Swaminathan N., Bray K.N.C. (eds.) Turbulent Premixed Flames

Подождите немного. Документ загружается.

References 363

[207] R. G. McKinney, D. Sepulveda, W. Sowa, A. K. Cheung, The Pratt & Whitney

talon × low emissions combustor: Revolutionary results with evolutionary

technology, persented at the 45th AIAA Aerospace Sciences Meeting and

Exhibit, Reno, NV, 8 – 11 January 2007.

[208] M. P. Spooner, Development of low emissions turbojet combustor, in Pro-

ceedings of the 2nd International Symposium on Japanese National Project for

Super/Hyper-sonic Transport Propulsion System (HYPR Association, Tokyo,

Japan, 1995) pp. 91–98.

[209] R. Tacina, C. Wey, P. Laing, and A. Mansour, Sector tests of a low nox, lean

direct injection multi point integrated module concept, ASME paper GT2002-

30089 (2002).

[210] R. Tacina, A. Mansour, P. Laing, and C. Wey, Experimental sector and ﬂame

tube evaluations of a multi point integrated module concept for low nox com-

bustors, IGTI-2004-53263 (2004).

[211] L. J. Spadaccini and J. A. TeVelde, Autoignition characteristics of aircraft-type

fuels, NASA paper CR-159888 (1980).

[212] S. Hayashi, H. Yamada, K. Shimodaira, T. Saitoh, S. Horiuchi, Spontaneous

ignition of atomised kerosene/air mixtures in a high pressure ﬂowing system.

[in Japanese], National Aerospace Laboratory Report TR966, Tokyo, Japan

(1988).

[213] C. M. Coates, Experiments on a high pressure premixing/prevaporising com-

bustor, Technical Memorandum P1099 RAE UK (1987).

[214] M. Rachner, J. Becker, C. Hassa, T. Doerr, Modeling of the atomisation of a

plain liquid jet in crossﬂow at gas turbine conditions, Aerospace Sc. Technol.

6, 495–506 (2002).

[215] J. C. Mankins, Technology readiness levels, A White Paper. Advanced Concepts

Ofﬁce, NASA (April 6, 1995).

[216] A. H. Lefebvre, Atomisation and Sprays (Hemisphere, Washington, D.C.,

1989).

[217] C. Cha, J. Zhu, N. K. Rizk, and M. S. Anand, A comprehensive liquid fuel

injection model for cfd simulation of gas turbine combustors, AIAA paper

2005-0349, presented at the 43rd AIAA Aerospace Sciences Meeting and Ex-

hibit, Reno, NV, (2005).

[218] E. De Villier, A. D. Gosman, and H. G. Weller, Large eddy simulation of

primary diesel spray atomisation, SAE paper 2004-01-0100 SAE 2004 World

Congress & Exhibition, Detroit, MI, March 2004.

[219] K. D

obbeling J. Hellat, and H. Koch, 25 Years of BBC/ABB/Alstom lean

premix combustion technologies, presented at the ASME Turbo Expocsition

2005: Power for Land, Sea and Air, Reno, Tahoe, NV, June 6-9, 2005.

[220] Integrated pollution, prevention and control. Reference document on best

available techniques for large combustion plant. European Commission, July

2006.

[221] Standards of performance for stationary combustion turbines: Final Rule.

40CFR, Part 60. Part 3, Environmental Protection Agency, 6 July 2006.

[222] Rule 413, Stationary gas turbines. Proposed Amendments. Sacramento

Metropolitan District Staff Report. Williams KJ. February 2005.

[223] San Joaquin Valley pollution control directive, Rule 4703, Stationary gas tur-

bines last amended September 20, 2007. Available at http://www.valleyair

.org/rules/currntrules/r4703.pdf.

[224] Directive 2001/80/EC of the European Parliament and of the Council of 23

October 2001 on the limitation of emissions of certain pollutants into the air

from large combustion plants.

[225] Tuning approaches for DLN combustor performance and reliability, EPRI,

Palo Alto, CA, 2005, 1005037. Available at http://mydocs.epri.com/docs/

public/000000000001005037.pdf.

364 Lean Flames in Practice

[226] S. Aoki, K. Matsumoto, Y. Douura, T. Oda, M. Ogata, and Y. Kinoshita, Up-

graded lineup of Kawasaki Green Gas Turbine combustion systems Proceed-

ings of the Conference on Power Engineering-09 (ICOPE-09), Kobe, Japan,

November 16-20, 2009.

[227] T. Scarinci, I. J. Day and F. Freeman, U.S. patent 6,698,206B2, March, 2004.

