THE AMERICAN SOCIETY OF MECHANICAL ENGINEERS

345 E. 47 St., New York. N.Y.10017

89-GT-254
A
E S The Society shall not be responsible for statements or opinions advanced in papers or in dis-
cussion at meetings of the Society or of its Divisions or Sections, or printed in its publications.
Discussion Is printed only if the paper is published in an ASME Journal. Papers are available
from ASME for fifteen months after the meeting.
Printed in USA.
Copyright © 1989 by ASME

Dry Low NOx Combustion Development for

Downloaded from http://asmedigitalcollection.asme.org/GT/proceedings-pdf/GT1989/79153/V003T06A008/2398364/v003t06a008-89-gt-254.pdf by guest on 23 July 2021

Electric Utility Gas Turbine Applications
A Status Report

LEONARD ANGELLO PHILLIP LOWE

Electric Power Research Institute INTECH Inc.
3412 Hillview Ave. 11316 Rouen Dr.

Palo Alto, CA 94303 Potomac, MD 20854

ABSTRACT dry, 15% oxygen exhaust conditions unless

otherwise stated. The units are volumetric, in
National standards for NOx control for gas parts per million (ppm), unless otherwise
turbines are being evaluated and modified stated.
throughout the world. These standards are
often supplemented by local or site specific SUMMARY
emission limits. The numeric value of any
emission limit is influenced by both the local It is apparent that extensive progress has been
environmental conditions, and, importantly, by made in commercializing dry, low NOx combustors
the actual NOx removal capacity of the specific for gas turbine applications. Almost all of the
technologies that can be applied to the available information deals with the development
processes and equipment that will be governed and performance of these combustors. NOx
by the emission standards. That is, emission emission levels as low as 10 ppm have been
limits can not be so restrictive that there is achieved under certain simulated engine
no technology available that can be applied to conditions, and it appears that emissions in the
meet the permit condition. However, as new range of 25-50 ppm will be routinely achievable
technologies are developed to meet the ever for the large, heavy duty turbines. Motivation
increasingly restrictive emission limitations, to achieve these performance conditions comes
they become the standard by which the next from two sources: an increasing regulatory
round of emission limit setting is guided. pressure to limit NOx emissions; and equipment
owner objections to the economic penalty
Current NOx control technologies for gas introduced by wet combustion controls. These
turbine applications employ fuel switching to penalties include items such as: a requirement
a low nitrogen content fuel, wet combustion for a large water supply, on the order of the
control (steam or water injection into the fuel rate; high level water purification to
combustor), or exhaust gas treatment (selective prevent corrosion; reduced thermal efficiency
catalytic reduction technology) or combinations expressed as a heat rate penalty; increases in
of those controls. CO emissions as the water injection increases,
often requiring the addition of a CO catalyst;
Recently, dry combustion controls have been, or and combustor pressure oscillations that cause
are being, developed by the gas turbine combustor wear.
manufacturers, and catalytic combustion control
processes are being researched. This paper it is also apparent that impacts and
presents a brief overview of the national NOx consequences of the fuel preparation system,
emission requirements of Europe, Japan, and the control systems, and combustion hardware (all
United States. A discussion of the new dry, of which are required by the dry, low NOx
low NOx controls presently emerging in the systems) upon the reliability of the combustion
electric utility, gas turbine market are then system and the gas turbine, itself, are a
reviewed, with an emphasis on the commercial critical area that has not been studied in
applications and typical operating results that sufficient detail to allow fully informed
have been experienced. purchase decisions to be made. This is
particularly true for the large engines, the
In this paper, all NOx values are referenced to equipment manufacturers can only provide short

Presented at the Gas Turbine and Aeroengine Congress and Exposition—June 4-8, 1989—Toronto, Ontario, Canada

0

term, part load tests at their fabrication NOx emission controls are installed in response
facilities. Also, reliability data requires to emission standards that the facility must
long-term evaluations that are only reasonably meet. In addition to the national emission
available from commercial operating systems. standards, local governmental authorities can
and do impose more stringent standards. In the
NITROGEN OXIDE (NOx) EMISSIONS AND STANDARDS U.S., for example, national standards require
something in the range of 75-140 ppm for
Examining the NOx emissions and emission limits electric utility gas turbine emissions,
is a convenient way to begin to understand the depending upon the plant efficiency and fuel
situation that is driving the development of used. However, several local authorities in
NOx mitigation technology. It needs to be California have imposed standards for

