EPA-460/3-74-006
January 1974
VARIABILITY
IN AIRCRAFT TURBINE ENGINE
EMISSION MEASUREMENTS
U.S. ENVIRONMENTAL PROTECTION AGENCY
Office of Air and Water Programs
Office of Mobile Source Air Pollution Control
Emission Control Technology Division
Ann Arbor, Michigan 48105
-------
EPA-460/3-74-006
VARIABILITY
IN AIRCRAFT TURBINE ENGINE
EMISSION MEASUREMENTS
Prepared by
Anthony F. Souza, Louis R. Reckner
Contract No. 68-01-0443
EPA Project Officer:
Dr. Robert E. Sampson
Prepared for
U.S. ENVIRONMENTAL PROTECTION AGENCY
Office of Air and Water Programs
Office of Mobile Source Air Pollution Control
Emission Control Technology Division
Ann Arbor, Michigan 48105
January 1974
-------
11
This report is issued by the Environmental Protection Agency to report
technical data of interest to a limited number of readers. Copies are
available free of charge to Federal employees, current contractors and
grantees, and nonprofit organizations - as supplies permit - from the
Air Pollution Technical Information Center, Environmental Protection
Agency, Research Triangle Park, North Carolina 27711, or from the
National Technical Information Service, 5285 Port Royal Road, Spring-
field, Virginia 22151.
This report was furnished to the Environmental Protection Agency by
Scott Research Laboratories, Inc., Plumsteadville, Pennsylvania, in
fulfillment of Contract No. 68-01-0443. The contents of this report are
reproduced herein as received from Scott Research Laboratories, Inc.
The opinions, findings, and conclusions expressed are those of the
author and not necessarily those of the Environmental Protection Agency.
Mention of company or product names is not to be considered as an endorse-
ment by the Environmental Protection Agency.
Publication No. EPA-460/3-74-006
-------
ill
TABLE OF CONTENTS
ABSTRACT
INTRODUCTION
Page
1.0 INTRODUCTION 1-1
2.0 SAMPLING AND ANALYSIS SYSTEM 2-1
3.0 SYSTEM EVALUATION 3-1
3.1 HYDROCARBON ANALYZER 3-1
3.2 CARBON MONOXIDE AND CARBON DIOXIDE ANALYZERS 3-3
3.3 NITROGEN OXIDES ANALYZER 3-5
3.4 MEASUREMENT SYSTEM ERROR 3-5
3.5 CALIBRATION GAS ACCURACY 3-6
4.0 SAMPLE LINE EVALUATION 4-1
4.1 EFFECT OF SAMPLE LINE TEMPERATURE 4-1
4.2 EFFECT OF SAMPLE LINE LENGTH 4-2
4.3 EFFECT OF SAMPLE FLOW RATE 4-6
4.4 EFFECT OF SAMPLE LINE MATERIAL 4-8
4.5 SMOKE SAMPLING 4-8
5.0 CORRELATION OF EMISSION MEASUREMENT INSTRUMENTATION 5-1
6.0 EVALUATION OF SAMPLING METHODS 6-1
6.1 DESIGN AND TESTING OF MIXING CHAMBER 6-1
6.2 TRAVERSING PROBE 6-3
6.3 MIXING PROBE EVALUATION 6-12
6.4 COMPARISON BETWEEN MIXING PROBES AND
TRAVERSE AVERAGES 6-16
6.5 PREDICTIONS FROM TRAVERSE DATA 6-18
6.6 TRAVERSE AVERAGES VERSUS FAA PROBE READINGS 6-22
6.7 CARBON BALANCE CALCULATION OF EXPECTED CARBON
DIOXIDE CONCENTRATIONS 6-24
7.0 BASELINE MEASUREMENTS 7-1
7.1 RUN-TO-RUN VARIATION 7-1
7.2 EFFECT OF APPROACH-TO-POWER 7-3
7.3 EFFECT OF VARIATIONS IN ENGINE THRUST AND FUEL FLOW
8.0 AMBIENT CONDITIONS 8-1
9.0 SUMMARY AND CONCLUSIONS 9-1
APPENDIX A' INSTRUMENT ERROR ANALYSIS A-l
APPENDIX B ENGINE EMISSIONS DATA B-l
APPENDIX C ENGINE OPERATING DATA C-l
-------
iv
ABSTRACT
Under Environmental Protection Agency Contract Number 68-01-
0443, Scott Research Laboratories, Inc., studied the causes of variability
in turbine engine emission measurements. A state-of-the-art analysis system
was built according to the specifications of the contract. The analysis
system was evaluated for reliability in the handling and accuracy in the
measurement of emissions.
Using the special analysis system, the variability in the gas
turbine emission measurements caused by the gas sample collection technique
was studied using a Pratt and Whitney JT8D engine. An exhaust gas mixer
and a detailed exhaust gas cross-section mapping technique were utilized
for the verification of average exhaust emission concentrations.
The effects of approach to power setting and the effects of varia-
tions in fuel flow and thrust on the measurement of mass emission rates are
discussed. The effects of ambient temperature and humidity on turbine engine
emissions are examined.
-------
FORWARD
This report was prepared by Scott Research Laboratories,
Inc., under Environmental Protection Agency Contract Number 68-01-0443.
The work reported herein was adminstered under the direction of the Emission
Control Technology Division, Office of Air and Water, Environmental Pro-
tection Agency with Dr. Robert E. Sampson serving as Project Officer. This
report covers work performed from June 1972 through July 1973. The authors
of this report are Anthony F. Souza and Louis R. Reckner.
The program described in this report was performed with the
cooperation of the Federal Aviation Agency at the National Aviation Facilities
Experimental Center, Atlantic City, New Jersey, and we wish to acknowledge
the special efforts of Mr. W. T. Westfield and Mr. G. Slusher of the F.A.A.
The engine used on this program was supplied to the F.A.A. by Pratt and
Whitney Aircraft.
-------
1-1
1.0 INTRODUCTION
This report describes the work performed by Scott Research
Laboratories, Inc. under Environmental Protection Agency Contract No.
68-01-0443. The major objective of the program was the determination of
the causes of variability in the measurement of aircraft turbine engine
emissions.
A state-of-the-art analysis system was designed and built according
to the specifications of the contract. The analysis system was evaluated for
reliability in the handling and accuracy in the measurement of emissions.
Difficulties encountered are documented and changes or compromises required
to produce a reliable system are discussed.
Using the special analysis system, the variability in gas turbine
engine emission measurements caused by the exhaust sample collection technique
was studied using a Pratt and Whitney JT8D gas turbine engine. An exhaust
gas mixing technique and a detailed exhaust gas cross section mapping
technique were utilized for the verification of average exhaust emission
concentrations. The virtues of these techniques are discussed.
The variability in exhaust emission measurements produced by the
direction of approach to a power setting and the effect of small variations
in thrust and fuel flow on the measurement of mass emission rates are
determined and compared to the actual run-to-run variations. All emission
data collected are examined for the effect of ambient temperature and
humidity on emissions from a gas turbine engine.
-------
2-1
2.0 SAMPLING AND ANALYSIS SYSTEM
A complete sampling and analysis system was constructed according
to the specifications in Attachment A-l of the contract. This attachment
is the test procedure published in the Federal Register of December 12,
1972 as the proposed aircraft turbine engine test procedure. Figure 2-1
illustrates the specified system. Figure 2-2 illustrates the details of
the CO, CO., and NO leg of the system as built by Scott. Any deviations
£ X
from Figure 2-1 are due to the requirements of the specific analyzers chosen
or modifications necessary in the original design made to improve sample
flow and sample integrity. The system features a heated sample train.
All lines carrying exhaust sample are heated from the probe to the analyzers.
The temperature of the sample gas is monitored at the analyzer exits in order
to be certain that a heated sample is passing through the analyzers.
Previous emission sampling systems have used various techniques
for the removal of water vapor from the exhaust sample. The water vapor,
chiefly a product of combustion but also that present in the ambient air
aspired by a fossil fueled engine, has been a problem both as optical inter-
ference to nondispersive infrared analyzers (NDIR), and as condensation in
sample lines whose temperature is below the dew point of the moist exhaust
gas sample. In previous exhaust analysis systems, the gas sample was dried
by refrigeration of the sample gas to condense out water vapor. Non-after-
burning gas turbine engines operate with considerable excess air. This
yields an exhaust gas concentration of water vapor which can easily be held
below saturation in moderately heated sample lines. At turbine exhaust
water concentration levels, the water vapor interference to carbon monoxide
NDIR analyzers can be controlled with optical filters. No refrigerators or
drying columns are used in the present system. The sampling system delivers
a hot moist sample to the analyzers. All measurements are made "wet," that
is, in concentrations based on actual values. In previous analysis systems
where dryers were used, the "dry" concentrations measured had to be converted
to "wet" values by calculating the proper dilution factor.
This sampling system makes use of a thermal converter in conjunc-
tion with a nitric oxide analyzer operating on the chemiluminescence principle.
This replaces troublesome nondispersive ultraviolet analyzers for NO- and
NDIR analyzers for NO which were previously used. The NO NDIR is quite
-------
Span
CO
Zero •--—•;
.5 liters/min.
Sample
\ Probe
Exhaust
Discharge j
Nozzle
»j,
Heated FID
160 + 5°C
/
Recorder
Pump
Filter
Zero '-5fi>- Span
J L
__ Sampling line
" heated to 150+5°C
Measure gas temperature at these outlet points
Minimum of 50°C maintained within + 2°C
Unrestricted outlet
Flow rate 2 liters/min.
nominal
Flowmeter
Unrestricted
outlet
Span
A\ Bypass
=^v,-/^-^i; Valve
* ' '&+—
Converter —
.owmeter
10
NJ
Recorder
-> Dump 20 liters/min.
Pump
Downstream of the heated to 65 +_ 5°C
FIGURE 2-1 SAMPLING SYSTEM AND INSTRUMENT ARRANGEMENT
To smoke measurement
system
-------
Q- &<|*
t
Z
3
)
NO, I»UET
Co
CO Ht-uo^«
foo, SouEuoiD
(00, COIQO
CO \taV«
CO £
MO, /MSt tt.0 MtUT 1 10 l_VT
U)
FLOW
- M^TEO
EJWUST
FIGURE 2-2
-------
2-4
sensitive to water vapor and requires additional dehydration treatment in
addition to the refrigeration drying technique acceptable for carbon
monoxide and carbon dioxide NDIR analyzers
The flow schematic in Figure 2-2 can be seen to contain 3 separate
legs with an analyzer in each leg: CO, C02 and N0-N0x- The total hydro-
carbon analyzer is mounted in a separate cabinet. The sample is driven by
a single pump with an electrically heated head. Additional pumps in the
heated hydrocarbon analyzer and chemiluminescence NO analyzer are used to
bypass sample around those instruments where only a small sample flows through
the detectors. Separate solenoid actuated valves are used in each leg to
admit zero, span and sample gases to the analyzers. A heated sample line
connects the heated hydrocarbon analyzer to the main heated sample line at
a tee. Flowmeters are used at the outlet of each analyzer to permit the
operator to set and maintain flow rates.
Departures and additions to the flow schematic of Attachment A-l (Figure
2-1) are summarized below.
1. The CO and C0? analyzers were placed in parallel legs rather than
series connected as specified to allow individual calibration.
2. One pump with a heated head was used to drive sample to the three
analyzers in parallel legs making independent flow adjustments
possible in each leg.
3. Three flowmeters were used instead of the two shown.
4. One heated prefilter was used to coincide with the single pump
mentioned in 2 above.
5. Zero and span gases were admitted to the hydrocarbon analyzer just
upstream of the detector. This is a built-in feature of the
instrument selected for use on this program. Admission of the
span gases directly to the detector provides speed of response
which is desirable in routine testing programs. However, injection
of zero and span gases at the probe end of the sample line offers
a check on the system heated sample line and prefilter condition
which is preferred. In the exhaust emission tests described later,
-------
2-5
gases were admitted to the system inlet at the sample probe as a
check on total system performance.
6. Flow control valves are used in each leg to adjust sample flow rates.
7. Flow control valves are used in each span and zero line to match flow
of zero and span gases to the sample flow rates.
8. A bypass vent with a pressure regulator set at 5 psi was added
downstream of the pump to act as a pressure relief. This protects
the analyzers from high pressure surges.
In the system, sample lines are heated by passing an electric
current through the stainless steel sample line tubing. The tubing is
thermally insulated to decrease the electrical power required and to
provide thermal stability. Electrical isolation from the analyzers was
provided by short lengths of Teflon tubing at the necessary points. The
solenoid valves which provide push button calibration were mounted on
heated aluminum blocks. Also, the system pump in contact with sample gas
had a heated head. Increased insulation around the solenoid valves and
around the sample lines inside the NDIR analyzers was necessary to obtain a
sample gas temperature of 50°C at the analyzer exits. A total of 17 thermo-
couples were used to insure that the sample line temperature was not
excessively hot at any point.
Instruments selected for use in the system included Beckman Model
315 nondispersive infrared analyzers for carbon monoxide and carbon dioxide.
The NO-NO analyzer is a Scott Model 125 chemiluminescence analyzer system
X
which includes a thermal converter for NO . The total hydrocarbon analyzer
is a Scott Model 215B heated hydrocarbon analyzer. It detects hydrocarbons
by a flame ionization technique and incorporates a heated detector and
pump assembly and a ten foot long heated sample line and heated prefilter
assembly.
The heated hydrocarbon analyzer is sensitive to sample inlet
pressure. It cannot draw a sample from a sample line whose Internal
pressure is below 750 mmHg. This translates in the case of a system
configured like Figure 2-1 into a maximum line length from the probe to
the heated hydrocarbon analyzer inlet of less than 40 feet and preferably
less than 30 feet. For longer lines a pump must be added to the inlet line.
-------
2-6
Eighty feet of heated 3/8-inch stainless steel sample line was
fabricated. Flow tests indicated that the sample line could be easily
resistance heated to 150°C using one 12.8 volt, 100 ampere transformer for
each 40 foot segment. The sample line is connected as a load resistor
across the secondary of the transformer and the primary voltage is controlled
by an autotransformer. The sample lines are insulated with Fibrefax and
encased in a vinyl zipper tube jacket.
-------
3-1
3.0 SYSTEM EVALUATION
The system described in the previous section was thoroughly
evaluated using both gas mixtures and turbine exhaust. The tests with gas
mixtures were performed primarily in Scott's Plumsteadville laboratory
while the tests utilizing turbine exhaust were conducted at the NAFEC
facility near Pomona, New Jersey. This section deals specifically with
the evaluation of the instrument system. The sampling system is discussed
in Section 4.0.
The first step in system evaluation is a check for leaks in the
system and the elimination of any that may be found. Since part of the
system operates below ambient pressure, leaks will cause sample dilution
and erroneous data. The leak test was performed by isolating the vacuum
side of the system, evacuating this portion of the system with a vacuum
pump, and monitoring on a vacuum gauge the time taken by the system to lose
vacuum after the pump was shut off and a valve in the pump line closed.
Points on the pressure side of the system were checked by pressurizing
the system and checking for leaks at the joints with Snoop leak detector.
Next, a careful evaluation of each instrument and its sub-system
was conducted in order to determine whether the specifications were met.
3.1 HYDROCARBON ANALYZER
The hydrocarbon concentration in turbine exhaust at some operating
conditions is less than 0.3 ppm carbon. Therefore, it is necessary that
the hydrocarbon analysis system have very low background. The zero gas
specified in Attachment A-l is air containing less than 2 ppm-C hydrocarbon.
This is clearly inadequate for use in measuring turbine engine exhaust
concentrations of 0.3 ppm-C. Air certified to contain less than 0.1 ppm-C
was used as a hydrocarbon zero gas in these tests.
