Commission Delegated Regulation (EU) 2017/654 of 19 December 2016 supplementing Regulation (EU) 2016/1628 of the European Parliament and of the Council with regard to technical and general requirements relating to emission limits and type-approval for internal combustion engines for non-road mobile machinery

where:

• T map is the unadjusted maximum torque for the respective test speed taken from the engine mapping performed in accordance with point 7.6.2

• T f is the torque required to drive auxiliaries that should have been fitted but were not installed for the test

• Tr is the torque required to drive auxiliaries that should have been removed for the test but were installed for the test

7.7.2.4.   Example of denormalization procedure

As an example, the following test point shall be denormalized:

% speed = 43 %

% torque = 82 %

Given the following values:

MTS = 2 200 min^{– 1}

n idle = 600 min^{– 1}

results in

Image omitted; see official source

With the maximum torque of 700 Nm observed from the mapping curve at 1 288 min^–1

Image omitted; see official source

7.8.   Specific test cycle running procedure

7.8.1.   Emission test sequence for discrete-mode NRSC

7.8.1.1.   Engine warming-up for steady state discrete-mode NRSC

Pre-test procedure according to point 7.3.1 shall be performed, including analyzer calibration. The engine shall be warmed-up using the pre-conditioning sequence in point 7.3.1.1.3. Immediately from this engine conditioning point, the test cycle measurement starts.

7.8.1.2.   Performing discrete-mode NRSC

• (a) The test shall be performed in ascending order of mode numbers as set out for the test cycle (see Appendix 1 of Annex XVII);

• (b) Each mode has a mode length of at least 10 minutes. In each mode the engine shall be stabilised for at least 5 minutes. Gaseous emissions, and, where applicable, PN, shall be sampled for 1 to 3 minutes at the end of each mode and PM emissions shall be sampled in accordance with point (c); Notwithstanding the previous paragraph, when either testing spark ignition engines using cycles G1, G2 or G3 or when conducting measurements in accordance with Annex V of this Regulation each mode has a mode length of at least 3 minutes. In this case gaseous emissions, and, where applicable, PN, shall be sampled for at least the last 2 minutes of each mode and PM emissions shall be sampled in accordance with point (c). The mode length and sampling time may be extended to improve accuracy; The mode length shall be recorded and reported.

• (c) For PM emissions, the PM sampling may be done either with the single filter method or with the multiple filter method. Since the results of the methods may differ slightly, the method used shall be declared with the results; For the single filter method the modal weighting factors specified in the test cycle procedure and the actual exhaust gas flow shall be taken into account during sampling by adjusting sample flow rate and/or sampling time, accordingly. It is required that the effective weighing factor of the PM sampling is within ± 0,005 of the weighing factor of the given mode; Sampling shall be conducted as late as possible within each mode. For the single filter method, the completion of PM sampling shall be coincident within ± 5 s with the completion of the gaseous emission measurement. The sampling time per mode shall be at least 20 s for the single filter method and at least 60 s for the multi-filter method. For systems without bypass capability, the sampling time per mode shall be at least 60 s for single and multiple filter methods;

• (d) The engine speed and load, intake air temperature, fuel flow and where applicable air or exhaust gas flow shall be measured for each mode at the same time interval which is used for the measurement of the gaseous concentrations; Any additional data required for calculation shall be recorded.

• (e) If the engine stalls or the emission sampling is interrupted at any time after emission sampling begins for a discrete-mode NRSC and the single filter method, the test shall be voided and be repeated beginning with the engine warm-up procedure. In the case of PM measurement utilizing the multi-filter method (one sampling filter for each operating mode), the test shall be continued by stabilizing the engine at the previous mode for engine temperature conditioning and then initiating measurement with the mode where the engine stalled;

• (f) Post-test procedures according to point 7.3.2 shall be performed.

7.8.1.3.   Validation criteria

During each mode of the given steady-state test cycle after the initial transition period, the measured speed shall not deviate from the reference speed for more than ± 1 % of rated speed or ± 3 min^{– 1}, whichever is greater except for idle which shall be within the tolerances declared by the manufacturer. The measured torque shall not deviate from the reference torque for more than ± 2 % of the maximum torque at the test speed.

7.8.2.   Emission test sequence for RMC

7.8.2.1.   Engine warming-up

Pre-test procedure according to point 7.3.1 shall be performed, including analyzer calibration. The engine shall be warmed-up using the pre-conditioning sequence in point 7.3.1.1.4. Immediately after this engine conditioning procedure, if the engine speed and torque are not already set for the first mode of the test they shall be changed in a linear ramp of 20 ± 1 s to the first mode of the test. In between 5 to 10 s after the end of the ramp, the test cycle measurement shall start.

7.8.2.2.   Performing an RMC

The test shall be performed in the order of mode numbers as set out for the test cycle (see Appendix 2 of Annex XVII) Where there is no RMC available for the specified NRSC the discrete-mode NRSC procedure set out in point 7.8.1 shall be followed.

The engine shall be operated for the prescribed time in each mode. The transition from one mode to the next shall be done linearly in 20 s ± 1 s following the tolerances prescribed in point 7.8.2.4.

For RMC, reference speed and torque values shall be generated at a minimum frequency of 1 Hz and this sequence of points shall be used to run the cycle. During the transition between modes, the denormalized reference speed and torque values shall be linearly ramped between modes to generate reference points. The normalized reference torque values shall not be linearly ramped between modes and then denormalized. If the speed and torque ramp runs through a point above the engine's torque curve, it shall be continued to command the reference torques and it shall be allowed for the operator demand to go to maximum.

Over the whole RMC (during each mode and including the ramps between the modes), the concentration of each gaseous pollutant shall be measured and where there is an applicable limit the PM and PN be sampled. The gaseous pollutants may be measured raw or diluted and be recorded continuously; if diluted, they can also be sampled into a sampling bag. The particulate sample shall be diluted with conditioned and clean air. One sample over the complete test procedure shall be taken, and, in the case of PM collected on a single PM sampling filter.

For calculation of the brake specific emissions, the actual cycle work shall be calculated by integrating actual engine power over the complete cycle.

7.8.2.3.   Emission test sequence

• (a) Execution of the RMC, sampling exhaust gas, recording data, and integrating measured values shall be started simultaneously;

• (b) Speed and torque shall be controlled to the first mode in the test cycle;

• (c) If the engine stalls anywhere during the RMC execution, the test shall be voided. The engine shall be pre-conditioned and the test repeated;

• (d) At the end of the RMC, sampling shall be continued, except for PM sampling, operating all systems to allow system response time to elapse. Then all sampling and recording shall be stopped, including the recording of background samples. Finally, any integrating devices shall be stopped and the end of the test cycle shall be indicated in the recorded data;

• (e) Post-test procedures according to point 7.3.2 shall be performed.

7.8.2.4.   Validation criteria

RMC tests shall be validated using the regression analysis as described in points 7.8.3.3 and 7.8.3.5. The allowed RMC tolerances are given in the following Table 6.1. Note that the RMC tolerances are different from the NRTC tolerances of Table 6.2.  When conducting testing of engines of reference power greater than 560 kW the regression line tolerances of Table 6.2 and the point deletion of Table 6.3 may be used.

