Lefebvre A.H., Ballal D.R. Gas Turbine Combustion: Alternative Fuels and Emissions

Подождите немного. Документ загружается.

410 Gas Turbine Combustion: Alternative Fuels and Emissions, Third Edition

15 stems with two premixers on each stem. Each stem incorporates two or

three separate fuel circuits for independently fueling the premixers.

A short annular liner was selected to minimize the amount of air needed

for wall cooling. Only backside cooling is used, so a thermal barrier coating

is applied to both the liner and in the dome area to keep the metal tempera-

tures within acceptable limits. The use of a multipass diffuser also permits

further reduction in overall combustor length. Of special importance to the

attainment of low emissions is the design of the premixers. Figure 9.32 shows

cross-sectional views of three different mixer designs that were subjected to

combustion testing. The double annular counter rotating swirler (DACRS)

was conceived to satisfy the restraints of autoignition and size. The duct

diameter is reduced toward the exit in order to create an accelerating ow,

thereby preventing ashback. The conical centerbody located along the cen-

terline of the premixer can be used to supply liquid fuel to an atomizer at its

tip, and gas passages for diffusion burning at low-power conditions [66].

The objective with this type of mixing device is to produce a completely

homogeneous mixture of fuel and air at the premixer exit. As the total area

of the fuel-injection holes is xed by the ow rate and the available fuel-

injection pressure, the design procedure is essentially one of nding the best

compromise between the desire for small injection holes to give a large num-

ber of fuel-injection points, and the equally important requirement of large

injection holes to allow the fuel jets to penetrate across the airstream.

In the premixer design designated as DACRS I in Figure 9.32a, the fuel is

injected radially outward into the airstream from holes in the centerbody

just downstream of the swirl vanes. This conguration suffered from unsat-

isfactory fuel jet penetration, so a modication was made by adding eight

Figure 9.31

Fuel nozzle assembly for LM6000 dry low-NO

combustor. (From Leonard, G. and Stegmaier, J.,

Journal of Engineering for Gas Turbines and Power, 116, 542–6, 1993. With permission.)

Emissions 411

radial spokes in the location of the holes in the centerbody, as shown in

Figure 9.32b. Each spoke has three holes to inject gaseous fuel perpendicular

to the owing airstream. Combustion testing of this DACRS II mixer showed

that single digit NO

emissions are attainable with this concept. A further

modication to the premixer designs described above was made by incor-

porating fuel-injection holes into the swirl vanes of the outer swirler, as illustrated

Outer

swirler

Outer

swirler

Outer

swirler

Inner

swirler

Inner

swirler

Inner

swirler

Fuel

Mixing

duct

Mixing

duct

Mixing

duct

Cooled

centerbody

(c)

(b)

(a)

Premixer

centerbod

Premixer

centerbod

Fuel injection spokes

Fuel injection holes

Circumferential

fuel manifold

Figure 9.32

Cross-sectional views of three mixer designs. (From Joshi, N.D., Epstein, M.J., Durlak, S.,

Marakovits, S., and Sabla, P. E., ASME Paper 94-GT-253, 1994. With permission.)

412 Gas Turbine Combustion: Alternative Fuels and Emissions, Third Edition

in Figure 9.32c. In this DACRS III conguration, the fuel is injected through

three holes in the trailing edge of each outer swirl vane and one hole in the

outer wall of the mixing duct in between each swirl vane. Fuel is fed to the

hollow outer vanes through a manifold on the outside of the premixing duct.

The NO

emissions obtained with the DACRS III mixer were very similar to

the DACRS II design.

A big advantage of the premixer module concept is that, once developed,

it has broad applications to a wide range of combustor sizes and congura-

tions, as discussed above in connection with the ABB-EV burner. The basic

module remains the same regardless of combustor size; only the number and

arrangement varies. Thus, according to Joshi et al. [66], the DACRS II and

DACRS III mixers could be applied to a range of GE engines, including the

LM1600, LM2500, and LM6000, because single digit NO

emissions have been

attained with both these mixers at test conditions encompassing the operat-

ing ranges of these engines.