Since the hydrocarbon levels in gas turbine exhaust are very low,
the sampling lines, prefliters, valves and other components in contact with
the sample before it reaches the hydrocarbon detector must not remove
hydrocarbons from the sampler or release incidental hydrocarbons. In a
new system the hydrocarbons which are present after manufacture of the
components must be driven off by operation of the system at elevated tempera-
ture for several hours. The background level is then checked periodically
-------
3-2
by flowing hydrocarbon free air alternately directly into the analyzer and
then through the complete sampling system. When the concentration measured
reaches zero the sampling system is ready for operation.
Initially, the hydrocarbon analyzer was optimized for fuel and
air flow rates and adjusted for minimum oxygen response. The oxygen response
was less than 0.5%. Low oxygen response is not as important in non-after-
burning turbine work as it is in diesel or automotive emission testing.
Non-afterburning turbine engines operate with up to 200% excess air and
the oxygen in the exhuast is never less than about 16%.
A comparison of instrument response to different classes of hydro-
carbon was made using propylene, toluene and n-hexane standards containing
approximately 35 ppm-C in nitrogen. The following table illustrates the
relative response obtained with the three test gases. Also included is the
response to the propane span gas and a second propane standard as a double
check on the instrument span adjustment.
TABLE 3-1 RELATIVE RESPONSE OF HYDROCARBON INSTRUMENT
TO SEVERAL HYDROCARBONS (BEFORE ADJUSTMENT)
FID
Carbon Cyl. Cone. Re:
Gas
Carbon
No.
3
3
3
7
6
Cyl. Cone.
ppm-C
147.3
15.24
41.7
31.1
36.4
Response
ppm-C
147.3
15.04
37.3
28.2
32.5
Relative
Response %
100
98.7
89.5
90.7
• 89.5
Propane in Air
Propane in NZ
Propylene
Toluene
n-Hexane
Average response to the three gases, propylene, toluene and
n-hexane, was 89.9% and the maximum deviation from the average response
was within 1.3%. Attachment A-l specified that the response to the
three gases should agree within 5%. This criterion was met. However,
the response to these three gases was 10% less than the response to
propane, indicating a problem of nonlinear!ty with carbon number. The
instrument responded well to the two propane standards, one in air and
the other in nitrogen, indicating very little oxygen response. The test
specification is deficient in this respect since the hydrocarbon
-------
3-3
instrument is spanned with propane and must respond linearly to the hydro-
carbons of higher carbon number in gas turbine exhaust.
Further investigation revealed that a small adjustment in the
air flow rate resolved this nonlinearity. Improvement was also noted in
the oxygen effect and at less than 0.4%, it was well within the allowable
limits. The following table illustrates the relative response obtained for
the three classes of hydrocarbons tested with the new air flow rate. Also
included are the two propane standards used to check the oxygen effect
during this test.
TABLE 3-2 RELATIVE RESPONSE OF HYDROCARBON INSTRUMENT
TO SEVERAL HYDROCARBONS (AFTER ADJUSTMENT)
Gas
Carbon
No.
3
3
3
7
6
Cyl. Cone.
ppm-C
8.82
15.24
41 .7
31.1
36.4
FID
Response
ppm-C
8.82
15.18
41.4
30.6
35.9
Relative
Response %
100
99.6
99.2
98.5
98.7
Propane in Air
Propane in N.
Propylene
Toluene
n-Hexane
The average response for the three gases was 98.8% and the maximum
deviation of 0.4% was well within the 5% criteria set in Attachment A-l.
3.2 CARBON MONOXIDE AND CARBON DIOXIDE ANALYZERS
The specifications for the NDIR analyzers for carbon monoxide and
carbon dioxide are as follows:
° Response time (electrical) - 90% of full scale in 0.5 seconds
or less.
o Zero drift - less than ±1% of full scale in 2 hours on the
most sensitve range.
o Span drift - less than ±1% of full scale in 2 hours on most
sensitive range.
o Repeatability - ±1% of full scale.
o Noise - Less than 1% of full scale on most sensitive range.
o Cell temperature - minimum 50°C maintained within ±2°C.
-------
3-4
Specified ranges and accuracies for the NDIR instruments are as follows for
non-afterburning engines:
Range Accuracy
Carbon monoxide 0-100 ppm ±2% of full scale
0-500 ppm ±1% of full scale
Carbon dioxide 0-2% ±1% of full scale
0-5% ±1% of full scale
Of all these specifications the most difficult to meet was the zero
drift limitation on the carbon monoxide analyzer. The carbon monoxide
analyzer is a Beckman Model 315 Analyzer with 13*s inch cells. In order to
obtain low noise performance, a solid state chopper assembly was substituted
for the original electromechanical assembly. Operation with less than ±1%
of full scale zero drift was possible in the laboratory where operation
of the instrument was on dry calibration gases and room air. In test cell
operation with engine exhaust, the zero drift exceeded the specification.
Zero drift appeared to be caused by both cyclic heating effects and soiling
of the cell windows by fine particulate which had passed through the
particulate filter. Coarse adjustments were made to the cell shutters to
keep the analyzer and reference cells in balance until no further adjustment
was possible. The cells were then removed and cleaned, restoring them to.
operation. The cells required cleaning after about 45 test runs. The cyclic
drift was approximately ±1% of full scale. A modification has been developed
recently at Scott to eliminate the cyclic drift. This eliminated any
problem due to zero drift.
The extent of water vapor and carbon dioxide interference on the
carbon monoxide analyzer was determined by passing a range of known concen-
trations of carbon dioxide and water vapor through the instrument and
observing the response. Water vapor present in concentrations found in gas
turbine exhaust can produce errors up to a few parts per million if not
taken into consideration. Carbon dioxide interference is slightly less.
Table 3- 3 lists the corrections to be applied to the carbon monoxide
readings obtained with the Scott analysis system.
-------
3-5
TABLE 3-3 CORRECTIONS TO CO ANALYZER FOR WATER VAPOR AND C02 INTERFERENCE
Exhaust C02 co Exhaust Water CO
Concentration Correction Concentration Correction
(%) (ppm) (%) (ppm)
0.0 0.0 0.0 0.0
1.0 1.0 1-0 1.5
2.0 2.0 2.0 2.5
3.0 3.0 3.0 3.5
4.0 3.5 4.0 4.0
5.0 4.0 5.0 4.5
6.0 4.5 6.0 4.7
7.0 5.0
3.3 NITROGEN OXIDES ANALYZER
No difficulty was experienced meeting the specification for the
NO analyzer. The procedure to check the thermal converter can be improved
A
by including the following points. Use an inert plastic bag such as Tedlar.
Admit the NO in nitrogen to the bag first, the air second. Evacuate
the bag before starting the test to remove leftover nitrogen dioxide.
An electrical timer of the type in which a small electric motor drives
a cam operated switch many be used to control two solenoid operated valves
which switch the sample flow either through or bypassing the converter.
The timer should be adjusted for one minute intervals. This device produces
even intervals and frees the instrument operator for other duties while
performing the converter check. A convenient daily check is one in which
an NO 2 in air standard is passed through the converter. The resulting
response should be consistent from day to day.
3.4 MEASUREMENT SYSTEM ERROR
The actual accuracy of error of a measuring system is a combination
of the individual accuracies or errors of the parts of the system. In a
typical NDIR system the major components which influence the accuracy are
the instrument accuracy of the NDIR analyzer, span and zero gas accuracy,
-------
3-6
and the analog recorder accuracy. The instrument accuracy is complicated
by the addition of zero and span drift errors to the basic comparative
accuracy or repeatability. A study was done on each of the analysis systems
to predict the overall system accuracy. This study is presented in
Appendix I.
3.5 CALIBRATION GAS ACCURACY
The magnitude of analysis errors resulting from inaccurately
labeled calibration gases can be estimated for a cross section of laboratories
by examining the data from Scott's Cross Reference Services. In these
services, essentially identical cylinder mixtures are submitted to participating
laboratories as unknowns every three months. The laboratories report their
analyses to Scott, and the data are tabulated and analyzed. The two
services which present mixtures typical of turbine exhaust are the diesel
service and the nitric oxide service. A summary of the data for a recent
group of cylinders is shown in Tables 3-4 and 3-5.
The standard deviation of the hydrocarbon data is seen to be
nearly 50% of the mean value. However, by excluding four outlying obser-
vations the standard deviation is reduced to 15.5 or 5% of the mean.
Similarly, the carbon monoxide and carbon dioxide data show standard
deviations of 7% and 9% of the mean values, respectively. Removal of one
outlier from each group reduces the deviations to 5% and 4%. Since all
of the cylinders gave identical instrument responses before shipment to
the participants, the data provides a good measure of error due to
instrument calibration.
The nitric oxide data show standard deviations of approximately
13% of the mean value for both NDIR and chemiluminescence measurements.
Removal of outliers still leaves the deviations at approximately 10%.
Allowing for a 1 to 2% deviation in the nitric oxide cylinders, the devia-
tions are still greater than for other components. The close agreement
between nitric oxide data from the NDIR and CL analyses indicates that the
errors are due to standards rather than instrument type. It must also be
noted that while the participants in the cross reference services represent
-------
3-7
TABLE 3-4 DIESEL CROSS REFERENCE SERVICE DATA
Hydrocarbons as
Methane (ppm)
by FID
Average (no. data points) 323 (21)
Median 327
Range 15-875
Estimate of Standard Deviation ±150.73
Carbon Monoxide Carbon Dioxide
(PPm) (Vol. %)
by NDIR by NDIR
Average (no. data points) 498.5 (24) 3.44 (22)
Median 499.5 3.39
Range 400-545 3.1-4.66
Estimate of Standard Deviation ±33.29 0.3
TABLE 3-5 NITRIC OXIDE CROSS REFERENCE SERVICE DATA
Nitric Oxide Nitric Oxide
(ppm) (ppm)
by NDIR by CL
Average (no. data points) 70.15 (28) 70.03 (23)
Median 69.00 68.00
Range 56-100 56-98
Estimate of Standard Deviation ±9.14 ±9.09
-------
3-8
a good sampling of laboratories in this country and abroad, they all
participated in a similar cross check three months earlier. Thus, the
data would be expected to show less deviation than if the laboratories had
never participated in a cross reference program.
-------
4-1
4.0 SAMPLE LINE EVALUATION
4.1 EFFECT OF SAMPLE LINE TEMPERATURE
The sample line must transport the exhaust from the probe to the
instruments without affecting the concentration of any pollutant being
measured. Sample lines up to 125 feet in length are in use at various
turbine engine test stands in the country. Sample line temperatures can
influence emission measurements in several ways. "Cold" sample lines can
cause the water of combustion contained in the exhaust gas sample to con-
dense. Liquid water absorbs nitrogen dioxide yielding a low reading for
that constituent. "Cold" sample lines will also condense hydrocarbons with
high boiling points. The hydrocarbon gases present in turbine exhaust are
mostly cracked products at high power levels (high temperatures) but at
the engine idle condition high boilers may be present which will condense
in a "cold" line. The typical result when hydrocarbons deposit inside a
sample line at idle, is that low readings are observed at idle. At high
power settings where the hydrocarbon level is substantially lower, the
hydrocarbon trace will slowly decrease, indicating that hydrocarbon deposited
on the sample line wall is slowly evaporating into the sample stream.
To prevent these problems, sample lines must be heated. The minimum
temperature to prevent water condensation is 50 C. The minimum temperature
to prevent hydrocarbon condensation is higher and becomes the minimum
allowable temperature. From experience, this temperature is about 150 C.
The sampling line was evaluated both in the laboratory and in
the test cell. The heated 3/8 inch stainless steel sample lines were test-
ed in the laboratory in lengths up to 80 feet. Approximately 600 watts is
required to maintain the sample line skin temperature at 150°C while
flowing room air at 36 liters per minute. This flow rate is the flow
required by the analysis system plus a bypass flow of 20 liters per
minute. Laboratory evaluations were performed using propane in nitrogen
standards as well as a jet fuel in nitrogen mixture. This mixture was
obtained by adding a few drops of JP-5 fuel to a cylinder of dry nitrogen.
The resulting gas had a measured hydrocarbon concentration of approximately
400 ppm-C. This mixture simulated the hydrocarbon emissions from a gas
turbine engine at idle power setting.
-------
4-2
No difference in reading was noted when propane was first admitted
directly to the.hydrocarbon analyzer and then passed through the 80 foot
heated sample line. This was repeated over a range of sample line temper-
atures centered on 150°C. The JP-5 in nitrogen mixture, however, suffered
losses in the sample line at all temperatures tried. The loss was minimum
at 150°C sample line temperature. At sample line temperatures of 200 C
and room temperature (25°C) approximately 15% of the hydrocarbons did not
survive the 80 foot of sample line. Some losses also occured in the
hydrocarbon analyzer prefilter where the concentration was reduced approximately
7% compared to direct injection into the hydrocarbon analyzer. When the
sample line was reduced to 40 feet in length, the hydrocarbon losses were
nearly the same as those experienced with the 80 foot line. It may be that
the hydrocarbons were lost in the first few feet of sample line which was
cooled by the gases being admitted from the relatively cool cylinder.
The test cell evaluations were performed using the configuration
shown in Figure 4-1. In the test cell evaluations gas turbine exhaust
was sampled using a system of four probes mounted inside a JT8D exhaust
nozzle. The probe configurations are illustrated in Figure 4-2.
This probe was designed by NAFEC personnel so that all sampling holes
would be within the projection of the fan air/core airbboundary. 'The
probes consist of four closed end tubes with three sample inlet holes in
each tube on circles concentric to the engine center line. The hole
spacing on each probe is different. The probes are oriented along nozzle
chord lines. The probes can be sampled individually or can be manifolded
together through a system of solenoid operated valves.
The hydrocarbon loss tests were repeated using exhaust gases from
a JT8D-11 engine. The results are shown in Table 4-1. No significant
difference in hydrocarbon concentration was noted over sample line tempera-
tures ranging from 150°F to 350°F, and sample line lengths from 20 to 60 feet.
4.2 EFFECT OF SAMPLE LINE LENGTH
The effect of sample line length was studied using a valving arrange-
ment in which the sample line length could be changed from 80 feet to 40 feet
total length. The sample line length between the heated hydrocarbon analyzer
-------
— HEATED SAMPLE LINE
SAMPLE PROBE
CONNECTOR RING
JT8D
NOZZLE
r-
1
I
i
' RECO
T * '
i
' HEATED
j HYDROS
1 ANAL
I
'Z:— TO FAA SYSTEM
RDER J RECORDER
TOTAL
ARBON C0» C°2 «• NOX
yZER ANALYZER CONSOLE
; A <\
I (CHEATED FILTER
"CHEATED FIL
t
i -
TER
_ ta SMOKE
METER
FLOWMETER
20 L/MIN. DUMP
PUMP
FIGURE 4-1 JT8D TURBIBE SAMPLING SYSTEM
-------
4-4
030" Diameter
ID
Looking Upstream
FIGURE 4-2 NAFEC PROBE AND HOLE PATTERN
-------
4-5
TABLE 4-1 EFFECT OF SAMPLE LINE TEMPERATURE
JT8D-11 Emission Measurements Made
Sample Line Bypass
Flow Rate 20
Tests made with up-stream sample
Engine at Idle
Line
Length
(ft)
80*
SO
80
80
80
80
80
40**
40
40
40
40
40
40
Line
Temp
(°F)
100
150
200
250
300
350
400
100
150
200
250
300
350
400
THC
(ppm-C)
165.0
171.0
169.5
169.5
168.0
168.9
166.5
172.5
172.5
171.0
171.0
171.0
171.0
171.0
CO
(ppm)
362
347
337
337
329
329
327
340
362
355
347
362
355
370
With
1/m.
pump.
CO.
!%L
1.48
1.48
1.48
1.48
1.48
1.46
1.48
1.48
1.48
1.48
1.48
1.48
1.48
1.46
NOX
(ppm)
12.1
12.1
12.1
12.1
12.1
11.9
11.7
12.1
12.1
12.1
12.1
12.1
12.1
12.1
* FID sample line - 60 ft.
