesηt,opt and the corresponding steam injection rate xjand humidification rate xh of the conventionalSTIG cycle and EvGT cycle, for TIT = 1300ECfor different pressure ratios β. To get the optimalefficiency, all the steam produced in the HRSG inthe STIG cycle is injected into the combustor atthe highest temperature allowed by the workingcondition of the HRSG, and all the parameters inthe EvGT cycle are optimized by taking thermalefficiency as the objective function. We can seefrom the figure that, within the parameter rangestudied, ηt,opt of the STIG cycle increases sub-stantially with β, while that of the EvGT cycle isnot so sensitive to β. The STIG cycle has higherinjection rate, from 0.195 to 0.306, comparedwith 0.16 to 0.22 for the EvGT cycle. This paperwill focus on the characteristics of the combinedsystems, and the readers are referred to references[2–4] to get detailed performance analysis andcomparison on the two power-only cycles.

As mentioned above, water production in bothSTIG- and EvGT-based systems is at the expenseof work-production efficiency. When β =30 andTIT = 1300EC, for instance, ηt,opt of the power-only STIG cycle is 52.8% with a xj of 0.195[which means that there is no water production,and actually net water consumption of 7 kg/(kgfuel)]. Reducing xj to 0.05, and using the extraenergy to produce steam for the METVC unit,reduces ηt to 44.85%, with a net water productionof 74.7 kg/(kg fuel).

It is noteworthy that the value and the trend ofηt of the STIG cycle and the EvGT cycle inFig. 10 are different from those in Fig. 13 in Part1 and Fig. 5 in this part, because the former isbased on the optimized xj or xh which can help getan optimal efficiency, while the latter is based ona specified xj, say 0.05 or 0.1.

Fig. 10. Performance of conventional power-only STIGand EvGT cycles.

The fuel energy allocation between power andwater in a dual-purpose system can be made inseveral ways [5,6], including the lost workmethod, exergy method, reference cycle method,etc. The reference cycle method [5] is used in thispaper to calculate the energy consumption ofwater production, but differently: instead of tak-ing a specified efficiency as the reference, theenergy allocation here is based on the optimalefficiency of a power-only STIG or EvGT cyclehaving the same pressure ratio and firing tem-perature as the corresponding combined power/water production system. In an optimal power-only cycle, mf kg fuel is needed to produce thework W:

(2),( )f t optm W q= ⋅η

where q is low heat value of fuel. Since part ofthe fuel energy is used to produce water in thecombined systems, which leads to a decreasedefficiency ηt, more fuel, mNf, is needed to generatethe same amount of work:

(3)( )f tm W q′ = ⋅η

The extra fuel needed can be considered as thefuel consumption of water production:


Fig. 11. Normalized fuel consumption of water pro-duction based on reference cycle method.

(4),1 ( / )f f f t t optm m m′ ′ ⎡ ⎤− = − η η⎣ ⎦

Based on this fuel allocation method, the fuelexergy consumption per kg produced fresh water,not including the desalination pumping work, iscalculated and shown in Fig. 11. These values arenormalized by the specific fuel consumption mf,0,6.059×10!3 kg/(kg distillate), of a 6-effectMETVC unit run by 2.5 MPa saturated steamfrom a boiler with an efficiency of 0.9. For theSTIG-based system, the fuel consumption isabout 45% of that of the water-only unit, and forthe EvGT-based system, it is 31%–54%, when xjand xh change from 0.05 to 0.1.

The energy saving of combined systems canalso be illustrated from another angle. Comparethe dual purpose system to two separate units,one producing just power at optimal efficiency(Fig. 10), and one METVC unit run by 2.5 MPasaturated steam from a boiler producing justwater, producing the same amount of net powerand fresh water as the dual-purpose system at thesame pressure ratio and firing temperature. Theresults are shown in Fig. 12. Thanks to the com-bination, the fuel saving of the STIG-basedsystem is 19.8%–27.8% and 11.9%–21.1% when

Fig. 12. Fuel saved compared with single-purpose powerand water production units.

xj = 0.05 and 0.1 respectively, and that of theEvGT-based system is 19.5%–20.8% and 10.3%–11.8% when xh = 0.05 and 0.1 respectively.

