Next Generation Dry Low NOx for Gas Turbines:

Environmental and Regulatory Impact

Leslie Witherspoon and Luke Cowell
Solar Turbines Incorporated, 9330 Sky Park Court, MZ:SP3-Q, San Diego, CA 92123-5398

INTRODUCTION
Since the early 90’s, nearly every gas turbine manufacturer has introduced a low emissions gas turbine
product line based on lean-premix [or dry low NOx (DLN) or dry low emissions (DLE)] combustion.
NOx emission levels achievable by the “1st generation of lean-premix” generally ranged from 25 to
42 ppm NO x on gas fuel. To meet the need for lower NOx , gas turbine manufacturers are in various
stages of development work that will culminate in the commercial introduction of the next
generation of lean-premix products. The next generation of lean-premix, for several manufacturers,
includes the implementation of advanced combustor liner technology and improved fuel injection
technology. This paper briefly explores the next generation of lean-premix technologies as they
affect emissions and focuses on the regulatory impact of lower NOx emissions.

LEAN-PREMIX COMBUSTION TECHNOLOGY

In 1980’s, gas turbine manufacturers began major development efforts to integrate dry low NOx
combustion technology into their product lines. Advantages of lean-premixed combustion include
the concepts proven potential for low NO x emissions and general similarity of combustion system
hardware to that used in conventional gas turbines.

Lean-premix combustion reduces the conversion of atmospheric nitrogen to NO x by reducing the

combustor flame temperature. Since NOx formation rates are strongly dependent on flame
temperature, lowering flame temperature (by lean operation) is an extremely effective strategy for
reducing NOx emissions. Lean combustion is enhanced by premixing the fuel and combustor airflows
upstream of the combustor primary zone. Premixing prevents local “hot-spots” within the
combustor volume that can lead to significant NOx formation.

LEAN-PREMIX TECHNICAL CHALLENGES

There are three aspects of lean-premixed combustion that warrant attention:

• CO/NOx tradeoff
• Combustor operating range
• Combustor pressure oscillations

While the three issues have been discussed for many years in technical papers, it is important
to mention them again since they represent significant constraints in efforts to develop
advanced combustion systems that will further reduce gas turbine NOx and CO emissions.

1
CO/NOx Tradeoff
Since the flame temperature of a lean-premixed combustor is designed to be near the
lean flammability limit, lean-premixed combustor performance is characterized by a
CO/NOx tradeoff (Figure 1). At the combustor design point, both CO and NO x are
below target levels; however, deviations from the design point flame temperature
cause emissions to increase. A reduction in temperature tends to increase CO
emissions due to incomplete combustion; an increase in temperature will increase
NOx. The tradeoff must be addressed during part-load turbine operation. Additional
air control is included to prevent operating leaner. The tradeoff also comes into play
in development efforts to reduce lean-premixed combustor NOx emissions by further
reducing the primary zone design point temperature.

Figure 1. CO/NO x tradeoff illustration.

Desired
Operating
CO
Range
Emissions

NOx

Lean Rich
Fuel/Air Ratio

Combustor Operating Range

In a gas turbine, the lean-premixed CO/NOx tradeoff is manifested as a limited load
range over which emissions limits can be satisfied. As a gas turbine moves away from
full-load operation, a lean-premixed combustor would eventually produce excessive
CO emissions. To broaden the operating range, low emissions gas turbines use
variable geometry or compressor air bleed to maintain the combustor primary zone at
its optimum low emissions point despite load changes.

Combustor Pressure Oscillations

The introduction of lean-premixed combustion systems for gas turbines has raised
manufacturer awareness of the consequences of large combustor pressure oscillations.
Simply put, lean flames have a greater tendency to cause pressure oscillations that
can lead to engine damage. Despite increased awareness, however, manufacturers are
still working to develop design methodologies and combustion system features that
prevent excessive combustor pressure oscillations. Combustor oscillations are not
acceptable in field operation and passive solutions are developed empirically on an
engine model basis.

2
COMBUSTION SYSTEM MODIFICATIONS PREDICATED BY LEAN-PREMIX
TECHNOLOGY
Depending on the existing combustor design, development of the lean-premix combustion
system may have required modifications to gas turbine components including:

• Combustor liner
• Fuel injectors
• Variable geometry systems
• Engine casing
• Control system
• Fuel system

Some of the gas turbine components listed above present opportunity for further
enhancement that may affect the emissions profile of the turbine.