[228] T. Schweiger, M. Underhill, and D. W. Livingston, Rolls-Royce RB211-DLE

combustor Diffuser Design Optimisation, paper 99-GT-237, presented at the

International Gas Turbine and Aero Engine Congress & Exhibition, Indi-

anapolis IN, June, 1999.

5 Future Directions

5.1 Utilization of Hot Burnt Gas for Better Control

of Combustion and Emissions

By S. Hayashi and Y. Mizobuchi

Recirculated hot burned gas has long been used to stabilize ﬂames in continuous

ﬂow combustion systems, such as furnaces, boilers, gas turbines, and afterburners in

jet engines [1]. The recirculation is created by employing swirl vanes, bluff bodies, or

bafﬂe plates. The heat and radicals in the hot combustion products act as the ignition

source for fresh mixtures of fuel and air ﬂowing into the ﬂame-stabilization region.

In the late 20th Century, air pollution by combustion-generated smoke, partially

burned hydrocarbons, carbon monoxide, and nitric oxides, NO

, became a public

concern. With the increasing stringency of the regulations for NO

emissions, re-

circulation of combustion products or exhaust-gas came to be used as one of the

measures for controlling NO

formation. Exhaust-gas recirculation (EGR), for ex-

ample, is extensively used in reciprocating engines for automobiles [2], where the

engine exhaust gas is recirculated into the intake manifold in order to mix with fresh

air. Another example is recirculation of combustion products in furnaces and boilers,

which is sometimes called internal EGR. Most of the reduction of NO

emissions

achieved by EGR is attributed to reduced peak ﬂame temperatures that are due to

increased speciﬁc heat capacity of the diluted mixtures, though some of the reduc-

tion is attributed to resulting slower NO formation that are due to reduced oxygen

concentrations and increased carbon dioxide and water vapor concentrations.

More recently, in response to the increasing demands for improving fuel ef-

ﬁciency while reducing emissions of pollutants, mainly NO

, a new area of com-

bustion, which is known as highly preheated combustion [3], has evolved. Auto-

ignition is the mechanism for supporting combustion in this new mode of combustion,

though ﬂame propagation is the mechanism in the conventional mode of combustion.

Major examples are homogeneous charge-compression ignition (HCCI) engines [4],

catalytically supported gas-phase reaction for gas turbine engines [5], and mild or

ﬂameless combustion for boilers and furnaces [6–8]. A feature common to these new

combustion strategies is that the mixture or the air is highly preheated by compres-

sion, heat generated by reactions on catalysts, waste heat recovery by heat exchang-

ers or recuperators, or recirculating hot combustion products in the combustion

365

366 Future Directions

chamber, so that the mixture can be continuously auto-ignited. Some waste heat

recovery recuperators made of heat-resistant ceramics can preheat combustion air

to temperatures higher than 1300 K, and therefore mixtures of most fuels and air

can be ignited even under highly diluted conditions.

As for HCCI engines, a prototype demonstration is in progress, but several

technical issues, involving the control of ignition timing by EGR for example, are

still to be resolved before commercial production can begin.

Mild or ﬂameless combustion has already been used in commercial material-

processing furnaces, boilers, and other combustion systems and successfully reduces

, CO, and soot emissions to very low levels. Additional merits of this new

combustion strategy are low combustion oscillations or noise and uniform wall tem-

perature distributions.

For gas turbines, lean premixed combustion is potentially the most promising ap-

proach for ultra-low-NO

emissions because the turbine inlet gas temperatures are a

few hundred degrees lower than the peak ﬂame temperature because of the thermal

limitation of the materials presently available for turbine blades. This approach has

already been extensively used worldwide in large-scale natural gas-ﬁred gas turbines

for combined cycle power generation with ultra-low-NO

emissions, meeting the

most stringent emissions regulations for this pollutant. In aviation, the regulations

for NO

emissions are increasingly stringent, and therefore premixed lean combus-

tion technologies for kerosene are being actively developed for the next-generation

aeroengines. Some technical issues, such as combustion oscillation, trade-off be-

tween CO and NO

emissions, and, for high-pressure-ratio engines, auto-ignition

and ﬂashback in mixture preparation, need to be resolved. However, it is worth

mentioning that, in high-pressure-ratio gas turbines, shorter ignition lags resulting

from higher air temperature and pressure conditions could make the application of

auto-ignition-supported combustion much easier.

