Zero gas as specified for the CO, CO2 , and HC analyzers shall be zero air (which includes "artificial" air with 20 to 22 per cent O2 blended with N2). For the NOX analyzer zero nitrogen shall be used as the zero gas. Impurities in both kinds of zero gas shall be restricted to be less than the following concentrations:
1 ppm C 1 ppm CO 100 ppm CO2 1 ppm NOx
The applicant shall ensure that commercial gases supplied to him do in fact meet this specification, or are so specified by the vendor.
Supplemental information Calibration and test gases are readily available, with traceability to the National Institute of Standards and Technology (NIST), from commercial suppliers. Gases with an accuracy of ± 1% are considered “primary” standards and intended for use where the highest accura cy and reliability are needed. Very tight controls over cylinder material, age, and history are considered good practice, if not essential. Periodic checks of the accuracy of assay for each cylinder is advisable. Gases such as CO, CO 2, and propane with a certified accuracy of ± 2% can be obtained as multi-component mixtures. This will reduce the number of gas cylinders needed for testing. The importance of this becomes obvious when considering the calibration requirements for each instrument and each range. Managing just the minimum number of calibration and test gases can be challenging. For example, an average certification test, when backup gases are counted, can require 100 two component cylinders, or more. The stability of the components in each mixture should be checked periodically. If the assay no longer agrees with the ce rtified analysis, the cylinder should be returned to the supplier or discarded as appropriate.
Because of the accuracy required and the sophistication of the techni ques necessary to produce and analyze gases to these standards, most testing laboratories rely on the commercial suppliers analysis and certification for primary accuracy and use in-house checks via instru ment response for consistency of assay. It is good practice to recheck all calibration and working standard gases immediately prior to a certification test.
All calibration gas monitoring data should be kept and made available to the certificating agency if required.
Although not a calibration gas or, strictly speaking, a test gas, the FID analyzer f uel and oxidant gases should also meet a hydrocarbon specification. For hydr ogen/nitrogen or hydrogen/helium fuel mixtures, total hydrocarbons present should be < 1 ppmC. The oxidant should be hydrocarbon free grade air, containing < 1 ppmC hydrocarbon. The SAE in ARP 1 256 restricts the FID fuel gas to a 40% hydrogen, 60% helium mixture which is required to meet this hydrocarbon level specification.
sample per mole of fuel sample per mole of fuel ATTACHMENT E TO ICAO MANUAL APPENDIX 3. THE CALCULATION OF THE EMISSIONS
PARAMETERS - BASIS, MEASUREMENT CORRECTIONS AND
ALTERNATIVE NUMERICAL METHOD
1. SYMBOLS P7 number of moles of NO 2 in the exhaust sample per mole of fuel AFR air/fuel ratio, the ratio of the mass flow rate of dry air to that of the fuel P8 number of moles of NO in the exhaust sample per mole of fuel El emission index; l03 x mass flow rate of gaseous emission product in exhaust per unit mass flow rate of fuel PT P1 + P2 + P3 + P4 + P5 + P6 + P7 + P8 K ratio of concentration measured wet to R concentration of O2 in dry air, by volume = 0.2095 normally S concentration of N 2 + rare gases in dry air, by volume = 0.7902 normally T concentration of CO2 in dry air, by volume = 0.0003 normally P0 number of moles of air per mole of fuel in initial air/fuel mixture Z symbol used and defined in 3.4 MHC molecular mass of exhaust [CO2] mean concentration of CO 2 in exhaust that measured dry (after cold trap) L, L' analyzer interference coefficient for interference by CO2 M, M' analyzer interference coefficient for interference by H2O MAIR molecular mass of dry air = 28.966 g or, where appropriate, = (32 R +
