119 E/ECE/374 } Rev.1/Add.82/Rev.3/Amend.2 E/ECE/I'RANS/505 Regulation No 83 page 40 2.2.       Calibration of the positive displacement pump (PDP) 2.2. I.    The following calibration procedure outlines the equipment, the test configuration and the various parameters that are measured to establish the flow-rate of the CVS pump. All the parameters related to the pump are simultaneously measured with the parameters related to the tlow-meter which is connected in series with the pump. 3 The calculated flow-rate (given in m /min at pump inlet, absolute pressure and temperature) can then be plotted versus a correlation function that is the value of a specific combination of pump parameters. The linear equation that relates the pump flow and the correlation function is then detennined. In the event that a CVS has a multiple speed drive, a calibration for each range used shall be perfonned. 2.2.2.     This calibration procedure is based on the measurement of the absolute values of the C....            pump and flow-meter parameters that relate the tlow rate at each point. Three conditions shall be maintained to ensure the accuracy and integrity of the calibration curve: 2.2.2.1.   The pump pressures shall be measured at tappings on the pump rather than at the external piping on the pump inlet and outlet. Pressure taps that are mounted at the top centre and bottom centre of the pump drive headplate are exposed to the actual pump cavity pressures, and therefore reflect the absolute pressure differentials; 2.2.2.2.   Temperature stability shall be maintained during the calibration. The laminar flow- meter is sensitive. to inlet temperature oscillations which cause the data points to be scattered. Gradual changes of ± 1 K in temperature are acceptable as long as they occur over a period of several minutes; 2:2.2.3.   All connections between the flow-meter and the CVS pump shall be free of any Ieakage. 2.2.3.     During an exhaust emission test, the measurement of these same pump parameters enables the user.to calculate the flow rate from the calibration e_quation. 2.2.4.     Figure 8 ofthis appendix shows one possible test set-up. Variationsare permissible,, provided that the Technical Service approves them as being of comparable accuracy. If the set-up shown in Figure 8 is used, the following data shall be found within the limits of precision given: Barometrie pressure (corrected)(Pb)              ±0.03 kPa Ambient temperature (T)                         ±0.2K Air temperature at LFE (ETI)                    ± 0.15 K Pressure depression upstream of LFE (EPI)        ± 0.01 kPa Pressure drop across the LFE matrix (EDP)       ± 0.0015 kPa Air temperature at CVS pump inlet (PTI)          ±0.2K Air temperature at CVS pump outlet (PTO)        ±0.2K

120 E/ECFJ324                } Rev. l/Add.82/Rev.3/Amend.2 E/ECE/TRANS/SOS Regulation No 83 page41         · Pressurc depression at CVS pump inlet (PPI)       ±0.22kPa Pressure head at CVS pump outlet (PPO)           ·±o.22kPa 1 Pump rcvolutions during tcst pcriod (n)          ± l min" Eiapscd time for pcriod (minimum 250 s) (t)      ±0.i s Figure 8: POP Calibration Configuration Hlllr r- Ia: .                                &..,Clllllllol >'IM(mDr) ..__ _ _ _ _ Tcnp,aflR   Im                           \ ~ pro 2.2.5.  After the systcm has bcen conncctcd as shown in Figurc 8 ·of this appcndix, set the variable restrictor in thc wide-opcn position and run the CVS pump for 20 minutes bcfore starting the calibration. 2.2.6.. Reset the rcstrictor valve to a more restrictcd condition in an increment of pump inlet depression (about l kPa) that will yield a minimum of six data points for the total calibration. Allow the system to stabiliz.e for tbree minu~ and repeat the data acquisition.                                            · 3 2.2.7.  The air flow rate (Q,) at cach test point is calcula~ in standard m /min from the ~ow-meter data using the manufacturcr's prescnl>ed method. 3 2.2.8.  The air flow-rate is then converted to pump flow· (V0) in m /rev at absolute pump inlet temperature and pressure. T. . 1 Vo =-• Q . _P__ 101.33 n   273.2    PP 1