9.7.8 Allison AgT100 Combustor

The main features of this combustor, shown schematically in Figure 9.33,

have been described by Rizk and Mongia [67]. It comprises a prechamber in

Main

chamber

Dilution

holes

Dome

Radial swirler

Starter

injector

Main fuel

injector

Axial

swirler

Ignitor

Pilot injector

Variable geometry

Figure 9.33

Allison AGT100 combustor. (From Rizk, N.K. and Mongia, H.C., Twenty-Third Symposium

(International) on Combustion, 1063–70, The Combustion Institute, Pittsburgh, PA, 1990.)

Emissions 413

which the fuel is vaporized and mixed with air, a pilot and ignition chamber,

and the main cylindrical chamber. Variable geometry is employed to control

the stoichiometry in the primary zone.

The prechamber contains a centerbody that houses both the main fuel injec-

tor and a pilot nozzle, which is employed only for lightup and acceleration to

engine idle speed. The main fuel is introduced from a manifold surrounding

the prechamber, just downstream of the prechamber axial swirler. Uniform

lming of the fuel is achieved by spraying it through eight tangential holes

onto the etched surface of the prechamber. The swirling air assists in the pre-

lming process. The high temperatures of the inlet air and the prechamber

walls combine to promote rapid vaporization of the fuel within the precham-

ber. At power modes higher than idle, additional air is admitted into the

prechamber through a radial swirler to merge and mix with the air owing

through the axial swirler.

Engine lightup is initiated in a small pilot chamber located on the side

of the main combustion chamber. This piloting device also serves as a sus-

tainer source when the combustor is operating at low inlet air temperatures

or at conditions that lie outside the normal lean blowout limits.

The swirling vaporized fuel–air mixture ows into the main chamber

through a round opening in the center of the dome. At high-power settings,

additional air is injected into the main chamber through eight holes that are

drilled in a manner designed to impart a swirling motion to the owing air.

Four simple rectangular dilution holes were chosen to ease fabrication of the

ceramic liner. Variable geometry, in the form of sliding bands, is used to vary

and control the ow areas of the dilution holes and the radial swirler in the

prechamber. At low-power modes, most of the air ows through the dilution

holes. As the fuel ow rate is increased above idle, the variable geometry is

moved to increase the airow through the radial swirler and to reduce, by a

corresponding amount, the airow through the dilution holes.

The use of variable geometry enabled the AGT100 combustor to meet the

program goals of 5.0 and 37 g/kg fuel for NO

and CO, respectively. Moreover,

the experimental data acquired in the course of this investigation was used

by Rizk and Mongia to develop a model for calculating NO

formation in

LPP combustors [68]. This model takes into account the effects of pressure,

residence time, and air distribution between different combustion zones. It

also provides useful insight into the contribution of the pilot chamber to the

total NO

emissions.

9.7.9 Developments in Japan

The strict NO

regulations in Japan have promoted several developments

in DLN combustion. Hosoi et al. [69] adopted a three-stage conguration in

their design of a DLN combustor for application to a 2 MW gas-red machine.

The arrangement is shown schematically in Figure 9.34. The burner assembly

consists of primary and secondary annular nozzles and a pilot nozzle at the

414 Gas Turbine Combustion: Alternative Fuels and Emissions, Third Edition

center. The fuel-injection schedule for the three coaxial burners is divided

into three modes. Mode 1 employs both pilot and primary nozzles with equal

fuel ow rates to each. It is used to sustain combustion from startup to 50%

of the maximum engine rpm. At this point, the scheduling system switches

to mode 2 in which the pilot fuel ow is held constant and the primary fuel

ow increases with an increase in load up to 50% base load. In mode 3, the

pilot and primary fuel ows both remain constant and the secondary fuel

ow increases with an increase in load from 50 to 100% base load. Tests

carried out on this combustor at a pressure of 1.18 MPa and an inlet air tem-

perature of 643 Κ indicated NO

levels of 10 ppmv (for 16% O

) at base load,

and combustion efciencies above 99.8% at all loads above 50%.