Investigation of the application of mild or ﬂameless combustion to gas turbine

combustors has continued. Model combustion chambers proposed by some investi-

gators [9, 10] are designed so that large strong vortex ﬂows of internally recirculating

burned gas can be produced to establish the conditions required for mild combus-

tion. The ratio of recirculating burned gas to fresh air in the previously mentioned

combustion chambers is much larger than that in the recirculation zone of conven-

tional swirl or bluff-body stabilized ﬂames, and therefore the mixture in the vortex

is highly preheated and diluted with burned gas.

The pressure to minimize CO

emissions from combustion systems for the miti-

gation of global warming has been increasing worldwide. Capture and sequestration

of CO

can be one of the major options for reducing CO

emissions from fossil

fuel combustion. For large combined cycle power generation plants with CO

se-

questration [11], combustion product recirculation is proposed to increase the CO

concentration in the working ﬂuid so that the efﬁciency for CO

separation can be

improved. As both combustion and emissions characteristics will be altered signiﬁ-

cantly by high CO

and H

O concentrations, the use of the combustion products to

optimally control the formation of emissions should be more actively investigated.

In gas turbines, the combustor operating conditions, such as inlet air tempera-

ture and pressure, equivalence ratio in the combustion region, and overall fuel–air

ratio of the mixture, vary widely with power or rotational speed. Therefore, an ideal

5.1 Utilization of Hot Burnt Gas for Better Control of Combustion 367

Figure 5.1. A schematic illustration of ﬂow of axially staged lean-mixture injection concept.

mild combustion mode can be established over only a limited range of engine op-

erations. The constraints imposed by the need for recirculating a large amount of

combustion products in the previously mentioned approaches would make the ﬂow

in the combustor complicated and therefore may result in an increase pressure loss

and a complicated combustor wall-cooling structure. Considering that staging of fuel

would be required for a combustor or to function satisfactorily over a wide range of

operation from engine start to full load, a multi-axially staged lean-mixture injection

concept [12–14] was proposed from our laboratory.

This section is devoted to axially staged lean mixture injection. Experimen-

tal observations with model combustors are presented and discussed based on the

predictions by simple numerical simulation of chemical reactions. A successful appli-

cation of the concept to a commercial gas turbine is presented, followed by comments

on more realistic simulations of the process to be utilized for the design of staged

lean-mixture injections.

5.1.1 Axially Staged Lean-Mixture Injection

In this combustion concept, hot combustion products prepared in the upstream

stage are used as the diluent for the downstream stage where lean to ultra-lean

mixtures, injected into and mixed with the hot burned gas, are reacted in the mild

combustion mode. Staged mixture injection in this simple straight-ﬂow system can

easily establish the conditions required for mild combustion, even in a gas turbine

whose combustor inlet air temperature is lower than the auto-ignition temperature

for the fuel. It should be noted that, even though they look similar, this concept is

different from the axially staged ﬂame combustion in that no ﬂame of the injected

mixture is stabilised. In this simple staging strategy, only the fuel ﬂow rates to each

stage are varied to modulate engine power or thrust while the air split between stages

is ﬁxed.

The combustion concept of a two-stage conﬁguration [14] is schematically il-

lustrated in Fig. 5.1 in which the secondary mixture is injected into the hot burned

368 Future Directions

Figure 5.2. A schematic illustration of gas turbine combustor with axially staged lean-mixture

injection.

gas produced by the conventional ﬂame combustion in the primary stage. Ideally, the

injected mixture is instantaneously and uniformly mixed with hot burned gas before

ignition so that reaction proceeds in a volumetric manner rather than in a ﬂame sheet

in the conventional combustion mode. The temperature and oxygen concentration

of the hot burned gas in the secondary stage are modiﬁed simply by varying the

fuel–air ratio in the ﬁrst stage.

The capability of this concept was experimentally veriﬁed by use of a three-

stage, model gas turbine combustor, shown schematically in Fig. 5.2 [15]. Tubular

ﬂame combustion of lean mixtures was employed for the ﬁrst stage because of its

inherently better stability at lean conditions and thermal reaction of lean to ultra-

lean mixtures was used for the second and third stages. In the ﬁrst stage the primary

mixture enters the cylindrical chamber from the four tangential slots opened on the

cylindrical wall to prepare a swirling ﬂow where a tubular ﬂame is stabilized. A pair

of mixture injection nozzles, opposing each other, was provided in the second and

third stages, respectively. Swirl was imparted to the mixture jets from the nozzles to

enhance the mixing of injected fresh mixtures and burned gas. Experiments were

conducted at an air ﬂow rate of 72 g/s and at an inlet air temperature of 700 K

and atmospheric pressure. Equivalence ratios of the secondary and tertiary mixtures

were varied from zero (just air) to about 0.7.