Downloaded from http://asmedigitalcollection.asme.org/GT/proceedings-pdf/GT1989/79153/V003T06A008/2398364/v003t06a008-89-gt-254.pdf by guest on 23 July 2021

cautioned that all the data bases (1,2,3 > are not application to the smaller turbines used in
necessarily consistent with each other. cogeneration facilities, ranging from 15-42 ppm
Different countries express their emissions in (some recent cogeneration facilities in
different units, some using volumetric and California have actually been permitted at 4.5
others using mass bases. These are not ppm), reflecting those regulators' understanding
directly convertible to one another, unless a of the lowest achievable emissions rate control
specific energy conversion process and fuel are available.
specified.
The type of fuel that is fired also strongly
The Japanese data (2 ) is just for stationary impacts the NOx emissions, and the emission
sources (no automobile sources were included), standards generally recognize this fact. Table
thus the total emissions for Japan are really 2 presents a comparison of the emission limits
much larger than those reported, since for oil and gas-firing, for selected countries.
automotive emissions are generally the dominant In developing this table, Dacey < 4 > drew many of
emission source in an industrialized country. the comparisons from standards that relate to
The U.S. energy base (e.g., the size of the boiler operations, as opposed to gas turbines.
U.S. energy production in MWhr.) is about 5 However the difference, in that oil-firing
times the size of the Japanese energy base, produces more NOx than does gas-firing, holds
suggesting that the U.S. emissions would be for either type of combustion operation.
about five time larger than the Japanese
emissions. However, the Japanese have applied Table 2
more extensively NOx controls to the combustion EMISSION LIMITS COMPARISONS
equipment in their energy base, as compared to FOR OIL AND GAS-FIRING
the U.S. Thus, the U.S. total NOx emissions
might be expected to be up to ten times the Country Ratio of Emission Limits
Japanese emissions, The 20 times difference in Oil-Firing Gas-Firing
NOx emissions indicated in the data in
references (1) and (2) and (3) suggests that a Australia 1 0.7
significantly different estimating procedure Austria 1 0.79
was used in the U.S. compared to the other Canada 1 0.67
countries. However, the data does show that FRG (after 1984) 1 0.67
significant emissions are attributable to Japan (after 1987) 1 0.46
utility and industrial operations, which Netherlands (before 1987) 1 0.79
provides the incentive for regulators to Switzerland 1 0.56
develop emission standards for those U.S. 1 0.67
activities. ECC (proposed up to 1995) 1 0.79
Table 1 presents a summary of selected national It is important to recall that many local
emission standards for NOx, expressed as NOz. jurisdictions have more stringent standards than
The table shows that the U.S. has the lowest those identified in Table 2. Also, normally the
national standards for the industrialized emission standard establishes the control
countries examined. technology that must be applied. But the
philosophy contained in the requirement to use
Table 1 the latest state of the technology means that
SELECTED NATIONAL NOx EMISSION STANDARDS if a new technology is developed that
(For Application to Gas Turbine Facilities) appreciably improves the ability to limit NOx
emissions, it effectively sets the emission
Country Turbine Ehission Limits standard that all subsequent plants must meet.
Capacity, MW ppm mg/a Since the new technology would then become the
lowest achievable emissions rate technology
FRG 60 & larger 147 300 available, it could also become the regulated
less than 60 172 350 control choice for new plants.
US 10 & larger 75 153
Japan (Natural gas) 10 & larger 100 205 DRY LOW NOx COMBUSTION FOR GAS TURBINES

Table Note: In the U.S. a correction for the Mitsubishi Heavy Industries, Ltd.
plant efficiency and fuel nitrogen can increase
the actual emission limits. For a modern, high Mitsubishi Heavy Industries, Ltd. (MHI)
efficiency, electric utility, combined-cycle presently manufactures the large gas turbines
plant the limits can increase to 110-140 ppm that are offered for sale by MHI or Westinghouse
(225-285 mg/Nm 3 ). Electric Co. Initially, MHI obtained a license
to manufacture gas turbines with the contains 0.01 % nitrogen. Tohoku would not
Westinghouse technology. It appears now that accept a wet NOx control system for this plant,
the combustor development work is being and required the development of a dry, low NOx
conducted by MHI, but there is a cooperative combustor. The development costs, from Tohoku's
relationship between the two companies. The viewpoint, were included in the plant purchase
new dry, low NOx combustor developed for their price. The dry, low NOx combustor design was
turbines (MHI markets the MW-701 series, and based, initially, on the two stage, low NOx
Westinghouse markets the similar W-501 series) combustor that MHI had developed for natural
has been installed at the Tohoku Electric Power gas-fired boilers. The additional development
Co. (Tohoku) 1090 MWe power plant (the Higashi took two years, and after the combustor was
Niigata plant).