Table 6.1

RMC Regression line tolerances

	Speed	Torque	Power
Standard error of estimate (SEE) of y on x	maximum 1 % of rated speed	maximum 2 % of maximum engine torque	maximum 2 % of maximum engine power
Slope of the regression line, a 1	0,99 to 1,01	0,98 - 1,02	0,98 - 1,02
Coefficient of determination, r 2	minimum 0,990	minimum 0,950	minimum 0,950
y intercept of the regression line, a 0	± 1 % of rated speed	± 20 Nm or 2 % of maximum torque whichever is greater	± 4 kW or 2 % of maximum power whichever is greater

In case of running the RMC test not on a transient test bed, where the second by second speed and torque values are not available, the following validation criteria shall be used.

At each mode the requirements for the speed and torque tolerances are given in point 7.8.1.3. For the 20 s linear speed and linear torque transitions between the RMC steady-state test modes (point 7.4.1.2) the following tolerances for speed and load shall be applied for the ramp:

- (a) the speed shall be held linear within ± 2 % of rated speed,

• (b) the torque shall be held linear within ± 5 % of the maximum torque at rated speed.

7.8.3.   Transient (NRTC and LSI-NRTC) test cycles

Reference speeds and torques commands shall be sequentially executed to perform the NRTC and LSI-NRTC. Speed and torque commands shall be issued at a frequency of at least 5 Hz. Because the reference test cycle is specified at 1 Hz, the in between speed and torque commands shall be linearly interpolated from the reference torque values generated from cycle generation.

Small denormalized speed values near warm idle speed may cause low-speed idle governors to activate and the engine torque to exceed the reference torque even though the operator demand is at a minimum. In such cases, it is recommended to control the dynamometer so it gives priority to follow the reference torque instead of the reference speed and let the engine govern the speed.

Under cold-start conditions engines may use an enhanced-idle device to quickly warm up the engine and the exhaust after-treatment system. Under these conditions, very low normalized speeds will generate reference speeds below this higher enhanced idle speed. In this case it is recommended controlling the dynamometer so it gives priority to follow the reference torque and let the engine govern the speed when the operator demand is at minimum.

During an emission test, reference speeds and torques and the feedback speeds and torques shall be recorded with a minimum frequency of 1 Hz, but preferably of 5 Hz or even 10 Hz. This larger recording frequency is important as it helps to minimize the biasing effect of the time lag between the reference and the measured feedback speed and torque values.

The reference and feedback speeds and torques maybe recorded at lower frequencies (as low as 1 Hz), if the average values over the time interval between recorded values are recorded. The average values shall be calculated based on feedback values updated at a frequency of at least 5 Hz. These recorded values shall be used to calculate cycle-validation statistics and total work.

7.8.3.1.   Performing an NRTC test

Pre-test procedures according to point 7.3.1 shall be performed, including pre-conditioning, cool-down and analyzer calibration.

Testing shall be started as follows:

The test sequence shall commence immediately after the engine has started from cooled down condition specified in point 7.3.1.2 in case of the cold-start NRTC or from hot soak condition in case of the hot-start NRTC. The sequence in point 7.4.2.1 shall be followed.

Data logging, sampling of exhaust gas and integrating measured values shall be initiated simultaneously at the start of the engine. The test cycle shall be initiated when the engine starts and shall be executed according to the schedule of Appendix 3 of Annex XVII.

At the end of the cycle, sampling shall be continued, operating all systems to allow system response time to elapse. Then all sampling and recording shall be stopped, including the recording of background samples. Finally, any integrating devices shall be stopped and the end of the test cycle shall be indicated in the recorded data.

Post-test procedures according to point 7.3.2 shall be performed.

7.8.3.2.   Performing an LSI-NRTC test

Pre-test procedures according to point 7.3.1 shall be performed, including pre-conditioning and analyzer calibration.

Testing shall be started as follows:

The test shall commence according to the sequence given in point 7.4.2.2.

Data logging, sampling of exhaust gas and integrating measured values shall be initiated simultaneously with the start of the LSI-NRTC at the end of the 30-second idle period specified in point 7.4.2.2(b). The test cycle shall be executed according to the schedule of Appendix 3 of Annex XVII.

At the end of the cycle, sampling shall be continued, operating all systems to allow system response time to elapse. Then all sampling and recording shall be stopped, including the recording of background samples. Finally, any integrating devices shall be stopped and the end of the test cycle shall be indicated in the recorded data.

Post-test procedures according to point 7.3.2 shall be performed.

7.8.3.3.   Cycle validation criteria for transient (NRTC and LSI-NRTC) test cycles

In order to check the validity of a test, the cycle-validation criteria in this point shall be applied to the reference and feedback values of speed, torque, power and overall work.

7.8.3.4.   Calculation of cycle work

Before calculating the cycle work, any speed and torque values recorded during engine starting shall be omitted. Points with negative torque values have to be accounted for as zero work. The actual cycle work W act (kWh) shall be calculated based on engine feedback speed and torque values. The reference cycle work W ref (kWh) shall be calculated based on engine reference speed and torque values. The actual cycle work W act is used for comparison to the reference cycle work W ref and for calculating the brake specific emissions (see point 7.2).

W act shall be between 85 % and 105 % of W ref.

7.8.3.5.   Validation statistics (see Appendix 2 of Annex VII)

Linear regression between the reference and the feedback values shall be calculated for speed, torque and power.

To minimize the biasing effect of the time lag between the reference and feedback cycle values, the entire engine speed and torque feedback signal sequence may be advanced or delayed in time with respect to the reference speed and torque sequence. If the feedback signals are shifted, both speed and torque shall be shifted by the same amount in the same direction.

The method of least squares shall be used, with the best-fit equation having the form set out in equation (6-19):

y = a 1 x + a 0	(6-19)

where:

• y is the feedback value of speed (min^{– 1}), torque (Nm), or power (kW)

• a 1 is the slope of the regression line

• x is the reference value of speed (min^{– 1}), torque (Nm), or power (kW)

• a 0 is the y intercept of the regression line.

The standard error of estimate (SEE) of y on x and the coefficient of determination (r

²) shall be calculated for each regression line in accordance with Appendix 3 of Annex VII.

It is recommended that this analysis be performed at 1 Hz. For a test to be considered valid, the criteria of Table 6.2 shall be met.

Table 6.2

Regression line tolerances

| | Speed | Torque | Power |

| --- | --- | --- | --- | | Standard error of estimate (SEE) of y on x | ≤ 5,0 percent of maximum test speed | ≤ 10,0 % of maximum mapped torque | ≤ 10,0 % of maximum mapped power | | Slope of the regression line, a 1 | 0,95 to 1,03 | 0,83 - 1,03 | 0,89 - 1,03 | | Coefficient of determination, r ²* | minimum 0,970 | minimum 0,850 | minimum 0,910 | | y intercept of the regression line, a 0* | ≤ 10 % of idle | ± 20 Nm or ± 2 % of maximum torque whichever is greater | ± 4 kW or ± 2 % of maximum power whichever is greater |

For regression purposes only, point deletions are permitted where noted in Table 6.3 before doing the regression calculation. However, those points shall not be deleted for the calculation of cycle work and emissions. An idle point is defined as a point having a normalized reference torque of 0 % and a normalized reference speed of 0 %. Point deletion may be applied to the whole or to any part of the cycle; points to which the point deletion is applied have to be specified.