28.156 4 S + 44.011 T) g MCO molecular mass of CO = 28.011 g hydrocarbon, taken as CH4 = 16.043 g sample, vol/vol MNO2 molecular mass of NO2 = 46.008 g [CO] mean concentration of CO in exhaust sample, vol/vol MC atomic mass of carbon = 12.011 g [HC] sample, vol C/vol mean concentration of HC in exhaust MH atomic mass of hydrogen = 1.008 g [NO] sample, vol/vol mean concentration of NO in exhaust P1 number of moles of CO2 in the exhaust [NO2] mean concentration of NO2 in exhaust sample, vol/vol P2 number of moles of N 2 in the exhaust [NOX] mean concentration of NO and NO 2 in exhaust sample, vol/vol P3 number of moles of O 2 in the exhaust [NOX]c mean concentration of NO in exhaust sample per mole of fuel sample. after passing through the N02/NO converter, vol/vol P4 sample per mole of fuel number of moles of H2O in the exhaust [NO2] mean = ([NOX]c -[NO]) η P5 number of moles of CO in the exhaust [ ] d mean concentration in exhaust sample after cold trap, vol/vol [ ] m mean concentration measurement indicated before instrument correction P6 sample per mole of fuel number of moles of C xHy in the exhaust sample per mole of fuel applied, vol/vol Page 74
h humidity of ambient air, vol water/vol dry air hd humidity of exhaust sample leaving "drier" or "cold trap", vol water/vol dry sample rn number of C atoms in characteristic fuel molecule n number of H atoms in characteristic fuel molecule x number of C atoms in characteristic exhaust hydrocarbon molecule y number of H atoms in characteristic exhaust hydrocarbon molecule η efficiency of NO2/NO converter
2. BASIS OF CALCULATION OF El AND AFR PARAMETERS
2.1 It is assumed that the balance between the orig inal fuel and air mixture and the resultant state of the exhaust emissions as sampled can be represented by the following equation:
CmHn + P0[R(O2) + S(N2) + T(CO2) + h(H2O)] = P1(CO2) + P2(N2) + P3(O2) + P4(H2O) + P 5(CO) + P6(CxHy) + P7(NO2) + P8(NO) from which the required parameters can, by definition, be expressed as ⎛ 103 ⎞ ⎜ MCO ⎟ EI(CO) = P5 ⎜ ⎟mM + nM ⎜ ⎟C H⎝ ⎠ ⎛ 103 ⎞ ⎜ MHC ⎟ EI(HC) = xP6 ⎜ ⎟mM + nM⎜ ⎟C H⎝ ⎠ expressed as methane equivalent ⎡ 103M ⎤ ⎢ NO ⎥2EI(NO ) = ( P + P ) ⎢ ⎥X 7 8 ⎢ mM + nM ⎥C H⎢ ⎥⎣ ⎦ expressed as NO2 equivalent ⎛ MAIR ⎞AFR = P0 ⎜ ⎟⎝ mMC = nMH ⎠ Page 75
Supplemental information This is a slightly different formulation for AFR than that stated in 7.1.2, “Basic parameters”. In this formulation m, the “number of C atoms in characteristic fuel molecule” is placed within the bracket. There is no particular advantage to using one formulation over the other.
2.2 Values for fuel hydrocarbon composition (rn, n) are assigned by fuel specification or analysis. If only the ratio n/m is so determined, the value m = 12 ma y be assigned. The mole fractions of the dry air constituents (R, S, 7) are normally taken to be the recommended standard values but alternative values may be assigned, subject to the restriction R + S + T = 1 and the approval of the certificating authority.
2.3 The ambient air humidity, h, is as measured at each test condition. It is recommended that, in the absence of contrary evidence as to the characterization (x, y) of the exhaust hydrocarbon, values of x = I and y = 4 are assigned.
2.4 Determination of the remaining unknowns requ ires the solution of the following set of linear simultaneous equations, where (1) to (4) derive from the fundamental atomic conservation relationships and (5) to (9) represent the gaseous product concentration relationships.
(1) M = TP0 = P1 + P5 + xP6
(2) n + 2hP0 = 2P4 + yP6
(3) (2R + 2T + h)P0 = 2P1 + 2P3 + P4 + P5 + 2P7 + P8
(4) 2SP0 = 2P2 + P7 + P8
(5) [CO2] PT = P1
(6) [CO] PT = P5
(7) [HC] PT = Xp6
(8) [NOX]C PT = �P7 + P8
(9) [NO] PT = P8
(10) PT = P1 + P2 + P3 + P4 + P5 + P6 + P7 + P8 The above set of conditional equations is for the ca se where all measured concentrations are true ones, that is, not subject to interference effects or to the need to correct for sample drying. In practice, interference effects are usually pres ent to a significant degree in the CO, NO X and NO measurements, and the option to measure CO2 and CO on a dry or partially dry basis is often used.
The necessary modifications to the relevant equations are described in 2.5 and 2.6.