Combustor Liner
The lean-premixed combustor liner is generally similar to a conventional liner in
terms of geometry, materials and construction. The most significant difference in
the lean-premixed liner is that most turbines required increased combustor volume.
The larger volume was required to ensure complete combustion and low CO and UHC
emissions at the lower overall flame temperature of the lean-premixed combustor.
Since combustor length was constrained by the engine “exchangeability” objective,
the increased combustor volume was typically achieved by adjusting liner diameter.

Another difference in the lean-premixed liner is the absence of large air injection
ports in the combustor primary zone. Most air used in the combustion process is
introduced through the air swirlers of the fuel injectors. The remaining compressor
delivery air is used for cooling the walls or for dilution to achieve the specified radial
temperature profile and pattern factors at the combustor exit.

Many lean-premix combustors use louvers on the inside of the liner to direct air
axially along the walls to produce a protective film of cooling air between the wall
and the hot combustion gases. Film cooling involves the passage of cooling air
through holes in the liner and the formation of a cooling film on the hot side of the
liner using internally positioned louvers. The louver method of liner cooling is
commonly used in industrial and aircraft gas turbine combustors. The louver method
of wall cooling uses relatively high levels of cooling air.

Effusion cooling of the combustor walls has been developed for some commercially
available lean-premix gas turbines in order to reduce the cooling air required. In turn,
CO emissions are reduced. Effusion cooling is obtained by starting a film of air with a
cooling louver at the front of the combustor and then continuously feeding the film
with additional air through a multitude of small diameter holes laser drilled at a
shallow angle to the wall surface.

Backside-cooled liners have been in use on some commercial products for several
years. Backside-cooled liners forego cooling air injection completely. Instead,
combustor wall temperatures are controlled solely through convective cooling by a
high velocity airstream on the cold side of the liner (Figure 2). In most instances, the
high heat flux from the flame requires augmenting of the backside convective process
to keep liner wall temperatures from becoming excessive. Turbulators in the form of

3
trip strips, fins, and pins act to increase the cooling flow turbulence at the liner wall
and augment the heat removal process. Although effective in reducing CO formation
through quenching, backside cooling is a challenge to implement due to the high
flame temperatures and heat fluxes associated with gas turbine combustors.

Figure 2. Backside-cooled liner illustration.

Fuel Injectors
Lean-premix fuel injectors are significantly larger than their conventional
counterparts due to the higher airflow through the injector air swirlers and the
required volume of the premixing chamber used to mix the fuel and air. Both axial
and radial swirlers have been used to swirl the premix air. Fuel is injected either
through the swirler vanes or fuel spokes.

Variable Geometry Systems

Several variable geometry systems have been employed to avoid lean extinction and
broaden the low emissions operating range of lean-premixed combustion systems.
Each technique ultimately provides control of the primary zone airflow to maintain
the primary zone fuel/air ratio near its optimum low emissions level during part-load
engine operation.

Casing Bleed. Some two-shaft gas turbines used for gas compression and
mechanical drives, bleed air from the combustor casing at part load. The case
bleed method of variable geometry has proved effective in controlling the CO
emission while using the production bleed valve of conventional engines. A
consequence of air bleed, however, is deterioration in engine part-load thermal
efficiency since compressed bleed air no longer enters the turbine section of the
engine to produce power.

Inlet Guide Vanes. Some single-shaft gas turbines used for power generation,
maintain optimum primary zone fuel/air ratios by modulating the compressor
inlet guide vanes (IGV). Closing the IGVs reduces the airflow through the engine

4
 compressor and combustor. No bleeding of the high pressure air is required.

Combustor Staging. Some lean-premix turbines use fuel injection in multiple

stages axially along the combustor. Airflow to the additional stages is variable.

Injector Staging. Some lean-premix designs use numerous injector “heads” that
are fired as a function of load.

Engine Casings
New combustor casings may have been required for the lean-premix system due to the
increased diameter of the combustor liner and larger fuel injectors. The larger
combustor case also required modification to the mating compressor diffuser and gas
producer turbine cases.