Table 6.3

Permitted point deletions from regression analysis

| Event | Conditions (n = engine speed, T = torque) | Permitted point deletions |

| --- | --- | --- | | Minimum operator demand (idle point) | n ref = n idle and T ref = 0 % and T act > (T ref – 0,02 T maxmappedtorque) and T act < (T ref + 0,02 T maxmappedtorque) | speed and power | | Minimum operator demand | n act ≤ 1,02 n ref and T act > T ref or n act > n ref and T act ≤ T ref' or n act > 1,02 n ref and T ref < T act ≤ (T ref + 0,02 T maxmappedtorque) | power and either torque or speed | | Maximum operator demand | n act < n ref and T act ≥ T ref or n act ≥ 0,98 n ref and T act < T ref or n act < 0,98 n ref and T ref > T act ≥ (T ref – 0,02 T maxmappedtorque) | power and either torque or speed | | Where: nref is the reference speed (see section 7.7.2.), nidle is the idle speed, nact is the actual (measured) speed, Tref is the reference torque (see section 7.7.2.), Tact is the actual (measured) torque, Tmaxmappedtorque is the highest value of torque on the full-load torque curve mapped in accordance with section 7.6. | Where: nref is the reference speed (see section 7.7.2.), nidle is the idle speed, nact is the actual (measured) speed, Tref is the reference torque (see section 7.7.2.), Tact is the actual (measured) torque, Tmaxmappedtorque is the highest value of torque on the full-load torque curve mapped in accordance with section 7.6. | Where: nref is the reference speed (see section 7.7.2.), nidle is the idle speed, nact is the actual (measured) speed, Tref is the reference torque (see section 7.7.2.), Tact is the actual (measured) torque, Tmaxmappedtorque is the highest value of torque on the full-load torque curve mapped in accordance with section 7.6. |

8. Measurement procedures

8.1.   Calibration and performance checks

8.1.1.   Introduction

This point describes required calibrations and verifications of measurement systems. See point 9.4 for specifications that apply to individual instruments.

Calibrations or verifications shall be generally performed over the complete measurement chain.

If a calibration or verification for a portion of a measurement system is not specified, that portion of the system shall be calibrated and its performance verified at a frequency consistent with any recommendations from the measurement system manufacturer and consistent with good engineering judgment.

Internationally recognized-traceable standards shall be used to meet the tolerances specified for calibrations and verifications.

8.1.2.   Summary of calibration and verification

Table 6.4 summarizes the calibrations and verifications described in section 8 and indicates when these have to be performed.

Table 6.4

Summary of Calibration and Verifications

Type of calibration or verification	Minimum frequency [^E0006]
8.1.3: accuracy, repeatability and noise	Accuracy: Not required, but recommended for initial installation. Repeatability: Not required, but recommended for initial installation. Noise: Not required, but recommended for initial installation.
8.1.4: linearity verification	Speed: Upon initial installation, within 370 days before testing and after major maintenance. Torque: Upon initial installation, within 370 days before testing and after major maintenance. Intake air, dilution air and diluted exhaust gas flows and batch sample flow rates: Upon initial installation, within 370 days before testing and after major maintenance, unless flow is verified by propane check or by carbon or oxygen balance. Raw exhaust gas flow: Upon initial installation, within 185 days before testing and after major maintenance, unless flow is verified by propane check or by carbon or oxygen balance. Gas dividers: Upon initial installation, within 370 days before testing and after major maintenance. Gas analyzers (unless otherwise noted): Upon initial installation, within 35 days before testing and after major maintenance. FTIR analyser: Upon installation, within 370 days before testing and after major maintenance. PM balance: Upon initial installation, within 370 days before testing and after major maintenance. Stand-alone pressure and temperature: Upon initial installation, within 370 days before testing and after major maintenance.
8.1.5: Continuous gas analyzer system response and updating-recording verification — for gas analyzers not continuously compensated for other gas species	Upon initial installation or after system modification that would affect response.
8.1.6: Continuous gas analyzer system response and updating-recording verification — for gas analyzers continuously compensated for other gas species	Upon initial installation or after system modification that would affect response.
8.1.7.1: torque	Upon initial installation and after major maintenance.
8.1.7.2: pressure, temperature, dew point	Upon initial installation and after major maintenance.
8.1.8.1: fuel flow	Upon initial installation and after major maintenance.
8.1.8.2: intake flow	Upon initial installation and after major maintenance.
8.1.8.3: exhaust gas flow	Upon initial installation and after major maintenance.
8.1.8.4: diluted exhaust gas flow (CVS and PFD)	Upon initial installation and after major maintenance.
8.1.8.5: CVS/PFD and batch sampler verification [^E0007]	Upon initial installation, within 35 days before testing, and after major maintenance. (Propane check)
8.1.8.8: vacuum leak	Upon installation of the sampling system. Before each laboratory test according to point 7.1: within 8 hours before the start of the first test interval of each duty cycle sequence and after maintenance such as pre-filter changes.
8.1.9.1: CO2 NDIR H2O interference	Upon initial installation and after major maintenance.
8.1.9.2: CO NDIR CO2 and H2O interference	Upon initial installation and after major maintenance.
8.1.10.1: FID calibration HC FID optimization and HC FID verification	Calibrate, optimize, and determine CH4 response: upon initial installation and after major maintenance. Verify CH4 response: upon initial installation, within 185 days before testing, and after major maintenance.
8.1.10.2: raw exhaust gas FID O2 interference	For all FID analyzers: upon initial installation, and after major maintenance. For THC FID analyzers: upon initial installation, after major maintenance, and after FID optimization according to 8.1.10.1.
8.1.11.1: CLD CO2 and H2O quench	Upon initial installation and after major maintenance.
8.1.11.3: NDUV HC and H2O interference	Upon initial installation and after major maintenance.
8.1.11.4: Sample dryer NO2 penetration (chiller)	Upon initial installation and after major maintenance.
8.1.11.5: NO2-to-NO converter conversion	Upon initial installation, within 35 days before testing, and after major maintenance.
8.1.12: Sample dryer verification	For thermal chillers: upon installation and after major maintenance. For osmotic membranes: upon installation, within 35 days of testing and after major maintenance.
8.1.13.1: PM balance and weighing	Independent verification: upon initial installation, within 370 days before testing, and after major maintenance. Zero, span, and reference sample verifications: within 12 hours of weighing, and after major maintenance.

[^E0006]: (¹)

Perform calibrations and verifications more frequently, according to measurement system manufacturer instructions and good engineering judgment.

The CVS verification is not required for systems that agree within ± 2 % based on a chemical balance of carbon or oxygen of the intake air, fuel, and diluted exhaust gas.

8.1.3.   Verifications for accuracy, repeatability, and noise

The performance values for individual instruments specified in Table 6.8 are the basis for the determination of the accuracy, repeatability, and noise of an instrument.

It is not required to verify instrument accuracy, repeatability, or noise. However, it may be useful to consider these verifications to define a specification for a new instrument, to verify the performance of a new instrument upon delivery, or to troubleshoot an existing instrument.