2.5 The interference effects are mainly caused by the presence of CO2 and H2O in the sample which can affect the CO and the NOX analysers in basically different ways. The CO analyzer is prone to a zero-shifting effect and the NOX analyzer to a sensitivity change, represented thus:
[CO] = [CO]m + L[CO2] + M[H2O] and [NOX]c = [NOX]cm(1 + L'[CO2] + M'[H2O]) Page 76
which transform into the following alternative equati ons to (6), (8) and (9), when interference effects require to be corrected, (6A) [CO]m PT + LP1 + MP4 = P5 (8A) [NOX]cm (PT +L'P1 +M'P4) = �P7 + P8 (9A) [NO]m (PT + L'P1 + M'P4) = P8 The option to measure CO2 and CO concentrations on a dry or partially dry sample basis, that is, with a sample humidity reduced to h,, requires the use of modified conditional equations as follows:
(5A) [CO2]d(PT - P4)(1 + hd) = P1 and [CO]d (PT - P4)(1 + hd) + P5 However, the CO analyzer may also be subject to interference effects as described in 2.5 above and so the complete alternative CO measurement concentration equation becomes (6B) [CO]md (PT - P4)(1 + hd) + LP1 + Mhd (PT - P4) = P5
3. ANALYTICAL FORMULATIONS
3.1 General Equations (1) to (10) can be reduced to yield the analytical formulations for the EI and AFR parameters, as given in 7.1 to this appendix. This reduction is a process of progressive e limination of the roots P 0, P1 through P8, PT making the assumptions that all concentrati on measurements are of the "wet" sample and do not require interference corrections or the like. In practice the option is often chosen to make the CO2 and CO concentration measurements on a "dry" or "semi-dry" basis; also it is often found necessary to make interference corrections. Formulations for use in these various circumstances are given in 3.2, 3.3 and 3.4 below.
3.2 Equation for conversion of dry concentration measurements to wet basis Concentration wet = K x concentration dry; that is.
[ ] = K [ ] d The following expression for K applies when CO and CO2 are determined on a "dry" basis:
{4+(nm T / ) +(| n m T | −2h ( N O 2 − 2 H C / x ) + 2+h yx | | | − n m )[H C ]}(1+h )) [ ] ( [ ] ) ( )( d K = (2+h){2+(n m ) ( 1+h )([C O ]d )} −(| | 2 )( − h [ ] d )/ d 2 n m T − h 1 | 1 + d | C O Page 77
3.3 Interference corrections The measurement of CO and/or NO may require corrections for interference by the sample CO2 and water concentrations before use in the above analytic al equations. Such corrections can normally be expressed n the following general ways:
[ ] [CO]m + L[CO2] + M [ 2 ]CO = H O ⎛ hd ⎞CO = CO + L CO + M[][] [ ] 1 hd md 2 d ⎜ ⎟ ⎝ + d ⎠ [NO] = [NO]m (1+ L '[CO ] + M '[H O ])2 2 η[NO ] = [NO ] −[NO] 1+ L '[CO ] + M '[H O ]2 X 2 2( cm m )( )
3.4 Equation for estimation of sample water content (|2 + | P 0 Water concentration in sample m |)([C O ]+ [C O ]+ [H C ])[HO ] = nmh 2 − ( yx2 H C )[ ]2 1+T (P0 m) Where 2Z n m− ( ) m =P0 ( +− | Z41 hT 2 | ) and 2C O −[ ]− (| 2 x | − | y x 2 | )[H C ] +[N O 2 ]Z = [ ] [] [] CO2 + CO + HC It should be noted that this estimate is a function of the various analyses concentration readings, which may themselves require water interference correction. For better accuracy an iterative procedure is required in these cases with successive recalculation of the water concentration until the requisite stability is obtained. The use of the alternative numerical solution methodology (4) avoids this difficulty.
4. ALTERNATIVE METHODOLOGY -NUMERICAL SOLUTION
4.1 As an alternative to the analytical procedures summarized in 3 above, it is possible to obtain readily the emissions indices, fuel/air ratio, co rrected wet concentrations, etc., by a numerical solution of equations (1) to (10) for each set of measurements, using a digital computer.
4.2 In the equation set (1) to (10) the actual concentration measurements are substituted using whichever of the alternative equations (5A), (6A), etc. applies for the particular measuring system, to take account of interference corrections and/or dried sample measurements.
