Gt Upgrades Janes-Asme-paper

8/8/2019 Gt Upgrades Janes-Asme-paper

1/12

INTERNA TION AL GA S TURBIN E INSTITUTEThe AM ERICAN SOCIETY of MECHAN ICAL ENGIN EERS

Atlanta , Georgia USA

MARINE COMMITTEE1996 Best Paper Award

A Fu lly Enh anced G as Turb ine For Surface Ship sASM E Paper 96-GT-527

JACK JANESCalifornia Energy Commission

Presented At The Interenational Gas Turbine And Aeroengine Congress & ExhibitionBirmingham, U.K. June 10-12, 1996


2/12

A.S.M.E. PAPER SUBMITTED FOR TURBO'96, MARIN E SESSION

"A FULLY ENH AN CED G AS TURBIN E FOR S URFACE SHIPS "

Jack Janes, P.E.California Energy Commission

1516 Ninth StreetSacramento, Calif. 95814

ABSTRACT:

Clearly, the adv anced gas tu rbine has center stage in the worldfor converting fuel to work. The power an d efficiencydelivered by the a dv anced gas turbine have m ade it thepredom inant prime m over in the air, and increasingly so onland and sea. This paper explores the full potential offered formarine applications by the advanced gas turbine, a potentialthat is fully enhanced by use of available engineering optionstha t are external to the gas generator. The en ha ncem entoptions are:1) intercooling 2) therma l recup eration 3) steaminjection 4) reheat 5) closed loop cooling 6) catalytic partialoxidation 7) wa ter recovery. Of the options studied reheatinvolves a unique approach. Reheat is postulated to beaccomplished with a new simplified technique. Anautoign itable hydrogen -rich fuel is injected into the air paththrou gh the cooling pas sages and from the trailing edge of blades and vanes of the low pressure tu rbine, reheating the air

prior to entry into the free pow er turbine.

CHRONICLE OF MARINE LOCOMOTION

Seafaring people have long searched for the "optimummean s" of getting from point A to p oint B on the su rface of theocean. In the beginning, oars were the only mean s and forcenturies oceans were so traversed . With evolution, the severalhund red m an galleys of Roman an d Chinese ships became awond er of marine engineering. Sail was introduced by afarsighted development program, no doubt, and sail had toha ve been preced ed by the development of suitable cloth.Despite the many disad vantages of wind pow er and sail , i tgradu ally mad e a place for itself in the world 's navies andeventually, with the development of canvas, became the"optimum means" of ocean exploration and transport for the

next millennia. Oars retained a niche in the "optimu m m eans"ma rket. Two centuries ago, James Watt inven ted the coalburning boiler/ reciprocating steam engine cycle for landapp lications. One century later, desp ite their many limitations,necessary adaptations, and burd ensome fuel requirement, thesteam engine, and later in the century, De Laval's steamturbine became a superior means for ship propulsion. Saildeclined. Also in the last centur y, Rud olph Diesel introduced anew oil fired reciprocating engine. While Mr. Diesel said itcould never be used in mobile applications, developmentsproved otherwise and this engine remains in widespread use atsea. The gas turbine, a non-reciprocating engine, wa s invented

early in this century by Hans Holzwarth, and it has

subsequently reached great heights in aircraft service.Adap tations of these advan ced aircraft engines are in operationon land and at sea. History suggests the search forimprovem ent goes on un abated. It is ineluctable that the highlysuccessful gas turbine, adapted and improved from the present air/ground application , is predestined to pr ovide the seafaringwith their newest "optimum means."

The Advan ced Aircraft Engine As a Prime Mover

In the last ha lf century remarkable gains in aircraft engineperformance have been achieved. These gains have beenaccomplished despite design weight restraints and theinordinate demand on reliability imposed by flightrequ irements. The gain in pow er and efficiency is the result of the inexorable increase in pressure, temp erature, and gas flowat turbine entry, together with red uction in aerodynam ic lossescompressing inlet air. Today's elevated firing temperatu re isachieved by comp lex turbine blade cooling strategies and theuse of sup er alloys with special castings and coatings, allowinggas tempera tures near the melting point of the metal! Despitethe rigorous du ty cycle and total depend ance on state-of-the-art high technology, man y millions of operating h ours, all onliquid fuels, have firmly established the long-term d urability,reliability, availability, and maintainability of the ad va ncedaircraft engine. The engine can be considered suitable for anyalternative duty imaginable. The advanced aircraft engine,suitably modified, can and is now sup plying the shaft powerrequired for prop ulsion of surface ships. This paper add ressesthe question: How can future marine gas turbines fully benefitfrom performance enhancements being developed for stationaryapp lications? The enhanced gas turbine schedu led for surfaceships is intercooled and regenerated (ICR) and constructed of aircraft engine m odu les: The Westing hou se-Rolls-Royce (WR21)ICR. An alternative aeroderivative gas turbine with aregener ator is the Genera l Electric LM2500R. A thirdaerod erivative gas turb ine cycle, a fully enha nced cycle, is thesubject of this paper. (See figure 1. for the cycle comparison)