8.1.4.   Linearity verification

8.1.4.1.   Scope and frequency

A linearity verification shall be performed on each measurement system listed in Table 6.5 at least as frequently as indicated in the Table, consistent with measurement system manufacturer recommendations and good engineering judgment. The intent of a linearity verification is to determine that a measurement system responds proportionally over the measurement range of interest. A linearity verification shall consist of introducing a series of at least 10 reference values to a measurement system, unless otherwise specified. The measurement system quantifies each reference value. The measured values shall be collectively compared to the reference values by using a least squares linear regression and the linearity criteria specified in Table 6.5.

8.1.4.2.   Performance requirements

If a measurement system does not meet the applicable linearity criteria in Table 6.5, the deficiency shall be corrected by re-calibrating, servicing, or replacing components as needed. The linearity verification shall be repeated after correcting the deficiency to ensure that the measurement system meets the linearity criteria.

8.1.4.3.   Procedure

The following linearity verification protocol shall be used:

• (a) A measurement system shall be operated at its specified temperatures, pressures, and flows;

• (b) The instrument shall be zeroed as it would before an emission test by introducing a zero signal. For gas analyzers, a zero gas shall be used that meets the specifications of point 9.5.1 and it shall be introduced directly at the analyzer port;

• (c) The instrument shall be spanned as it would before an emission test by introducing a span signal. For gas analyzers, a span gas shall be used that meets the specifications of point 9.5.1 and it shall be introduced directly at the analyzer port;

• (d) After spanning the instrument, zero shall be checked with the same signal which has been used in paragraph (b) of this point. Based on the zero reading, good engineering judgment shall be used to determine whether or not to re-zero and or re-span the instrument before proceeding to the next step;

• (e) For all measured quantities manufacturer recommendations and good engineering judgment shall be used to select the reference values, y ref i, that cover the full range of values that are expected during emission testing, thus avoiding the need of extrapolation beyond these values. A zero reference signal shall be selected as one of the reference values of the linearity verification. For stand-alone pressure and temperature linearity verifications, at least three reference values shall be selected. For all other linearity verifications, at least ten reference values shall be selected;

• (f) Instrument manufacturer recommendations and good engineering judgment shall be used to select the order in which the series of reference values will be introduced;

• (g) Reference quantities shall be generated and introduced as described in point 8.1.4.4. For gas analyzers, gas concentrations known to be within the specifications of point 9.5.1 shall be used and they shall be introduced directly at the analyzer port;

• (h) Time for the instrument to stabilize while it measures the reference value shall be allowed;

• (i) Image omitted; see official source

• (j) Steps in paragraphs (g) to (i) of this point shall be repeated until all reference quantities are measured;

• (k) The arithmetic means

Image omitted; see official source , and reference values, y ref i, shall be used to calculate least-squares linear regression parameters and statistical values to compare to the minimum performance criteria specified in Table 6.5. The calculations described in Appendix 3 of Annex VII shall be used.

8.1.4.4.   Reference signals

This point describes recommended methods for generating reference values for the linearity-verification protocol in point 8.1.4.3. Reference values shall be used that simulate actual values, or an actual value shall be introduced and measured with a reference-measurement system. In the latter case, the reference value is the value reported by the reference-measurement system. Reference values and reference-measurement systems shall be internationally traceable.

For temperature measurement systems with sensors like thermocouples, RTDs, and thermistors, the linearity verification may be performed by removing the sensor from the system and using a simulator in its place. A simulator that is independently calibrated and cold junction compensated, as necessary shall be used. The internationally traceable simulator uncertainty scaled to temperature shall be less than 0,5 % of maximum operating temperature T max. If this option is used, it is necessary to use sensors that the supplier states are accurate to better than 0,5 % of T max compared to their standard calibration curve.

8.1.4.5.   Measurement systems that require linearity verification

Table 6.5 indicates measurement systems that require linearity verifications. For this Table the following provisions shall apply:

• (a) a linearity verification shall be performed more frequently if the instrument manufacturer recommends it or based on good engineering judgment;

• (b) ‘min’ refers to the minimum reference value used during the linearity verification; Note that this value may be zero or a negative value depending on the signal;

• (c) ‘max’ generally refers to the maximum reference value used during the linearity verification. For example for gas dividers, x max is the undivided, undiluted, span gas concentration. The following are special cases where ‘max’ refers to a different value:
• (i) For PM balance linearity verification, m max refers to the typical mass of a PM filter;
• (ii) For torque linearity verification, T max refers to the manufacturer's specified engine torque peak value of the highest torque engine to be tested;

• (d) the specified ranges are inclusive. For example, a specified range of 0,98-1,02 for the slope a 1 means 0,98 ≤ a 1 ≤ 1,02;

• (e) these linearity verifications are not required for systems that pass the flow-rate verification for diluted exhaust gas as described in point 8.1.8.5 for the propane check or for systems that agree within ± 2 % based on a chemical balance of carbon or oxygen of the intake air, fuel, and exhaust gas;

- (f) a

1 criteria for these quantities shall be met only if the absolute value of the quantity is required, as opposed to a signal that is only linearly proportional to the actual value;

- (g) stand-alone temperatures include engine temperatures and ambient conditions used to set or verify engine conditions; temperatures used to set or verify critical conditions in the test system; and temperatures used in emissions calculations:

• (i) these temperature linearity checks are required. Air intake; after-treatment bed(s) (for engines tested with exhaust after-treatment systems on cycles with cold start criteria); dilution air for PM sampling (CVS, double dilution, and partial flow systems); PM sample; and chiller sample (for gaseous sampling systems that use chillers to dry samples);
• (ii) these temperature linearity checks are only required if specified by the engine manufacturer. Fuel inlet; test cell charge air cooler air outlet (for engines tested with a test cell heat exchanger simulating a non-road mobile machinery charge air cooler); test cell charge air cooler coolant inlet (for engines tested with a test cell heat exchanger simulating a non-road mobile machinery charge air cooler); and oil in the sump/pan; coolant before the thermostat (for liquid cooled engines);

• (h) stand-alone pressures include engine pressures and ambient conditions used to set or verify engine conditions; pressures used to set or verify critical conditions in the test system; and pressures used in emissions calculations:
• (i) required pressure linearity checks are: intake air pressure restriction; exhaust gas back-pressure; barometer; CVS inlet gage pressure (if measurement using CVS); chiller sample (for gaseous sampling systems that use chillers to dry samples);
• (ii) pressure linearity checks that are required only if specified by the engine manufacturer: test cell charge air cooler and interconnecting pipe pressure drop (for turbo-charged engines tested with a test cell heat exchanger simulating a non-road mobile machinery charge air cooler) fuel inlet; and fuel outlet.