Downloaded from http://asmedigitalcollection.asme.org/GT/proceedings-pdf/GT1989/79153/V003T06A008/2398364/v003t06a008-89-gt-254.pdf by guest on 23 July 2021

installed at the plant it took another three or
four months to complete the final adjustments.
Since Tohoku was required to install Selective Figure 2 (,6 % is an outline of the combustor,
Catalytic Reduction (SCR) NOx control including the variable geometry, air by-pass
technology to achieve 15 ppm NOx emissions from system.
the stack, the initial target for NOx emissions
for the 701D series turbine combustor was 75 Figure 2
ppm (an 80% efficient SCR NOx control would MHI DRY LOW NOx FOR 701D ENGINE: SCHEMATIC
then achieve the required 15 ppm emission).
They are continuing the development process
with a 25-42 ppm emissions goal for the 501F
series machines, which will have a 1260 °C
firing temperature (the 701D series combustor
fires at 1093 °C). Figure 1 gives the range of
test rig NOx emissions for the 501F combustor.
Additional work also is being directed at
developing a catalytic-based combustor.

Figure 1
MHI DRY LOW NOx COMBUSTOR TEST RESULTS
150

1. Main fwI 6. Variable ring

2. Pilot fuel 7. Bypass valve
100 3. Premix combustor 6. bypass Air
4. Air for combustion 9. Transition piece
E 5. Compress discharge air 10. T1x01ns
n
n

0
Z 50
The pilot burner is a standard diffusion burner.
Representatives of Tohoku have stated that it
has a venturi throat. The pilot burner is used
to carry the load from 0 to 40% of rated load.
It is noted that the pilot burner would produce
about 140 ppm NOx at full power if it were
0 1 carrying the full load. However, at about 40%
400 600 800 1000 1200 1400 of rated load (typically 38%), the pre-mix
FIRING TEMP, °C burners pick up the load and the pilot burner
is used to provide preheat to the pre-mix fuel.
The MHI development program is performed in An air by-pass flow control valve admits air to
steps, like all combustor development programs, the transition piece and is used to adjust the
with each step adding complexity. The present fuel:air ratio for the pre-mix burner. The
combustor is "developed" for natural gas by-pass valve is fully closed until the 40% load
firing, fairly extensive additional development is reached. It is then opened wide. As the
is needed to prove it for oil firing. Initial load increases further, the valve is partially
screening tests are used to select concepts for closed until the final setting (about 25% open)
further development. These are subscale, is reached at full load.
atmospheric tests. They have three high
temperature test rigs (maximum temperature of The liner contains about 12 cooling ports, each
1260 °C) that test full scale combustors: a one at two axial locations, to provide cooling and
combustor atmospheric pressure rig; a one combustion air to the primary or pilot burner.
combustor 3.5-atmosphere rig; and a two The design of the pre-mix fuel entry ports and
combustor 15-atmosphere rig. Engine tests in the combustion/cooling air entry ports into the
the shop can be tested up to 132 MW loads for combustor was a major development problem.
short periods of time. About 40 small holes located around the
circumference of the burner are used. There are
The Tohoku plant has six Mitsubishi Heavy four separate axial locations along the pre-mix
Industries, Ltd. (MHI) MW-701D gas turbines, combustion zone where the combustion and/or
operated in a base load mode. The fuel, LNG, cooling air is introduced in addition to the air
provided with the pre-mixed fuel. The NOx (TEPCO) Futtsu Station. Initially they
discharge to the turbine is also effected by experienced flame out problems at low loads when
the by-pass valve setting (which controls the the steam injection was initiated. Subsequent
fuel:air ratio in the pre-mix combustion zone) modifications to the drains and steam
and by the ambient air temperature. Initially, manifolding were reported as correcting this
if the ambient temperature changed more than 15 problem. However, TEPCO, for a variety of
°C, a new by-pass valve setting was required. reasons, appears to have required that the
After the first year of plant operation the second 1000 MW of capacity be based upon dry,
control valve had to be modified < 7 > because of low NOx combustion control. They have joined
the operational reliability problems. Now the GE as a sponsor of the development work needed