THE MARINE AERODERIVATIVE GAS TURBINE

The mu lti-shaft design of the ad van ced aircraft enginereadily lends itself to additional engineering techniques andprocesses routinely employed outside the aircraft industry.When integrated into the cycle, these techniques significantlyincrea se the shaft pow er delivered and thermal efficiency


3/12


4/12

An Option for Recoupin g Low Grade Heat the Cycle N ow Rejects

The high pressure heat exchanger redu ces the compr essedair tempera ture to < 200 oF, or to near ambient t empera turewith a sea wa ter aftercooler. This ultra cool comp ressed airwou ld be the choice for the coolant for the turbine hot sections.This exchanger produ ces a performan ce dividend that can be

divided between less comp ressor bleed air and/ or raising thefiring temp erature. The source of cooling air may be deemed as"too cold" or at least below op timum for the very hot m etalsurfaces. There is a p reheat solution for this "too cold" coolant.This is because low grade heat the cycle is norma lly forced toreject can n ow be recuperated.

The hot lubricating oil carries the heat lost in the bearingsand the very large heat generated in the gearing dow n of thegas turbine shaft speed from 10000 RPM to th e 100 RPM of the prop eller. If the gas turbine d rives a generator, the heatgenerated by the "copper losses" and the heat prod uced in the"iron losses" in the transformer are both eligible for recuperationto the cycle by the cool comp ressed air generated by the highpressu re heat exchanger. The cool air, so preheated, is ready toact as the optimum turbine coolant.

3. STEAM INJECTION

Ano ther of the long-identified m ethods to enh ance gasturbine performan ce, even before employing a recuperator torecover heat in the exha ust, is to increase the volume flow of gas at the inlet to the power turbine. Particularly, if the gas,unlike air, requires no comp ressor work, steam is such a gas.Injecting steam raised from the gas turbine exhaust into thecombustor has long been emp loyed to augment gas turbineperforma nce. To preserve the design surge m argin, gas turbineman ufacturers often l imit the amou nt of steam that wil l beperm itted to be injected to 5 percent of inlet air flow at baseload. The surge limit can be increased if the first vane throatdim ension is increased by "restaggering" the blades. Anyquantity of steam can be injected if the tu rbine is redesigned.

Steam is often used for NO x control. For instance, a steam

flow of ap proximately 2.8 percent air flow is requ ired to red uceNO x to 42ppmv from a f ir ing temperatu re of 2350oF. At this

temperatu re level, poun d for poun d steam injected in theadvanced aeroderivative gas turbine combustor isapp roximately 3.5 times more effective than air in pr odu cingnet pow er, i .e . the 5 percent steam flow w ill increase pow eroutput by 17.5 percent. The steam-to-fuel ratio is 2.5 / 1 a t asteam injection rate of 5 percent of air flow. The work prod uced by the steam injected in the gas turbine can farexceed the work p roduced in the alternative disposition of thesteam in a steam tu rbine. The difference in the two tu rbineinlet steam tempera tures can be over 1500 oF. The gas tur bine,steam injected, will produ ce 1 kilowatt of pow er for every 6 to10 poun ds p er hour of steam injected.

The steam injected gas tu rbine (or Cheng cycle) is not acombined cycle. It is a quasi-simple cycle. Essentially, the steam

injected gas turbine plant has two principal components 1) thegas turbine an d 2) the heat r ecovery steam generator (HRSG).The reliability and availability factors of the HRSG, particularlyof subcritical once-throu gh designs with almost no m ovingmecha nical parts, ha s prov en to be close to 1.000. This impliesthat the overall plant availability is set almost entirely by that of the gas tu rbine ava ilability (.96 to.98), and that oth er pow erplant configuration or arrangem ent with the sam e gas turbinecould be expected to exceed this overall availability.

Th e HR SG capital cost is a precise commod ity-class of purchase given the sizing param eters. {unfin ished, more tobe said}

What can be said of pa rt load efficiency? A critical featu reof economic operation of the ship prop ulsion gas turbines, aspreviously stated, is to maintain therm al efficiency at par t load.Mr. Stanley C. Keller of GE Evand ale Marketing Depa rtment,in a recent a rticle in the spr ing of 1994 issue of, Cogeneration and Co mp etitive Power Journal , suggests that steam injectionprovides an excellent method of fulfilling this primaryrequirement.