Table 6.5

Measurement systems that require linearity verifications

Measurement System	Quantity	Minimum verification frequency	Linearity Criteria	Linearity Criteria	Linearity Criteria	Linearity Criteria
Measurement System	Quantity	Minimum verification frequency	Image omitted; see official source	α	SEE	r 2
Engine speed	n	Within 370 days before testing	≤ 0,05 % n max	0,98-1,02	≤ 2 % n max	≥ 0,990
Engine torque	T	Within 370 days before testing	≤ 1 % T max	0,98-1,02	≤ 2 % T max	≥ 0,990
Fuel flow rate	qm	Within 370 days before testing	≤ 1 % qm , max	0,98-1,02	≤ 2 % qm , max	≥ 0,990
Intake-air flow rate [^E0008]	qV	Within 370 days before testing	≤ 1 % qV , max	0,98-1,02	≤ 2 % qV , max	≥ 0,990
Dilution air flow rate [^E0008]	qV	Within 370 days before testing	≤ 1 % qV , max	0,98-1,02	≤ 2 % qV , max	≥ 0,990
Diluted exhaust gas flow rate [^E0008]	qV	Within 370 days before testing	≤ 1 % qV , max	0,98-1,02	≤ 2 % qV , max	≥ 0,990
Raw exhaust gas flow rate [^E0008]	qV	Within 185 days before testing	≤ 1 % qV , max	0,98-1,02	≤ 2 % qV , max	≥ 0,990
Batch sampler flow rates [^E0008]	qV	Within 370 days before testing	≤ 1 % qV , max	0,98-1,02	≤ 2 % qV , max	≥ 0,990
Gas dividers	x/x span	Within 370 days before testing	≤ 0,5 % x max	0,98-1,02	≤ 2 % x max	≥ 0,990
Gas analyzers	x	Within 35 days before testing	≤ 0,5 % x max	0,99-1,01	≤ 1 % x max	≥ 0,998
PM balance	m	Within 370 days before testing	≤ 1 % m max	0,99-1,01	≤ 1 % m max	≥ 0,998
Stand-alone pressures	p	Within 370 days before testing	≤ 1 % p max	0,99-1,01	≤ 1 % p max	≥ 0,998
Analog-to-digital conversion of stand-alone temperature signals	T	Within 370 days before testing	≤ 1 % T max	0,99-1,01	≤ 1 % T max	≥ 0,998

Molar flow rate may be used instead of standard volumetric flow rate as the term representing ‘quantity’. In this case maximum molar flow rate may be used instead of the maximum standard volumetric flow rate in the corresponding linearity criteria.

8.1.5.   Continuous gas analyser system-response and updating-recording verification

This section describes a general verification procedure for continuous gas analyzer system response and update recording. See point 8.1.6 for verification procedures for compensation type analysers.

8.1.5.1.   Scope and frequency

This verification shall be performed after installing or replacing a gas analyzer that is used for continuous sampling. Also this verification shall be performed if the system is reconfigured in a way that would change system response. This verification is needed for continuous gas analysers used for transient (NRTC and LSI-NRTC) test cycles or RMC but is not needed for batch gas analyzer systems or for continuous gas analyzer systems used only for testing with a discrete-mode NRSC.

8.1.5.2.   Measurement principles

This test verifies that the updating and recording frequencies match the overall system response to a rapid change in the value of concentrations at the sample probe. Gas analyzer systems shall be optimized such that their overall response to a rapid change in concentration is updated and recorded at an appropriate frequency to prevent loss of information. This test also verifies that continuous gas analyzer systems meet a minimum response time.

The system settings for the response time evaluation shall be exactly the same as during measurement of the test run (i.e. pressure, flow rates, filter settings on the analyzers and all other response time influences). The response time determination shall be done with gas switching directly at the inlet of the sample probe. The devices for gas switching shall have a specification to perform the switching in less than 0,1 s. The gases used for the test shall cause a concentration change of at least 60 % full scale (FS).

The concentration trace of each single gas component shall be recorded.

8.1.5.3.   System requirements

• (a) The system response time shall be ≤ 10 s with a rise time of ≤ 5 s for all measured components (CO, NOx, 2 and HC) and all ranges used. All data (concentration, fuel and air flows) have to be shifted by their measured response times before performing the emission calculations given in Annex VII.

• (b) To demonstrate acceptable updating and recording with respect to the system's overall response, the system shall meet one of the following criteria:
• (i) The product of the mean rise time and the frequency at which the system records an updated concentration shall be at least 5. In any case the mean rise time shall be no more than 10 s;
• (ii) The frequency at which the system records the concentration shall be at least 2 Hz (see also Table 6.7).

8.1.5.4.   Procedure

The following procedure shall be used to verify the response of each continuous gas analyzer system:

• (a) The analyzer system manufacturer's start-up and operating instructions for the instrument setup shall be followed. The measurement system shall be adjusted as needed to optimize performance. This verification shall be run with the analyzer operating in the same manner as used for emission testing. If the analyzer shares its sampling system with other analyzers, and if gas flow to the other analyzers will affect the system response time, then the other analyzers shall be started up and operated while running this verification test. This verification test may be run on multiple analyzers sharing the same sampling system at the same time. If analogue or real-time digital filters are used during emission testing, those filters shall be operated in the same manner during this verification;

• (b) For equipment used to validate system response time, minimal gas transfer line lengths between all connections are recommended to be used, a zero-air source shall be connected to one inlet of a fast-acting 3-way valve (2 inlets, 1 outlet) in order to control the flow of zero and blended span gases to the sample system's probe inlet or a tee near the outlet of the probe. Normally the gas flow rate is higher than the probe sample flow rate and the excess is overflowed out the inlet of the probe. If the gas flow rate is lower than the probe flow rate, the gas concentrations shall be adjusted to account for the dilution from ambient air drawn into the probe. Binary or multi-gas span gases may be used. A gas blending or mixing device may be used to blend span gases. A gas blending or mixing device is recommended when blending span gases diluted in N2 with span gases diluted in air; Using a gas divider, an NO–CO–CO2–C3H8–CH4 (balance N2) span gas shall be equally blended with a span gas of NO2, balance purified synthetic air. Standard binary span gases may be also be used, where applicable, in place of blended NO-CO-CO2-C3H8-CH4, balance N2 span gas; in this case separate response tests shall be run for each analyzer. The gas divider outlet shall be connected to the other inlet of the 3-way valve. The valve outlet shall be connected to an overflow at the gas analyzer system's probe or to an overflow fitting between the probe and transfer line to all the analyzers being verified. A setup that avoids pressure pulsations due to stopping the flow through the gas blending device shall be used. Any of these gas constituents if they are not relevant to the analyzers for this verification shall be omitted. Alternatively the use of gas bottles with single gases and a separate measurement of response times is allowed;

• (c) Data collection shall be done as follows:
• (i) The valve shall be switched to start the flow of zero gas;
• (ii) Stabilization shall be allowed for, accounting for transport delays and the slowest analyzer's full response;
• (iii) Data recording shall be started at the frequency used during emission testing. Each recorded value shall be a unique updated concentration measured by the analyzer; interpolation or filtering may not be used to alter recorded values;
• (iv) The valve shall be switched to allow the blended span gases to flow to the analyzers. This time shall be recorded as t 0;
• (v) Transport delays and the slowest analyzer's full response shall be allowed for;
• (vi) The flow shall be switched to allow zero gas to flow to the analyzer. This time shall be recorded as t 100;
• (vii) Transport delays and the slowest analyzer's full response shall be allowed for;
• (viii) The steps in paragraphs (c)(iv) to (vii) of this point shall be repeated to record seven full cycles, ending with zero gas flowing to the analyzers;
• (ix) Recording shall be stopped.