Control System
The control system for lean-premix engines modulates the air management systems
to keep the combustion primary zone temperature within a specified range. Accurate
control of the primary zone temperature is critical to controlling NOx and CO
emissions; however, direct measurement of the primary zone temperature over an
extended period of time, which is greater than 1540°C (2800°F), is impractical.
Standard gas turbines use the power turbine inlet temperature as an indirect
measurement of the combustor exit or turbine inlet temperature. The initial release
of the lean-premix gas turbine also used turbine inlet temperature for control.

To control the lean-premix engine primary zone temperature (Tpz), the combustion
zone temperature is derived from a thermodynamic heat balance across the
combustion system. The parameters used in the calculation are the compressor
discharge temperature, the power turbine inlet temperature, the flow split between the
combustor primary zone air and the total combustor airflow, and the ratio of the
power turbine inlet temperature to the first-stage turbine inlet temperature.

Fuel System
The natural gas fuel system for lean-premix gas turbines can include multiple fuel
circuits. On some systems, separate fuel manifolds are used to supply pilot and main
gas to the respective fuel circuits of each fuel injector. During start-up and low load
operation, both fuel circuits are active. When the engine is in the low emissions
mode, a pilot fuel shutoff valve closes. A fixed percentage of the total fuel continues
to flow through the pilot circuit via an orifice in parallel with the pilot shutoff valve.
The fixed pilot flow is used to stabilize the flame.

Fuel staging used on some systems delivers fuel to controlled injector “heads” or fuel
injection ports or staged combustors on a predetermined load schedule.

ACHIEVABLE NOx LEVELS

Early lean-premix production units offered NOx warranty levels as low as 42 ppmv.
Improvements in the initial product offering lean-premix design allowed several
manufacturers to lower the NOx warranty level on some units to 25 ppm within a few years
of the initial offering.

5
NEXT GENERATION LEAN-PREMIX COMBUSTION
To remain competitive, and in response to the trend toward more stringent emissions regulations, gas
turbine manufacturers are assessing their current lean-premixed systems to establish viable system
enhancements. The areas that exhibit the greatest potential for lower emissions are optimizing the
air management system via advancing combustor liner and fuel injector technology, control system
improvements, and tighter quality control practices.

ADVANCES IN COMBUSTOR LINERS

Research has shown that the method used to cool a lean-premixed combustor liner can have a
significant effect on emissions. The current generation of lean-premixed combustors
available in the market today uses a variety of liner cooling methodologies. Most 1 st
generation lean-premixed turbines use film cooling to maintain acceptably low combustor
wall temperatures. Some 1 st generation lean-premixed turbines use augmented backside
cooling (ABC) technology.

Film cooling can lead to reaction quenching at the combustor primary zone wall. The
quenching process leads to high CO emissions because the CO, a combustion intermediate, is
prevented from oxidizing to CO2 . The quenching is traceable to the injection of a relatively
large flow of cooling air into the primary zone. Development of an advanced liner that does
not promote reaction quenching, such as the augmented backside cooling line, provides a two-
fold benefit in terms of emissions. First, of course, CO emissions will be reduced.
Additionally, the lower CO levels will allow combustor reoptimization to a lower flame
temperature. The reoptimization will produce lower NO x levels along with the lower CO
concentrations.

Many manufacturers will make the move to augmented backside cooled technology with their
next generation of lean-premix gas turbines. Those who already utilize the technology will
optimize its design in order to be able to warranty lower NOx levels. Other combustor liner
related technologies include ceramic combustor liners and thermal barrier coatings.

Ceramic Combustor Liners

A ceramic combustor addresses the CO quenching issue in the same manner as the
ABC liner. Cooling air injection through the liner is avoided, thus providing potential
emissions benefits. The emissions benefits have been found to be very similar to
those of the ABC combustor.

Thermal Barrier Coatings

An additional degree of liner protection can be achieved through the application of a
thermal barrier coating (TBC) on the hot sides of the liner walls. The TBCs are
frequently composed of zirconia-based materials that are plasma sprayed on the liner.

ADVANCED FUEL INJECTOR DEVELOPMENT

Fuel injectors are being optimized to reduce local “hot spots” while improving flame stability.
Achieving an optimum fuel/air temperature profile exiting the injector is key. Design
modifications to the fuel injection points and the air swirler are being investigated both
computationally and experimentally.