"Steam injection can also improve part-load efficiency. Singleshaft gas turbines usu ally have poor efficiency at redu ced load. With steam injection, steam can displace fuel demand and raise the part load efficiency. "

The Cheng cycle has been in commercial operation for about15 year s and has close to a million hours of operation. TheGeneral Electric steam injected aeroderivative plants havebeen in w orld-wide operation for about eight years. These gasturbines have been p roposed for utility operations inunm anned plants. The only negative aspect of steam injectionand of the subcritical once-throug h boilers is the dem and for

100 percent makeu p w ater of nearly the highest qu ality. Thisattribute prohibits consideration of such p lants in water shortareas or w here the p rice or use restrictions of water makes anopen cycle facility imp ractical. This paper will address th evexing w ater problem in a later discussion, with a p roposedsolution that seems particularly well suited for surface shippropu lsion app lications.

NO x Abatement Alternative to SCR/NH 3 and to Dry Low NO x Combustor

In addition to providing both augm entation in gas turbinepow er and an increase in efficiency, steam injection has longbeen employed to dilute peak flame temp eratures and red ucethe formation of oxides of nitrogen (the Zeldovich NO x) t h a tare otherwise formed in the gas turbine's combustion process.Unfortunately, steam dilution inhibits the combustion processand unburned products are formed such as carbon monoxide inrapidly increasing amou nts as the NO x is being redu ced. Steamor water injection has long been the gas turbine m anufacturer'srecommendation for NO x emission red uction. General Electricwill gua rantee 25 ppmv, with natural gas fueling, by thismeans. Further redu ction from this guaranteed level has beenthe target of a major developm ent effort by al l gas turbineman ufacturers. In the interim, for those areas in the world thathave a critical air qua lity problem, Selective Catalytic Reduction(SCR) of NO x with amm onia, with the catalyst bed suitablypositioned in the gas turbine exhaust, has been deemed th e bestavailable N O x control technology. Steam injection plus SCR willyield less than 9 ppm v of NO x in the stack gas. Often the CO isalso reduced by means of an oxidation catalyst positioned aheadof the SCR catalyst bed in the h igh temp erature exhau st flow.Recently, the gas turbine manu facturers have developed, atconsiderable expense, and some are now op erating, a dry lowNO x combustor. By premixing fuel and excess air (to a carefullycontrolled d egree) the peak flame temperatures are red uced,enabling the lower NO x concentrations ( 25 ppmv) to beachieved w ithout steam injection or SCR.

If steam injection is employed for performanceenhancement only, what are the coincident N O x and CO levelsthat could result , without the new dry low NO x combustor oruse of SCR? On Ma y 13, 1988, Dr. Donald W. Bahr , Manager o f Combu stion and H eat Transfer, Gener al Electric M &I,Evandale, Ohio gave a presentation to the South Coast AirQua lity Man agem ent District (SCAQMD) in El Monte,


5/12

California. Dr Bahr p resented this data for the two p rincipalsteam injected gas turbines GE is selling, the LM2500 and theLM5000, natural ga s fueled:

FULL LOAD, WITH MAXIMUMSTEAM INJECTION RATES

Steam Steam No x Percent of AccompanyingG as In je ct ed Fu el P pm U nco nt ro lle d C o C on c.P pm v

Tu rb in e Lb s/ H r Ratio @15%O 2 NOx @15% O 2---------------------------------------------------------------------------------------------LM2500 22,000 2.02 12 8 170LM5000 37,000 2.41 13 6 167

The Figure 3. data show s that 1) NO x can be redu ced to 6 to8 per cent of the u ncontrolled concentra tion and 2) theresulting CO concentration is una cceptable. The "maximumstea m injection rat es" ar e set by th e combustibility limits of methane in this steam/ air oxidizer. The exponential increase inCO is evidence of increasing combu stion instability. If moresteam were injected, combustion would be u nsustainable andflame ou t wou ld be imm inent because the so called "lower leanlimit" has been reached for this particular fuel . Gaseous fuels arecategorize d, on one basis, by their calorific value (Btu / ft 3).What is the effective calorific value of the steam diluted fuelburned in the above gas turbines? If the steam and methane( 1000 Btu/ ft 3) were supp lied to the combustor p remixed as a

medium Btu fue l , would then have a d i lu ted LHV hea t ingvalue calculated as follows:

= 1000 Btu pe r ft 3 / (2.41( 16 )+1) = 318 Btu/ Ft 3 18

The foregoing combu stion d ata on the GE engines speaks foritself. What is not said, how ever, is that chemical comp ositionof the fuel is a factor, the gas tur bine will suppor t stable andefficient combustion with an even more dilute fuel


6/12

wh ose blade s and vanes employ the most ad vanced alloys,castings and coatings and are also cooled by steam . Additionally,the steam, after cooling the internal hot sections of the pow erturb ine, is recovered and , with the ad ditional superheat, isrouted to an LP steam turbine, making wh at GE terms "athermodynamically seamless combined cycle." GE says the rotorsteam d elivery system meets extremely tight limits for leakage.Steam is a sup erior coolant compared to air and the GE gasturbine firing temperature can now be raised from 2350 oF t o2600 oF. This allows specific work to be raised from an alreadyremarkable .26 to an u nprecedented .33 MW-sec per pou nd of inlet air. This together w ith net p lant therm al efficiency over 60percent.