Downloaded from http://asmedigitalcollection.asme.org/GT/proceedings-pdf/GT1989/79153/V003T06A008/2398364/v003t06a008-89-gt-254.pdf by guest on 23 July 2021

control system for the air by-pass valve to commercialize the new combustor. The stated
provides fine adjustment following ambient goal for the combustor is 30 ppm, however, the
temperature changes. Tohoku now reports that TEPCO plant will also install SCR controls.
the modified valve is working satisfactorily. Thus, any dry control of 50 ppm (or even as high
as 75 ppm) should be adequate for the
During the initial plant operations it took 3 installation, although the lower the NOx level
or 4 months to establish the by-pass valve the lower will be the SCR system and operating
adjustments needed to control the turbine costs.
outlet NOx to 60-85 ppm at 15% oxygen. An SCR
system was used to control the plant NOx The GE program is building upon their experience
discharge to 10-15 ppm. Since then it has been in the EPRI program conducted at the Wharton
stated that a seasonal adjustment to the Station of the Houston Lighting and Power
by-pass valve setting is required, and that the Company, where the GE "quiet" combustor was
adjustment can be made automatically. demonstrated < 9) . That combustor uses a primary
Examination( 7 ) of operational strip charts of zone fueled by six annular fuel nozzles, with a
actual NOx and unburned hydrocarbon (UHC) secondary combustion zone fueled by a single,
readings as a function of the turbine power central fuel nozzle and air swirler. In the
level showed that five to ten percent daily current program, GE has apparently tested at
variations are possible. Although the least three separate designs in a full-pressure,
manufacturer has indicated that there are no single-burner test rig. An atmospheric pressure
restrictions on the turbine operating range, it test rig was also used to evaluate the fuel/air
is recommended that the power range abe mixture profile at the exit of the first
maintained between 90 and 130 MW, from the combustion stage and to map the NOx and CO
point of view of emissions performance. Since relationship at various combustor discharge
there are three turbines in a train and two temperatures. Although the combustor has a
trains at the Tohoku power plant, there is reported operating condition of 30 ppm NOx at a
reasonable flexibility present to provide for firing temperature of 1104 °C, the data < 8 >
their operational needs. indicates that the unit operates at 40 ppm NOx
and 25 ppm CO at a firing temperature of 1085
Tohoku inspected (7 > the plant after about 8,000 °C. Figure 3( 8) shows the asymptotic
hours of operation. Their conclusions were relationship of the CO emission at the proposed
that the combustion system would operate at firing temperature. The figure suggests that
least 16,000 hours before "wear" becomes a operational control at 1085 °C would be very
problem (perhaps causing the need to replace difficult, and CO catalysts might be required.
the cooling shroud), but the fuel nozzles were
expected to last much longer. This is Figure 3
consistent with U.S. data on the life of most GE DRY LOW NOx COMBUSTOR CO EMISSION TEST DATA
dry combustion control systems. Water
injection typically reduces the combustor
lifetime by 50-75 percent. At the inspection
it was estimated that the transition piece
should last three or more years. However, they
did observe that the high temperature, thermal
barrier coating on the combustor liner had E
d
peeled off and had to be reapplied. Special a
instrumentation used during the start up period 0
had indicated that the liner wall near the V
pilot burner had reached about 600 °C, and at
the end of the pre-mix burn region it had
reached about 550 °C. The metal temperature at
the leading edge of the first stage turbine
vane had reached about 840 °C. These facts
indicate an over temperature problem probably 00
occurred. FIRING TEMP, °F
General Electric Company (GE)
The test program is continuing using a variety
GE had used wet combustion control in a number of test rigs. For analysis purposes( 9 ), the
of U.S. applications involving their MS7001E, combustor is segmented into three zones: a
60 Hz turbines, and they applied the concept < 8 > primary zone, a secondary zone, and a dilution
to the first 1000 MW installed for the MS9001E zone. The primary zone extends to the end of
turbines of the Tokyo Electric Power Company the centerbody and includes the six primary fuel
 Li

nozzles located on the circumference of the the cooling air. Its wear pattern is
combustor. Premixed fuel is injected through acceptable, but it produces about 199 ppm NOx
these six nozzles. The secondary zone extends at full load and 88 ppm with steam injection.
from the end of the centerbody to the plane of
the dilution holes. The secondary fuel nozzle Hitachi has continued its low NOx burner
extends through the center of the centerbody. development, examining a two stage combustor
The dilution zone is the region immediately with pre-mix and also catalytic combustion. The
downstream of the dilution holes. two stage combustor is their reference dry, low
NOx combustor. It uses diffusion combustion for
Combustion light-off occurs with the six the primary burner and pre-mixed fuel, with an
primary nozzles operated as diffusion flame axial swirl device (see Figure 4) , for the