** FID sample line - 20 ft.
-------
4-6
and the remaining instruments was 20 feet. This meant that the
sample line length from the probe to the hydrocarbon analyzer could be
changed from 60 feet to 20 feet. In order to perform this test, a
configuration change was made to the system. A pump was located just
downstream of the probe so as to pressurize the sample line at engine idle
and prevent problems with the hydrocarbon analyzer which will not operate
properly with sample line pressures of less than 750 mmHg Abs. At engine
power settings above idle, the pump is unnecessary and can be bypassed.
The results of these tests are shown in Table 4-2. No effect
could be determined on the exhaust concentration measurements due to sample
line length.
4.3 EFFECT OF SAMPLE FLOW RATE
The effect of sample flow rate was evaluated for lines of 20, 60
and 80 feet. Because the final 20 feet of sample line was between the
hydrocarbon analyzer and the main gas analyzer cabinet, the actual sample
line lengths from the exhaust sample probe to the hydrocarbon analyzer
were 20 and 60 feet. The maximum flow rate allowable with the test system
was determined by proper operation of the heated hydrocarbon detector which
cannot operate with sample line pressure below 750 mmHg Abs. The maximum
flow rate is also a function of sample line length since the pressure drop
through a sample line increases with the length of the sample line. For
short lines higher flow rates could be tolerated by the hydrocarbon detector.
With the engine at idle power setting, the maximum line length
possible when using 20 liters per minute bypass flow was 20 feet from the
sample probe to the hydrocarbon analyzer. At higher flow rates the sample
line pressure was too low to allow normal hydrocarbon analyzer operation.
A boost pump was added at the probe end of the sample line. It was a Teflon
lined diaphragm type pump. Using this boost pump line lengths up to 80 feet
were tested at bypass flow rates up to 40 liters per minute (53 liters per
minute total flow).
No effect was noted on measured exhaust levels over a range of
bypass rates from 10 to 40 liters per minute. Twenty liters per minute appears
to be a reasonable operating point with line lengths up to 60 feet. With
-------
4-7
TABLE 4-2 EFFECT OF SAMPLE LINE LENGTH
JT8D-11 Emission Measurements Made Through Heated
Stainless Steel Sample Line at 150°C and Bypass Rate 20 1/min
Sample
Line THC
Length (ppm-C)
80 167 -
40 173
80 168
40 168.
CO
(ppm)
327
327
329
329
C02
SSL.
1.48
1.48
1.48
1.48
NOX
(ppm)
11.7
11.9
11.9
12.1
-------
4-8
3/8" line lengths longer than 60 feet the maximum allowable transport time
of two seconds becomes the governing criteria and bypass flow rates greater
than 20 liters per minute are required.
Flow rate was varied from a maximum of 53 liters per minute total
flow to minimum flow rate to satisfy the analyzers only, which was 13 liters
per minute. Total flow rate was varied by changing the bypass flow rate.
All tests were performed at engine idle condition. Table 4-3 presents some
test results taken on two different days.
Hydrocarbon concentration remained unchanged with increasing flow
rate, up until instrument performance was hampered by excessively low sample
line pressure. The hydrocarbon response then decreased due to decreased
sample pressure within the hydrocarbon detector.
4.4 EFFECT OF SAMPLE LINE MATERIAL
The effect of sample line material was evaluated by comparing
two heated stainless steel lines, one of which was internally sleeved with
Teflon tubing. These sample lines were operated over a range of flow rates
and temperatures. Table 4-4 lists the results of these tests at different
engine power settings. No significant differences could be detected between
the measurements of hydrocarbons, carbon monoxide, carbon dioxide and nitrogen
oxides made with the two lines. Calibration gas standards were passed through
the two lines both before and after operation on gas turbine exhaust. No
differences were noted between the instrument response to the gases passed
through the lines and the response to the same gas injected directly into
the instruments.
4.5 SMOKE SAMPLING
Exhaust gas was transported to the smoke meter through the
stainless steel sample line during all tests where sampling was performed
on gas turbine exhaust. Some difficulty was experienced on engine test
runs 23 through 45 where the smoke numbers obtained did not correspond to
those numbers obtained using a separate smoke sampling line. Dilution in
the sample line was suspected, but the difficulty was traced to a problem
-------
4-9
TABLE 4-3 EFFECT OF SAMPLE LINE FLOW RATE FOR DIFFERENT LENGTHS
JT8D-11 Emission Measurements Made Through Heated
Stainless Steel Sample Line at 300°F
Engine at Idle
Bypass
Flow Rate
(1/min.)
10
20
30
40
10
20
30
40
THC
(ppm-C)
171.0
172.5
170.1
169.5
195.0
193.5
184.5
193.5
CO
(ppm)
40 Foot Length*
362
370
375
370
80 Foot Length**
385
385
382
382
CO,
uL
1.46
1.46
1.46
1.46
1.51
1.51
1.48
1.48
N0x
(PPm)
12.5
12.5
12.5
12.5
11.4
11.4
11.4
11.7
* Length from probe to hydrocarbon analyzer - 20 ft.
** Length from probe to hydrocarbon analyzer - 60 ft.
-------
TABLE 4-4 EFFECT OF SAMPLE LINE MATERIAL
JT8D-11 Emission Measurements Made Through Heated
Stainless Steel & Teflon Sample Lines
THC
Read.
No.
24
27
38
36
33
Power
Condition
Idle
Approach
Land
Cruise
Max. Cruise
(ppm-C)
Teflon
141.0
8.4
0.6
0.3
0.3
S. Steel
144.0
7.2
0.6
0.3
0.3
CO
(ppm)
Teflon
336.0
64.7
24.1
16.2
11.0
S. Steel
336.0
64.7
23.7
15.7
10.9
C02
(
Teflon
1.48
1.75
2.12
2.25
2.53
%)
S. Steel
1.48
1.79
2.12
2.30
2.60
NOX
(ppm)
Teflon
13.0
38.2
71.6
92.2
121.2
S. Steel
13.2
38.2
71.6
92.2
123.4
-------
4-11
within the smoke sampler itself. As a result of all the engine test runs
to date, a light film of soot lines the sample line. This soot lining
appeared to have no effect on the gas analysis being performed on the
exhaust sample. However, later in the program, an apparent effect of
soot lining was noted at low hydrocarbon concentrations during traverse
testing. The smoke meter flow requirements and operation appeared not
to be affected due to using the same sample lines as the gas analysis
equipment. When no smoke samples were being taken using the gas sample
line, a heated prefilter was added upstream of the sample line to keep
particulate out of the line.
In general, separation of the smoke and gas sampling lines at
the exit of the sampling probe is more convenient. This allows the use
of a filter to prevent particulate from reaching the gas analyzers
and minimum interaction in the flow requirements between the smoke and
gas analysis systems.
-------
5-1
5.0 CORRELATION OF EMISSION MEASUREMENT INSTRUMENTATION
The analysis system discussed in Sections 2, 3 and 4 was compared
to an analysis system previously assembled by NAFEC. The NAFEC system
consisted of Beckman NDIR analyzers for carbon dioxide and nitric oxide, a
Beckman NDUV analyzer for nitrogen dioxide, a Beckman heated hydrocarbon
detector, and an MSA NDIR for carbon monoxide. The system employed refriger-
ation drying of the sample before it was passed through the NDIR analyzers.
The nitrogen dioxide and hydrocarbon measurements were made wet. Indicating
Drierite was used as an additional drying agent upstream of the nitric
oxide analyzer. The NAFEC exhaust analysis system was operated by NAFEC
personnel in coincidence with the Scott system.
This provided two independent analyses of the same exhaust sample.
The two systems were used to compare measurements made on the JT8D gas
turbine engine exhaust emissions over a range of engine power settings.
Table 5-1 compares the data obtained during the first series of
engine tests. There are obvious differences in the data for hydrocarbons
and carbon monoxide. The exhaust analyses of hydrocarbons made by NAFEC
were higher than Scott's analyses at all conditions and the difference
increased with hydrocarbon level. An investigation showed that the differ-
ences between the systems were primarily caused by an inaccurately analyzed
standard gas mixture. The NAFEC hydrocarbon standard marked 70 ppm propane
contained only 50 ppm propane when analyzed against Scott standards. The
NAFEC system could not measure less than 1 ppm-C. The Scott system was
able to measure as low as 0.1 ppm-C. This still exceeded the low hydro-
carbon levels emitted by the JT8D at maximum continuous power.
NAFEC measurements of carbon monoxide were either higher or lower
than the Scott measurements depending on the concentration level. The
NEFEC measurements were lower at high levels and higher at low levels. The
disagreement was traced to a NAFEC standard labeled 71 ppm CO which was
found to contain 83 ppm.
-------
TABLE 5-1 JT8D TURBINE ENGINE EMISSIONS - WET
Lead.
No
2
4
5
8
11
21
22
Power
Cond •
Idle
App.
Cru.
Max.
T.O.
THC(ppm-C)
EPR
1.042
1.042
1.076
1.297
1.671
1.879
1.979
Scott
130.0
97.8
35.4
7.2
<0.1
<0.1
<0.1
FAA
180.0
140.0
62.0
7.0
1.0
0.0
-
CO(ppm)
Scotc
334.0
284.0
160.0
63.0
16.0
13.2
7.1
FAA
327.0
282.0
163.0
72.0
22.0
15.9
11.8
C0o(%)
Scotc
1.52
1.49
1.59
1.75
2.35
2.50
2.70
FAA
1.60
1.58
1.65
1.76
2.38
2.59
2.75
N0x(ppm)
Scott
13.3
13.6
18.3
38.8
91.0
121.4
145.0
FAA
11.8
15.3
21.2
43.5
100.0
135.8
162.8
N02(ppm)
Scott
-
8.2
7.9
10.0
3.8
3.6
2.0
FAA
7.0
8.0
10.0
12.5
5.0
5.0
5.0
NO(ppm)
Scott
-
5.4
10.4
28.8
87.2
117.8
143.0
FAA
4.8
7.3
11.2
31.0
95.0
130.8
157.8
in
I
to
-------
5-3
The carbon dioxide concentration measurements agree within 1 to
6%. This spread, though more than desirable, is not surprising in light
of the error analysis detailed in Appendix A-l which predicts a system
measurement error of 2.86% for carbon dioxide using NDIR. The carbon
dioxide standards in use by NAFEC and Scott checked within 1.5% of each
other.
The NAFEC nitric oxide analyses were consistently 10% higher
than the Scott analyses. The span gases in use by the two measurement
systems agreed within 2.5%. It is believed that the difference was due to
the indicating Drierite used to dry the NAFEC sample. Nitric oxide analysis
by NDIR using Drierite yields high results due to N(>2 conversion to NO by
the cobalt chloride indicator. Non-indicating Drierite is preferred but is
still known to give high readings. Since water interference to nitric
oxide'NDIR analyzers is severe, refrigeration drying and drying agents
must be used to remove water vapor from the sample gas.
Agreement in the nitrogen dioxide measurements is poor. The
concentration levels were always less than 15 ppm. The Scott measurements
were obtained from the difference between the NO reading, obtained by
passing the exhaust sample through the thermal converter which converts NO.
to NO, and the NO reading which is obtained when the sample is admitted to
the chemiluminescence analyzer directly. The subtraction of the two
similar numbers (NO and NO) can lead to substantial error. Small errors
X
in the NO and NO data result in large percentage differences in the NO.
X fc
measurements. The NO differences ranged from 10 to 15% with the NAFEC
x °
readings higher than the Scott readings. Since NO present in the gas
A
turbine exhaust is mostly NO, the errors present in the NO measurement
controled the error in the NO measurement.
x
A cylinder containing reference standards of propane and carbon
monoxide in nitrogen was provided by the EPA project monitor. This referee
gas standard was used to reference both measuring systems to the Environmental
Protection Agency laboratories and to each other. The referee standard
confirmed that the analyses of the NAFEC hydrocarbon and carbon monoxide
calibration gases were in error.
-------
6-1
6.0 EVALUATION OF SAMPLING METHODS
The effect of the exhaust sample collection procedure on the
accuracy of the exhaust emission measurement was evaluated. The composition
of exhaust gas sample must be an accurate representation of the turbine
exhaust stream. The exhaust stream of a mixed-flow turbine engine, such
as the JT8D, has a complex non-uniform gas concentration distribution.
The pattern of exhaust gases consists of the core engine flow
which occupies Che central portion of the exhaust plume and the fan air
which has bypassed the combustion process and occupies an imaginary concen-
tric outer cylinder. Within the core engine exhaust region the distribution
of exhaust gas concentrations is further influenced by the distribution of
combustion chambers, turbine blades and flow straightening vanes, and by
other aerodynamic influences. A sampling probe which is to deliver a
representative exhaust gas sample to the analyzers must coordinate the
various regions of the exhaust by a judicious location of sample inlet
ports.
In order to evaluate the sampling accuracy and representativeness
of sampling probe design, it is first necessary to know the true average
concentration present in the exhaust stream. The first attempt at
measuring this true engine exhaust emission level used a mixer located
behind the gas turbine engine. It was hoped that this mixer would mix
the exhaust gases into a homogeneous product such that the gas concentra-
tions would be the same at any point in the mixer exhaust stream. Due to
difficulties encountered with this technique a second technique was
evaluated wherein measurements were made point by point across the gas
turbine engine exhaust plume.
6.1 DESIGN AND TESTING OF MIXING CHAMBER
A mixer which would swirl the exhaust and produce a uniform
mixture through turbulence was considered first. It soon became apparent
that tolerable pressure drops were not attainable with this type of mixer.
-------
6-2
6.1.1 Kenics Mixer
Specifications were given to the Kenics Corporation to enable
them to finalize the design and quotation of a mixer to completely mix
the exhaust emissions from a JT8D gas turbine engine. The full power
exhaust flow, temperature and pressure were used as the design points. A
pressure drop through the mixer of less than 1.8 pounds per square inch
was specified as being acceptable. This value was arrived at from conver-
sations with Pratt and Whitney aircraft engineers and the EPA project
officer. Although the effect of back pressure on the engine was unknown,
this value represented a low value which was initially considered attainable
by Kenics. In the final analysis, however, the pressure drop calculated
by Kenics could not be obtained in a 36 inch diameter, 3 element mixer. A
48 inch diameter, 2 element mixer was quoted which would have a pressure
drop of 2.9 psi.
The exhaust gas parameters used by Kenics to calculate the expected
pressure drop included a value of 29.5 psia gas pressure. This is the
turbine discharge pressure in a normal engine. The actual pressure in the
mixer would be a decreasing amount, decreasing from 29.5 psia to 14.7 psia,
the ambient value. Using the value of 14.7 psia in the Kenics calculation
of mixer pressure drop, yields a pressure drop of approximately 6 psig or
twice that calculated by Kenics for the 48 inch mixer. These values of
pressure drop were believed to be too high to allow normal operation of the
JT8D. Although some emission data may have been obtained at low power
settings, it did not seem to be worth the time and expenditure involved.
6.1.2 Straight Pipe Mixer
The EPA project officer agreed with Scott's recommendation to
abandon the Kenics mixer and to proceed with a straight pipe mixer 36 inches
in diameter. The pipe was made up of a 5 foot adapter section to fasten
onto the jet engine followed by a 6 inch long section which housed experi-
mental probes and three 20 foot sections providing a total mixing tube
length of over 60 feet. Suitable pipe stands were fabricated to hold
the mixer in place and a 1/8" thick silicone rubber sheet was used as a
flexible coupling between the adapter section and the engine exhaust nozzle.
This flexible pressure seal was held in place by three steel hoops 1/16"
thick, fastened together at the ends by a 3/8" bolt.