4.3 Suitable simple two-dimensional array equation-solving computer programmes are widely available and their use for this purpose is convenient and flexible, allowing ready incorporation and identification of any sample drying options and interference or other corrections.
Supplemental information Details explaining various calculation procedures can be found in AeroSpace Recommended Practice (ARP) 1533, Procedure for the Calculation of Gaseous Emissions from Aircraft Turbine Engines. ARP 1533 includes, among other things, the combustion chemical equations, the basic matrix solution with computer programs, derivation of equations and a sample calculation.
[THIS PAGE INTENTIONALLY LEFT BLANK]
XI. ICAO MANUAL APPENDIX 5
INSTRUMENTATION AND MEASUREMENT TECHNIQUES
FOR GASEOUS EMISSIONS FROM
AFTERBURNING GAS TURBINE ENGINES
1. INTRODUCTION Note.- The procedures specified in this appendix are concerned with the acquisition of representative exhaust samples and their tran smission to, and analysis by, the emissions measuring system. These procedures only appl y when afierburning is employed. The methods proposed are representative of the best readily available and most established modern practice.
The need to correct for ambient conditions is recognized and a method will be specified when one becomes available. Meanwhile any correction methods used when afterburning is employed should be approved by the certificating authority.
Variations in the procedure contained in this appendix shall only be allowed after prior application to and approval by the certificating authority.
Supplemental information This sampling and analytical procedure only applies to afterburning engines, while operating in the afterburning mode. Smoke and gaseous emissions levels in the exhaust plume of these engines, while in the non-afterburning mode, are determined using the standard procedures described in ICAO Annex 16, Volume II, Appendices 2 and 3. Smoke is not measured during afterburning operation.
Experience in sampling and analyzing emissions in the plume of afterburning engines is sparse, and the resultant methodology not as well proven as for non-afterburning engines. In fact no effort, similar to that which resulted in SAE AeroSpace Recomm ended Practice ARP 1256 , for non-afterburning engines, has emerged for the afterburning case.
Sampling in the same region as required for non-afterburning engines would involve highly cooled and robust sampling probes, an understanding of the reaction chemistry in the exhaust plume and a detailed history of the gaseous species as they proceeded downstream to a point where the reaction could be considered complete or frozen. An alternative approach, and that adopted in ICAO Appendix 5, is to sample sufficiently far downstream to be in the well mixed, turbulent and relatively cool region of the plume where all reactions, for all practical purposes, have ceased. Having taken this approach, ICAO does make allowance for al ternatives, providing their viab ility can be demonstrated to the certificating authority.
The analytical instrumentation used for non-a fterburning engines has been found suitable for measuring gaseous emissions from afterburning engines with due care to be taken for measuring low concentration levels resulting from mixing the engine exhaust plume with ambient air. This mixing, with its resulting low concentrations suggests that close attention be paid to the ambient levels of CO, CO2, HC, and NO X. Proven methods for incorporating ambient corrections into these calculations have not been fully developed and are left to the testing agency to develop, use and validate.
The sampling and analytical methods recommended in Appendix 5 were likely influenced by the work of T. F. Lyon, et al, on military engines, (under U.S. Air Fo rce contract – AFAPL-TR-75-52, “Development of Emissions Measurement Techniques for Afterburning Turbine Engines”, T. F. Lyons, et al, General Electric Company, 1 1 Apr 73 – 31 Mar 75). Considerable insight can be gained from this report particularly in regard to sampling and validating the data. Validating data obtained in the far plume requires the assumption that the individual species CO, HC, and NO X, relative to CO2, form Page 81
linearly related, bivariate Gaussian normal data sets to a high degree and for which correlation coefficient calculations are meaningful. The Lyon, et al, data seems to bear out this assumption.
2. DEFINITIONS Where the following expressions are used without further explanation in this appendix, they have the meanings ascribed to them below:
Accuracy. The closeness with which a measurement approaches the true value established independently.
Calibration gas. A high accuracy reference gas to be used for alignment, adjustment and periodic checks of instruments.
Concentration. The volume fraction of the component of interest in the gas mixture - expressed as volume percentage or as parts per million.
Flame ionization detector. A hydrogen-air diffusion flame detector that produces a signal nominally proportional to the mass-flow rate of hydrocarbons entering the flame per unit of time - generally assumed responsive to the number of carbon atoms entering the flame.