Downloaded from http://asmedigitalcollection.asme.org/GT/proceedings-pdf/GT1989/79153/V003T06A008/2398364/v003t06a008-89-gt-254.pdf by guest on 23 July 2021

combustors. At a preselected transfer point, secondary combustion. At present they are only
fuel is transferred from the primary nozzles to looking at natural gas and light distillate
the secondary nozzle in the centerbody, and a fuels. Their development program involves
lean diffusion flame stabilizes in the subscale screening tests to help them decide
centerbody recirculation zone. In this mode what full scale tests should be conducted.
the NOx emission are reduced but the CO and UHC Temperature measurements are the key test item,
emissions increase, compared to the light-off the results do not equate to full scale results
conditions. In a third mode, fuel is provided but can be used to select concepts for further
to both the primary and the secondary study. The remaining testing is done in one and
combustors and they operate in a lean-lean two burner test rigs. Full scale testing is
mode. Operation in this third mode is a trade critical and the inlet temperature match to
off between NOx and CO and UHC emissions. In turbine operating conditions is also critical.
the fourth mode of operation lean pre-mix fuel Since only one combustor has an ignitor (others
is introduced through the six primary nozzles, ignite by cross-fire) , it is also important that
and the secondary nozzles is operated with a they conduct multiple combustor tests to confirm
minimal pilot flame. Fuel combustion begins flame stability and the ability to light off the
when the fuel reaches the vicinity of the pilot other combustors. Hitachi has indicated that
flame. In this final operational mode, NOx several more years are required to complete the
emissions are lowest and CO emissions are development of this combustor system.
highest for lower firing temperatures, and the
reverse is true for the higher firing Reliability testing can take several years to
temperatures. The actual emission levels can accumulate the required operational hours.
be controlled by scheduling the fuel fraction Since fuel costs are high, engine testing is
to the pre-mix and the pilot nozzles. The air limited to the preshipping tests normally
staging or regulation, using inlet guide vane conducted on the engine. They can test up to 17
modulation, is external to the combustor in MW at 13.6 atmospheres in a large shop facility.
order to increase the combustor reliability and Combustor development work occurs at their
to enhance the control of. the burning process. Mechanical Engineering Research Laboratory and
Test rig data for this combustor(e> shows step at the manufacturing facility at Hitachi Works.
changes in the NOx emissions when the load is Hitachi is also working on a catalytic combustor
transferred from one combustion process to (three stages of catalysts) , but that program is
another. It should be recognized that this is reported as being 7 to 10 years from having a
laboratory data, field confirmation is still commercial product.
needed.
Results from the Hitachi current dry, low NOx
Hitachi Works (Hitachi) combustor are shown in Figures 4 and 5( 11 )• The
figures present data from their present two
Hitachi turbines are based upon a GE license. stage combustor. Figure 4 shows the gas
They manufacture the MS6001B (37 MW), the temperature and NOx contours for firing with the
MS7001B (61 MW), the MS7001E (77 MW), the primary or first stage burner only. It suggests
MS9001B (85 MW) and the MS9001E (113 MW) that the 75 ppm NOx goal is achievable. Figure
machines. Although they have a new two stage, 5 indicates that the final operational NOx
dry, low NOx combustor undergoing reliability condition will be near 75 ppm, since the
testing, the first field application is combustor produces about 37% of the NOx that
expected to be in 1990 at the Chugoku or Kyushu their standard combustor produces, and their
power stations. Hitachi's first dry, low NOx standard combustor produces about 180 ppm NOx at
application occurred in 1981( 10 ) at the Japanese full load (10 >. Data on CO emissions suggests
National Railway Kawasaki plant. They that load following with acceptable combustion
installed a lean firing combustor using liner characteristics will be very difficult.
1908 (10 . Since the combustor produced 155 ppm
)

at full load (176 ppm at peak load), and since ASEA Brown Boveri (BBC)
the local governmental requirement was 59 ppm,
Hitachi used steam injection (ratio of 1:1 to BBC initiated their dry, low NOx combustor
fuel) to meet the requirement. At normal development for their type 11 and 13 gas
turbine inlet temperatures of 1,062 °C the test turbines which employ a single, silo-type
data seems to indicate that they really achieve combustor. Although the new combustor is
about 70 ppm NOx output. After one year of somewhat larger than the original combustion
operation the plant inspection revealed that chamber, it fits to the same flange on the top
there was excessive liner wear( 1 °>. The liner of the gas turbine casing. Work has begun on
was replaced with the 6406 type (10 ) another developing a can type, dry, low NOx combustor
lean burn combustor with radial distribution of for the type 8 gas turbine, however that
 a

development schedule has not been released. emission performance.