In the small engine size range, the Japan Automobile Research Institute is

collaborating with the Toyota Central Research and Development Laboratory

in the development of a low-emissions combustor for a 100 kW automotive

ceramic gas turbine. Some of the studies carried out in a premix-prevaporize

system, operating both at atmospheric pressure and on the engine at high

pressures, have been described by Kumakura et al. [70] and Ohkubo et al.

[71]. The aim of this work is to provide quantitative data on the inu-

ence of fuel drop size and fuel distribution on the degree of vaporization

achieved.

Trapezoid shaped liner

(Convection promoted liner)

Secondary fuel nozzle

Primary nozzle

Pilot fuel nozzle

Pilot swirler

Primary swirler

Secondary swirler

Secondary nozzle

Primary fuel nozzle

Ignitor

Dilution air holes

Cooling air

Figure 9.34

Dry low-NO

combustor for a 2 MW class gas turbine. (From Hosoi, J., Watanabe, T., Toh, H.,

Mori, M., Sato, H., and Ishizuka, A., ASME Paper 96-GT-53, 1996. With permission.)

Emissions 415

9.8 Lean Premix Prevaporize Combustion

A common feature of all the DLN combustors described above is that posi-

tive efforts are made to eliminate local regions of high temperature within

the ame by mixing the fuel and air upstream of the combustion zone. The

LPP concept represents the ultimate in this regard. Its underlying principle

is to supply the combustion zone with a completely homogeneous mixture of

fuel and air, and then to operate the combustion zone at an equivalence ratio

that is very close to the lean blowout limit. The smaller the margin between

stable combustion and ame blowout, the lower the output of NO

A typical LPP combustor can be divided into three main regions. The rst

region is for fuel injection, fuel vaporization, and fuel–air mixing. Its func-

tion is to achieve complete evaporation and complete mixing of fuel and air

before combustion. By eliminating droplet combustion and supplying the

combustion zone with a homogeneous mixture of low equivalence ratio,

the combustion process proceeds at a uniformly low temperature and very

little NO

is formed. In the second region, the ame is stabilized by the cre-

ation of one or more recirculation zones. Combustion is completed in this

region and the resulting products then ow into region three, which may

comprise a fairly conventional dilution zone.

A useful by-product of LPP combustion is that it is essentially free from

carbon formation, especially when gaseous fuels are used, in which case the

description “lean premixed” or “LPM” is more appropriate. The absence of car-

bon not only eliminates soot emissions, but also greatly reduces the amount of

heat transferred to the liner walls by radiation, thereby reducing the amount

of air needed for liner wall cooling. This is an important consideration because

it means that more air is made available for lowering the temperature of the

combustion zone and improving the combustor pattern factor.

Another important advantage of LPP systems is that for ames in which

the temperature does not exceeds 1900 K, the amount of NO

formed does

not increase with an increase in residence time [18,19]. This means that LPP

systems can be designed with long residence times to achieve low CO and

UHC, while maintaining low NO

levels. This nding is especially signicant

for industrial engines, where size is less important than for aero engines.

As noted above, this approach leads to an LPM combustor volume that is

approximately twice that of a conventional combustor [19].

The main problem with the LPP concept is that the long time required for

fuel evaporation and fuel–air premixing upstream of the combustion zone

may result in the occurrence of autoignition at the high inlet air temperatures

and pressures associated with operation at high-power settings. Appropriate

equations for calculating evaporation and autoignition delay times over wide

ranges of mixture temperature, pressure, and equivalence ratio are given in

Chapter 2. They may be used in the design of LPP combustors to ensure that

at no operating condition does the sum of the fuel evaporation and mixing

416 Gas Turbine Combustion: Alternative Fuels and Emissions, Third Edition

times exceed the autoignition delay time. Another problem associated with

all well-mixed combustion systems is that of acoustic resonance, which

occurs when the combustion process becomes coupled with the acoustics of

the combustor. Lean-premixed systems are especially prone to this problem,

as discussed in Chapter 7.

In summary, LPP combustion has considerable potential for ultralow NO

emissions. NO

levels below 10 ppmv have been reported by Poeschl et al.