-------
6-3
Preliminary runs were made with the first 5-foot adapter section
attached to the nozzle. This section increased the effective nozzle length
by 15 inches. The engine was operated at idle, high idle, and approach
power in this configuration. At approach power the engine vibration limit
was exceeded, so no higher power settings were attempted. Examination of
the engine operating data revealed that the engine was operating off its
design operating line. Concern was caused by the fact that the NI and N rotor
speeds were diverging. The' next step was to check the engine operation
with the first 20-foot section added to the existing 5-foot adapter. In
this configuration the engine was run at idle and high idle. The results
were the same as with the 5-foot adapter. Consultations with the engine
manufacturer resulted in a decision to discontinue operation with the mixer
attached.
6.2 TRAVERSING PROBE
Pratt and Whitney Aircraft Corp. reported that the total pressure
profile across the JT8D exhaust was flat within 4% at idle and at maximum
continuous power it was essentially flat with one pocket which deviated
by 6%. If the density is also flat, an area weighted average concentration
should produce a result which is representative of actual emission rates.
A traversing probe was fabricated and placed so that the probe
inlet was located two inches behind the nozzle exhaust plane. An array of
177 sample points were laid out.
6.2.1 Design and Operation of Traversing Probe
The traversing probe designed and constructed by Scott for use
on this program can position a probe anywhere in a vertical plane measuring
39 inches wide and 34 inches high. It was mounted on a welded frame so
that the center of the traverse area coincided with the center of the JT8D
engine mounted on the NAFEC test stand. Figure 6-1 illustrates the
traversing probe.
The probe was positioned by electric motor driven ball screw
jacks which moved the probe in both the horizontal and vertical directions.
-------
FIGURE 6-1 TRAVERSING PROBE
-------
6-5
The probe carriage rode on two horizontal steel rods using Thompson linear
ball bushings. The horizontal carriage supported the bottom end of two
vertical rods which guided the vertical probe carriage. The top end of
the rods rode along a horizontal track. The exhaust sampling probe was
fastened to the vertical carriage and was thus capable of being moved both
horizontally and vertically.
Probe position was monitored with a digital voltmeter connected to
a potentiometer voltage divider. The potentiometer was a ten-turn device
gear driven by the electric motor actuators. The system was calibrated so
that a one millivolt signal equaled 0.1 inches. The probe position was
repeatable and known within one tenth of an inch. Operation was smooth and
controlled. Variable speed motor control enabled fast traversing or slow
approach to a precise position. No problems were experienced with the
traversing probe except for the loosening of a few screws due to the severe
environment. The probe worked well at all power settings including take-off
power.
The sampling probe which was used to make detailed measurements
of the jet engine exhaust plane is shown in Figure 6-2. The inlet to the
probe is a 1/4 inch diameter stainless steel tube.
6.2.2 Nozzle Traverse Data
Detailed traverse data were obtained on the JT8D engine at four
power settings: Idle, holding, approach and maximum continuous power.
Some cursory traverses were also made at take-off power. Take-off power
may only be maintained for five minute periods. It must be followed by
sufficient time at reduced power after the take-off run so as to allow the
engine to cool down to operating temperature.
The array was a rectangular grid circumscribed by the exhaust
nozzle diameter. The sample points were located two inches apart. The
concentration averages were calculated at the center of two inch squares
except for those sample points in the outer periphery. These areas were
assumed to be combinations of triangles, rectangles and sections of a
circle with the sample"point concentrations assumed to be representative of
the whole area. The products of the concentration x area for each section
were summed and divided by the total area to obtain the area weighted average.
-------
FIGURE 6-2 SAMPLING PROBE
-------
6-7
Maps of exhaust plane gas concentrations were constructed using
the traverse data. Concentration maps for carbon monoxide, carbon dioxide,
total hydrocarbons and nitrogen oxides were drawn for each of the four power
settings tested. These maps are presented in Figures 6-3 through 6-6. The
maps were constructed by interpolation between the 177 measured data points.
Similarity among the concentration patterns for carbon monoxide,
carbon dioxide and nitrogen oxide is striking at all power settings. The
total hydrocarbon pattern is variable and seemingly unrelated to the patterns
of the other gaseous exhaust components. The concentration pattern of
hydrocarbons at the higher power settings is influenced by the atmospheric
background level of 2 ppm or greater. The concentration levels around the
periphery of the exhaust nozzle are the levels found in the air which bypass
the combustion process. On the other hand, the air which passes through the
hot combustion zones of the engine is reduced in hydrocarbon level by
incineration.
The hydrocarbon map for maximum continuous power clearly shows the
influence of the slow bleed off of hydrocarbons from the sampling lines.
The hydrocarbon concentration pattern is different from the hydrocarbon
patterns at all other power settings and no longer shows the characteristic
symmetry. The effect is large because the concentrations are so low that
small changes are magnified. Traverses were begun at the left center of the
nozzle cross-section. The traverse probe then was moved horizontally to
the right stopping at all the sampling points along the horizontal center
line. The data obtained along this initial traverse line is higher than
the remainder of the data in the cross-section due to the slow bleed-off
of hydrocarbons adsorbed on the interior surfaces of the sampling system.
At the time the traverse data were taken, an apparent stable reading was
obtained at each sampling point.
Background air influence is also observed in the CO. concentration
pattern at the idle power level. The lowest level found at idle is in the
bypass air region around the outer periphery where the CO- level is
approximately 200 ppm. The resolution of the CO. analyzer at these low
levels is poor since the instrument was optimized for the CO- concentration
range of one to five percent. However, the fact remains that CO. background
level does influence the concentration pattern.
-------
6-8
THC(ppm)
CO(ppm)
C02(%) N0x(ppm)
FIGURE 6-3
NOZZLE TRAVERSE - P&W JT8D WITH OLD STYLE COMBUSTORS
IDLE
-------
6-9
THC(ppm)
co
CO(ppm)
NO ppm
x *v
FIGURE 6-4
NOZZLE TRAVERSE - P&W JT8D WITH OLD STYLE COMBUSTORS
HOLDING POWER
-------
6-10
THC(ppm)
CO(ppm)
co2(%)
NOZZLE TRAVERSE -
NO ppm
FIGURE 6-5 X
P&W JT8D WITH OLD STYLE COMBUSTORS
APPROACH POWER
-------
6-11
THC(ppm)
CO(ppm)
\
co2(%)
N0x(ppm)
FIGURE 6-6
NOZZLE TRAVERSE - P&W JT8D WITH OLD STYLE COMBUSTORS
MAXIMUM CONTINUOUS POWER
-------
6-12
6.2.3 Mixer Traverses
Twenty feet of the mixer pipe described in Section 6.1.2 was
placed behind the engine. The pipe was not coupled to the engine. A
special 6 inch flanged test section was bolted on the engine end of the
mixer pipe. This section held the probes under test.
Traverses were made of the exhaust from the 20 foot mixer pipe
by locating the traversing probe at the mixer exit. Traverse data on the
mixer exhaust at 61 points were area weighted and concentration averages
calculated. Figure 6-7 is a map of the exhaust concentration at idle as
emitted from the mixer section.
The mixer was not coupled to the engine directly and a 3 inch
annular gap existed between the engine exhaust nozzle and the mixer inlet.
Dilution air was induced into this gap resulting in concentration measure-
ments at the mixer outlet which were diluted approximately 20%.
6.3 MIXING PROBE EVALUATION
Several mixing probe designs were evaluated using the test
configuration described above. The probes were mounted in the 6 inch test
section located at the forward end of the three foot diameter mixer pipe.
The mixer pipe was not coupled to the engine. The inlet to the test probes
and mixer pipe was positioned from 2 inches to 12 inches behind the engine
exhaust nozzle. The traversing probe was located at the mixer pipe outlet.
Throughout the probe evaluation, exhaust emission levels at the NAFEC probes
mounted in the engine exhaust nozzle were continuously monitored. There
were therefore three sampling points and techniques which could be compared
simultaneously: 1) The NAFEC probes in the exhaust nozzle area which
sampled core engine exhaust, 2) The traverse probe located at the mixer
outlet, and 3) The mixing probe located at the engine exhaust plane which
was under evaluation to determine if it sampled the, average exhaust concen-
trations at its cross-section.
The mixer pipe, though detached, was located behind the engine.
In order to prevent any possibility of engine damage due to backpressure,
the engine was not operated above idle. All mixing probe trials were
conducted at this operating condition.
-------
6-13
THC(ppm)
CO (ppm)
co2(%)
NO (ppm)
FIGURE 6-7
MIXER TRAVERSE - P&W JT8D WITH OLD STYLE COMBUSTORS
IDLE
-------
6-14
6.3.1 Cruciform Probe
A cruciform shaped probe was built according to the specification
of Attachment A-l. Two one-inch diameter extra heavy wall stainless steel
tubes were welded together in the center to form a cross. A hole was
drilled through one tube into the other tube at the center joint. A
tubing fitting was welded into this hole. A web*of 1/4 inch thick stainless
steel four inches wide was welded along the downstream side of the rear tube.
Twelve holes were made with a //43 drill on the upstream side of the tubes.
The holes were located at centers of equal area based on an "effective"
tailpipe diameter of 26 inches. This "effective" diameter had been used
previously by EPA in JT8D studies. The cruciform probe was mounted on four
angle brackets inside the 6 inch test section which was attached to the
mixer pipe inlet. A 1/4 inch stainless steel tube connected the probe
central sample port to a bulkhead fitting on the wall of the 6 inch flange
section.
Varations of hole patterns were evaluated in addition to the
12 point pattern described above. The hole patterns evaluated are described
below and illustrated in Figure 6-8.
TABLE 6-1 HOLE PATTERNS TESTED
1. 12 point pattern on center of equal areas based on an effective
engine diameter of 26 inches.
2. 8 point probe made up of the innermost two holes of the 12 point
pattern (1).
3. 24 point pattern on center of equal areas based on a 26 inch effective
nozzle diameter
4. 24 point combination - made up of the 24 inside holes of the 12 point
and 24 point hole patterns above [(1) & (3)1.
5. 16 point combination - made up of the sixteen inside points of the
24 point and 12 point patterns.
HOLE POSITIONS ON CRUCIFORM PROBE
Radial Position Radial Position
Hole (Inches) Hole (Inches)
12 pt. probe A.D.G.K 5.30 24 pt. probe N,T,Z,FF 3.80
B,E,H,L 9.20 O.V.AA.GG 6.50
C.F.J.M 11.85 P,V,BB,HH 8.40
Q.W.CC.JJ 9.95
R,X,DD,KK 11.25
S.Y.EE.LL 12.45
-------
6-15
30" Nozzle Dia.
26" Dia.
P B 0 R C S
DD DD H BB AA G Z
Holes A-M 12 Pt. Probe
Holes N-LL 24 Pt. Probe
Looking Upstream
FIGURE 6-8 CRUCIFORM PROBE HOLE PATTERN
-------
6-16
6.4 COMPARISON BETWEEN MIXING PROBES AND TRAVERSE AVERAGES
Comparisons between the performance of probe types at idle can be
obtained from: 1) the data obtained using the traverse probe at the engine
exhaust plane, 2) the data obtained using the traverse probe at the mixer
exhaust plane, 3) the data obtained with the NAFEC probes which were
mounted in the engine exhaust nozzle, and 4) the data obtained from the
cruciform probes which were being evaluated.
In the tests involving the cruciform probes, simultaneous measure-
ments were made with the NAFEC probes. During runs 137, 139, 143 and 144
mixer traverses were also made in addition to the cruciform probe and
NAFEC probe readings. Indirect comparisons can be made to the engine
traverse averages which were obtained previously.
The probe comparison data is compiled in Table 6-2. The first
column lists the probe from which the sample was obtained. The second
column describes the position of the cruciform probe which was varied in
angular orientation and in distance behind the exhaust plane. The reading
number in the third column references the emission data to the corresponding
set of engine operational data. More than one set of emission data may have
been taken during the time period ascribed to a single reading number.
This is followed by the measured emission concentrations and the mass
emission rates calculated from a carbon balance as specified in Attachment
A-l.
Little may be learned from the concentration data because bypass
air entering the outer holes of the cruciform probe dilutes the exhaust
sample. Thus, direct comparisons cannot be made with either the concen-
trations measured with the NAFEC probes which were located in the core
engine flow or the average concentrations calculated from the traverse
data. The data is best compared by referring to the emission index values
(Ib/thousand Ibs. of fuel).
The percent deviation of each emission index from the index
calculated for the traverse average in reading number 124 is listed for
the three pollutants. The hydrocarbons and carbon monoxide indices for
-------
TABLE 6-2 PROBE COMPARISON DATA
ALL AT IDLE
Probe
Nozzle Traverse
Nozzle Traverse
12 PC. Probe
12 PC. Probe
Mixer Traverse
12 PC. Probe
8 PC. Probe
NAFEC Probe
NAFEC Probe
8 PC. Probe
Mixer Traverse
8 PC. Probe
NAFEC
8 PC. Probe
12 PC. Probe-
NAFEC
12 PC. Probe
12 PC. Probe
NAFEC
12 PC. Probe
NAFEC
12 PC. Probe
NAFEC
12 PC. Probe
NAFEC
24 PC. Probe
NAFEC
24 PC. Comb.*
NAFEC +
16 PC. Comb.
NAFEC
16 PC. Comb.
NAFEC
16 PC. Comb.
NAFEC
Disc, from
Exh.(in.)
Angular
Posicion
Degrees
8/0
8/0
-
8/0
8/0
_
_
8/0
-
8/45
8/45
8/45
_
2/45
2/22>s
_
2/0
_
2/90°
_
12/0
_
12/0
_
12/0
_
12/0
_
12/45
_
12/22>s
-
Corres.
Reading
9
124
119
137
137
136, 137
138, 139
139
140
140
141
141
142, 143
144
143
142
144
145
145
146
147
147
148
148
149
149
150
150
151
151
152
152
153
153
154
154
155
155
Concencraclons "Wee"
THC
ppm-C
121
127
169
168
119
154
238
318
329
244
118
193
328
184
102
357
112
79
374
178
346
196
380
166
337
149
362
201
366
252
366
198
360
194
371
CO
Ppm
233
237
245
241
181
241
397
489
501
409
187
295
489
273
134
488
153
100
493
245
496
263
506
253
493
216
501
300
493
414
501
300
501
294
501
C02
0.71
0.71
0.67
0.67
0.55
0.67
1.20
1.42
1.40
1.20
0.53
0.80
1.37
0.77
0.37
1.40
0.40
0.27
1.40
0.65
1.35
0.67
1.40
0.67
1.37
0.51
1.42
0.82
1.40
1.15
1.40
0.82
1.40
0.77
1.40
NOX
Ppm
5.7
5.8
4.6
5.1
4.0
5.1
9.1
10.4
9.7
8.7
4.1
5.8
9.6
5.8
3.0
9.6
2.9
2.2
9.6
4.6
9.1
4.8
9.6
4.8
9.1
3.6
9.2
5.7
9.2
8.2
9.6
5.8
9.1
5.4
9.3
Fuel
Flo-/
tf/Hr.