Figure 4 Figure 6
HITACHI DRY LOW NOx COMBUSTOR BBC DRY LOW NOx SILO COMBUSTOR: SCHEMATIC
TEMPERATURE AND NOx CONTOURS
1

(a) Temperature Contours III III

(K) A 1 — Fuel supply
te-, 2 - Combustion chamber cover
3 — Outer shell

Downloaded from http://asmedigitalcollection.asme.org/GT/proceedings-pdf/GT1989/79153/V003T06A008/2398364/v003t06a008-89-gt-254.pdf by guest on 23 July 2021

• 4 — Intermediate shell
,toG ty00\ 5 — Finned segment carrier
6 - Finned segments
7 - Trunnion
70 1^ II/ 4 s/
• - ^ ] 30 8 — Air from compressor
9 — Hot gas flowing to turbine
10 - Manhole
(b) NO. Contours 11 — Air nozzle
(Pam) 12 — Sight glass
13 — inner liner
14 — Supporting plate

Figure 5
HITACHI COMBUSTOR NO2 EMISSIONS

8T
O
'.100 Figure 7
BBC PROTOTYPE COMBUSTOR OPERATIONAL DATA
o
J
0 o0 80
0
„^ 0
o
0 OppmUHC
60 2 1ppmCO
2 27 9ppmNOx
4 2 ppm NO2
2E 350 1 75 2
0 a. 41 53 ppmNO X
0 40 n5
d 150
7
6
13
12
O ^ 0 910 65 13 22
04 690
255 8

i 20 509 1 8 11 3
Z3 11 43
10
2 120 65
m2 30 245 36
0 60 25 27 3Q
0
Z 0 0 25 50 75 100 0 450 320 5O 65
2 1 491 136 27
Gas Turbine Load (%) 292 331 25 34 KEY
6 2658 74 insitu flue gas
5 25 ..151.02
24 70
Figure 6(12,13) shows the outline of the new
silo combustor. In the BBC design, there are n re 20 30 40 50 60 70
37 mini-burners in the type 13 combustor and 34
in the type 11 combustor. As the load is POWER, MW
increased, groups of mini-burners are ignited
as a unit. Thus, it is reported, that there is
no abrupt load shift as the turbine output is
increased or decreased. The emissions do, Table 3
however, experience a step change in emission COMMERCIAL APPLICATIONS OF THE
levels (see Figure 7) as mini-burner groups are BBC DRY LOW NOx COMBUSTOR
added or removed from operation. Figure
7( 12,14 ) indicates that the NOx emissions range Plant, Country Thrbine Rating Date in NOx Emission
between 22 and 74 ppm, but the CO can vary Model MWe Service mg/>n' pas, ave.
between 1 and 690 ppm, depending upon the load
and the number of mini-burner groups that are Lauswaard, Germany Type 13C 70 1984 100 13-74
fired to meet that load requirements. Full Lauswaard, Germany Type 13B 60 1987 150(oil) 75
power operation is very good, indeed, but CO Kornenburg, Australia Type 13B 60 1987 150 75
and unburned hydrocarbon emissions can become Lange Welde, Netherlands Type 11D 66 1986 102 27-62
large at part loads. Merwede Kanaal, N.lands Type 13E 140 1988 112 56
Hemreg, Netherlands Type 13E 140 NA 150 75
The prototype of the original BBC design was
tested at the Lauswaard plant of Stadtwerke The Lauswaard and Kornenburg applications were
Dusseldorf. Initial tests were conducted in retrofits, the others are new installations.
1984. Table 3(14) gives the schedule of The NOx emission reported is for the gas turbine
applications since then as well as their NOx only; all of the above applications are for

6
 L

combined cycle plants that have a fired waste assure the ability to meet new NOx limits for
heat recovery boiler. Thus, depending upon the fuel oil operation. Natural gas operation is
operation, the reported NOx emissions are not expected to require the use of wet
referenced to 15 or 3% oxygen. The values combustion controls in addition to the low NOx,
given in the table are for dry exhaust gas and premix operation.
are referenced to a 15% oxygen concentration.
Also all the values are for natural gas firing Figure 8
except for one unit at Lauswaard. BBC is now KWU LOW NOx BURNER OUTLINE
guaranteeing a NOx output of 100 g/GJ,
equivalent to 56 ppm at 15% oxygen.

Downloaded from http://asmedigitalcollection.asme.org/GT/proceedings-pdf/GT1989/79153/V003T06A008/2398364/v003t06a008-89-gt-254.pdf by guest on 23 July 2021

The most detailed performance data reported is
from the first Lauswaard plant retrofit
application. Kassebohm(14> reported the NOx
emissions before and after the retrofit, for Diffusion burn.r

P-1
fuel oil inlet
both fuel oil and natural gas firing. The data Pilot fuel pas nozzle
indicate that at full load a reduction factor Steam or water
of about 70% was achieved for natural gas Main fuel pas nozzla—.

firing and of about 60% for fuel oil firing.