[72], even for ame temperatures higher than 2000 K. However, many for-

midable problems remain, the principal being that of achieving complete

evaporation of the fuel and thorough mixing of fuel and air within the autoi-

gnition delay time and without risk of acoustic resonance or ashback.

9.8.1 Fuel–Air Premixing

Most types of ultra-low-emissions combustors rely on the attainment of

near-perfect mixture homogeneity before combustion for their success. A

homogeneous combustible mixture has the added advantage that it greatly

reduces the possibility of autoignition. Although fuel-lean mixtures tend to

have long autoignition delay times, imperfections in mixing result in local

regions in which the equivalence ratio is higher than the average value, and

ignition delay times are thereby greatly reduced. Thus, a high degree of mix-

ture homogeneity is essential, not only for the attainment of low NO

emis-

sions, but also to alleviate the problem of autoignition.

The inuence of mixture inhomogeneity on NO

formation has been exam-

ined by several workers, both theoretically and experimentally. Lyons [73]

used a multipoint fuel injector spraying Jet A fuel to achieve different equiv-

alence ratio proles across the diameter of the ametube. The results showed

that spatial nonuniformity in the equivalence ratio resulted in increased NO

emissions for equivalence ratios below 0.7 and decreased NO

emissions for

near-stoichiometric mixtures. Flanagan et al. [74] used a simple mixing tube

tted with a bluff-body ameholder at its exit. By changing the location of the

natural gas fuel injector along the length of the tube, the degree of fuel–air

mixing in the mixture approaching the stabilizer could be varied. When the

system was operating at an equivalence ratio of 0.66, a nearly vefold increase

in NO

emissions was recorded when going from well-mixed to incompletely

mixed conditions. Fric [75] used an experimental apparatus very similar to

that employed by Flanagan et al., to examine the NO

emissions produced

when burning natural gas in air at normal atmospheric pressure. He found

that temporal uctuations in the equivalence ratio can also raise NO

emis-

sions, in addition to spatial nonuniformities. For example, temporal uctua-

tions of 10% resulted in a doubling of NO

Leonard and Stegmaier [19] used a gas-red GE LM6000 combustor to

examine the effects of premixing on NO

formation. The results obtained are

given in Figure 9.35, which shows NO

as a function of average ame tem-

perature for various degrees of premixing. Nonuniformities are the result of

Emissions 417

uctuations in time as well as variations in space. Figure 9.35 contains data

for a nearly perfect premixer, a well-designed premixer, and a nonoptimized

premixer. It clearly illustrates the tremendous advantage to be gained from

thorough mixing of air and fuel. Leonard and Stegmaier also noted that the

amount of NO

formed in a nonoptimized premixer increased with increas-

ing pressure. They attributed this result to the fact that NO

is formed in the

hot spots (>2000 K) of poorly premixed ames by the thermal mechanism,

which is pressure dependent.

The only exception to the general rule that better premixing yields less NO

are the results obtained by Santavicca et al. [76]. These workers examined

the effects of incomplete fuel–air mixing on the emissions characteristics

of an LPP coaxial mixing tube combustor. Contrary to expectations, it was

found that an improvement in fuel–air mixing resulted in comparable NO

emissions for the same conditions of inlet temperature and equivalence ratio.

However, the better mixed device was able to demonstrate lower NO

emis-

sions due to its ability to operate at lower equivalence ratios.

A number of studies have been carried out on the use of mechanical

mixers for achieving a satisfactory degree of fuel–air premixing. Static mix-

ers are widely used in process engineering for mixing of both gases and

liquids, but they appear to have evoked little interest for combustion appli-

cations. Poeschl et al. [72] examined the mixing capability of a commercially

available static mixer and observed excellent homogeneity, with a standard

deviation of lower than 5% for a 2% loss in total pressure. According to

these workers, ashback should not be a problem if the velocity through

the mixer is higher than 20 m/s. Also, by attening the velocity prole, the

mixer eliminates the boundary layer along which ashback is most prone

Corrected NO

, ppmv

1650 1700 1750

Nearly perfect

premixer

Poor mixer

Good mixer

Flame temperature, K

1800 1850 1900 1950

Figure 9.35

Ef fe c ts of non un i fo rm fu el –a i r pr em i xi n g o n NO

formation. (From Leonard, G. and Stegmaier,

J., Journal of Engineering for Gas Turbines and Power, 116, 542–6, 1993. With permission.)