Figure 5.3 shows typical pictures of the combustion phenomena recorded for

ultra-lean to lean secondary mixtures injected into the hot burned gas of a temper-

ature of 1230 K. The hot burned gas was prepared by burning a premixed mixture

of an equivalence ratio of 0.46 in a tubular ﬂame. The blue ring in the photo (see

coloured plate) corresponds to the cross section of the tubular ﬂame in the ﬁrst

stage, and a pair of mushroom-shaped plumes is formed by the secondary mixture

5.1 Utilization of Hot Burnt Gas for Better Control of Combustion 369

Figure 5.3. Photograph of ﬂames in primary and secondary stages. (See colour plate.)

jets injected into the primary burned gas. The end wall is red-hot. In the third stage,

not mixture but air was injected from the tertiary mixture nozzles into the hot com-

bustion product. It was found that the secondary stage combustion was supported

by the hot burned gas from the primary stage. No ﬂame was seen with ultra-lean

secondary mixtures (φ

≤ 0.25), and a pair of mushroom-shaped reaction volumes

gradually becomes visible with increasing fuel supply to the secondary stage. Com-

bustion efﬁciencies greater than 99.9 were achieved over a wide range of φ

with

low-NO

emissions for ultra-lean mixtures.

The NO

emissions, corrected to 15% O

, and combustion efﬁciencies for two-

and three-stage modes are shown in Fig. 5.4 as functions of overall equivalence ratio,

, determined by gas analysis. The data with single-stage ﬂame combustion are also

shown for comparison. Auxiliary scales for equivalence ratios of the secondary and

tertiary mixtures are attached to the abscissa for convenience. The levels of NO

emissions in the staged mixture injection mode are maintained at very low levels,

provided that the equivalence ratios of secondary or tertiary mixtures are smaller

than about 0.5. It is found that most of the NO formed in the upstream stages is

converted into NO

probably by the fuel in the mixtures injected. Small combustion

inefﬁciencies seen when shifting from single- to two-stage modes and from two- to

three-stage modes can be avoided by slightly increasing φ

and φ

or by increasing

the residence times in these stages.

Visualization of the reaction zones of the mixtures injected into hot burned

gas was conducted by use of planer laser-induced ﬂuorescence (PLIF) imaging. A

rectangular combustor with quartz windows was used for ease of optical access. A

nozzle was attached to one of the side walls of the combustor to inject the secondary

mixture horizontally into the vertical ﬂow of combustion products of a temperature

of about 2000 K. Swirling motion was imparted to the secondary mixture jet to

enhance the mixing of the secondary mixture and the combustion products. The

equivalence ratio of the secondary mixture was varied from 0.1 to 0.8 and that of the

primary mixture was 0.6.

370 Future Directions

120

100

0.04

0.2

0.4

0.6

0.2 0.4

0.6

0.8 1

0 0.2 0.4 0.6 0.8 1

0.08 0.12 0.16

Single

Mode NO

C.E.

0.46

0.46 0.43

: Varied

=72 g/s,

=700 K

Two

Three

0.2 0.24 0.28

0.04 0.08 0.12 0.16 0.2 0.24 0.28

Overall Equivalence Ratio, φ

Secondary Mixture Equivalence Ratio, φ

Tertiary Mixture Equivalence Ratio, φ

(a) NO

concentration corrected to 15 % O

(b) Raw concentrations of NO and NO

Tertiary Mixture Equivalence Ratio, φ

=0.46

=0.43

(1397K)

=0.43

(1397K)

(1230K)

=0.46

(1230K)

concentration, ppm (15% O

)

NO, NO

concentration, ppm

Combustion Efficiency, %

Single

Mode NO

0.46

0.46 0.43

: Varied

=72 g/s,

=700 K

Two

Three

Figure 5.4. NO

and CO emissions, and combustion efﬁciency for single-stage combustion

and lean-mixture injection in the second and third stages.

Figure 5.5 shows typical instantaneous OH images obtained by PLIF for the

reaction of an ultra-lean mixture in a hot burned gas. The mixture was injected into

the cross-ﬂowing hot burned gas from a nozzle ﬁtted on the wall of the rectangular

combustor. No visible ﬂames are seen to be stabilised on the nozzle exit. Distributed

unburned mixture volumes are seen in the bottom-right corner of the photo. A

close-up of this region is presented in Fig. 5.5(b). Reacting volumes of small scales

are seen to be spatially distributed in the unburned mixture, suggesting that the

reaction occurs in a volumetric manner rather than in a ﬂame sheet. For mixtures

of equivalence ratios greater than about 0.4, a conventional premixed jet ﬂame is

always stabilised on the nozzle exit and no distributed unburned mixture volumes

are observed in the hot burned gas.