As an example of the effect of ultra high sp ecific powerfrom the newest land based plant , at this specific power, the9.3 MW of ship-board cruise power per engine could besupp lied by 9.3/ .33 or 28 pound s per second of air. At cruisepower the WR21 ICR requires 86 pounds of inlet air per second.

The question is: "What specific work can a fully enha ncedcycle be expected to deliver?" Such a cycle would m ake use of the GE developm ent, the closed loop cooling of the powerturbine as one of the performance enha ncing techniques of afully enhanced gas turbine.

Closed Circuit Cooling With An Alternate Coolant

The fully enhan ced gas turbine cycle proposed in this pa perwou ld make u se of a variat ion of the GE development. Thepow er turbine w ould be cooled by a closed loop. The coolantneed n ot be composed of steam only. The coolant wou ld be theeffluent from th e heat recovery u nit operating off of the gasturbine exhaust. For stationary pow er plant use, the effluentwould be a reformed steam/ methane mixture. For shippropu lsion, the effluent w ould be a sup erheated m ixture of distillate and steam . The HRSG feed would be an emu lsion of distillate and water. The shipboard cycle would have acatalytic par tial oxidation un it over wh ich the superheatedsteam/ hydrocarbon mixture would p ass. With the addition of acertain amou nt of air, the chemical reaction partial oxidationwould produ ce a hydrogen-rich fuel gas that wou ld fuel both

the reheat combustor and the pr imary combustor. This fuelwou ld allow operation of the reheat combustor as previouslydescribed.

In summ ary, the power tu rbine can be cooled closed loop,and th e reheat combustor mad e to work efficiently, asdescribed, when the cycle is fueled w ith distillate.

6. CHEMICAL RECUPERATION

Recuperation of the gas turbine exhaust heat is normallyaccomplished by steam-raising. Sensible and latent heat is takenup by an in -tube counter cur ren t f low of water. The s teamproduced is heated further (superheated) to a temperature inclose proximity to the gas turbine exhau st tempera ture. If thein-tube flow is a blend of water and hydrocarbon fuel, a processcalled steam distillation takes place. If the hydrocarbon is

natural gas, as in the case of a power p lant , the superheatedmixtur e can, with th e aid of a nickel catalyst, react chemicallyi.e. steam reform -- to produce a hydrogen-rich, highlycombu stible fuel gas. In the surface ship application of gasturbines, the standa rd fuel is a d istillate petroleum fraction thatcannot be easily steam reformed. The superheatedsteam/ distillate vapor effluent can, however, react with asmall amoun t of air in the p resence of a suitable catalyst and, bypartial oxidation, prod uce the same h ydrogen-rich fuel gas.

Two Phase Feed to the HRSG

The gas turbine exhaust heat-recovery would em ploy a sub-critical once-through boiler. The steam distillation processtaking place in the tubes not only pr odu ces an ideal feed forcatalytic partial oxidation to prod uce hyd rogen-rich fuel, butalso has a thermodynamic advantage over use of water only.The water and hydrocarbon are imm iscible, and each phase

exerts it's own independent vapor pressure. As thetemper ature of the flow in the tube rises, steam and distillatevapor are formed at a temp erature significantly less than w ithpure w ater or distillate. Unlike pure water, the temperatu re of the fluid mix continues to rise during th e whole evapor ationprocess. The "pinch p oint" restriction normally seen w ith steamis significantly lowered. Thus, the two phase feed recuperates aportion of the exhau st heat with th e distillate flow and offers athermodynam ic benefi t which is realized in the reduction inlatent hea t losses in the stack gas. (See figure 2)

Mu lti-Pressure H eat Recovery of Gas Tu rbine Ex hau st Heat

Norm ally, a heat recovery steam generator may operatewith one to three p ressure levels and, in the case of a combinedcycle, often the H P steam tu rbine exhaust steam is reheatedbefore going to th e IP steam tu rbine. The multi-pressure (and

saturation temp erature) levels allow the m aximum h eat to beremoved from th e gas turbine exhaust flow thereby minimizingthe stack temperature and minimizing the three steamtemp erature ap proaches (to the exhaust gas) exiting the HRSG.The steam reheat redu ces the overall water f low required bythe H RSG to recover the said heat. Reheat effectively reducesthe cycle losses associated with th e latent heat of water w hen,in a combined cycle the steam turbine exhaust flow iscondensed or, in the steam injected gas turbine cycle, the steamleaves up the stack along with the latent heat. The dow n sideof the mu lti-pressur e HRSG wh en employed w ith a steaminjected gas tur bine is that the steam flows associated with thetwo lower pressure levels, IP and LP, do not h ave sufficientpressure to enter the primary ga s turbine combustor. Some 10 to20 percen t of the total steam raised in the u ltra efficient multi-pressure H RSG mu st find a home in the cycle in the less thanthe optimum position, i.e. other than the h igh temperature gas

turbine combustor.