The results of the above two methods indicatesubstantial improvements in fuel utilization ofSTIG- and EvGT-based combined systems,although with different rate and trend.

3.2. Power-water ratio of STIG- and EvGT-basedcombined systems

Fig. 13 shows the power and water ratio Rpw ofthe two combined systems. Within the parameterrange studied, the trend of Rpw of the combinedsystems is the same as that of the ηt or ge of thetwo cycles. Higher β and higher xj are beneficialfor STIG cycle to get a higher ηt and ge (Figs. 11and 13 in Part 1), and then tend to a higher Rpw(Fig. 13), and to EvGT cycle, lower β and higherxh are helpful to an increased ηt, ge and Rpw(Figs. 3, 5 and 13).

Compared with the typical designed dual-purpose system [7], including steam turbine-MSFplant with a Rpw of 4–19, gas turbine-MSF plantwith a Rpw of 6–13, and combined cycle-MSFplant with a Rpw of 9–18, STIG-based and EvGT-


Fig. 13. Power and water ratio of combined power andwater systems.

based system are seen to have higher Rpw , 12–44and 20–56 respectively, when xj and xh changefrom 0.05 to 0.1, mainly because of the higherefficiency of the base cycles.

One advantage of the combined systemsstudied in this paper is the great design andoperational flexibility of water and power pro-duction, because Rpw can be regulated in a widerange by changing xj or xh (Fig. 13). For instance,for a STIG-based system with β of 20 and TIT of1300EC, varying xj from 0 to 0.1, Rpw changesfrom 11.6 to 30.7.

3.3. Energy and exergy utilization of STIG- andEvGT-based combined systems

To make the exergy and energy utilization ofSTIG- and EvGT-based combined systems andways to improve them clearer, the exergy andenergy flow diagrams of the two systems for β =20, TIT = 1300EC, and xj or xh = 0.1 are shown inFigs. 14–17.

As well known, exergy and energy analysesgive different results. For instance, in the STIG-based system, the exergy loss rate, χe, in thecombustor is 29.5%, ranking as the highest

among the exergy losses of the components,while the energy loss rate, χt, is only 1%; χe in thestack is 6.5%, while χt is 30.3%, ranking as thehighest among the energy losses of thecomponents; χe of the compressors, turbine andHRSG are 1.9%, 3.1% and 6.9% respectively,while the χt magnitudes are negligible. Detailedexergy and energy utilization as well as perfor-mance improvement approaches, such as inter-cooling, reheating, recuperation, etc, of STIG andEvGT cycles can be found in many references,and a summary can be found in [8]. Here wefocus only on the interface of the power andwater production subsystems.

In a STIG-based system, the exergy loss in theHRSG is large, the second among the exergylosses of the components, because of the largeheat-transfer temperature difference between theexhaust gas and water/steam, especially in thewater evaporation process (Fig. 2). Comingdirectly from the gas turbine, the temperature ofthe exhaust gas used to produce steam in theHRSG is usually high, 630EC in the case shownin Figs. 14 and 15, while the evaporationtemperature of water is much lower, about 224EC(2.5 MPa) in the said case, resulting in a largeheat-transfer irreversibility, and thus a largeexergy loss. Obviously, increasing the steampressure and correspondingly the evaporationtemperature can help reduce the exergy loss(Fig. 2), at the expense of an increase of the heat-transfer area owing to the reduced heat-transfertemperature difference. In a conventional STIGsystem, however, no benefit can be obtained bythe production of higher pressure steam, becausethe steam must be throttled to a pressure close tothat of the combustor before its injection into thecombustor, and thus that extra exergy gained bythe steam in the HRSG would be uselessly lost inthe throttling process. Moreover, such high pres-sure steam may cause unstable operation of thesteam jet ejector. Even if the operation would bestable, the gain in water production would not besignificant because the performance ratio, PR, of


Fig. 14. Exergy flow diagram of the STIG-METVC system.