1008
1033
998
998
1003
1021
998
998
998
998
998
998
998
998
1021
1021
998
998
998
975
975
998
998
1021
1021
998
998
998
993
998
998
998
998
975
975
I Diff . Relative
tl/Kll Fuel
THC
18.7
19.5
27.3
27.2
23.7
25.0
21.7
24.4
25.5
22.2
24.1
26.1
26.0
25.9
29.8
27.7
30.3
31.7
29.0
29.7
27.8
31.5
29.4
26.8
26.7
31.3
27.7
26.6
28.4
23.8
28. *
26.2
27.9
27.3
28.7
CO
62.9
63.7
69.2
68.2
62.7
68.3
63.2
65.5
67.9
65.0
67.1
69.9
67.7
67.3
68.4
66.1
72.1
69.8
66.7
71.1
69.5
73.9
68.3
71.4
68.2
79.5
66.9
69.3
66.7
68.4
67.8
69.3
67.8
72.2
67.7
NO
2.52
2.56
2.13
2.37
2.28
2.37
2.38
2.29
2.16
2.27
2.42
2.25
2.18
2.35
2.52
2.14
2.25
2.52
2.13
2.19
2.09
2.21
2.13
2.23
2.07
2.17
2.02
2.16
2.04
2.22
2.13
2.20
2.02
2.18
2.06
To
THC
0
4.6
46.4
45.6
26.7
34.0
16.2
30.6
38.6
18.9
29.3
40.0
39.5
38.8
59.8
48.3
62.2
70.0
55.1
58.9
48.7
68.8
57.4
43.7
43.0
67.8
48.2
42.4
51.9
27.6
51.8
40.4
49.4
46.2
53.8
0124 Traverse
CO
0
1.3
10.1
8.3
0.3
8.5
n /
4.1
7.9
3.3
6.6
11.0
7.6
6.9
8.8
5.0
14.6
10.9
6.0
13.0
10.5
17.4
8.6
13.6
8.4
26.3
6.3
10.2
6.0
8.7
7.7
10.2
7.7
14.7
7.6
NO
0
1.6
-15.5
-5.9
-9.5
-5.9
-5 6
-9.1
-14.3
-9.9
-4.0
-10.3
-13.5
-6.7
0
-15.0
-10.7
0
-15.5
-13.1
-17.1
-12.3
-15.5
-11.5
-17.8
-13.9
-19.8
-14.3
-19.0
-11.9
-15.5
-12.7
-19.8
-13.5
-18.2
ConcenCracion Ratio
THC/CO,
£
170
178
252
250
217
230
198
224
225
203
222
241
240
238
276
255
280
294
267
275
256
292
271
248
246
291
255
245
261
219
261
241
257
252
265
co/cOg
328
334
366
360
329
360
331
344
358
341
353
369
357
354
362
348
382
370
352
377
367
392
361
378
360
423
352
366
352
360
358
366
358
382
358
HO /CO-
8.0
8.1
6.8
7.6
7.2
7.6
7.5
7.3
6.9
7.2
7.7
7.2
7.0
7.5
8.1
6.8
7.2
8.1
6.8
7.0
6.7
7.1
6.8
7.1
6.6
7.0
6.5
6.9
6.5
7.0
6.5
7.0
6.5
7.0
6.6
* 24 Inside polncs of 24 PC. & 12 PC. Paccern
+ 16 Inside poincs of 24 PC. & 12 PC. Paccern
@ Average during traverse
-------
6-18
the mixing probes were consistently higher than for the nozzle traverses.
The reverse was true for nitrogen oxides. The greatest differences were
found in the hydrocarbon data.
Table 6-3 ranks the cruciform probes in the order in which they
come closest to the nozzle traverse average. In general the probes whose hole
patterns are well within the core engine exhaust agreed best and as the probe hole
patterns include more of the outer exhaust stream the more the derived
emission rates differ from the nozzle traverse averages.
In Table 6-4 the cruciform probes are compared with the mixer
traverse average and ranked in order of closest agreement. The probes with
the central patterns again agree more closely with the mixer traverse
averages than the probes which include holes in the outer portion of the
exhaust. The overall agreement between the cruciform probes and mixer
traverse is much better than between the cruciform probes and the nozzle
traverses.
6.5 PREDICTIONS FROM TRAVERSE DATA
Predictions based on the point by point nozzle traverse data
were made of the emission index which would be calculated from concentrations
obtained with multi-point cruciform probes. The predicted values are
compared to the actual values obtained with cruciform probes in Table 6-5.
All predictions and measurements are at idle power.
The predicted cruciform probe values for carbon monoxide and
nitrogen oxides are all within 7% of the nozzle traverse. The predicted
hydrocarbon indices for all cruciform probes are lower than the nozzle
traverse averages. The measured values for carbon monoxide are from 6 to
25% higher than predicted. The measured hydrocarbons are from 37 to 63
higher, and the measured nitrogen oxides are up to 14% lower than predicted.
It appears that more exhaust was sampled through the outer
probe holes where the ratios of hydrocarbons and carbon monoxide
to carbon dioxide were greater than at the inner hole locations.
This can be seen more clearly from the point by point concentra-
tion data for the 12 point probe in Table 6-6. The data were obtained
-------
6-19
TABLE 6-3 COMPARISON OF CRUCIFORM PROBES
TO NOZZLE TRAVERSE AVERAGES
% Difference in Emission
Rank
1
2
3
4
5
6
7
8
9
10
Probe
8 Pt.
16 Pt. Comb.
8 Pt.
16 Pt. Comb.
24 Pt. Comb.
FAA
12 Pt.
16 Pt. Comb.
12 Pt.
12 Pt.
24 Pt.
TABLE
Orientation
0°
0°
45°
45°
0°
-
0°
22%°
45°
224°
0°
No. of
Runs
Averaged
2
1
2
1
1
13
6
1
2
1
1
Index From Exhaust Average
Traverse Average Diff.
THC
17.6
27.6
39.4
40.4
42.4
47.3
49.6
46.2
61.0
70.0
67.8
6-4 COMPARISON OF CRUCIFORM
CO
1.8
8.7
8.9
10.2
10.2
7.2
11.8
14.7
11.7
10.9
26.3
PROBES
NO
-7.8
-11.9
-8.5
-12.7
-14.3
-16.1
-10.7
-13.5
-5.4
0
-13.9
9.1
16.1
18.9
21.1
22.3
23.5
24.0
24.8
26.0
26.9
36.0
TO MIXER TRAVERSE AVERAGES
% Difference
Rank
1
2
3
4
5
6
7
8
9
10
Probe
8 Pt.
16 Pt. Comb.
8 Pt. Comb.
16 Pt. Comb.
24 Pt. Comb.
FAA
12 Pt.
16 Pt. Comb.
12 Pt.
12 Pt.
24 Pt.
Orientation
0°
0°
45°
45°
0°
-
0°
22%°
45°
22Jj°
0°
No. of
Runs
Averaged
2
1
2
1
1
13
6
1
2
1
sion
Mixer
THC
-8.16
-0.29
8.87
9.66
11.26
14.73
16.86
14.28
25.77
32.84
31.05
in Emis-
Index from
Average
CO
-1.26
5.41
5.69
6.84
6.81
3.93
8.41
11.23
8.29
7.58
22.43
NO
-1.28
-5.53
-2.12
-6.38
-8.08
-10.21
-4.25
-7.23
1.28
7.23
-7.66
Average
Diff.
3.57
3.74
5.56
7.62
8.72
9.62
9.84
10.90
11.78
15.88
20.38
-------
6-20
TABLE 6-5 PREDICTED VERSUS MEASURED
CRUCIFORM PROBE PERFORMANCE
Mode: Idle
All Probes at 0° (arms vertical and horizontal)
Probe
(All
Cruciform)
12 Pt.
12 Pt.
8 Pt.
24 Pt.
24 Pt. Comb.
16 Pt. Comb.
Nozzle Trav.
Mixer Trav. j
Emission Index ///Ktf fuel
Reading
No.
137
148
140
151
152
153
Avg. 124
\vg.l42, 143
Measured
THC
27.33
29.68
21.69
31.32
26.59
23.83
18.67
24.1
CO
69.24
71.14
63.17
79.46
69.32
68.41
62.92
67.1
NO
2.13
2.19
2.38
2.17
2.16
2.22
2.52
2.42
prpiH r.t-pcl from Nnxxli* Travpree
THC
17.72
17.72
15.86
18.01
16.89
15.56
-
CO
63.88
63.88
59.26
64.64
63.00
61.28
-
^x-
2.46
2.46
2.36
2.47
2.50
2.53
-
144
-------
6-21
TABLE 6-6 PREDICTED CONCENTRATIONS FOR
TWELVE POINT PROBE AT IDLE BASED ON NOZZLE
TRAVERSE DATA (RUN 124)
Hole
C
F
J
M
Ave.
THC
PPm
148
127
73
53
100
CO
PPm
235
195
188
91
177
N0x
PPm
4.4
3.5
3.6
2.3
3.5
C02
%
OUTER RING
0.59
0.49
0.53
0.25
0.47
THC/CO?
252
260
137
211
215
co/co2
401
399
353
362
381
N0../C00
7.6
7.2
6.7
9.2
7.4
MIDDLE RING
B
E
H
L
Ave.
A
D
G
K
Ave.
Overall
Average
Measured
Run 137
Run 148
181
223
157
131
173
221
260
218
199
225
166
with 12
169
179
325
346
319
309
325
513
545
505
540
526
342
point
245
245
7.3
6.7
6.2
7.9
7.0
13.4
13.9
12.3
15.0
13.6
8.0
probe
4.6
4.6
0.90
0.95
0.93
0.93
0.93
INNER RING
1.64
1.77
1.62
1.72
1.69
1.03
0.67
0.64
201
235
169
141
186
135
147
134
116
154
162
253
279
361
364
343
332
349
313
308
312
314
312
334
367
383
8.1
7.1
6.6
8.5
7.6
8.1
7.8
7.6
8.7
8.1
7.8
6.9
7.2
-------
6-22
from the nozzle traverse of Run 124 by interpolation. This assumes that the
flow of gases from the nozzle traverse location (2 inches downstream of the
plane of the nozzle) to the mixing probe location (2 to 12 inches down-
stream of the plane of the nozzle) was perfectly horizontal.
The carbon dioxide concentration obtained with the mixing probe
is approximately midway between the concentrations predicted for the outer
and middle rings. This appears to confirm that greater gas volumes were
sampled at the outer holes as compared to the middle and inner holes.
However, this does not explain the high hydrocarbon and carbon monoxide
concentrations found since it is seen that the ratios of these pollutants
to carbon dioxide are greater than would be expected for any point at or
within the outer ring. The explanation offered for this anomalous data
relates to the position of the mixing probe in the 36 inch diameter mixer
tube. It is known that dilution air' entered the mixer tube because
the tube was not coupled to the nozzle. The volume of dilution air was
estimated to be 20% of the turbine exhaust volume by comparing the mixer
and nozzle traverse carbon dioxide concentrations. It is now proposed that
as this dilution air entered the mixer tube it distorted the plume coming
from the nozzle. The net effect was that some of the gas in the extreme
outer portion of the nozzle plume was driven or diffused inward to the outer
sampling holes of the 12 point cruciform probe. Since the outer portion of
the nozzle plume was composed of exhaust with higher hydrocarbon and carbon
monoxide to carbon dioxide ratios than the inner portion, the ratios
found in the sample from the cruciform probe were higher than predicted.
While there is no firm evidence that this proposed explanation is correct,
it appears to be the only way to explain the measured data.
6.6 TRAVERSE AVERAGES VERSUS NAFEC PROBE READINGS
The nozzle traverse averages are compared to the NAFEC probe
readings at several load conditions in Table 6-7. The data for the NAFEC
probe are given for test runs other than the nozzle traverse runs because
leaks in the NAFEC probe sample system gave erroneous data for that probe
during the nozzle traverse runs. The emission indices show the greatest
-------
TABLE 6-7 TRAVERSE AVERAGES VERSUS NAFEC PROBE
Mode
Idle
Holding
Approach
MCP
Probe
TA
FAA
TA
FAA
TA
FAA
TA
FAA
Read.
No.
124
110
125
97
121
116
123
113
Concentration
EPR
1.042
1.041
1.0705
1.065
1.302
1.284
1.873
1.887
THC
120.9
345
30.8
148.5
.2.52
4.3
1.2
2.0
CO
237
508
147
345
35
82
18
27
CO.
£.
.71
1.42
.75
1.40
.83
1.62
1.29
2.38
NOX
5.8
9.8
7.1
9.0
14.6
26.2
43.1
80.5
///K// fuel
THC
18.64
26.35
4.61
11.78
.348
.304
.107
.097
CO
63.80
67.76
38.44
47.79
8.43
10.11
2.80
2.28
NO
2.56
2.15
3.05
2.05
5.77
5.31
11.00
11.14
N)
U)
-------
6-24
differences for all pollutants at the holding condition. Except for the
41% difference in hydrocarbons indices at idle, the differences in the
emission indices for the three pollutants at conditions other than holding
did not exceed 20%.
6.7 CARBON BALANCE CALCULATION OF EXPECTED CARBON DIOXIDE CONCENTRATIONS
As a check on the representativeness of the exhaust concentration
data collected using the various probes described above carbon balance
calculations were made of the expected exhaust carbon dioxide concentration
based on the measured engine fuel and air consumption. The ratio of the
mass of carbon dioxide produced to fuel burned from a CxHy fuel is:
*CO..
44x
# fuel 12-. X + y
The fuel used during the reported program had a measured hydrogen/carbon
mole ratio of 1.9. The carbon dioxide produced from this fuel calculated
from the above equation will be 3.165 # CO./fuel. Now since:
0 „ ..
2 v fuel _
_
// fuel ' //air
and the volume fraction is: //C°2 < 29_ = m°1 C°2
9 air * 44 mol air
then the average concentration of C02 in percent by volume which we would
expect to find in the gas turbine exhaust would be:
where: Wp is the engine fuel flow in Ibs/hr. , and W is the engine air
flow J.n Ibs/sec. The engine air flow was measured at the bell mouth inlet
to the JT8D gas turbine.
Table 6- 8 compares the expected average C02 concentrations with
the actual measured concentratiocs for several probes and probe positions
tested. The nozzle traverse averages agree very well with the calculated
C02 levels. This lends credence to the validity of the assumption of equal
flow at each of the 177 nozzle traverse points.
-------
TABLE 6-8 CALCULATED AVERAGE CO EXHAUST CONCENTRATIONS
COMPARED TO MEASURED VALUES
Power
Setting
Idle
Holding
Approach
MCP
Reading
No.
124
137
139
140
141
142
143
144
145
146
147
148
149
150
151
152
153
154
155
125
121
123
WF
ff/Hr.
1008
997.8
998.5
1021.4
997.9
997.9
998.5
998.5
1021.4
998.5
998.5
975.0
997.9
1020.8
997.9
997.9
997.9
997.9
975.0
1347.4
2957.2
6916.3
wa
ff/Sec.
81.5*
81
81
81
81
83
81
85
83
83
81
83
83
83
81
81
81
81
81
109.9
212
303.8
Calc. Measured CO? Concentration (%)
CO, 12 Pt. Nozzle Probe
(%) NAFEC 12 Pt. Position Traverse Other Position
.71 .71
.71 .67 0°
.71 .67 0°
.73 1.42 1.20 ^5^-
.71 1.40 1.20 ^j!^-
.70
.71 1.37 .80 1^-
.68 .77 J-Jt.
.71 1.40 .37 45°
.70 .40 45°
.70 1.40 .27 22Ji°
.68 1.35 .65 0°
.70 1.40 .67 90°
.71 1.37 .67 0°
.71 1.42 .51 2*0Pt-
7i i /.n a* 24 Pt. Comb.
-82 Q0
71 i /,o * 15 16 Pt. Comb.
'71 ! /|0 Qo 16 Pt. Comb.
.82 450
70 i /,0 -,-. 16 Pt. Comb.
.71 .71
.81 .83
1.32 1.29
S3
U1
* Average during run
-------
7-1
7.0 BASELINE MEASUREMENTS
The contributions of operating conditions, the operating technique
and the instrumentation to variations in emissions measurements were deter-
mined from the data acquired in the tests on the JT8D engine. The effect
of approach to power was determined by examining the data acquired during
engine calibration runs where the engine was operated at distinct power
levels over its operating range and the direction of approach to the
power setting was noted. The run-to-run variation at the same power setting
was calculated from the emissions data. This run-to-run variation illustrates
the combined uncertainty level of the instrumentation and engine operation.
The effect of variation in fuel flow was illustrated by its effect on the
uncertainty of the calculated mass per mode values.