418 Gas Turbine Combustion: Alternative Fuels and Emissions, Third Edition

to occur. According to Valk [77], the static mixer system has good potential

for engines of modest compression ratio, but more work is needed to reduce

length, residence time, and pressure loss, while maintaining good mixing

performance.

9.9 Rich-Burn, Quick-Quench, Lean-Burn Combustor

The underlying principle behind the RQL combustor concept is illustrated

in Figure 9.36. Combustion is initiated in a fuel-rich primary zone and NO

formation rates are low due to the combined effects of low temperature and

oxygen depletion. A gradual and continuous admission of air into the com-

bustion products emanating from the primary zone would raise both their

temperature and oxygen content, thereby greatly accelerating the rate of

formation, as indicated by the high NO

route in Figure 9.36. If, how-

ever, the additional air required to complete the combustion process and

reduce the gas temperature to the desired predilution zone level could be

mixed uniformly and instantaneously with the primary-zone efux, the

combustion process would then follow the low NO

route shown in Figure

9.36. This gure serves to demonstrate that the design of a rapid and effec-

tive quick-quench mixing section is of decisive importance to the success of

the RQL concept.

A practical form of RQL combustor is shown schematically in Figure 9.37.

Combustion is initiated in a fuel-rich primary zone operating at an equiva-

lence ratio of between 1.2 and 1.6. A higher equivalence ratio would be even

more efcacious in reducing NO

, but could lead to excessive soot formation

formation (ppm/ms)

1,000

100

Lean

burn

Low NO

route

Rich

burn

High NO

route

0 0.5

Equivalence ratio

1.0 1.5

Figure 9.36

Graph to illustrate the principle of rich-burn, quick-quench, lean-burn (RQL) combustion.

Emissions 419

and smoke. For RQL combustion to be fully effective, the fuel must be

nely atomized and uniformly distributed throughout the fuel-rich zone.

Moreover, the primary-zone airow pattern must be designed to prevent the

occurrence of localized ow recirculation zones, which could increase resi-

dence times, thereby increasing the production of NO

[44].

As well as reducing thermal NO

, this initial fuel-rich combustion pro-

cess also discourages NO

formation from FBN by converting a large frac-

tion of the FBN into nonreactive N

[78]. A further advantage of an initial

fuel-rich stage in the combustion of low heating value (LHV) fuels contain-

ing ammonia (NH

) is that it can greatly reduce the conversion of NH

into

[79].

As the fuel-rich combustion products ow out of the primary zone, they

encounter jets of air that rapidly reduce their temperature to a level at which

formation is negligibly small. As mentioned above, this transition from

a rich zone to a lean zone must take place quickly to prevent the formation of

near-stoichiometric, high NO

-forming streaks.

If the temperature of the lean-burn zone is too high, the production of

thermal NO

becomes excessive. On the other hand, the temperature must

be high enough to consume any remaining CO, UHC, and soot. Thus, the

equivalence ratio for the lean-burn zone must be carefully selected to sat-

isfy all emissions requirements. Typically, lean-burn combustion occurs at

equivalence ratios between 0.5 and 0.7 [80]. After the requirements of com-

bustion and liner-wall cooling have been satised, any remaining air can

be used as dilution air to tailor the exit temperature pattern for maximum

turbine durability.

In some designs, the atomizing air is arranged to ow over the outside of the

liner wall in the rich zone before entering the fuel nozzle. This regenerative

backside convective cooling is an important design feature because conven-

tional lm cooling in the rich zone would create local near-stoichiometric

mixtures that would produce high levels of NO

Work has been in progress on RQL combustors since the late 1970s [81]. In

early experimental studies by Novick et al. [82], NO

emissions appeared to be

controlled only by inlet temperature and rich-zone equivalence ratio, whereas

CO and smoke emissions were inuenced markedly by both rich-zone and

Fuel

Air

Figure 9.37

Schematic diagram of RQL combustor.