121 E/ECE/324 } Rev.l/Add.82/Rev.3/Amend.2 E/ECE/fRANS/505 Regulation No 83 page42 where: 3 Vo = pump flow rate at Tp and Pp (m /rev), 3 Q5 = air flow at 101.33 kPa and 273.2 K (m /rnin), TP = pump inlet temperature (K), Pp = absolute pump inlet pressure (kPa), 1 N = pump speed (min" ).                 . 2.2.9.     To compensate for the interaction of pump speed pressure variations at the pump and the pump slip rate, the correlation function (Xo) between the pump speed (n), the pressure differential from pump inlet to pump outlet and the absolute pump outlet pressure is then calculated as follows: where~ Xo = correlation function, Af>p = pressure differentiaJ from pump inlet to pump outlet (kPa), Pe= absolute outlet pre~sure (PPO + Pb) (kPa). A linear least-square fit is performed to generate the calibration equations which have the formula: Vo= Do-M (Xo) n =A-B (Af>p) D0, M, A and B are the slope-intercept constants describing the lines. 2.2.10.    A CVS system that has multiple speeds shall be calibrated on each speed used. The calibration curves generated for the ranges shall be approximately parallel and the intercept values (D0) shall increase as the pump flow range decreases. 2.2.11.   lf the calibration has been performed carefully, .the calculated values from the equation will be within 0.5 per cent ofthe measured value ofV0 • VaJues ofM will vary from one pump to another. Calibration is performed at pump start-up and after major maintenance.                                                                    / 2.3.      Calibration ofthe critical-flow venturi (CFV) 2.3.1.    Calibration ofthe CFV is based upon the flow equation for a critical venturi:

122 E/ECE/324 E/ECE/fRANS/505            } Rev.1/Add.82/Rev.3/Amend.2 Regulation No 83 page 43 where: Q.= flow, Kv = calibration coefficient, P = absoiute pressure (kPa), T = absolute temperature (K). Gas flow is a function of inlet pressure and temperature. The calibration procedure described below establishes the value of the calibration coefficient at measured values of pressure, temperature and air flow. 2.3.2.  The manufacturer's recommended .procedure shall be followed for calibrating electronic portions ofthe CFV. 2.3.3. Measurements for flow calibration of the critical flow venturi are required and the following data shall be found within the limits ofprecision given: Barometrie pressure (corrected) (Pb)             ±0.03 kPa, LFE air temperature, flow-meter (ETI)             ± 0.15 K, Pressure depression upstream of LFE (EPI)         ±0.01 kPa, · Pressure drop across (EDP) LFE matrix             ± 0.0015 kPa, Air flow (Qs)                                     ± 0.5 per cent, CFV inlet depression (PPI)                        ±0.02 kPa, Temperature at venturi inlet (Tv)                 ±0.2K. 2.3.4. Tue equipment shall be set up as shown in Figure 9 of this appendix and checked for leaks. Any leaks between the flow-measuring device and the critical-flow venturi will seriously affect the accura~y of the calibration.

123 ) E/ECE/324 } Rev .UAdd.82/Rev.3/Amend.2 E/ECEffRANS/505 Regulation No 83 page 44 Figure 9: CFV Calibration Configuration EIJ> I                 mt l'Clllricu ~ Enl~ 2.3.S.     The variable-flow restrictor shall be set to the open position, the blower shall be started and the system stabilized. Data from all instruments shall be recorded. 2.3.6.     The flow restrictor shall be varied and at least eight readings across the critical flow range of the venturi shall be made. 2.3.7.     The data recorded during the calibration shall be used in the following calculations. The air flow-rate (Q,) at each test point is calculated from the flow-meter data using the manufacturer's prescribed method. Calculate values of the calibration coefficient for each test point: where: 3 Q,  = flow-rate in m /min at 273.2 K and 101.33 kPa, T v = temperature at the venturi inlet (K), Pv = absolute pressure at the venturi inlet (kPa). Plot Kv as a function of venturi inlet pressure. For sonic flow, Kv will have a relatively constant value. As pressure decreases (vacuum. increases) the venturi becomes unchoked and Kv decreases. The resultant Kv changes are not permissible.