7.1 RUN-TO-RUN VARIATION
The overall variability of the emissions measurements can be
demonstrated using the data acquired during the probe evaluation series of
tests. All these tests were made at idle power. The engine was shut down,
the probe configuration changed and the engine restarted between each test.
The time the engine was shut down varied from 15 minutes to overnight
(approximately 16 hours). At each test point emissions data were obtained
from the NAFEC probes which remained unchanged in position or configuration
through the series of tests. The ambient conditions varied over only a
small range of winter levels.
Table 7-1 contains the emission index values calculated from the
data taken on 13 consecutive engine runs at idle power. Reading number
142, 144 and 146 correspond to additional data taken during mixer traverses.
The greatest uncertainty was found in the hydrocarbon data
where the coefficient of variation was 5.3%. The coefficients of variation
for carbon monoxide and nitrogen oxides were 1.5% and 3.5%.
The engine operating parameters (thrust, fuel flow and EPR)
exhibited similar variability with the thrust coefficient of variation
largest at 4.8%. Removal of the obvious thrust outlier of Run 147 would
reduce the variation considerably. The fuel flow coefficient of variation
was low (1.3%) and the EPR coefficient of variation was very low (.008%).
-------
7-2
TABLE 7-1 RUN-TO-RUN VARIATION IDLE POWER SETTING
FAA Probe Data - Scott Readings
Run
No.
140
141
143
145
147
148
149
150
151
152
153
154
155
X
2
0
o
CV %
///K# Fuel
THC
24.39
25.50
26.05
27.69
28.96
27.77
29.39
26.70
27.67
28.36
28.34
27.89
28.72
27.49
2.11
1.45
5.3
CO
65.51
67.92
67.72
66.09
66.67
69.53
68.34
68.21
66.87
66.70
67.75
67.78
67.73
67.45
1.12
1.05
1.5
_NOX_
2.29
2.16
2.18
2.14
2.13
2.09
2.13
2.07
2.02
2.04
2.13
2.02
2.06
2.11
.005
.075
3.5
EPR
1.040
1.040
1.041
1.041
1.040
1.040
1.040
1.040
1.040
1.040
1.041
1.043
1.043
1.040
.72xlO"6
.85xlO~3
.008
Thrus t
//'s
757.5
782.7
808
808
707
797.9
797.9
828.2
828.2
828.2
808
848.4
848.4
803.7
1487.2
38.6
4.8
Obs.
Fuel
Flow
///Hr.'
997.9
997.9
998.5
1021.4
998.5
975.0
997.9
1020.8
997.9
997.9
997.9
997.9
975.0
998.7
178.4
13.4
1.3
-------
7-3
7.2 EFFECT OF APPROACH-TO-POWER
The engine power setting Just previous to that at which a measure-
ment is made may influence the emissions data at that test point. Measured
engine emission rates of carbon monoxide, nitrogen oxides and total hydro-
carbons are plotted against engine pressure ratio (EPR) in Figures 7-1,
7-2 and 7-3. EPR is a convenient measure of engine power. Emission data
from both smoky and smokeless combustors are included. The curves drawn
through the data points are intended for reference only.
The "up" symbols designate those data points which were obtained
when the engine power setting was increased up to the data point level.
The "down" symbols designate data points where the engine power was decreased
from a higher power setting to arrive at the correct level for the data
point.
The direction of approach to power did not have a significant
effect on the carbon monoxide and nitrogen oxide emission indices (Figures
7-1 and 7-2). However, the total hydrocarbon indices were heavily depen-
dent on the approach-to-power. At the higher power settings there is con-
siderable scatter. Since the carbon monoxide and nitrogen oxides show
no effect of approach-to-power, it is possible that the apparent hydrocarbon
effect is not caused by changes in the engine emissions themselves. The
hydrocarbon measurements would be affected by the condition of the sample
line. The concentrations of hydrocarbons increase by a factor of 1000
from high power settings to low power settings. The heated sample lines
absorb hydrocarbons during the low power sampling and require some time
to desorb those hydrocarbons when the power setting is increased and the
exhaust hydrocarbon level drops sharply. A possible cure for this problem
is to use very clean lines and to wait until the hydrocarbon level is stable
for a few minutes before logging the concentration data for that data point.
7.3 EFFECT OF VARIATIONS IN ENGINE THRUST AND FUEL FLOW
The measurements of engine thrust and fuel flow were< performed at
NAFEC with transducers whose accuracy is - 0.25%. The accuracy of the actual
-------
7-4
LT.
1
EFFECT OF APPROACH TO POWER
--CARBON MDNOXIDE--
O UP "X SMOKY
O DOWN J COMBUSTORS
SMOKELESS
COMBUSTORS
UP
O DOWN
FIGURE 7-1
-------
7-5
EFFECT OF APPRAOCH TO POWER
--NITROGEN OXIDES--
O UP
,7 O DOWN
SMOKY
COMBUSTORS
SMOKELESS
COMBUSTORS
/C 9 UP
O DOWN
-------
7-6
EFFECT OF APPROADi TO POWER
—TOTAL HYDROCARBONS—
SMOKY
COMSU--VTORS
SMOKELESS
COMBUSTORS
UP
O DOV^N
FIGURE 7-3
-------
7-7
measurement of thrust was determined at NAFEC to be ±3%* over all thrust
levels. The degradation of accuracy in the measurement is due to the
error in the conversion of the transducer signals into electrical outputs
and the errors associated with recording those signals as well as the effects
of engine thrust unsteadiness due to the hunting in the fuel control
system. The fuel flow accuracy was determined at NAFEC to be * 2.25% mostly
due to the hunting action in the fuel control system.
The uncertainty in the calculation of mass per mode may be
estimated from the experimentally determined variances shown in Table 7-1
by standard statistical methods. The mass per mode for the taxi/idle
out mode is calculated from the equation:
Mass/mode = emission rate(///// fuel) x fuel flow (///Hr) x_TIM (Hr)
The variance in the mass/mode calculation is:
V(mass/mode) - [TIM- (CO)]2. V(F) + [TIM-(F)] 2. V(CO)
where: TIM = time in mode
CO, THC, NO = emission rates in ///// fuel of CO, THC and NO
X X
F = fuel flow rate
V(CO) = variance in the CO measurement
V(F) = the variance in the fuel flow measurement.
Using the mean values of fuel flow and emission rates and the
variances in fuel flow and emission rates in Table 7-1, the variance in
•
the calculated mass per mode values was determined. The variance of the mass
per mode values is tabulated in Table 7-2. Also tabulated in Table 7-2 are
the mean value of mass per mode which was calculated using the mean emission
rates and mean fuel flow and the standard deviation which is the square root
of the variance. The coefficient of variation was calculated from the
* Personal communication with G. Slusher, NAFEC.
+ Statistical Methods in Research and Production, Halfner Publishing Co.,
New York, New York, 1972
-------
7-8
mean and standard deviation values for each emission parameter.
The coefficient of variation of the product of fuel flow and
emission rates of hydrocarbons in pounds per mode increased 14.5% over
the coefficient of variation of the emission rate of total hydrocarbons in
pounds per thousand pounds of fuel. Similarly the coefficient of variation
in the pounds per mode of carbon monoxide and nitrogen oxides increases
36.7% and 8.5% over the coefficient of variation in the values of pounds
per thousand pounds of fuel.
1ABLE 7-2 VARIANCE IN THE CALCULATION OF MASS PER MODE (IDLE)
Mean
Std. Deviation
Variance
C.V.
X
o
2
a
%
THC
7.75
.47
.221
6.07
CO
21.2
.44
.190
2.05
NO
.67
.025
6.4xlO~4
3.8
-------
8-1
8.0 AMBIENT CONDITIONS
It is necessary to determine the effect of ambient temperature
and humidity on emission rates from turbine engines so that emission
measurements made at various ambient conditions can be corrected to
equivalent values at a standard condition. Temperature and humidity
effects are best determined in a controlled atmosphere test cell. This was
not available at NAFEC, but indications of ambient effects can be gleaned
from plots of data obtained during the various test runs. Figures 8-1
through 8-3 depict emission indices as a function of humidity at idle,
approach and maximum continuous power. Temperature effects are shown
similarly in Figures 8-4 through 8-6. At idle hot and cold runs are denoted.
A "cold" idle is one in which the run was first idle after starting the
engine and was the only power setting at which the engine had been operated
after the start of the engine. A "hot" idle occurred when the engine was
brought down to idle from some higher power setting. All data shown are
from the NAFEC probes because these were common to all test runs. It is
apparent that with increasing temperature and/or humidity there were
usually decreases in hydrocarbons and carbon monoxide and increases in
nitrogen oxides. The greatest change was noted at approach power. The
observed effect ranged from a 2.5 to one change in hydrocarbons with a
1.5 to one change in carbon monoxide and a 1.15 to one change in nitrogen
oxides at approach to a slight change at maximum continuous power over the
range of temperatures and humidities sampled.
The primary problem in separating temperature and humidity
effects results from the fact that temperature and humidity followed similar
patterns of increase or decrease from run to run. Thus, it is not clear
which parameter caused the variation in emission indices. Based on experience
it would seem that temperature of the inlet air should have a greater
effect on combustion than humidity.
-------
00
NJ
FIGURE 8 -1
-------
00
CJ
-------
00
FIGURE 8-3
-------
' • i
f-H-!-T- -4-4-
±i
22
ogr
_j L
i
•4 -(••
I
FIGURE 8-4
-------
FIGURE 8-5
-------
FUEL.
-------
9-1
9.0 SUMMARY AND CONCLUSIONS
Variability in the measurement of gas turbine emissions is
caused by instrument error, calibration gas uncertainty and deficiencies
in exhaust gas sampling and handling techniques. Calibration gas uncertainty
was found to be the largest contributor to the variability between two
measurement systems.
The EPA specified exhaust gas sample handling and conditioning
system was found to be adequate and reliable for the handling of gas
turbine exhaust samples.
Exhaust sampling methods were studied using both an exhaust
mixing technique and a detailed concentration mapping technique using a
traversing probe. The mixer technique could not be used at power levels
above idle. The detailed concentration maps may require integration with
a mass flow profile in order to enable the calculation of true exhaust
emission rates.
-------
A-l
01 0374
APPENDIX A
INSTRUMENT ERROR ANALYSIS
A.1 INTRODUCTION
The error analysis of the individual sub-systems of the turbine
system was done by the application of statistical methods to the different
analog instruments comprising a sub-system. The analysis essentially gives
an "engineering estimate" with a reasonable degree of confidence, for a
steady-state accuracy of any proposed analog computing system.
A.1.1 CO Sub-system
The NDIR CO analyzer is essentially a comparitive instrument in
that it compares (with some "instrument error") the sample CO level against
a reference or standard CO level (which in turn has an error). Thus, the
error in a CO sub-system would be that due to the instrument itself, the
calibration CO (or no CO) gases and the recorder. Figure A-l describes
the modular structure of the CO sub-system. The notation on the signal arrows
denote the signal level and its associated uncertainty; for instance, the
signal from the span gas has a level B and an uncertainty in the measurement
of that level of ±x or ±y or ±z depending on what point in the error model
the measurement is made. The instrument errors which cause this uncertainty
are labeled e., e2, etc. P is the signal level at the output of the
system. P±f is the recorded response to a sample where ±f is the uncertainty
due to the span gases above. P±h is the recorded response to a sample gas
where ±h includes both the span gas uncertainty and the recorder and analyzer
uncertainty in comparing the sample gas to the span gas.
-------
FIGURE A-l SYSTEM "ERROR FLOW" FOR CO SUB-SYSTEM
Zero gas
el
Span gas
e2
A±a ^
B±x ^ V
. „*-
>*•
CO Analyzer
G3
/
\
P±5 ^
B±y
Recorder
• ••
. i
, i
i l
1 1
l
P±f
Sample
A±c
w
i-j
oo
o
o
I-1
w
r
IS3
-------
A-3
1318 06 0173
The following errors were determined for the CO sub-system
instrumentation:
ej: ±1%
e«: ±2% of span value
e_: ±1% based on repeatability of instrument
e,: ±0.25% manufacturer's specifications
It should be remembered that the sample gas will be "seen" by
the CO analyzer with an error present due to the span and zero gases as
well as its own error (repeatability). Referring to Figure A-l.
x = ±2%; y
z =
z=
f =
g
I'+l = ±1.414%
±2.236%
2 + 0.25 = ±1.436%
±-/5 + 0.252 = ±2.25%
±-Jc2+z2 - -/1.4362 + 2.252 = ±2.
67%
±2.850%
h = ±-/g2+e4Z =-/2T8502 + 0.252 - ±2.
861%
Therefore, the CO sub-system error involved in the system
read-out as indicated by the recorder pen is ±2.861%. This is the
three standard deviation system error (confidence level of 99.7%) of the
output range.
-------
A-4
1318 06 0173
A.1.2 CQ2 Sub-system
The CO sulv-system is entirely analagous to the CO sub-system
wjth tlie sairc individual instrument errors. Thus, the steady-state system
error as Indicated by Llic recorder pen is also ±2.861%.
A.1.3 NO Sub-system
^^^^vt ~~
Tlie NO system error-f.low is essentially the. same as the CO and
X
CO sub-sysLcri except for the inclusion of a NO- - NO converter (see
Fig. A-2). The converter has an error associated with it due to its
efficiency in converting all the N0_ to NO. Thus, if the converter is only
99% efficient, there is a possible maximum error of ±1% error associated
with it. For the NO. sub-system the following instrument error values
f*.
have been determined:
c^: ±1%
C2: ±2%
c,: ±]% manufacturer supplied
cf : ±0.25% manufacturer supplied
cr: ±1% assumed
FoI lowing in a manner similar to the error analysis for the CO
sub-system,
f * ±2.67%
±2.85%
h == V&+C.2 - V2-852 + I2 = ±3.02%
-------
FIGURE A-2 SYSTEM "ERROR Fi-OK1' FOR NO SUB-SYSTEM
X
Zero gas
el
Span gas
e2
A 4- «
i B±x
NO Chenulumines-
X
cence
63
/
NO + NO.
P±8
(as NO)
OJ
i— i
00
o
P.th ^
B ;y _ n>
Converter
e5
-3 P±f
Recorder
e4
1 o
P±i ' d
i
1 i
i
i
B±z ' ' A±c
1 i
1 1
r '
Sample
NO + N02
i
; £
-------
A-6
1318 06 0173
If Cp i'j ;-3ssuinct! - i2%:
f = ±2.67%
^ 22 = ±3.336%
A /*} ^ *'' ' *
h = V3.336 + 1 = ±3.483%
i = A/3.4832 + 0.252 = ±3.492%
A.1.4 Total Hydrocarbon Sub-system
The total hydrocarbon (THC) analyzer is also a comparitive
instrument, wherein the response and accuracy of the instrument output
is dependent on the pressure setting and zero and Cull scale adjustments
from standardized rases. The instrument- is also sensitive to errors
induced by oxygen in the sample. This "oxygen response" of the instru-
ment is determined prior to its use in any system and if the "oxygen
response" is less than 2%, the instrument is considered reliable. The
THC annlyzer tested for the turbine system was well within this 2% level
and hence it ±s assumed that the "error-flow" of the THC analyzer will
pattern iLut-.l f closely to that of the CO or CO- sub-system "error-flow",
with the following instrument errors:
eL: ±1%
e»: ±2% of span value
e~: ±1% of Cull scale
e4: ±0.25%
Hence, the three standard deviation system error for the total
hydrocarbon analyzer is also ±2.861% of the output range.
-------
APPENDIX B
JT8D
Engine Emission Data
Read.
No.
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
Probe
Used
NAFEC
NAFEC
NAFEC
NAFEC
NAFEC
NAFEC
NAFEC
NAFEC
HAFEC
NAFEC
NAFEC
NAFEC
NAFEC
NAFEC
NAFEC
NAFEC
NAFEC
NAFEC
NAFEC
NAFEC
NAFEC
NAFEC
NAFEC
NAFEC
NAFEC
Power
Condition
Approach
Approach
Cruise
Cruise.