Fig. 15. Energy flow diagram of the STIG-METVC system.

the METVC unit has a diminishing rate of anincrease as pm is raised (Fig. 5 in Part 1). Forinstance, for the 6-effect METVC unit used inthis paper, PR increases from 8.9 to 9.2 when pmis raised from 1.5 MPa to 2.5 MPa, but only from9.2 to 9.3 when pm is raised further from 2.5 MPato 3.5 MPa. Obviously, using steam at a pressureof, say, 10 MPa, would thus not raise the waterproduction rate much when compared with that at3.5 MPa.

The exergy analysis results point to severalways for performance improvement of the STIG-based system by more efficient use of the highpressure steam that could be produced in theHRSG. For instance, (1) that steam can beexpanded in a steam turbine to the pressureneeded for supplying the heat for the desalinationand thus produce power in that intermediateprocess. From the exergy viewpoint, expandingthe steam to a suitable pressure, e.g. 0.034MPa, to


Fig. 16. Exergy flow diagram of the EvGT-METVC system.

Fig. 17. Energy flow diagram of the EvGT-METVC system.

run an MEE unit is more efficient than expandingit to a relatively high pressure, e.g. 2 MPa, to runa METVC unit, mainly because the steam jetejector is not an efficient compressor (ejectors arecommonly used in industrial processes anywaybecause they are simple devices without movingparts, and of low maintenance and cost), (2) thesteam can be expanded in a steam turbine to thelowest pressure allowed by the environmentalcoolant to produce more power, which can thenbe used for running a reverse osmosis (RO)

desalination unit, or a mechanical vapor com-pression one in which the thermal vapor com-pressor (i.e. the steam jet ejector in a METVCunit) is replaced by a mechanical compressor;mechanical compressors are much more efficientthan ejectors. Detailed study of these and otherimprovement options are under way.

In an EvGT-based system within the para-meters range studied in this paper, the pressure ofthe steam produced in the HRSG is lower than2 MPa when the minimum pinch point tempera-


ture difference of the HRSG ΔTp = 15EC becauseof the relatively low temperature (310EC in thecase shown in Figs. 16 and 17) of the exhaust gasfrom the regenerator in which part of its exergyhas been recovered by the moist air. Higher pmand then higher exergy utilization efficiency ofHRSG and higher water production could beobtained by further decreasing ΔTp throughincreasing the heat transfer area in the HRSG, butbecause that ΔTp is already very low, little roomis left for improvement. Although pm in an EvGT-based system is usually much lower than that ina STIG-based system, it is still much higher thanthe 0.034 MPa required for the MEE unit in thispaper, so the approaches proposed for improvingthe STIG-based system can also be used in theEvGT-based one.

It is noteworthy that in both combined systemsthe highest energy losses are in the stack, 30.3%and 33.1%, respectively, for the sample casesshown in Figs. 14–17, and the exergy losses aremuch lower, but still 6.5% and 7%. The largelatent heat of condensation of the water vapor inthe flue gas is one of the main reasons for thehigh energy/exergy losses. Three methods werediscussed in Part 1 to recover this low-level heat,and addition of a gas-saline water direct contactheat exchanger to preheat the feed saline water byusing the stack gas was found to provide adistinct gain in water production for the same fuelinput as that without stack heat recovery. Morepumping work will, however, be consumedbecause (1) more saline water and produced freshwater need to be pumped, (2) the pressure of partof the saline water needs to be raised to a valuehigher than that without stack heat recovery so asto make it possible to be sprayed into the exhaustgas in the heat exchanger. Further study is neededto evaluate the advantages and disadvantages ofthis modification.

This paper focuses only on the thermodynamicperformance of the STIG- and EvGT-basedcombined systems. Further study of other aspectssuch as operation, size, cost and environmental

impact, is necessary to determine which con-figuration is more favorable.

4. ConclusionsCombined power and water systems based on

the STIG cycle and on the EvGT cycle wereconfigured, modeled, and analyzed in detail inthis two-part paper. Both energy and exergyperformance criteria were employed. The mainresults are:

1. Great synergy, reflected in energy savings,is accomplished by combining the humidifiedpower cycle with the METVC desalination unit.The fuel consumption of water production in aSTIG-based combined system is, based on areference-cycle method, about 45% of a water-only unit, and that in an EvGT-based system it is31–54%. Compared with the individual power-only and water-only unit, the fuel savings of thetwo combined system are 11.9%–27.8% and10.3%–20.8%, respectively.