Figure 7 gives more details on the emissions.
It shows the NOx, the contribution from NO to
the NOx value, the carbon monoxide (CO) and the
unburned hydrocarbon (UHC) emissions, all
measured in the wet flue gas. The figure also Steam injection Water Injection
shows the value of NOx, corrected to dry, 15% Fuel gas for
oxygen conditions. The data is presented as a premix burner Main sir flow
function of the number of mini-burner groups
that are fired. The data also show the impact
of initiating the third or fourth burner group
at an earlier power level. At power levels
above 55 MW, all the emission levels are low
and the NOx is essentially composed of NO.
Between 40 and 55 MW the emissions, with the —/I
Fuel oil
Full gas
diffusion burner
exception of CO, are relatively low. Below
about 30 MW the CO and UHC emissions are high.
The data show that the maximum NOx emission is
about 75 ppm (dry, 15% oxygen), and levels as
low as 13 ppm are achieved at part load power.
The contribution of NOz is important, since
this gas is visible at concentrations above 20 KWU has found that in their large geometry
ppm. combustor, lean firing, especially with a lean,
premix fuel, attains low NOx while keeping CO
Siemens/Kraftwerke Union (KWU) and UHC emissions low. This is similar to the
results EPRI( 15 ) projects have achieved.
The KWU design incorporates two large, silo, However, some KWU test data indicated that from
external combustors for their utility-sized gas start up to idle, the NOx, while low, is
turbines (the V82 - 60 MW machine; the V93.0, composed of up to 90% NO2. At concentration
V93.2 and V94.1 - 60 MW machines; and the V94.2 above 20-25 ppm, NO2 can be seen as a yellow-
- a 120 to 130 MW machine). Their new dry, low brown color, which causes public concern even
NOx combustor is similar to the earlier silo though the emission level may be in compliance
combustor, except that the single large with the local environmental emissions limit.
combustion nozzle has been replaced with a NOz formation is countered by controlling the
hybrid combustor composed of six smaller pre- fuel supply to some of the mini-burners, causing
mixed burners (mini-burners) and a single, more burning at fewer nozzles. Their measure of
center, diffusion pilot burner. merit for the combustor is the air coefficient,
lambda, which is the actual air to fuel ration
They use a small diffusion pilot burner to divided by the stoichiometric air to fuel ratio.
stabilize the combustor at very lean fuel air At lambda = 1 the maximum NOx is produced, and
mixtures and to increase the turn down range of a lambda average (averaged throughout the entire
the combustor. Figure 8( 16 ) shows an individual combustion region) of about 1.8 leads to flame
burner. Fuel is injected through a central out in a single premix flame. The data also
nozzle (a "conventional burner" concept), mixed indicate that at a lambda greater than about 2.5
with air injected at an angle to the fuel flow, CO emissions increase at exponential rates.
and a lean, premixed fuel is also injected in
a ring surrounding the conventional burner The burner performance has been measured in a
zone. During startup, the some of the mini- test rig (16 >. At part load operation the ratio
burners are operated essentially as diffusion of the premix burner fuel mass flow to the total
burners, e.g., no premix fuel addition, and the fuel mass flow is decreased in order to achieve
low NOx operation is initiated when a flame stability. The data shows that the NOx
predetermined exhaust gas temperature is and CO will increase at part load conditions.
reached. The figure also shows that the nozzle The test rig data of 30-50 mg/m 3 for NOx is
is equipped for steam or water injection to expected to increase to about 100 mg/m 3 in an
 E

actual operating gas turbine. The conversion Figure 9

from mg/m 3 to ppm is 2.0536 mg/m 3 = 1 ppm for KWU LOW NOX BURNER OPERATIONAL DATA
NOx, 1.25 mg/m 3 = 1 ppm for CO. The part load
performance is due to the interactions between
the numerous burners and the actual pressure
conditions in the silo combustor. Actual
operations should be constrained to an air fuel
mixture represented by a lambda of 1.8 to 2.2.
Flow control by variable geometry (a moveable
ring is used to control the available flow NO,

opening) can increase the operational range.

Downloaded from http://asmedigitalcollection.asme.org/GT/proceedings-pdf/GT1989/79153/V003T06A008/2398364/v003t06a008-89-gt-254.pdf by guest on 23 July 2021

Figure 9 presents data( 16 > from an actual
installation at the Simmering plant. This
plant has a turbine with a low firing
temperature of about 980 oC. Most dry, low NOx
combustors provide less NOx control as the
firing temperature increases. The newer KWU
machines (the V84 and 94 series) fire at about
1090 oC. The Simmering plant low NOx operation
occurs during power operations of 50-100%. If
two turbines are attached to a single steam secondary SIr wcon0.ry Si,
turbine, the effective operating range would
become 25-100%. 1 Dlnwlon f4nu
^Dm