The prop osed scheme will allow all steam raised in theefficient multi-pressure boiler to go to the gas turbinecombu stor: 1) the high p ressure steam will be raised at apressure far in excess of the pressure requ ired to enter thecombustor 2) the intermediate pressure steam will be raised atthe pr essure required to enter the combustor 3) the lowpressure steam will be raised at the pressure designed toma ximiz e the hea t recovery. Two op tions are available forcombining the high and low pressure steam flows to therequired pressur e for combustor entr y. An efficient single shaftsteam turbine/ steam compressor arrangement, or a lessefficient thermocomp ressor with no m oving parts, may be usedto combine the three flows an d m ake all the steam available tothe cycle at the combustor pressu re.

In the fully enhanced cycle being proposed , the liquid feedto the hea t recovery un i t i s a two ph ase feed composed of hyd rocarbo n (distillate) and water as d iscussed earlier. Thethermod ynam ic adva ntages of the two phase feed will becombined with the multi-pressure heat recovery and vaporcompression discussed in the previous paragraph.


7/12

7. RECOVERY OF CYCLE WATER

As previously d iscussed, the steam injected gas turbinerequires treated makeup water of nearly the highest quali ty.Depen ding on location, this water can be expensive, or evenprohibitive if the availability and quality of the source water arenot favorable. For example, ships on Siberian rivers, floating ondistilled ice water , wou ld pay the lowest cost for makeupwa ter. Ironically, from the very beginning of sea travel, alloceangoing ships have been forced to pa y any pr ice in meetingthe vital demands for water of the right quality for bo th menand machines.

In the fully enhan ced gas turbine, the techniques andoptions of recovering both heat an d w ater are ev aluatedagain st an econom ic yard stick. Since all of th e ultra-qualitywater requ ired by the cycle is contained in th e stack gas flow,one need look no further for the first possible alternative source.If this source of makeup water is at all economically viable, thesteam injected gas turbine p ower p lant can then be located inplaces on land or sea that w ould otherwise be unsu itable forsuch a p lant. There are m any obvious trad eoffs in the severalmethod s of recovering the wa ter. All are beyond the scope of this paper. On the other hand , the physical requirements forcondensa tion to occur in the stack gas are not at all ambigu ous

and are read ily identified an d eva luated as to achievability. Briefly, the stack gas in th e steam injected gas tu rbine is

perhaps one third by volume water vap or, the partial pressureof water vapor is 1/ 3 of an atmosphere and the dew p oint isfound to be 161 oF. Cooling the stack gas further to 101 oF willhave condensed all of the necessary cycle makeup water. Atitanium-tubed surface condenser w ith ocean water as coolantis one solution. A "dry" cooling tow er is another. A th ird w ouldbe a d irect cold or chilled water contact condenser. (See Figures4 & 5) What is the qu ality of the condensed water? For a firstapp roximation, at worst, the entire sulfur content of the fuelwill be in the condensate as su lfurous or sulfuric acid. At 1/ 4percent sulfur in the distillate the condensate w ould be 445ppm sulfur. Passage through an anion bed and a polishing beddemineralizer should, in p rinciple, restore the water tofeedwater condition. Alternatively, the sulfuric acid could be

neutral ized and form a h ighly insoluble (

1 PPM) filterableprecipitate such as Barium Sulfate, negating the d emineralizer.Can any water r ecovery process be certified for sea du ty? Thefirst requirement, that it work effectively in land based plants,has been met.

The first land based plant em ploying a steam injected gasturbine with recovery of the injected water began operation inJanu ary, 1993, near Turin, Italy. The design of the p lant an doper ating featu res are covered in deta il in the 1994 ASMEpap er, 94-GT-17, by Ennio Macchi and Aurelio Poggio given inthe H ague in June, 1994. In this plant, the next add itional andmore complex step is taken in that the latent heat of condensation is recovered and used in a cogeneration m ode.This is also possible for surface ships. Other shipboar d ener gyneeds, space heating and hot water could be m et in a similarmanner.

FULLY ENHANCED GAS TURBINE SPECIFIED FOR ASURFACE SHIP

From a review of the available options for enhancing theper forma nce of a gas turbine, certifiable for sea du ty, thefollowing cycle is proposed as the fully enhanced gas turbine.