6
REGULATORY IMPACT OF LEAN-PREMIX NOX ADVANCEMENTS
Despite the significant improvement in gas turbine emissions over the last decade, regulatory
agencies continue to consider and implement more stringent emissions regulation. In the last five
years, mid-range gas turbine users have witnessed a large step-change in required emission levels from
gas turbines due to the application of add-on control technologies such as Selective Catalytic
Reduction (SCR) and EMx GT. While the add-on control requirements are focused in the power
generation market, compressor stations nationwide have seen increased pressure to evaluate add-on
technologies.

CURRENT BEST AVAILABLE CONTROL TECHNOLOGY (BACT)

The federal air permitting program, New Source Review (NSR) program, is divided into two
primary programs: Prevention of Significant Deterioration (PSD) and Non-Attainment NSR.
The PSD program applies to major sources (100 or 250 tpy NO x ) and major modifications
(40 tpy NO x ) in attainment areas [areas of the country that meet the National Ambient Air
Quality Standards (NAAQS)]. In areas that do not meet the NAAQS, the Non-Attainment
NSR program applies to major sources (25, 50 or 100 tpy NOx ) and major modifications (25
or 40 tpy NO x ).

Any major source or major modification subject to PSD must conduct an analysis to ensure
the application of BACT. The BACT requirement is defined as “… an emissions limitation
… based on the maximum degree of reduction for each pollutant … which the [permitting
authority] … on a case-by-case basis, taking into account energy, environmental, and
economic impacts … determines is achievable …” 40 CFR 52.51(b)(12).

The federal BACT for mid-range gas turbines is generally lean-premix technology. The level
most commonly associated with the BACT is 25 ppm NOx . Solar has consistently seen the
25 ppm NOx level permitted in both power generation and oil and gas applications in
attainment areas. Add-on control technologies such as SCR have yet to be deemed federal
BACT, in the mid-range turbine size class, due to their high cost. Table 1 illustrates the
varying cost effectiveness calculations for SCR applied to mid-range gas turbines.

Table 1. Estimated NOx removal costs for SCR technology.

(25 ppm NO x baseline)

Approximate Turbine Size

6,000 hp 15,000 hp 20,000 hp
Simple-Cycle
at 5 ppm NO x $28,000/ton $18,800/ton $17,100/ton
Cogen Cycle
with duct firing $20,100/ton $13,300/ton $12,300/ton
at 2.5 ppm NO x
While lean-premix technology at the 25 ppm NO x level usually represents BACT in PSD
permitting, it is becoming a greater challenge to demonstrate such since many states are
raising their cost-effectiveness thresholds. NOx cost-effectiveness thresholds range from
$5,000/ton NO x to more than $20,000/ton NOx . On simple-cycle (and non-duct fired
cogeneration) applications, it is fairly straightforward to demonstrate that SCR is not cost
effective. On the other hand, when cogeneration applications are duct fired, the add-on
control cost is spread over both the turbine and the duct burner NOx emissions, significantly

7
affecting the $/ton removal costs (refer to Table 1).

Most states do not have a state level BACT program. However, in some states that have
ozone non-attainment areas, state BACT programs, affecting projects that do not trigger
federal non-attainment NSR, have been developed. State BACT programs often represent a
significant air-permitting hurdle. Where states have BACT programs, BACT levels range
from 9 to 25 ppm NO x . States that have BACT programs typically have very high cost-
effectiveness thresholds. Even lower state BACT NOx levels (2 to 5 ppm) have occurred in
power generation applications in California and Massachusetts.

California has a state BACT program that requires very similar emission levels as the federal
Non-Attainment NSR program. The federal Lowest Achievable Emission Rate (LAER) is
described in the next section. Most of the 35 California air districts do not allow for cost
consideration in determining BACT. California BACT for gas turbines in compressor
applications is the turbine plus an SCR at 8 ppm NO x . However, several air districts, the
California Air Resources Board (CARB), and EPA Region 9 are pushing for a 5 ppm NO x
level; a level they say has been achieved in practice for simple-cycle applications.

CURRENT LOWEST AVAILABLE EMISSION RATE

For major sources or major modifications in areas of the country that do not meet the
NAAQS, Non-Attainment NSR is triggered, thus triggering a LAER analysis. From an end
user perspective, the unfortunate facet of LAER is not being able to consider costs when
evaluating technologies. For power generation applications, federal LAER is generally 2 to 3
ppm NOx depending on equipment configuration and averaging period. There are two
compressor stations operating gas turbines that represent LAER depending on how “other
environmental impacts” are considered. There is a compressor station with gas turbine plus
SCR in California permitted at the 8 ppm NO x level. There is a compressor station in
Massachusetts utilizing the next generation lean-premix technologies permitted at the 15
ppm NO x level. As DLN technology moves to <15 ppm NO x , the goal is to avoid the
requirement for add-on control and have next generation lean-premix represent LAER.