The potential of staged lean-mixture injection can also be demonstrated by

a simple 1D simulation. The schematic of an example is presented in Fig. 5.6.A

methane–air mixture of equivalence ratio of φ = 0.8, temperature of 300 K, and

pressure of 1 atm ﬂows into the computational domain from the left side. The mixture

burns through a premixed laminar ﬂame at the ﬁrst stage just after the inlet. The

secondary and tertiary mixtures are injected at 2-cm and 4-cm downstream locations,

respectively. In this 1D simulation, the ﬂow speed is chosen to be very low so that the

ﬁrst-stage premixed ﬂame, which is normal to the ﬂow direction, can be stabilised in

the combustor. Therefore the combustor size is correspondingly small to adjust the

time scales of the axial ﬂow to be those in 3D combustors.

5.1 Utilization of Hot Burnt Gas for Better Control of Combustion 371

Figure 5.5. PLIF images of OH in the reaction of ultra-lean mixture (0.3 equivalence ratio)

injected into ﬂowing hot combustion product: (a) overall view, (b) close-up of the region in (a)

where unburned mixture volumes are spatially distributed. The grey scale is arbitrary: light

patches represent high OH concentration whereas dark ones represent low concentrations.

In the following discussion, results for staged injection of a very lean mixture

at equivalence ratio φ = 0.4, which is leaner than the lower ﬂammability limit are

compared with those for staged air injection, which is currently used in practical

low-emission engine systems.

In Fig. 5.7(a), a very lean mixture at φ = 0.4 is injected at the secondary and

tertiary injections. The temperature and pressure of injected mixtures are 300 K and

1 atm, respectively. At the secondary injection, the same mass of the primary stage

mixture is injected and the overall equivalence ratio is reduced to about 0.6. One and

Figure 5.6. Schematic of 1D analysis of

staged lean-mixture injection.

372 Future Directions

(

)(

)

Figure 5.7. NO

reduction and operation range extension by staged lean-mixture injection:

(a) lean-mixture injection, (b) air injection.

a half times the primary stage mass is injected at the tertiary stage, and the overall

equivalence ratio becomes about 0.52.

Rapid mixing of burnt gas and injected lean mixture is assumed at each staged

injection. The burnt gas of the primary stage combustion is obtained by PREMIX

[16]. Then a zero-dimensional (0D) analysis at constant pressure is applied down-

stream of the secondary injection, where the reacting mixture is advected with the

ﬂow velocity. The detailed chemical kinetics is quoted from GRI-MECH 3.0 [17].

In the ﬁgure, the solid lines, dashed lines, and dot–dashed lines are the proﬁles of

temperature and NO and CO concentrations in ppm, respectively. The NO and CO

proﬁles for the single-stage combustion are represented by thin lines.

The temperature of the mixture rises to about 2050 K at the primary stage of

combustion, and drops down to about 1250 K by the secondary mixture injection.

The injected very lean mixture can burn without leaving a signiﬁcant amount of CO,

thanks to the relatively high temperature and radicals produced by the primary stage

combustion. On the other hand, the NO formation is suppressed or almost frozen,

and the NO emission is much smaller than that of the single-stage combustion case.

At the tertiary injection, there is an entry region upstream of the substantial com-

bustion region. Just after the injection, H, OH, and O decrease rapidly and then

increase gradually in the entry region until the substantial combustion reactions

start. The concentration of NO decreases in the entry region, where the injected O

converts NO to NO

. The conversion is analogous to the NO-to-NO

conversion ex-

perimentally observed in the very lean staged injection cases, as previously shown in

Fig. 5.4.NO

increases in the entry region and decreases rapidly once the substantial

combustion reactions start. Finally, the injected very lean mixture burns, and lower

emissions of NO

and CO are achieved than in the single-stage combustion.

In Fig. 5.7(b), the results for air injection are shown for comparison. The mass

ratios of staged injections are the same as those in the lean-mixture injection case

shown in Fig. 5.7(a). The overall equivalence ratio goes down to 0.41 and 0.24 at the

secondary and tertiary injections, respectively. When one compares the two types of

staged injections with respect to NO formation, the lean-mixture injection produces