Hot Idle
Hot Idle
Take Off
Max.Cont.
Max. Con t.
Cruise
Landing
Approach
Hot Idle
Idle Cold
Take Off
Cruise
Approach
Take Off
Idle
Approach
Landing
Cruise
P.P.
Smoke
No.
23.1
23.8
35.1
—
3.7
4.9
32.3
33.7
33.1
35.1
35.5
33.8
32.3
24.1
3.1
3.1
32.2
33.9
23.7
31.8
6.2
29.6
34.2
36.8
36.0
THC
CO
pounds/K pounds
0.53
0.46
<0.01
<0.01
—
—
0.06
0.01
.0.01
0.01
0.02
0.02
0.03
0.45
—
14.01
0.01
0.01
0.74
0.02
8.09
7.26
1.38
1.55
—
—
0.63
0.84
0.97
1.06
1.38
2.23
2.25
7.63
—
48.69
0.72
1.66
9.59
0.67
NO
X
fuel
7.32
7.04
13.31
13.04
—
—
19.74
15.72
16.03
15.31
13.27
11.77
11.18
7.74
—
2.54
17.94
12.52
7.02
18.52
THC
CO
NO
X
pounds/hour
1.63
1.42
0.03
0.03
—
—
0.53
0.10
-0.03
<0.03
0.10
0.09
0.17
1.47
13.49
0.85
0.09
2.21
0.21
24.96
22.29
8.56
9.68
—
—
5.65
6.21
7.15
7.40
8.65
12.10
11.50
25.05
—
46.89
6.13
10.22
28.54
5.65
22.59
21.61
82.56
81.21
—
—
177.50
115.56
117.82
106.46
83.43
63.94
57.08
25.41
—
2.45
152.53
77.27
20.91
157.30
THC
CO
nu
pounds/K pound thrust - hrs.
0.31
0.27
0.01
0.01
—
—
0.04
0.01
,'0.01
-"O.Ol
0.01
0.01
0.02
0.26
—
13.98
0.06
0.01
0.40
0.01
4.73
4.22
0.78
0.88
—
—
0.38
0.48
0.55
0.60
0.78
1.25
1.26
4.39
—
48.59
0.42
0.92
5.21
0.39
4.28
4.09
7.54
7.42
—
—
11.99
8.96
9.06
8.62
7.48
6.59
6.27
4.46
—
2.54
10.58
6.99
3.82
10.78
-------
Read.
No.
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
Power
Condition
Idle
Idle Cold
Idle Cold
Approach
Landing
Cruise
Idle Cold
Approach
Landing
Cruise
Max. Con t.
Idle Hot
Idle Cold
Idle Cold
Probe
Used
NAFEC
NAFEC
NAFEC
NAFEC
NAFEC
NAFEC
NAFEC
NAFEC
NAFEC
NAFEC
NAFEC
NAFEC
NAFEC
NAFEC
NAFEC
NAFEC
NAFEC
NAFEC
NAFEC
NAFEC
NAFEC
NAFEC
NAFEC
NAFEC
NAFEC
Smoke
No.
—
—
—
—
—
—
5.8
26.8
30.8
32.6
37.5
4.9
7.3
22.3
26.1
31.5
29.8
35.7
37.5
38.0
38.9
38.7
5.2
3.4
3.0
THC
CO
pounds /K pounds
-
9.59
7.94
7.39
2.54
2.01
11.05
0.47
0.13
0.02
<0.01
13.83
5.07
1.54
0.58
0.15
0.02
0.01
-------
Read.
No.
51
52
53
54**
55**
56**
5^***
58***
59
60*
61
62
63*
64
65
66
67
68
69
70
71
72*
73
Jit
75*
* PT
Power
Condition
Idle
Holding
Approach
Approach
Idle Hot
Idle
Idle
Idle
PP
Approach
Approach
Approach
PP
Land
Land
Land
Probes
Probe Smoke
Used No.
NAFEC 34.9
NAFEC
NAFEC
NAFEC
NAFEC
NAFEC
NAFEC
NAFEC
NAFEC
PT
NAFEC
NAFEC
PT?
NAFEC
NAFEC
PT?
NAFEC
NAFEC
PT?
NAFEC
NAFEC
PT7
NAFEC
NAFEC
PT 29.03
** Five Foot Mixer
THC CO NO
X
pounds/K pounds fuel
1.28
0.66
0.85
13.13
7.02
13.68
20.64
14.50
2.54
2.40
2.18
1.10
1.30
1.14
0.50
0.54
0.37
0.14
0.20
0.11
0.13
0.11
***
13.71
8.33
9.10
45.57
29.34
53.46
56.67
53.88
17.07
17.57
16.74
11.22
12.17
11.30
6.73
7.53
6.72
4.13
4.75
4.13
3.12
3.60
Twenty
7.13
8.30
7.17
2.57
3.19
2.60
2.32
2.47
5.56
5.55
5.52
6.46
6.42
6.58
8.08
7.98
7.91
9.43
9.64
9.43
10.84
11.47
Foot Mixer
THC CO
pounds/hour
3.04
2.00
2.55
13.88
9.38
13.71
20.72
14.53
5.98
5.66
5.13
3.29
3.87
3.37
1.89
2.08
1.43
0.64
0.96
0.53
0.68
0.59
32.44
25.23
27.36
48.17
39.23
53.57
56.90
53.99
40.21
41.37
39.43
33.54
36.31
33.48
25.55
28.77
25.69
19.43
22.47
19.43
16.13
18.82
NO
X
16.88
25.16
21.56
2.72
4.26
2.61
2.33
2.47
13.09
13.08
13.00
19.30
19.15
19.51
30.69
30.49
30.25
44.40
45.65
44.40
56.13
60.05
THC CO NO
X
pounds/K pound thrust -
0.80
0.40
0.54
18.63
8.30
14.13
19.70
13.21
1.46
1.38
1.24
0.63
0.74
0.65
0.28
0.31
0.21
0.08
0.12
0.06
0.08
0.07
8.54
5.05
5.81
64.66
34.72
55.23
54.09
49.08
9.81
10.09
9.55
6.39
6.98
6.44
3.81
4.26
3.81
2.34
2.71
2.34
1.79
2.08
4.44
5.03
4.58
3.65
3.77
2.69
2.21
2.25
3.19
3.19
3.15
3.68
3.68
3.75
4.58
4.52
4.48
5.35
5.50
5.35
6.24
6.64
w
-------
Read.
No.
76
77*
78
79
80
81*
82
83*
84
85
86
87
88
89
90
91
92
93
94
95
96
97
98
99
» 00
Power
Condition
Cruise
Cruise
M.C.P.
M.C.P.
Take Off
Take Off
Take Off
Take Off
M.C.P.
M.C.P.
Cruise
Cruise
Landing
Landing
Approach
Approach
Holding
Holding
Idle
Idle
Idle
Holding
P.P.
Approach
Idle
Probe
Used
NAFEC
PT?
NAFEC
"7
NAFEC
PT?
NAFEC
PT
NAFEC
PT7
NAFEC
PT?
NAFEC
PT?
NAFEC
PT?
NAFEC
PT7
NAFEC
PT?
NAFEC
NAFEC
NAFEC
NAFEC
NAFEC
Smoke
No,
31.74
30.83
33.06
32.13
31.27
30.90
28.53
31.06
31.67
31.41
32.42
30.78
29.03
26.86
20.31
20.48
—
wm
43.56
48.73
64.23
68.71
—
THC CO
pounds/K pounds
0.03
0.02
0.01
<0.01
0.09
0.01
0.01
<0.01
\-o.oi
^0.01
^0.01
^0.01
0.01
0.12
0.69
0.86
6.33
6.92
12.41
13.97
15.60
11.78
1.12
.55
24.18
2.16
2.34
1.17
1.16
0.77
0.83
0.70
0.73
1.16
1.21
1.72
1.75
3.25
3.57
10.53
11.53
32.92
35.27
50.83
54.06
63.36
47.79
15.13
10.30
65.59
NO
X
fuel
12.26
12.58
15.10
15.23
17.97
18.68
18.35
18.89
15.49
16.11
12.33
13.45
10.38
10.78
7.09
6.97
3.35
3.47
2.59
2.68
1.83
2.05
4.41
5.18
2.26
THC CO
pounds/hour
0.16
0.11
0.10
<0.03
0.73
0.10
0.10
<:0.03
<0.03
<0.03
<0.03
^0.03
0.05
0.56
2.03
2.51
8.12
8.87
11.86
13.36
16.01
15.11
2.65
1.61
24.81
13.02
14.09
8.13
8.11
6.26
6.82
5.82
6.09
8.06
8.46
10.36
10.66
15.30
16.81
31.18
33.60
42.20
45.21
48.59
51.68
65.01
61.31
35.65
30.03
67.30
NO
X
73.74
75.68
104.90
106.20
146.68
154.20
151.77
157.90
107.50
113.05
74.36
81.84
48.89
50.76
21.10
2 .31
4.29
4.45
2.48
2.56
1.88
2.63
10.39
15.10
2.32
THC CO NO
X
pounds/K pound thrust - hrs.
0.02
0.01
0.01
<0.01
0.05
0.01
0.01
<0.01
< 0.01
<0.01
0.01
<0.01
0.01
0.07
0.39
0.49
6.01
6.57
14.12
15.90
17.80
11.34
.68
.32
27.29
1.25
1.35
0.64
0.64
0.44
0.48
0.41
0.42
0.64
0.67
0.94
0.97
1.78
1.95
5.97
6.49
31.26
33.49
57.85
61.52
72.31
46.00
9.19
6.01
74.03
7.06
7.28
8.29
8.36
10.26
10.78
10.58
10.98
8.50
8.90
6.77
7.45 w
5.68 **
5.90
4.04
3.92
3.18
3.30
2.95
3.05
2.09
1.97
2.68
3.02
2.55
-------
Read.
No.
101
102
103
104
105
106
107
108
109
110
111
112
113
114
115
116
117
118
119
120
121
122
123
124
125
Power Probe
Condition Used
Idle NAFEC
P.P. NAFEC
Landing NAFEC
P.P. NAFEC
Cruise NAFEC
Max.Cont. NAFEC
Take Off NAFEC
Take Off NAFEC
Max. Cont. NAFEC
Idle NAFEC
Take Off NAFEC
Max. Cont. NAFEC
Max. Cont. NAFEC
Cruise NAFEC
Landing NAFEC
Approach NAFEC
Holding NAFEC
Idle NAFEC
Idle Traverse
Approach Traverse
Approach Traverse
Max.Cont. Traverse
Max.Cont. Traverse
Idle Traverse
Holding Traverse
Smoke
No.
—
71.48
74.49
73.75
75.23
76.70
83.29
78.78
77.61
45.06
78.11
80.50
77.41
76.70
75.90
69.36
51.50
48.35
—
—
—
—
—
—
—
THC
CO
pounds/K pounds
26.16
.76
.31
.21
.09
.19
18
.29
.45
26.35
.61
.26
.09
.06
.14
.30
• 8.92
24.25
19.50
.34
.34
10
.10
18.70
4.60
68.17
7.01
4.87
4.01
3.29
2.68
1.85
1.97
2.64
—
1.89
2.05
2.27
3.07
4.44
10.11
46.15
65.95
63.70
8.44
8.44
2.80
2.80
62.90
38.40
NO
X
fuel
2.08
6.19
7.48
7.95
9.04
10.01
12.61
12.32
10.26
2.15
12.27
11.47
11.14
9.03
7.46
5.30
2.81
2.37
2.56
5.66
5.66
11.02
11.02
2.52
3.05
THC
27.47
2.87
1.49
1.11
.55
1.35
1.60
2.54
3.18
25.80
5.37
2.11
.73
.39
.71
9.15
11.64
25.44
19.15
1.03
1.03
.74
.74
18.85
6.20
CO
pounds/hour
71.58
26.33
23.06
20.96
19.88
18.76
15.96
17.00
18.53
—
16.44
16.36
17.25
9.62
21.53
30.42
60.23
69.18
62.55
25.12
25.12
19.36
19.36
63.40
51.72
NO
X
2.18
23.25
35.43
41.55
54.63
70.06
108.79
106.28
72.00
2.10
106.71
91.52
84.68
55.37
36.18
15.95
3.67
2.49
2.51
16.84
16.84
76.21
76.21
2.54
44.12
THC
pounds/K
32.01
.44
..18
.12
.05
.11
.11
.17
.26
29.69
.37
.15
.05
.03
.08
1.83
9.04
29.65
22.61
.20
.20
.05
.05
21.97
4.41
CO
pound thrust
83.42
4.04
2.80
2.29
1.89
1.55
1.11
1.18
1.53
__
1.14
1.21
1.33
.90
2.60
6.08
46.76
80.63
73.85
4.82
4.82
1.55
1.55
73.90
36.84
NO
X
- hrs.
2.55
3.57
4.30
4.55
5.20
5.78
7.57
7.41
5.94
2.42
7.39
6.78
6.52
5.22
4.37
3.19
2.85
2.90
2.97
3.23
3.23
6.09
6.09
2.96
2.93
7
-------
Read.
No.
126
127
128
129
130
131
132
133
134
135
136
Power
Condition
Take Off
Take Off
Take Off
Take Off
Take Off
Take Off
Take Off
Approach
Idle
Idle
Idle
Probe Smoke
Used No.
NAFEC
NAFEC
NAFEC —
NAFEC
Traverse
Traverse
Traverse
Traverse
Traverse
Traverse
Traverse
THC
CO
NO
pounds/K pounds fuel
THC CO NO THC CO NO
x x
pounds/hour pounds/K pound thrust-hrs.
w
For Emissions Data obtained during Reading Numbers 137-155, see Table 6-2, Probe Comparison Data
-------
APPENDIX C
JT8D
Engine Operating Data
Read
NO.
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
Power
Setting
Idle
Idle
—
—
—
—
Idle
Approach
—
Landing
Cruise
Idle
—
—
Approach
Landing
—
Cruise
—
Max.Cont.
—
Idle Hot
Idle
Idle
Thrust
(pounds)
745
750
805
850
1485
1740
740
5050
6680
8100
10450
760
1380
3920
5150
6650
8460
9620
11000
12350
12610
1391
930
800
800
EPR
1.041
1.042
1.042
1.042
1.076
1.093
—
1.297
1.550
1.503
1.671
1.041
1.073
1.222
1.304
1.412
1.544
1.629
1.729
1.845
1.879
1.979
1.041
1.041
1.041
Fuel
Flow
Ibs./hr
998
994
998
1042
1412
1582
991
2980
3872
4602
5828
985
1352
2334
2984
3846
4773
5367
6164
6945
7189
8051
984
1001
985
WA Air
Flow
Ibs/sec
22
17
24
15
59
66
4
196
213
247
272
69
110
183
209
243
262
277
288
301
160
313
90
81
80
Nl
rpm
2468
2516
2591
2654
3407
3646
2486
5577
6172
6612
7107
2434
3207
4990
5556
6119
6630
6865
7162
7442
7523
7815
2468
2509
2507
N2
rpm
6433
6528
6636
6730
7715
8000
6451
9568
10001
10295
10670
6337
7398
9063
8480
9334
10231
10415
10617
10805
10860
11039
6371
6475
6474
PT7
(in.Hg)
1.25
1.25
1.25
1.25
2.25
2.75
1.25
10.50
22.00
22.00
22.00
1.25
2.25
6.75
9.25
12.50
16.50
19.00
22.00
25.50
26.50
29.50
1.25
1.25
1.25
TT
650
685
680
680
700
700
650
300
330
350
660
680
610
650
715
760
800
840
875
880
920
650
690
690
TT
55
70
67
67
67
65
55
55
57
55
55
43
43
43
45
45
45
45
45
45
45
45
45
55
55
Barometric
Pressure
(in.Hg)
30.56
29.79
30.06
30.03
30.02
30.02
29.99
29.99
29.97
29.97
29.97
30.46 ~
30.46
30.46
30.46
30.46
30.46
30.44
30.44
30.44
30.44
30.44
30. 44
30.09
30.09
-------
Read
No.