8.1.5.5.   Performance evaluation

The data from point 8.1.5.4(c) shall be used to calculate the mean rise time for each of the analyzers.

• (a) If it is chosen to demonstrate compliance with point 8.1.5.3(b)(i) the following procedure has to be applied: The rise times (in s) shall be multiplied by their respective recording frequencies in Hertz (1/s). The value for each result shall be at least 5. If the value is less than 5, the recording frequency shall be increased or the flows adjusted or the design of the sampling system shall be changed to increase the rise time as needed. Also digital filters may be configured to increase rise time;

• (b) If it is chosen to demonstrate compliance with point 8.1.5.3(b)(ii), the demonstration of compliance with the requirements of point 8.1.5.3(b)(ii) is sufficient.

8.1.6.   Response time verification for compensation type analysers

8.1.6.1.   Scope and frequency

This verification shall be performed to determine a continuous gas analyzer's response, where one analyzer's response is compensated by another's to quantify a gaseous emission. For this check water vapour shall be considered to be a gaseous constituent. This verification is required for continuous gas analyzers used for transient (NRTC and LSI-NRTC) test cycles or RMC. This verification is not needed for batch gas analyzers or for continuous gas analyzers that are used only for testing with a discrete-mode NRSC. This verification does not apply to correction for water removed from the sample done in post-processing. This verification shall be performed after initial installation (i.e. test cell commissioning). After major maintenance, point 8.1.5 may be used to verify uniform response provided that any replaced components have gone through a humidified uniform response verification at some point.

8.1.6.2.   Measurement principles

This procedure verifies the time-alignment and uniform response of continuously combined gas measurements. For this procedure, it is necessary to ensure that all compensation algorithms and humidity corrections are turned on.

8.1.6.3.   System requirements

The general response time and rise time requirement set out in point 8.1.5.3(a) is also valid for compensation type analysers. Additionally, if the recording frequency is different than the update frequency of the continuously combined/compensated signal, the lower of these two frequencies shall be used for the verification required by point 8.1.5.3(b)(i).

8.1.6.4.   Procedure

All procedures set out in point 8.1.5.4(a) to (c) shall be used. Additionally also the response and rise time of water vapour has to be measured, if a compensation algorithm based on measured water vapour is used. In this case at least one of the used calibration gases (but not NO2) has to be humidified as follows:

If the system does not use a sample dryer to remove water from the sample gas, the span gas shall be humidified by flowing the gas mixture through a sealed vessel that humidifies the gas to the highest sample dew point that is estimated during emission sampling by bubbling it through distilled water. If the system uses a sample dryer during testing that has passed the sample dryer verification check, the humidified gas mixture may be introduced downstream of the sample dryer by bubbling it through distilled water in a sealed vessel at 298 ± 10 K (25 ± 10 °C), or a temperature greater than the dew point. In all cases, downstream of the vessel, the humidified gas shall be maintained at a temperature of at least 5 K (5 °C) above its local dew point in the line. Note that it is possible to omit any of these gas constituents if they are not relevant to the analyzers for this verification. If any of the gas constituents are not susceptible to water compensation, the response check for these analyzers may be performed without humidification.

8.1.7.   Measurement of engine parameters and ambient conditions

Internal quality procedures traceable to recognised national or international standards shall be applied. Otherwise the following procedures apply.

8.1.7.1.   Torque calibration

8.1.7.1.1.   Scope and frequency

All torque-measurement systems including dynamometer torque measurement transducers and systems shall be calibrated upon initial installation and after major maintenance using, among others, reference force or lever-arm length coupled with dead weight. Good engineering judgment shall be used to repeat the calibration. The torque transducer manufacturer's instructions shall be followed for linearizing the torque sensor's output. Other calibration methods are permitted.

8.1.7.1.2.   Dead-weight calibration

This technique applies a known force by hanging known weights at a known distance along a lever arm. It shall be made sure that the weights' lever arm is perpendicular to gravity (i.e., horizontal) and perpendicular to the dynamometer's rotational axis. At least six calibration-weight combinations shall be applied for each applicable torque-measuring range, spacing the weight quantities about equally over the range. The dynamometer shall be oscillated or rotated during calibration to reduce frictional static hysteresis. Each weight's force shall be determined by multiplying its internationally-traceable mass by the local acceleration of Earth's gravity.

8.1.7.1.3.   Strain gage or proving ring calibration

This technique applies force either by hanging weights on a lever arm (these weights and their lever arm length are not used as part of the reference torque determination) or by operating the dynamometer at different torques. At least six force combinations shall be applied for each applicable torque-measuring range, spacing the force quantities about equally over the range. The dynamometer shall be oscillated or rotated during calibration to reduce frictional static hysteresis. In this case, the reference torque is determined by multiplying the force output from the reference meter (such as a strain gage or proving ring) by its effective lever-arm length, which is measured from the point where the force measurement is made to the dynamometer's rotational axis. It shall be made sure that this length is measured perpendicular to the reference meter's measurement axis and perpendicular to the dynamometer's rotational axis.

8.1.7.2.   Pressure, temperature, and dew point calibration

Instruments shall be calibrated for measuring pressure, temperature, and dew point upon initial installation. The instrument manufacturer's instructions shall be followed and good engineering judgment shall be used to repeat the calibration.

For temperature measurement systems with thermocouple, RTD, or thermistor sensors, the calibration of the system shall be performed as described in point 8.1.4.4 for linearity verification.

8.1.8.   Flow-related measurements

8.1.8.1.   Fuel flow calibration

Fuel flow meters shall be calibrated upon initial installation. The instrument manufacturer's instructions shall be followed and good engineering judgment shall be used to repeat the calibration.

8.1.8.2.   Intake air flow calibration

Intake air flow meters shall be calibrated upon initial installation. The instrument manufacturer's instructions shall be followed and good engineering judgment shall be used to repeat the calibration.

8.1.8.3.   Exhaust gas flow calibration

Exhaust flow meters shall be calibrated upon initial installation. The instrument manufacturer's instructions shall be followed and good engineering judgment shall be used to repeat the calibration.

8.1.8.4.   Diluted exhaust gas flow (CVS) calibration

8.1.8.4.1.   Overview

• (a) This section describes how to calibrate flow meters for diluted exhaust gas constant-volume sampling (CVS) systems;

• (b) This calibration shall be performed while the flow meter is installed in its permanent position. This calibration shall be performed after any part of the flow configuration upstream or downstream of the flow meter has been changed that may affect the flow-meter calibration. This calibration shall be performed upon initial CVS installation and whenever corrective action does not resolve a failure to meet the diluted exhaust gas flow verification (i.e., propane check) in point 8.1.8.5;

• (c) A CVS flow meter shall be calibrated using a reference flow meter such as a subsonic venturi flow meter, a long-radius flow nozzle, a smooth approach orifice, a laminar flow element, a set of critical flow venturis, or an ultrasonic flow meter. A reference flow meter shall be used that reports quantities that are internationally-traceable within ± 1 % uncertainty. This reference flow meter's response to flow shall be used as the reference value for CVS flow-meter calibration;

• (d) An upstream screen or other pressure restriction that could affect the flow ahead of the reference flow meter may not be used, unless the flow meter has been calibrated with such a pressure restriction;

• (e) The calibration sequence described under this point 8.1.8.4 refers to the molar based approach. For the corresponding sequence used in the mass based approach, see point 2.5 of Annex VII.