2. In a STIG-based system, higher pressureratios are advantageous to power generation, butdisadvantageous to water production, and theopposite is true for the EvGT-based system. Inboth systems, higher injection or humidificationrate can result in a higher power/water ratio.

3. The steam injection rate in the STIG-basedsystem and the humidification rate in the EvGT-based system have strong influence on the powerand water production of the two systems, pro-viding greater flexibility for design and operation.

4. A distinctly positive effect is that a waterproduction gain of more than 15% can be ob-tained by using a direct-contact saline water-gasheat exchanger to recover the stack heat.

5. Compared with the currently used typicaldual-purpose systems, these two combined sys-tems have higher power-to-water ratios becauseof the higher efficiency of the base cycles.

6. The exergy analysis gave guidance forsome potentially promising methods for furtherefficiency improvements.


5. Symbols

E — Exergy, kJm — Mass, kgpm — Pressure of motive steam, MPaPR — Performance ratio of thermal desali-

nationq — Low heat value of fuel, kJ/kgQ — Energy, kJRpw — Power-to-water ratio, MW/MIGDTm — Temperature of motive steamTIT — Turbine inlet temperature (or called

firing temperature), ECw — Specific work, kJ/kgW — Power output, kJxj — Steam injection rate in STIG cyclexh — Humidification rate in EvGT cycle

Greek

β — Pressure ratio of gas turbine cyclege — Exergy efficiency of power gene-

ration, %ηt — Thermal efficiency of power gene-

ration, %χt — Heat loss rate, %χe — Exergy loss rate, %ξt — Heat recovery rate, %ξe — Exergy recovery rate, %ΔTp — Pinch point temperature difference of

HRSG, ECΔTph — Temperature rise of saline water in

preheaters of METVC unit, EC

Subscriptsa — AirD — Desalinationf — Fuell — Lossnet — Net productionr — Recoveryw — Fresh water

opt — Optimal0 — Reference parameter for power and

water production

Acknowledgement

The authors gratefully acknowledge thesupport of the State Scholarship Fund of China tothe first author.

References

[1] Y. Wang and N. Lior, Performance analysis ofcombined humidified gas turbine power generationand multi-effect thermal vapor compression desali-nation systems — Part 1: The desalination unit andits combination with a steam-injected gas turbinepower system, Desalination, 196 (2006) 84–104.

[2] P. Chiesa, G. Lozza and S. Consonni, An assessmentof the thermodynamic performance of mixed gas-steam cycles, Part A: Intercooled and steam-injectedcycle; part B: Water-injected and HAT cycles,ASME J. Eng. Gas Turbine Power, 117 (1995) 489–498; 499–508.

[3] Y. Wang, A. Chen and J. Yan, Performance com-parison of combined cycle, STIG cycle, HAT cycleand other relevant cycles, J. Eng. Thermophysics,(1998) 545–548 [in Chinese].

[4] M.A. Bartlett and M.O. Westermark, A study ofhumidified gas turbine for short-term realization inmidsized power generation Part I: Nonintercooledcycle analysis, ASME J. Eng. Gas Turbine Power,127 (2005) 91–99.

[5] N.M. Wade, Energy and cost allocation in dual-purpose power and desalination plants, Desalination,123 (1999) 115–125.

[6] M.A.K. Al-Sofi and M.M. Srouji, Fuel allocation indual-purpose plants, Desalination, 100 (1995) 65–70.

[7] A.M. El-Nasgar, Cogeneration for power anddesalination — state of the art review, Desalination,134 (2001) 7–28.

[8] M. Jonsson and J. Yan, Humidified gas turbines —a review of proposed and implemented cycles,Energy, 30 (2005) 1013–1078.

Performance analysis of combined humidified gas turbine ...

Documents

Transcript of Performance analysis of combined humidified gas turbine ...