124 E/ECE/324 } . Rev.l/Add.82/Rev.3/Amend.2 E/ECE/TRANS/505 Regulation No 83 page 45 For a minimum of eight points in the critical region, calculate an average Kv and the standard deviation. If the standard deviation exceeds 0.3 per Cent of the average Kv, take corrective action. 3.       SYSTEM VERIFICATION PROCEDURE 3.1.     General Requirements The total accuracy of the CVS sampling system and analytical system shall be determined by introducing a known mass of a J>Pllutant gas into the system whilst it is being operated as if during a normal test and then analysing and calculating the     ( ) pollutant ~ass according to the formula in paragraph 6.6. of Annex 4a except that the density ofpropane shall be taken as 1.967 grams per litre at standard conditions. The following two techniques are known to give sufficient accuracy. The maximum permissible deviation between the quantity of gas introduced and the quantity of gas measured is 5 per cent. 3.2.     CFO Method 3 .2.1.                                                             a Metering a constant flow of pure gas (CO or C3H8) using critical flow orifice device. 3.2.2.   A known quantity of pure gas (CO or C3H8) is fed into the CVS system through the calibrated critical orifice. If the inlet pressure is high enough, the flow-rate (q), which is adjusted by means of the critical flow orifice, is independent of orifice outlet pressure (critical flow). If deviations exceeding 5 per cent occur, the cause of the malfunction shall be detemrined and corrected. The CVS system is operated as in an exhaust emission test for about 5 to I Ominutes. The gas collected in the sampling bag is analysed by the usual equipment and the results compared to the ,,..         concentration of the gas samples which was known beforehand. 3.3.    Gravimetrie Method 3.3.1.  Metering a limited quantity of pure gas (CO or C3H8) by means of a gravimetric technique. 3.3.2.  The following gravimetric procedure may be used to verify the CVS system. Tue weight of a small cylinder filled with either carbon monoxide or propane is determined with a precision of ± 0.0 i g. For about 5 to 10 minutes, the CVS system is operated as in a normal exhaust emission test, while CO or propane is injected into the system. · The quantity of pure gas involved is detennined by means of differential weighing. The gas accumulated · in the oag is ilien analysed · by means of the equipment normally used for exhaust-gas analysis. The results are then compared to the concentration figures computed previously.

.125 E/ECE/324 } Rev.l/Add.82/Rev.3/Amend.2 E/ECF./TRANS/505 Regulation No 83 page46 Appendix 3 GASEOUS EMISSIONS MEASUREMENT EQUIPMENT 1;         SPECIFICATION 1. i.      System Overview A continuously proportional sample of the diluted exhaust gases anlthe dilution air shall be collected for analysis. Mass gaseous emissions shall be. determined from the proportional sample concc;ntrations and the total volume measured during the test. The sample Q             concentrations shall be corrected to take account of the pollutant content of the · ambient air. 1.2.       Sampling System Requirements 1.2.1.     The sample of dilute exhaust gases shall be taken upstream from the suction device but downstream from the conditioning devices (if any). 1.2.2.     The flow rate shall not deviate from the 'average by more than ± 2 per cent. 1.2.3.     The sampling rate shall not fall below 5 litres per minute and shall not exceed 0.2 per cent of the flow rate of the dilute exhaust gases. An equivalent limit shall apply.to constant-mass sampling systems. 1.2.4.     A sample ofthe dilution air shall be taken at a constant flow rate near the ambient air- inlet (after the filter if one is fitted). 1.2.5.     The dilution air sample shall not be contaminated by exhaust gases from the mixing 1 area. c~ 1.2.6.     The sampling rate for the dilution air shall be comparable to that used in the case of the dilute exhaust gases. 1.2.7.     The materials used for the sampling Operations shall be such as not to change the pollutant concentration.        ·                                    · 1.2.8.     Filters may be used in order to extract the solid particles from the sample. 1.2.9.     The various valves used to direct the exhaust gases shall be of a quick-adjustment, quick-acting type.