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
Power
Setting
Approach
Approach
Cruise
Cruise
Idle Hot
Idle
Take Off
Max.Cont.
Max. Con t.
—
Cruise
—
Landing
Approach
Idle Hot
Idle
Take Off
Cruise
Approach
Take Off
Idle
Approach
Landing
Cruise
FiP.
Thrust
(pounds)
5280
5280
10950
1D950
760
775
14800
12900
13000
12350
11150
9700
9100
5700
850
965
14422
11055
5477
14593
1015
5399
8885
11137
12731
EPR
1.308
1.308
1.744
1.744
1.033
1.033
2.384
1.899
1.892
1.849
1.707
1.616
1.574
1.323
1.025
1.042
2.046
1.714
1.302
2.064
1.042
1.300
1.554
1.732
1.876
Fuel
Flow
Ibs/hr.
3086
3069
6201
6230
961
-986
8994
7352
7352
6953
6286
5431
5107
3285
985
963
8500
6171
2977
8494
8494
2942
4763
5975
6938
WAAir
Flow
Ibs/sec.
213
212
288
—
85
74
321
307
306
301
289
271
264
215
80
76
321
293
211
322
85
210
264
286
303
Nl
ppm
5673
5676
7232
2448
2482
8117
7600
7588
7457
7200
6892
6784
5776
2490
2465
7950
7965
5566
7994
2511
5538
6595
7085
7418
N2
ppm
9628
9634
10720
—
6374
6423
11205
10922
10917
10829
10666
10453
10361
9667
6427
6374-
. 11081
10564
9470
11118
6438
9412
10160
10517
10721
PT
(in-.Hg)
9.25
9.25
22.25
—
it-oo
1.00
32.50
27.00
26.77
25.50
21.25
18.50
17.25
9.75
0.75
1.25
31.00
22.00
9.00
31.50
1.25
9.00
16.50
21.75
26.00
TT
(°F)
685
685
660
—
690
680
980
900
000
000
000
000
000
000
000
660
940
830
650
940
645
640
750
810
850
TT2
(°F)
55
55
55
—
55
45
47
47
47
45
47
4?
47
48
48
40
40
40
40
40
30
30
30
30
30
Barometric
Pressure
(in. Hg)
30.14
30.14
30.14
—
30.14
30.31
30.31
30.31
30.30
30.30
30.30
30.28
30.28
30.28
30.28
29.82
29.85
29.85
29.87
29.87
30.07
30.07
29.98
29.98
29.98
o
-------
Read
No.
-•*-"* *-
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
Power
Spff Ing
Take Off
Idle
Holding
P.P.
Approach
Approach
Idle
Holding
Idle
Idle
Idle
P.P.
P.P
P.P.
Approach
Approach
Approach
P.P.
P.P.
P.P.
Landing
Landing
Landing
P.P.
P.P.
Thrust
(pounds)
8592
996
1375
3800
5000
4710
745
1130
970
1052
1100
4100
4100
4130
5250
5200
5200
6700
6750
6750
8300
8300
8300
9000
9050
EPR
2.038
1.033
1.051
1.188
1.281
1.255
1.028
1.044
1.040
1.040
1.040
1.208
1.208
1.208
1.285
1.285
1.282
1.394
1.395
1.397
1.512
1.512
1.512
1.577
1.577
Fuel
Flow
Ibs/hr.
8244
983
1356
2367
3030
3006
1057
1337
1002
1004
1002
2355
2355
2355
2988
2984
2964
3796
3819
3823
4709
4733
4709
5176
5234
W. Air
A
Flow
Ibs/sec.
319
82
114
173
155
94
114
80
80
77
178
179
179
204
204
203
231
232
232
257
257
255
260
264
Nl
ppm
7888
2638
3306
5200
5760
5810
2657
3220
2438
2430
2434
4856
4868
4866
5414
5292
5391
5990
5999
6006
6466
6467
646
6689
6703
N2
ppm
11020
6620
7489
7174
9600
9590
6677
7395
6315
6304
6307
8891
8906
8900
9317
9305
9303
9730
9736
9736
10057
10055
10053
10215
10232
PT
(in.Hg)
30.80
1.00
1.55
5.60
8.30
7.60
0.85
1.30
1.20
1.20
1.20
6.25
6.25
6.25
8.55
8.55
8.45
11.80
11.85
11.90
15.35
15.35
15.35
17.20
17.20
TT?
(°F)
915
650
660
628
665
557
660
670
650
650
650
595
610
610
650
640
640
685
690
690
730
725
730
750
760
TT2
(°F)
30
42
42
53
53
50
50
48
33
33
33
32
32
32
29
29
30
30
30
28
28
30
30
26
30
Barometric
Pressure
29.98
30.16
30.16
30.04
30.04
29.95
29.84
29.84
30.15
30.15
30.15
30.15
30.15
30.15
30.15
30.15
30.15
30.15
30.15
30.15
30.15
30.15
30.19
30.01
30.01
-------
Read,
No.
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
93
94
95
96
97
98
99
100
Power
Setting
Cruise
Cruise
Max.Cont.
Max.Cont.
Take. Off
Take Off
Take Off
Take Off
Max.Cont.
Max.Cont.
Cruise
Cruise
Landing
Landing
Approach
Approach
Holding
Holding
Idle
Idle
Idle
Holding
P.P.
Approach
Idle
Thrust
(pounds)
10450
10400
12650
12700
14300
14300
14350
14380
12650
12700
10990
10990
8600
8600
5220
5180
1350
1350
840
840
899
1333
3878
5000
909
EPR
1.680
1.681
1.885
1.885
2.048
2.047
2.049
2.049
1.887
1.888
1.738
1.738
1.550
1.551
1.303
1.303
1.070
1.070
1.059
1.059
1.043
1.065
1.209
1.281
1.043
Fuel
Flow
Ibs./hr,
6015
6015
6948
6971
8161
8254
8270
8358
6940
7018
6031
6085
4709
4709
2961
2915
1282
1282
956
956
1026
1283
2356
2916
1026
WAAir
Flow Nt
. Ibs/sec. ppm
287
280
303
305
319
318
319
319
304
303
287
288
257
256
218
217
114
113
92
91
N/A
N/A
N/A
N/A
N/A
6960
6976
7467
7490
7908
7907
7962
7948
7503
7518
- 7122
7137
6611
6613
5550
5523
3148
3140
2513
2497
2476
3020
4845
5362
2527
N2
Ppm
10427
10438
10756
10773
11049
11034
11034
11080
10802
10808
10556
10568
10187
10179
9442
9420
7299
7289
6433
6419
6375
7092
8846
9247
6432
PT7
(in.Hg)
20.25
20.25
26.30
26.30
31.00
31.00
31.00
31.00
26.25
26.25
21.85
21.85
16.30
16.30
9.10
9.10
2.10
2.10
1.75
1.75
1.30
1,95
6.30
8.45
1.30
TT
(°F)
800
800
865
870
930
925
930
940
875
875
820
830
750
750
645
645
665
665
650
645
650
640
610
625
645
H2
(°F)
30
30
31
31
30
32
35
32
34
32
32
33
33
32
32
32
33
35
35
35
25
25
25
25
25
Barometric
Pressure
30.01
30.01
30.01
30.01
29.93
29.93
29.88
29.88
29.88
29.88
29.88
29.88
29.88
29.88
29.85
29.85
29.85
29.85
29.85
29.85
30.17
30.17
30.17
30.17
30.17
0
-------
Read.
No.
101
102
103
104
105
106
107
108
109
110
111
112
113
114
115
116
117
118
119
120
121
122
123
124
125
Power
Setting
Idle
P.P.
Landing
P.P.
Cruise
Max.Cont.
Take Off
Take Off
Max.Cont.
Idle
Take Off
Max.Cont.
Max.Cont.
Cruise
Landing
Approach
Holding
Idle
Idle1
Approach
Approach
Max.Cont.
Max.Cont.
Idle
Holding
Thrust
(pounds)
858
6514
8232
9140
10504
12120
14362
14342
12120
869
14443
13484
12978
10605
8282
5000
1288
858
847
5207
5213
12519
12519
858
1404
EPR
1.043
1.397
1.518
1.585
1.690
1.833
2.025
2.022
1.819
1.041
2.012
1.934
1.887
1.690
1.513
1.284
1.064
1.041
1.041
1.291
1.302
1.873
1.873
1.042
1.070
Fuel
Flow
Ibs./hr.
1050
3756
4736
5226
6043
6999
8627
8627
7018
979
8697
7979
7601
6132
4850
3009
1305
1049
982
2991
1961
6916
6916
1008
1347
WAAir
Flow
Ibs. /sec.
N/A
N/A
N/A
N/A
N/A
N/A
N/A
N/A
N/A
N/A
N/A
N/A
N/A
N/A
N/A
N/A
N/A
N/A
N/A
N/A
212.0
303.8
303.8
81.5
110.0
Nl
ppm
2498
5957
6483
6680
6963
7320
7835
7833
7310
2510
7866
7627
7495
6985
6508
5407
3034
2506
2509
5507
5604
7440
7440
2540
3256
N2
ppm
6405
9686
10051
10208
10419
10646
10974
10974
10644
6454
11024
10871
10794
10470
10112
9339
7148
6435
6485
9435
9580
10743
10743
6514
7475
PT?
( in.Hg)
1.30
11.70
15.50
17.50
20.60
24.90
30.60
30.50
24.50
1.25
30.50
28.10
26.70
20.80
15.50
8.60
1.95
1.25
1.26
8.77
8.94
25.62
25.62
1.25
2.11
TT
(°F)
655
680
730
760
805
840
910
910
835
680
930
900
880
820
745
650
660
670
648
621
652
824
824
663
661
TT2
(°F)
28
27
27
27
30
28
30
30
28
37
35
34
34
35
35
35
35
35
42
36
54
29
29
46
47
Barometric
Pressure
30.10
30.10
30.10
30.10
30.10
30.18
30.18
30.18
30.18
30.38
30.39
30.39
30.39
30.39
30.41
30.41
30.41
30.41
30.23
30.16
29.63
29.68
29.68
30.13
30.07
1) Traverse
-------
Read.
No.
126
127
128
129
130
131
132
133
134
135
136
137
138
139
140
141
142
143
144
145
146
147
148
149
150
151
152
153
154
155
Power
Setting
Take Off1
Take Off
Take Off
Take Off
Take Off
Take Off
Take Off
Approach
Idle
Idle
Idle
Idle
Idle
Idle
Idle
Idle
Idle
Idle
Idle
Idle
Idle
Idle
Idle
Idle
Idle
Idle
Idle
Idle
Idle
Idle
Thrust
(pounds)
143062
143062
143062
143062
143062
143062
143062
4646
808
859
808
808
808
808
758
783
808
808
859
808
782
707
798
798
828
828
828
808
848
848
EPR
2.0192
2.0192
2.0192
2.0192
2.0192
2.0192
2.0192
1.269
1.040
1.043
1.040
1.040
1.040
1.040
1.040
1.040
1.041
1.041
1.040
1.041
1.040
1.040
1.040
1.040
1.040
1.040
1.040
1.041
1.043
1.043
Fuel
Flow
Ibs./hr.
82942
82942
82942
82942
82942
82942
82942
2671
975
998
998
998
1003
1021
998
998
998
998
998
1021
998
998
975
998
1021
998
998
998
998
975
WA Air
Flow
Ibs./sec.
3182
3182
3182
3182
3182
3182
3182
200
83
85
81
81
81
81
81
81
83
81
85
83
83
81
83
83
83
81
81
81
81
81
Nl
ppra
79452
79452
79452
79452
70452
79452
79452
5327
2468
2619
2532
2519 '
2537
2547
2339
2530
2523
2526
2522
2515
2471
2487
2477
2487
2522
2509
2497
2462
2485
2465
N2
ppm
111672
111672
111672
111672
111672
111672
111672
9271
6384
6575
6499
6480
6512
6524
6513
6483
6454
6458
6454
6439
6373
6403
6379
6398
6453
6431
6440
6377
6412
6384
PT?
(in.Hg)
3D.272
30. 272
30. 272
30. 272
30. 272
30. 272
30. 272
7.95
1.20
1.27
1.20
1.20
1.20
1.20
1.20
1.20
1.25
1.25
1.25
1.25
1.20
1.20
1.20
1.20
1.25
1.20
1.25
1.20
1.25
1.25
TT_
(°F)
9292
9292
9292
9292
9292
9292
9292
600
640
630
655
655
660
660
660
655
630
630
630
630
630
630
620
625
630
650
650
650
650
650
TT2
(°F)
462
/,62
462
462
462
462
462
31
32
38
45
45
48
47
44
44
33
33
34
34
32
33
33
34
33
43
42
42
42
41
Barometric
Pressure
30. OO2
30. OO2
30. OO2
30. OO2
30. OO2
30. OO2
30. OO2
29.72
29.72
29.67
30.23
30.23
30.23
30.23
30.20
30.23
30.25
30.25
30.24
30.24
30.19
30.19
30.20
30.20
30.33
30.32
30.32
30.30
30.30
30.30
n
i
1. Traverse
2. Average for traverse made during reading numbers 126-132.
-------
TECHNICAL REPORT DATA
(Please read Instructions on the reverse be/ore completing)
REPORT NO.
EPA-460/3-74-006
2.
3 RECIPIENT'S ACCESSIOWNO.
4 TITLE AND SUBTITLE
Variability in Aircraft Turbine Engine Emission
Measurements
lfanuazyTi974
6. PERFORMING ORGANIZATION CODE
7 AUTHOR(S)
8 PERFORMING ORGANIZATION REPORT NO.
Anthony F. Souza and Louis R. Reckner
9 PERFORMING ORGANIZATION NAME AND ADDRESS
Scott Research Laboratories, Inc.
Plumsteadville, Pa. 18949
10 PROGRAM ELEMENT NO.
11 CONTRACT/GRANT NO
68-01-0443
12 SPONSORING AGENCY NAME AND ADDRESS
U.S. Environmental Protection Agency
Office of Air and Water Programs
Ann Arbor, Mich. 48105
13. TYPE OF REPORT AND PERIOD COVERED
TYPE OF
Final
14. SPONSORING AGENCY CODE
15. SUPPLEMENTARY NOTES
16. ABSTRACT
Under Environmental Protection Agency Contract Number 68-01-0443, Scott Research
Laboratories, Inc., studied the causes of variability in turbine engine emission
measurements. A state-of-the-art analysis system was built according to the specifi-
cations of the contract. The analysis system was evaluated for reliability in the
handling and accuracy in the measurement of emissions.
Using the special analysis system, the variability in the gas turbine emission
measurements caused by the gas sample collection technique was studied using a Pratt
and Whitney JT8D engine. An exhaust gas mixer and a detailed exhaust gas cross-
section mapping technique were utilized for the verification of average exhaust
emission concentrations.
The effects of approach to power setting and the effects of variations in fuel
flow and thrust on the measurement of mass emission rates are discussed. The effects
of ambient temperature and humidity on turbine engine emissions are examined.
17.
KEY WORDS AND DOCUMENT ANALYSIS
DESCRIPTORS
b.IDENTIFIERS/OPEN ENDED TERMS C. COSATI Field/Group
Turbine engine
Aircraft emissions
18. DISTRIBUTION STATEMENT
Unlimited
19. SECURITY CLASS (ThisReport)
Unclassified
21. NO OF PAGES
96
20 SECURITY CLASS (This page)
Unclassified
22. PRICE
EPA Form 2220-1 (9-73)
D-l
------- |