Intercooled Compression

The decision on the u se of an intercooler requires a carefulanalysis. Is the initial compression to em ploy an intercooled as in

the WR21 unit, or operate withou t an intercooler as the 2500R?The heat rem oved in the intercooler is norm ally too cool to berecovered for a useful pur pose and will be rejected in a seawater exchanger, and will be an energy debit to the cycle. Theintercooler may still be appropr iate depend ing on the gasturbine selected.

The High Pressure Heat Exchanger

The cycle will use a two-section high pressure heatexchanger (HPQX). The high temperature, high pressurecompressor (HPC) discharge air w ill be diverted to the HPQXwhere th e air will be cooled to w ithin the limits imposed by theHP QX "pinch poin t". The high p ressure air w ill be cooledfurther by a final seawater cooled aftercooler section within theHPQX, the heat removed being rejected. Ideally, therecuperator w ill then allow the gas tu rbine exhaust flow to becooled dow n to the dew point, returning the heat removed tothe cycle and in anticipation of condensing the w ater vapor ina following step. The d iversion of the H PC air is identical inconcept with the WR21 and the LM2500R gas turbine cycles. Inboth of those cycles, however , the HPC discharge air is routeddirectly to the recup erator. The HPQX is the new equ ipmen t inthe cycle, positioned before the recup erator.

If an intercooler is used, the HPC d ischarge air temperatu rewill be several hundred degrees lower than if no intercooler isemp loyed. Due to the "pinch p oint" limitation, the final HPCair temperature is actually lower when the initial HPCtemperature is greater. In addition to avoiding rejecting anycycle heat , the heat transfer du ty (Btu/ hr) will be greaterwithout the intercooler. The opportu nity to return more h eat tothe cycle in the form of steam is therefore greater. Furtherm ore,the cooler the HPC air is before entering the sea-wa ter cooledaftercooler section, the less heat is lost to the cycle. Wheth er anintercooler is used or not, the colder sea water temp erature w illthen read ily al low the HPC discharge air temperatu re to becooled furth er in the after cooler section, down to around 135 oF. This, in turn, will then allow the recupera tor to cool thegas turbine exhau st to a f inal stack gas temper ature to the 165 oF dew point. This allows the recuperator an assumedreasonable terminal temp erature d ifference of 30oF. A portion

of the 135o

F HPC discharge air wil l be employed to cool thetur bine hot sections. If the cool-air temperature is deemedbelow the optimum temperature, the plant ancillary low gradeheat loads (hot lube a nd transformer oils) now being rejectedwou ld be used to preh eat the cold air al located for turbinecooling.

The Primary Surface Recuperator

In the case of natural gas fueling, more particularlymetha ne, the total moles in the exhaust flow are equ al to thesum of the moles of inlet air, recuperation water, and m ethane.Thus, if the air flow on ly is to recuperate hea t from the total gasflow there is mismatch ed m olar (or volume) flow. The coolingair dedu ction ad ds to the regenerator molar f low imbalance.The thermodynamics are improved if the gas flows arebalanced, part icularly if the difference is al lowed for in the

average specific heats. The heat-carrying capacities of the twostreams are equal and the resulting hot-end and cold-endtempera ture app roaches are approximately equal. To achievethe desired m olar balance, a portion of the tu rbine exhaust gasflow do es not flow through the recuperator. Instead, tha tportion (the remaining unmatched exhaust gas) flows througha separa te parallel heat exchanger. This once-through tu bularexchanger employs the same fuel/ water feed as the HPQX torecup erate the remaining turbine exhaust heat in theunm atched exhaust flow.


8/12

The Parallel Exhaust Heat Recovery Unit

The gas turbine exhaust flow is precisely divided betweenthe recuperator and a second, conventional once-throughtubular heat recovery unit operating in parallel with therecuper ator. The heat in this port ion of the exhaust f low isrecupera ted by th e in-tube flow of a m ixture of distillate andwater (the sam e heat recovery fluid as used in the HPQ X). In is

exp ected tha t the reheat combustion will raise the powerturbine inlet temperatu re 500 o to 900 oF. The tu rbine exhaustflow is then anticipated to be 200 o to 400 o h o t t e r t h a n t h ecompressor discharge temperature, the effluent from theHPQX is eligible to be superh eated fu rther by effecting a closeapproach to the exhaust temp erature. The heat remaining inthis portion of the exhau st flow after sup erheating the H PQXeffluent is the heat recovered by the fuel/ water m ix. The twoflows of steam/ distillate vapor ar e joined to form the total fuelflow to the partial oxidation u nit prior to injection into the gasturbine combustors. The cooled exhaust gas exiting theexchanger joins the cooled exhaust gas exiting the recuperatorwhich rep resents the total stack gas flow. This exhau st gas isthen read y to condense, filter and recycled the water conten t.