126 E/ECE/324 } Rev.l/Add.82/Rev.3/Amend.2 E/ECE/TRANS/505 Regulation No 83 page47 1.2.10.   Quick-fastening gas-tight connections may be used between the three-way valves and the sampling bags, the connections sealing themselves automatically on the bag side. O~her systems may be used for conveying the samples to the analyser (three-way stop valves, for example). 1.2.J l.  Storage of the sample The gas samples shall be collected in sampling bags of sufficient capacity not to impede the sample flow; the . bag material shall be such as to affect neither the measurements themselves nor the chemical composition of the gas samples by more than ± 2 per cent after 20 minutes (for instance: laminated polyethylene/polyamide films, or fluorinated polyhydrocarbons). _., 1 1 :1.2.12.   Hydrocarbon Sampling System - Diesel Engines '-' 1.2.12.l. The hydrocarbon sampling system shall consist of a heated sampling probe, line, filter and pump. The sampling probe shall be installed at the same distance from the exhaust gas inlet as the particulate sampling probe, in such a way that neither interferes with samples taken by the other. lt shall have a minimum intemal diameter of4mm. 1.2.12.2.  All heated parts shall be maintained at a temperature of 463 K ( 190 °C) ± 10 K by the heating system. 1.2.12.3.  The average concentration of the measured hydrocarbons shall be determined by integration. L2.12.4.   The heated sampling line shall be fitted with a heated filter (FH) 99 per cent efficient with particles-~ 0.3 µm, to extract any solid particles fri>m the continuous flow of gas required for analysis. 1.2.12.5.  The sampling system response time (from the probe to the analyser inlet) shall be no ~ more than four seconds.                                                                      ) 1.2.12.6.  The HFID shall be used with a constant flow (heat exchanger) system to ensure a representative s~mple, unless compensation for varying CFV or CFO flow is made. 1.3.       Gas Analysis Requirements 1.3. l.    Carbon monoxide (CO) and carbon dioxide (C02) analyses: Analysers shall be of the non-dispersive infra-red (NDIR) absorption type.

127 E/ECE/324 } Rev.1/Add.82/Rev.3/Amend.2 E/ECEffRANS/505 Regulation No 83 page 48 1.3.2.    . Hydrocarbons (HC) analysis - spark-ignition engines: The analyser shall be of the flame ionisation (FID) type calibrated with propane gas expressed equivalent to carbon atoms (C 1).          ' .1.3.3 . .   Hydrocarbons (HC) analysis - compression-ignition engines: The analyser shall be of the flame ionisation type with detector, valves, pipework, etc., heated to 463 K ( 190 °C) ± l 0 K (HFID). lt shall be calibrated with propane gas expressed equivalent to carbon atoms (C1). 1.3.4.      Nitrogen oxide (NOx) analysis: Q               The analyser shall be either of the chemi-luminescent (CLA) or of the non-dispersive ultra-violet resonance absorption (NDUVR) type, both with NOx-NO converters. · 1.3.5.      The analysers shall have a measuring range compatible with the accuracy required to measure the concentrations ofthe exhaust gas sample pollutants. 1.3.6.      Measurement error shall not exceed ± 2 per cent (intrinsic error of analyser) disregarding the true value for the calibration gases. 1.3.7.      Fot concentrations of less than 100 ppm; the measurement error shall not exceed ± 2 ppm. 1.3.8.      The ambient air sample s.hall be measured on the same analyser with an appropriate range. · 1.3.9.      No gas drying device shall be used before the analysers unless shown to have no effect on the pollutant content of the-gas streani. 1.4.        Recommended System Descriptions 0               Figure 10 is a schematic drawing of the system for gaseous emissions sampling.

128 E/ECE/324 } Rev.l/Add.82/Rev.3/Amend.2 E/ECFJI'RANS/505 Regulation No 83 page49 . Figure 10: Gaseous Emissions Sampling Schematic . Air DIiution Tunnel 1- FlgurN I 51 a 7) S3 S2/SV l from vehlcle •xh-t F p N p N                       tovent or optional contlnuous      V Q FL                      anelyHr townt    V        Q The components of the system are as follows: 1.4. l.  Two sampling probes (St and S2) for continuous sampling of the dilution air and of the diluted exhaust-gas/air mixture; 1.4.2.   A filter (F), to extract solid particles from the flows of gas collected for analysis; 1.4.3.   Pumps (P), to collect a constant flow of the dilution air as weil as of the diluted exhaust-gas/air mixture during the test; ,,,. b-   1.4.4.   Flow controller (N), to ensure a constant uniform flow of the gas samples taken during the course of the test from sampling probes St and S2 (for PDP-CVS) and flow of the gas samples shall be such that, at the end of each test, the quantity of the samples is sufficient for analysis (approximately 10 litres per minute); 1.4.5.   Flow meters (FL), for adj1,1Sting and monitoring the constant flow of gas samples during the test; 1.4.6.   Quick-acting valves (V), to divert a constant flow of gas samples into the sampling bags or to the outside vent;