REGULATORY ADVANTAGES OF NEXT GENERATION LEAN-PREMIX

The primary benefits of a lower NO x warranty level is that many projects will be able to
avoid triggering PSD review, Non-Attainment NSR review, and, in some cases, Title V
permitting. At next generation lean-premix NOx levels, for many projects, the federal major
modification threshold or federal major source thresholds will not be triggered. The ability to
stay under the thresholds (without add-on control) and, therefore, avoid BACT/LAER review
is extremely advantageous.

8
 Figure 3 shows the approximate ton per year emission rate at 25 ppm NO x (most common
commercially available lean-premix warranty level today), 15 pm NOx (warranty level
available on limited commercial units), and 9 ppm NOx (level advertised by one manufacturer
on two models1 ). Figure 3 illustrates that as NOx levels drop, depending on the unit size and
geographic location (attainment status), lean-premix units may be able to avoid federal air
permitting review. In fact, for some smaller units, multiple unit projects may be able to stay
below the trigger thresholds.

Figure 3. Approximate tpy NOx levels for three NOx emission rates.

70

60

50
Tons Per Year

40 tpy trigger level 25 ppm

40
15 ppm
30 25 tpy trigger level 9 ppm
20

10

0
~6,000 hp ~15,000 hp ~20,000 hp
Approximate hp rating

When a state BACT program is triggered, lower NOx levels may avoid add-on controls
depending on how the state’s BACT program is structured. In many cases, the lower NO x
level associated with next generation lean-premix will be an advantage because the emissions
will stay below trigger thresholds and, therefore, will not trigger BACT review.

In the cases where a project triggers federal BACT (or state BACT), another benefit of a
lower lean-premix NOx emission level is evidenced in the BACT review. The lower NO x
levels achievable by the next generation of lean-premix should be instrumental in avoiding
add-on control requirements. The lower lean-premix baseline will result in even higher $/ton
NOx removed levels than at the 25 ppm NOx level. Table 2 illustrates the NOx cost-
effectiveness calculations for SCR applied to mid-range gas turbines using 15 ppm NOx as the
baseline.

1
Alstom advertises <10 ppm NOx for the Cyclone and Tempest products at full-load ISO conditions. Reference
Gas Turbine Users Association, 2000; Compressor Tech, May-June 2000.

9
 Table 2. Estimated NOx removal costs for SCR technology.
(15 ppm NO x baseline)

Approximate Turbine Size

6,000 hp 15,000 hp 20,000 hp

Simple-Cycle
at 5 ppm NO x $54,600/ton $36,600/ton $33,300/ton

Cogen Cycle
with duct firing $30,600/ton $20,000/ton $18,400/ton
at
2.5 ppm NOx

In California, the state BACT program is triggered by a lb/day NOx emission rate; therefore,
the lower NOx emission level will not avoid triggering an add-on control requirement.

REGULATORY DISADVANTAGES OF NEXT GENERATION LEAN-PREMIX

In today’s gas turbine air permitting arena, most everyone “presumptively” assumes that
BACT is 25 ppm NO x . While 25 ppm NO x is the most common emission level associated
with the lean-premix BACT technology, the level should not be “presumed” to represent
BACT for all gas turbine applications. BACT by definition is to be determined on a case-by-
case basis. Specific applications, operating profiles, fuel type and quality, and technology
generation can all affect the emission level achievable with BACT technology.

As the industry transitions to the next generation of lean-premix, there is a danger of a

“presumptive” BACT emission level for mid-range gas turbines at a NOx emission level below
what is widely commercially available. In accordance with guidelines, a user selects equipment
and goes through the permitting process for that equipment. If the equipment selected has
not yet incorporated, will not incorporate, or cannot incorporate next generation lean-
premix technology, then by BACT procedure guidelines that piece of equipment should be
evaluated on a case-by-case basis based on the equipment’s capabilities. There is concern that
regulators will presumptively assume that all lean-premix gas turbines are capable of
achieving the same emission level when a lower emission level may not be available on the
particular make and model being permitted. The regulator in effect may “specify” equipment
by stating a specific emission level. Regulators do not have the authority to “specify”
equipment.