• (f) CFV or SSV may alternatively be removed from its permanent position for calibration as long as the following requirements are met when installed in the CVS:
• (1) Upon installation of the CFV or SSV into the CVS, good engineering judgment shall be applied to verify that any leaks have not been introduced between the CVS inlet and the venturi.
• (2) After ex-situ venturi calibration, all venturi flow combinations must be verified for CFVs or at minimum of 10 flow points for an SSV using the propane check as described in point 8.1.8.5. The result of the propane check for each venturi flow point may not exceed the tolerance in point 8.1.8.5.6.
• (3) In order to verify the ex-situ calibration for a CVS with more than a single CFV, the following verification shall be conducted:
• (i) A constant flow device shall be used to deliver a constant flow of propane to the dilution tunnel.
• (ii) The hydrocarbon concentrations shall be measured at a minimum of 10 separate flow rates for an SSV flow meter, or at all possible flow combinations for a CFV flow meter, while keeping the flow of propane constant.
• (iii) The concentration of hydrocarbon background in the dilution air shall be measured at the beginning and end of this test. The average background concentration from each measurement at each flow point must be subtracted before performing the regression analysis in paragraph (iv).
• (iv) A power regression has to be performed using all the paired values of flow rate and corrected concentration to obtain a relationship in the form of y = a × x^b, using the concentration as the independent variable and the flow rate as the dependent variable. For each data point, the calculation of the difference between the measured flow rate and the value represented by the curve fit is required. The difference at each point must be less than ± 1 % of the appropriate regression value. The value of b must be between – 1,005 and – 0,995. If the results do not meet these limits, corrective actions consistent with point 8.1.8.5.1(a) must be taken.

8.1.8.4.2.   PDP calibration

A positive-displacement pump (PDP) shall be calibrated to determine a flow-versus-PDP speed equation that accounts for flow leakage across sealing surfaces in the PDP as a function of PDP inlet pressure. Unique equation coefficients shall be determined for each speed at which the PDP is operated. A PDP flow meter shall be calibrated as follows:

• (a) The system shall be connected as shown in Figure 6.5;

• (b) Leaks between the calibration flow meter and the PDP shall be less than 0,3 % of the total flow at the lowest calibrated flow point; for example, at the highest pressure restriction and lowest PDP-speed point;

• (c) While the PDP operates, a constant temperature at the PDP inlet shall be maintained within ± 2 % of the mean absolute inlet temperature, T in;

• (d) The PDP speed is set to the first speed point at which it is intended to calibrate;

• (e) The variable restrictor is set to its wide-open position;

• (f) The PDP is operated for at least 3 minutes to stabilize the system. Then by continuously operating the PDP, the mean values of at least 30 s of sampled data of each of the following quantities are recorded:
• (i) Image omitted; see official source
• (ii) The mean temperature at the PDP inlet, T in;
• (iii) The mean static absolute pressure at the PDP inlet, p in;
• (iv) The mean static absolute pressure at the PDP outlet, p out;
• (v) The mean PDP speed, n PDP;

• (g) The restrictor valve shall be incrementally closed to decrease the absolute pressure at the inlet to the PDP, p in;

• (h) The steps in paragraphs 8.1.8.4.2(f) and (g) shall be repeated to record data at a minimum of six restrictor positions reflecting the full range of possible in-use pressures at the PDP inlet;

• (i) The PDP shall be calibrated by using the collected data and the equations set out in Annex VII;

• (j) The steps in paragraphs (f) to (i) of this point shall be repeated for each speed at which the PDP is operated;

• (k) The equations in section 3 of Annex VII (molar based approach) or section 2 of Annex VII (mass based approach) shall be used to determine the PDP flow equation for emission testing;

• (l) The calibration shall be verified by performing a CVS verification (i.e., propane check) as described in point 8.1.8.5;

• (m) The PDP may not be used below the lowest inlet pressure tested during calibration.

8.1.8.4.3.   CFV calibration

A critical-flow venturi (CFV) shall be calibrated to verify its discharge coefficient, C d, at the lowest expected static differential pressure between the CFV inlet and outlet. A CFV flow meter shall be calibrated as follows:

• (a) The system shall be connected as shown in Figure 6.5;

• (b) The blower shall be started downstream of the CFV;

• (c) While the CFV operates, a constant temperature at the CFV inlet shall be maintained within ± 2 % of the mean absolute inlet temperature, T in;

• (d) Leaks between the calibration flow meter and the CFV shall be less than 0,3 % of the total flow at the highest pressure restriction;

• (e) The variable restrictor shall be set to its wide-open position. In lieu of a variable restrictor the pressure downstream of the CFV may be varied by varying blower speed or by introducing a controlled leak. Note that some blowers have limitations on non-loaded conditions;

• (f) The CFV shall be operated for at least 3 minutes to stabilize the system. The CFV shall continue operating and the mean values of at least 30 s of sampled data of each of the following quantities shall be recorded:
• (i) Image omitted; see official source
• (ii) Optionally, the mean dew point of the calibration air, T dew. See Annex VII for permissible assumptions during emission measurements;
• (iii) The mean temperature at the venturi inlet, T in;
• (iv) The mean static absolute pressure at the venturi inlet, p in;
• (v) The mean static differential pressure between the CFV inlet and the CFV outlet, Δp CFV;

• (g) The restrictor valve shall be incrementally closed to decrease the absolute pressure at the inlet to the CFV, p in;

• (h) The steps in paragraphs (f) and (g) of this point shall be repeated to record mean data at a minimum of ten restrictor positions, such that the fullest practical range of Δp CFV expected during testing is tested. It is not required to remove calibration components or CVS components to calibrate at the lowest possible pressure restrictions;

• (i) Cd and the highest allowable pressure ratio r shall be determined as described in Annex VII;

• (j) Cd shall be used to determine CFV flow during an emission test. The CFV shall not be used above the highest allowed r, as determined in Annex VII;

• (k) The calibration shall be verified by performing a CVS verification (i.e., propane check) as described in point 8.1.z8.5;

• (l) If the CVS is configured to operate more than one CFV at a time in parallel, the CVS shall be calibrated by one of the following:
• (i) Every combination of CFVs shall be calibrated according to this section and with Annex VII. See Annex VII for instructions on calculating flow rates for this option;
• (ii) Each CFV shall be calibrated according to this point and Annex VII. See Annex VII for instructions on calculating flow rates for this option.