The Free Power Turb ine, Closed loop cooled

As previously described, the free power tu rbine is expectedto be cooled by using th e flow of fuel gas composed of steamand sup erheated distillate. The fuel gas will be recovered fromthe p ower turbine after effecting the necessary cooling. Thefuel flow is more than ad equate to provide the requ ired coolingwith a relatively small rise in tem perature. This is predicated ona fir ing temperature ap proximately the same as the prima rycombustor. The large coolant f low together w ith blades andvanes made of superalloys, special castings and gas-sidecoatings will sup port a considerably higher firing temperatu re.This coolant is superior to air at this elevated tem peratu re inthat antioxidation coatings on the internal m etal surfaces willnot be required. Stoichiometric combustion in the reheatcombustor is a distinct possibility, particularly with sufficientsteam injection. Power turbine exhaust temp eratures will notlimit stoichiometric firing temp eratu res in the rehea tcombu stor. Existing alloys and mat erials w ill allow close

approach temperatures.

The Partial Oxidation Unit

The high temperature flow of steam and superheateddist i llate exit ing the pow er turbine w ill be prem ixed w ith asmall stream of high temperature air and p assed through a bedof partial oxidation catalyst. The effluent from this bed will be ahigh temperature, Hydrogen-rich fuel gas capable of autoignition and yielding ultra low NO x . This fuel gas will fireboth the primary combustor and th e reheat combu stor.

The Reheat Combustor

This reheat combustor will rely on the existing technologyassociated with blade and va ne cooling and the uniquecombustibility of Hydrogen. There will be no hardw are such astha t associated with the primary combustor. The hightempera ture Hyd rogen-rich auto-ignitable, low NO x fuel gaswill be introdu ced into the main air path from th e trailing edgeof the blades and van es of the last turbine stage prior to entryinto the free power tu rbine. The fuel flow will also provid e thecooling for the last stage turbine blade and vanes.

Cycle Water Recovery

Cycle water w ill be recovered by a d irect-contact cold w aterfog condenser with a secondary warm condensate-to-sea waterTitanium tubed heat exchanger (see Figures 4.& 5.) The dilute

sulfuric acid in the cond ensate will be neutralized w ith BariumCarbonate an d th e precipitate filtered ou t. The Sulfate can beconverted continuously (57 lb/ hr) to the carbonate.

PROJECTED PERFORMANCE OF FULLY ENHANCEDGAS TURBINE

The ultimate performan ce achieved by the fully enhancedcycle dep end s in part on the basic level of the gas turbinetechnology employed (firing temperatures and compressionratios). This requires turbine blades and vanes configured w ithstate-of-the-art aerodynamics, constructed of the mostadva nced alloys cast as a single crystal, coated w ith the mosteffective therm al barrier coatings, employing ad vanced coolingstrate gies. The cycle steam-raising capacity and the steam-assimilat ing capacity of the gas turbine must match and beoptimized , i.e. the "passing capacity" design of the turbinesmust recognize the combined air/ steam volume flow expectedin the optimized steam/ gas turbine engine. The gain in theenhanced cycle performance, starting with the uniqueadvanced gas turbine as described, can be significant. Forinstance, the very ad vanced (2600F and 23.2/ 1 compressionratio) and highly performing GE H cycle, has a net cyclespecific work of .3252 Mwe-sec per poun d of inlet air an d yetrequires 241 percent mor e air to be processed tha n the chem ical

stoichiometry requires. That is, if the oxygen w ere burned outby steam dilution and/ or a second burn in series or both, thetherm al pow er released wou ld be 2.41 times as great. Thespecific work is a prod uct of therm al efficiency and oxygenburnout efficiency times the stoichiometric heat release of 1.308Mwt-sec per poun d of air.

Can the oxygen be burned ou t? Can the efficiency bemaintained with increasing steam injection? Initially, theinjection of steam into the combu stor simultaneou sly raises thepow er output and efficiency (steam requires no ded uct forcompressor work, and with maximum preheat beforecombustor entry) and bu rns up add itional portion of theremaining oxygen. Eventually, as more of the steam is raisedinefficiently (no superhea t) with greater therm odyna mic lossthe cycle thermal efficiency peaks and declines with furthersteam injection. Power continues to increase. The several

rhetorical question posed: can the GE H cycle offer greaterperformance as a steam injected gas turbine than inconjunction with a steam turbine? If reheat w ere an option(which it is not in this engine) could the oxygen be burn ed outwith steam injection and a mod est second firing temperatu re?Would the efficiency exceed the 60 percent cycle efficiencynow achieved ? And a pertinent qu estion for GE: wh at wou ldbe the effect on over all pla nt econom ics of eliminating thesteam cycle by th e alternative of steam injection? If the oxygencan be consumed by steam injection and reheat in a marineengine, what size engine (air f low) would sup ply the cruisepow er of 9.3 Mws? At an assum ed efficiency of 50 percen t theair flow r equired = 9.3/ .5/ 1.308 = 14.22 lbs/ sec, a far sma llerengine. The following stu dy rep orts on use of an earlier lessadva nced engine emp loyed in a less enhan ced steam injectedand reheat cycle. The peak cycle efficiency was d etermined tobe 43.68 per cent, the stoichiome tric efficiency at maximum

pow er w as 35.17 percent.