In addition to showing the very low NOx and CO 1 1

emission performance attained at full load s

1
operation, the operational data of Figure 9 1
1
indicates another important item. In order to 1
1 1
operate the combustor at low emissions and at 1 b
high power levels, several air flow schedules ' ^ 1
are required. This is indicated by the 1 ^ ^
0 ^
different operating curves for minimum and Dnmi..E f4m^^
secondary air (as controlled by the moveable
ring) and the two different emission curves for 0 30 ^0 00
Po.n outpm
^0 % tf
......
premixed fuel firing. The data does not answer m1,m
the question of how capable the air and fuel M condaly IIr MC01141y .1r

control system is in load following, e.g., for OOZ F414 rrw•.unnynl eI V53.2 lln.,*anp
• IV Cry .,n.u.1 y...115% 07 by 701.
what rate of change of load can the system
perform as shown, and what impact this
complexity will have on plant reliability?
However, for steady state and slowly changing
loads, the performance appears to be REFERENCES
remarkable.
1. Rentz, O. "Overview of Recent Developments
KWU has retrofit the new burner in several in NOx Control in Europe," EPRI CS-5361, Vol.
plants, and Table 8 presents a listing of 2, August, 1987.
currently planned applications.
2. Ando, J., "NOx Abatement for Stationary
Sources in Japan," EPA 600/7-83-027, May, 1983.
Table 8
APPLICATIONS OF THE KWU IOW NOx COMBUSTOR 3. EPA Report "National Air Pollutant Emission
Estimates, 1940-86," January, 1988.
Turbine Power Utility Location Installed NO^c
Model MW Date rqq/d 4. Dacey, P., "An Overview of International NOx
Control Regulation," EPRI CS-4360, Vol. 1,
V93.2 60 Wiener Stadtwerke Austria Oct. 1986 150 January, 1986.

V93.0 60 VEW Germany March 1987 25 5. "1,090 MW Combined Cycle Power Plant for
Higashi-Niigata Thermal Power Station No. 3",
V93.0 60 VEW Germany 1987 100 undated publication by Tohoku Electric Power
Co., Inc. and also "Mitsubishi 1,090 MW Combined
V94.2 120 Wiener Stadtwerke Austria 1987 NA Cycle Power Plant for Higashi-Niigata Thermal
Power Station No. 3," Mitsubishi Heavy
V94.2 130 Eeaszenbrale Netherlands 1987/1988 NA Industries, Ltd., H400P06394, January, 1987.

V94.1 60 Stadtwerke Munchen Germany June 1988 150 6. Aoyama, K. et. al., "Development of a Dry
Low NOx Combustor for a 120-MW Gas Turbine,"
V93.0 60 VEW Germany 1988 100 Journal Engineering for Gas Turbines and Power,
Vol. 106, October, 1984.

7. Private material supplied by Tohoku Electric

Power Co., Inc.

8
 Li

8. Sato, T. and Hilt, M.B., "NOx Abatement

Measures for a Combined-Cycle Plant,"
Proceedings of the 1987 Tokyo International Gas
Turbine Congress, October, 1987.
9. Washam, R., "Dry Low NOx Combustion System
for Utility Gas Turbine," ASME paper 83-JPGC-
13, September, 1983.

10. Asano, M. et. al., "Low NOx Combustor

Experiences MS9001 100 MW Gas Turbine for

Downloaded from http://asmedigitalcollection.asme.org/GT/proceedings-pdf/GT1989/79153/V003T06A008/2398364/v003t06a008-89-gt-254.pdf by guest on 23 July 2021

Kawasaki Power Station, The Japanese National
Railways," undated paper from Hitachi, Inc.
11. Kuroda, M., et. al., "Development of Two-
Stage Dry Low NOx Combustion for Gas Turbines,"
presented at the 32nd International Gas Turbine
Conference, Anaheim, CA, June, 1987.

12. Marnet, C., et. al., "Reduction of NOx in

the Lauswaard Combined Cycle Plant," Brown
Boveri Review, March, 1985.

13. Jeffs, E., "New Low-NOx Combustors in

European Service," Gas Turbine World, Vol. 17,
No. 5, September-October, 1987.

14. Kassebohm, B., "NOx-Reduction of Combined

Power Plants By Low-NOx Gas Turbine
Combustors," presented at the 1987 Joint Power
Conference, Miami, FL, 1987.
15. Electric Power Research Institute Report,
NOx Reduction for Small Gas Turbine Power
Plants, EPRI AT-5347, September, 1987.

16. "Brown Boveri Gas Turbines," Brown Boveri

report CH-KW 1059875.

17. Becker, B., et. al., "Premixing Gas and Air

to Reduce NOx Emissions with Existing Proven
Gas Turbine Combustion Chambers," ASME paper
86-GT-157.