POTENTIAL REGULATORY AFFECT OF CATALYTIC COMBUSTION

PRODUCTS
The limited availability of catalytic combustion on single shaft (generator drive) gas turbines
such as the Kawasaki M1A and the GE10B 2 (late ’02 availability) with a low NOx ppm
warranty may have a similar “presumptive” BACT effect. Regulators, environmentalists,
and potential end users will have to be educated on the technology, the applications, and the
limitations.

2
Merrill Lynch Global Securities Research & Economics Group Global Fundamental Equity Research Department,
Catalytica Energy Systems Inc., August 9, 2002.

10
 OUTPUT BASED STANDARDS
Effort is being made on various levels across the nation to develop new emission standards
for power generation (primarily focused on distributed generation) on an output basis. While
in theory output based standards are an excellent idea, they are geared to benefit/encourage
combined heat and power applications. Simple-cycle power generation applications will be
challenged over time to meet the standards without the use of add-on control. While the
standards developed to date have only been for the power generation market, users of gas
turbines in compressor/mechanical drive applications should pay close attention to the
development of such standards to ensure they do not crossover to the gas compression
market.

Table 3 contains a summary of three recent output based standard development efforts. The
emissions of simple-cycle gas turbines (6,000, 15,000, and 20,000 hp turbines) at 25 ppm, 15
ppm, and 9 ppm NOx are compared to the recent standards on a lb/MW-hr basis. As shown
in Table 3, simple-cycle gas turbines at 9 ppm can not meet the mid-term and long-term
standards without add-on control.

Table 3. Power generation output based NOx standards compared to simple-cycle gas turbine
emission rates in lb/MW-hr.

RAP*
East-Texas California (5/02 DRAFT)
Non-attainment
NOx Emission Standards (lb/MW-hr)
Phase I 0.47 (2003) 0.5 (2003) 0.6 (2004)
Phase II 0.14 (2005) -- 0.3 (2008)
Phase III None 0.07 (2008) 0.15 (2012)

Simple-Cycle Gas
Turbine @ 25 ppm NOx 1.1 lb/MW-hr
Simple-Cycle Gas
Turbine @ 15 ppm NOx 0.64 lb/MW-hr
Simple-Cycle Gas
Turbine @ 9 ppm NOx 0.38 lb/MW-hr
*Regulatory Assistance Project

SUMMARY

Manufacturers of mid-range gas turbines continue to improve lean-premix technology.

Simultaneously, regulators continue to push for lower and lower NOx emissions. The current
regulatory push is for add-on control on the back-end of an already very low emitting gas turbine. It
is fair to say that the gas pipeline industry sector prefers to keep add-on control out of the industry
sector due to the added cost for relatively small environmental benefit. The next generation of lean-
premix will hopefully help achieve this goal and at the very least be instrumental in delaying
widespread add-on control implementation.

11
Bibliography

Igoe, B., and McGurry, M., 2002, “Design, Development and Operational Experience of Alstom’s
13.4 MW Cyclone Gas Turbine,” ASME # GT-2002-30254, ASME Turbo Expo, Amsterdam, The
Netherlands.

Alstom Power, 2002, “Cyclone Status Report”, GTUA Conference 2001, Banff, Alberta Canada.

Cowell, L, and Smith, K., 2001, “SoLoNOx II, Low Emissions Systems,” Solar Turbines,
Incorporated, San Diego, California.

Bianchi, D., Noccioni, P., and Silvestri, C., 2000, “The New PGT5B/2, A State-of-the-Art 6 MW
Two-Shaft Gas Turbine,” ASME # 2000GT-0561, ASME Turbo Expo, Munich, Germany.

Goodman, D., Igoe, B., and Stocker, A., 1998, “Gas Turbine Technology for Today’s CHP Market”,
In-Power ’98 Conference, Wembley, London.

Rawlins, D., and Smith, K., 1998, “Developments in SoLoNOx Low Emissions Systems,” TTS082,
Turbomachinery Technology Seminar, San Diego, California.

U.S. EPA, Office of Air Quality and Planning Standards, Draft October 1990, “New Source Review
Workshop Manual,” Research Triangle Park, North Carolina.

12