9/12

Herm an B. Urbach of the Nava l Surface Warfare Center,Anna polis, Maryland in a 1993 ASME pap er, Titled, " A Stud yof the Feasibility of Steam-Augmented Gas tu rbines for SurfaceShips" employed a reheat combustor and massive steaminjection to achiev e stoichiom etric combustion, the physicallimit on therm al pow er. The thermal and shaft power formethan e and ambient air @20.734 percent Oxygen:

@59o

FReaction CH 4 +2O 2 --> 2H 2O + CO 2 -345,210 Btu/ lb m ole

Air req uired = 2/ .20734 = 9.646 mo les = 278.335 lbs.

Hea t/ lb air= 345,210/ 278.335 = 1240.27 Btu/ lb=1754.77 Hp -Sec/ lb

Air/ distillate ratio at stoichiometric comb.= 15.186 lb/ lb

Distillate fuel= 18300 Btu / lb = 1205.06 Btu/ lb of air15.186 lb/ lb

= 1704.96 Hp -Sec/ lb of air

In the aforementioned stud y, stoichiometric combu stion wasachieved with a steam injection flow equal to 49.47 percent of inlet air flow. The firing tem pera ture of the gas turbine

combustor and the reheat combustor w ere 2200 o F and 1900 oFrespectively with a 16 atmosphere gas turbine combustorpressu re. The cycle thermal efficiency was found to be 35.17percent. The shaft work per p ound of air for stoichiometricsteam injection is given by:

= .3517(1704.96) = 599.63 Hp-Sec/ lb of air

The gas turbine air flow was 84.1 lb/ sec, the shaft pow er is := 84.1(599.63) = 50,429 Horse po we r

REFERENCES

1. Parker, T.E. et. al.,1992,"Feasibility of Reheat Combustor forChem ically Recuperat ed Gas Tur bine Cycle". Technical report

by Phy sical Science Inc., And over, Mass.

2. Kam ali, K., Taw ney , R.,1990, " Aircra ft-Derive d SteamInjected Gas Turbines For Power Plant Applications", BechtelCorp . at Pow er-Gen "90, December 4-6, 1990

3. Moeller,D.J., Kolp,D.A., 1988,"World's First Full Stig LM5000Installed at Simpson Paper Company", ASME paper 88-GT-198

4. Macchi, E and Poggio, A, 1994, "A Cogeneration Plant Basedon Steam In jection Gas Tur bine With Recovery of the In jectedWater: Design Criteria and Oper ating Exper ience", ASMEpap er 94-GT-17

5. Fulton,K.,1992, "US Na vy ICR Engin e is Rated at 26,400hpand 42% Efficiency", Gas Turbine World, Novem ber-December1992

6. Maughan,J.R., et al, 1993, "Evaluation of Reducing GasTurbine Emissions through Hydrogen-Enhanced Steam-Injected Combustion." General Electric Corporate Research.

7. Cheng,D.Y. 1978, " Regenerative Parallel Comp oun d D ualFluid H eat Engin e", U.S. Pat ent 4,128,994 1978

8. Urbach, H .B. et al 1993 "A Study of the Feasibility of Steam -Augm ented Gas Turbines for Surface Ships", 1993 ASME paper93-GT-251

9. Bahr ,D.W., 1988, "LM2500 &LM5000 Ga s Turb ine Stea mInjection N O x Abatem ent Alternat ive to SCR", Presentation toSouth Coast Air Q uality Man agemen t District, May 13, 1988 inEl Monte, CA

10. Horn er, J.E. et al, 1994,"The Dev elopm ent an d Testing of the MFT8 Gas Turbine" ASME paper 94-GT-96. The Hague,Netherlands, 1994

APPENDICES

Conversion of the English Units for temp erature, poun ds andBritish Therm al Units (Btu) to the SU I units is accomplished asfollows:

C = (F - 32) * 5/ 9Kilogram s = 2.2046 (poun ds)Kcalorie = Btu/ .252


10/12


11/12


12/12

Gt Upgrades Janes-Asme-paper

Documents

Transcript of Gt Upgrades Janes-Asme-paper