8.1.8.4.4.   SSV calibration

A subsonic venturi (SSV) shall be calibrated to determine its calibration coefficient, C d, for the expected range of inlet pressures. An SSV flow meter shall be calibrated as follows:

• (a) The system shall be connected as shown in Figure 6.5;

• (b) The blower shall be started downstream of the SSV;

• (c) Leaks between the calibration flow meter and the SSV shall be less than 0,3 % of the total flow at the highest pressure restriction;

• (d) While the SSV operates, a constant temperature at the SSV inlet shall be maintained within ± 2 % of the mean absolute inlet temperature, T in;

• (e) The variable restrictor or variable-speed blower shall be set to a flow rate greater than the greatest flow rate expected during testing. Flow rates may not be extrapolated beyond calibrated values, so it is recommended that it is made certain that a Reynolds number, Re, at the SSV throat at the greatest calibrated flow rate is greater than the maximum Re expected during testing;

• (f) The SSV shall be operated for at least 3 min to stabilize the system. The SSV shall continue operating and the mean of at least 30 s of sampled data of each of the following quantities shall be recorded:
• (i) Image omitted; see official source
• (ii) Optionally, the mean dew point of the calibration air, T dew. See Annex VII for permissible assumptions;
• (iii) The mean temperature at the venturi inlet, T in;
• (iv) The mean static absolute pressure at the venturi inlet, p in;
• (v) Static differential pressure between the static pressure at the venturi inlet and the static pressure at the venturi throat, Δp SSV;

• (g) The restrictor valve shall be incrementally closed or the blower speed decreased to decrease the flow rate;

• (h) The steps in paragraphs (f) and (g) of this point shall be repeated to record data at a minimum of ten flow rates;

• (i) A functional form of C d versus Re shall be determined by using the collected data and the equations in Annex VII;

• (j) The calibration shall be verified by performing a CVS verification (i.e., propane check) as described in point 8.1.8.5 using the new C d versus Re equation;

• (k) The SSV shall be used only between the minimum and maximum calibrated flow rates;

• (l) The equations in section 3 of Annex VII (molar based approach) or section 2 of Annex VII (mass based approach) shall be used to determine SSV flow during a test.

8.1.8.4.5.   Ultrasonic calibration (reserved)

Figure 6.5

Schematic diagrams for diluted exhaust gas flow CVS calibration

Image omitted; see official source

8.1.8.5.   CVS and batch sampler verification (propane check)

8.1.8.5.1.   Introduction

• (a) A propane check serves as a CVS verification to determine if there is a discrepancy in measured values of diluted exhaust gas flow. A propane check also serves as a batch-sampler verification to determine if there is a discrepancy in a batch sampling system that extracts a sample from a CVS, as described in paragraph (f) of this point. Using good engineering judgment and safe practices, this check may be performed using a gas other than propane, such as CO2 or CO. A failed propane check might indicate one or more problems that may require corrective action, as follows:
• (i) Incorrect analyzer calibration. The FID analyzer shall be re-calibrated, repaired, or replaced;
• (ii) Leak checks shall be performed on CVS tunnel, connections, fasteners, and HC sampling system according to point 8.1.8.7;
• (iii) The verification for poor mixing shall be performed in accordance with point 9.2.2;
• (iv) The hydrocarbon contamination verification in the sample system shall be performed as described in point 7.3.1.3;
• (v) Change in CVS calibration. An in-situ calibration of the CVS flow meter shall be performed as described in point 8.1.8.4;
• (vi) Other problems with the CVS or sampling verification hardware or software. The CVS system, CVS verification hardware, and software shall be inspected for discrepancies;

• (b) A propane check uses either a reference mass or a reference flow rate of C3H8 as a tracer gas in a CVS. If a reference flow rate is used, any non-ideal gas behaviour of C3H8 in the reference flow meter shall be accounted for. See section 2 of Annex VII (mass based approach) or section 3 of Annex VII (molar based approach), which describe how to calibrate and use certain flow meters. No ideal gas assumption may be used in point 8.1.8.5 and Annex VII. The propane check compares the calculated mass of injected C3H8 using HC measurements and CVS flow rate measurements with the reference value.

8.1.8.5.2.   Method of introducing a known amount of propane into the CVS system

The total accuracy of the CVS sampling system and analytical system shall be determined by introducing a known mass of a pollutant gas into the system while it is being operated in the normal manner. The pollutant is analyzed, and the mass calculated in accordance with Annex VII. Either of the following two techniques shall be used:

• (a) Metering by means of a gravimetric technique shall be done as follows: A mass of a small cylinder filled with carbon monoxide or propane shall be determined with a precision of ± 0,01 g. For about 5 to 10 minutes, the CVS system shall be operated as in a normal exhaust emissions test, while carbon monoxide or propane is injected into the system. The quantity of pure gas discharged shall be determined by means of differential weighing. A gas sample shall be analyzed with the usual equipment (sampling bag or integrating method), and the mass of the gas calculated;

• (b) Metering with a critical flow orifice shall be done as follows: A known quantity of pure gas (carbon monoxide or propane) shall be fed into the CVS system through a calibrated critical orifice. If the inlet pressure is high enough, the flow rate, which is adjusted by means of the critical flow orifice, is independent of the orifice outlet pressure (critical flow). The CVS system shall be operated as in a normal exhaust emissions test for about 5 to 10 minutes. A gas sample shall be analyzed with the usual equipment (sampling bag or integrating method), and the mass of the gas calculated.

8.1.8.5.3.   Preparation of the propane check

The propane check shall be prepared as follows:

• (a) If a reference mass of C3H8 is used instead of a reference flow rate, a cylinder charged with C3H8 shall be obtained. The reference cylinder's mass of C3H8 shall be determined within ± 0,5 % of the amount of C3H8 that is expected to be used;

• (b) Appropriate flow rates shall be selected for the CVS and C3H8;

• (c) A C3H8 injection port shall be selected in the CVS. The port location shall be selected to be as close as practical to the location where engine exhaust system is introduced into the CVS. The C3H8 cylinder shall be connected to the injection system;

• (d) The CVS shall be operated and stabilized;

• (e) Any heat exchangers in the sampling system shall be pre-heated or pre-cooled;

• (f) Heated and cooled components such as sample lines, filters, chillers, and pumps shall be allowed to stabilize at operating temperature;

• (g) If applicable, a vacuum side leak verification of the HC sampling system shall be performed as described in point 8.1.8.7.

8.1.8.5.4.   Preparation of the HC sampling system for the propane check

Vacuum side leak check verification of the HC sampling system may be performed in accordance with point (g). If this procedure is used, the HC contamination procedure set out in point 7.3.1.3 may be used. If the vacuum side leak check is not performed according to paragraph (g), then the HC sampling system shall be zeroed, spanned, and verified for contamination, as follows:

• (a) The lowest HC analyzer range that can measure the C3H8 concentration expected for the CVS and C3H8 flow rates shall be selected;

• (b) The HC analyzer shall be zeroed using zero air introduced at the analyzer port;

• (c) The HC analyzer shall be spanned using C3H8 span gas introduced at the analyzer port;

• (d) Zero air shall be overflowed at the HC probe or into a fitting between the HC probe and the transfer line;

• (e) The stable HC concentration of the HC sampling system shall be measured as overflow zero air flows. For batch HC measurement, the batch container (such as a bag) shall be filled and the HC overflow concentration measured;

• (f) If the overflow HC concentration exceeds 2 μmol/mol, the procedure may not be advanced until contamination is eliminated. The source of the contamination shall be determined and corrective action taken, such as cleaning the system or replacing contaminated portions;

• (g) When the overflow HC concentration does not exceed 2 μmol/mol, this value shall be recorded as x HCinit and it shall be used to correct for HC contamination as described in section 2 of Annex VII (mass based approach) or section 3 of Annex VII (molar based approach).

8.1.8.5.5.   Propane check performance