Interference. Instrument response due to presence of components other than the gas (or vapour) that is to be measured.
Noise. Random variation in instrument output not associated with characteristics of the sample to which the instrument is responding, and distinguishable from its drift characteristics Non-dispersive infra-red analyzer. An instrument that by absorption of infra-red energy selectively measures specific components.
Parts per million (ppm). The unit volume concentration of a gas per million unit volume of the gas mixture of which it is a part.
Parts per million carbon (ppmC). The mole fraction of hydrocarbon multiplied by 10 6 measured on a methane-equivalence basis. Thus, 1 ppm of methane is indicated as 1 ppmC. To convert ppm concentration of any hydrocarbon to an equivalent ppmC value, multiply ppm concentration by the number of carbon atoms per molecule of the gas. For example, 1 ppm propane translates as 3 ppmC hydrocarbon; 1 ppm hexane as 6 ppmC hydrocarbon Plume. Total external engine exhaust flow, including any ambient air with which the exhaust mixes.
Reference gas. A mixture of gases of specified and kn own composition used as the basis for interpreting instrument response in terms of the concentration of the gas to which the instrument is responding.
Repeatability. The closeness with which a measurement upon a given, invariant sample can be reproduced in short-term repetitions of t he measurement with no intervening instrument adjustment.
Resolution. The smallest change in a measurement which can be detected.
Response. The change in instrument output signal that occurs with change in sample concentration. Also the output signal corresponding to a given sample concentration.
Stability. The closeness with which repeated measurements upon a given invariant sample can be maintained over a given period of time.
Zero drift Time-related deviation of instrument output from zero set point when it is operating on gas free of the component to be measured.
Zero gas. A gas to be used in establishing the zero, or no-response, adjustment of an instrument.
3.1 Gaseous emissions Concentrations of the following emissions shall be determined:
a) Hydrocarbons (HC): a combined estimate of all hydrocarbon compounds present in the exhaust gas.
b) Carbon monoxide (CO).
c) Carbon dioxide (CO2) Note.- CO2 is not considered a pollutant but its concentration is required for calculation and check purposes.
d) Oxides of nitrogen (NOx): an estimate of the sum of the two oxides, nitric oxide (NO) and nitrogen dioxide (NO2).
e) Nitric oxide (NO).
3.2 Other information In order to normalize the emissions measurem ent data and to quantify the engine test characteristics, other information in addition to the requirements of Chapter 3, 3.4 shall be provided as follows:
- inlet temperature;
- inlet humidity;
- atmospheric pressure;
- wind vectors relative to engine exhaust axis;
- hydrogen/carbon ratio of fuel;
- engine installation details;
- other required engine parameters (for ex ample, thrust, rotor speeds, turbine temperatures);
- pollutant concentration data and statistical validation parameters.
This data shall be obtained either by direct meas urement or by calculation, as presented in Attachment F to this appendix.
4. GENERAL ARRANGEMENT OF THE SYSTEM Owing to the reactive nature of the exhaust plume from engines using afterburning, it is necessary to ensure that the measured emissions do in fact correspond to those actually emitted into the surrounding atmosphere. This is achi eved by sampling the plume sufficiently far downstream from the engine that the exhaust gases have cooled to a temperature where reactions have ceased. No desiccants, dryers, water traps or related equipment shall be used to treat the exhaust sample flowing to the oxides of nitrogen and the hydrocarbon analysis instrumentation. Requirements for the various component sub-systems are given in 5, but the Page 83
following list gives some qualifications and variations:
a) it is assumed that each of the various in dividual subsystems includes the necessary flow control, conditioning and measurement facilities;
b) the necessity for a dump and/or a hot-sample pump will depend on the ability to meet the sample transfer time and analysis sub-system sample flow rate requirements. This in turn depends on the exhaust sample-driving pressure and line losses. It is considered that these pumps usually will be necessary at certain engine running conditions; and
c) the position of the hot pump, relative to t he gas analysis sub-systems, may be varied as required. (For example, some HC analysers contain hot pumps and so may be judged capable of being used upstream of the system hot pump) Note.- Figures A5-1 and A5-2 are schematic drawings of the exhaust gas sampling and analytical system and typify the basic requirements for emissions testing Figure A5-1. Exhaust gas sampling system, schematic Page 84