ANNEX 4B

Test procedure for compression-ignition (CI) engines and positive-ignition (PI) engines fuelled with natural gas (NG) or liquefied petroleum gas (LPG) incorporating the world-wide harmonised heavy duty certification (WHDC, global technical regulation (gtr) No 4)

1.   APPLICABILITY

This annex is not applicable for the purpose of type approval according to this Regulation for the time being. It will be made applicable in the future.

2.   RESERVED. (1)

3.   DEFINITIONS, SYMBOLS AND ABBREVIATIONS

3.1.   Definitions

For the purpose of this Regulation,

3.1.1.   “continuous regeneration” means the regeneration process of an exhaust after-treatment system that occurs either permanently or at least once per WHTC hot start test. Such a regeneration process will not require a special test procedure.

3.1.2.   “delay time” means the difference in time between the change of the component to be measured at the reference point and a system response of 10 per cent of the final reading (t10) with the sampling probe being defined as the reference point. For the gaseous components, this is the transport time of the measured component from the sampling probe to the detector.

3.1.3.   “deNOx system” means an exhaust after-treatment system designed to reduce emissions of oxides of nitrogen (NOx) (e.g. passive and active lean NOx catalysts, NOx adsorbers and selective catalytic reduction (SCR) systems).

3.1.4.   “diesel engine” means an engine which works on the compression-ignition principle.

3.1.5.   “drift” means the difference between the zero or span responses of the measurement instrument after and before an emissions test.

3.1.6.   “engine family” means a manufacturers grouping of engines which, through their design as defined in paragraph 5.2 of this Annex, have similar exhaust emission characteristics; all members of the family shall comply with the applicable emission limit values.

3.1.7.   “engine system” means the engine, the emission control system and the communication interface (hardware and messages) between the engine system electronic control unit(s) (ECU) and any other powertrain or vehicle control unit.

3.1.8.   “engine type” means a category of engines which do not differ in essential engine characteristics.

3.1.9.   “exhaust after-treatment system” means a catalyst (oxidation or 3-way), particulate filter, deNOx system, combined deNOx particulate filter or any other emission-reducing device that is installed downstream of the engine. This definition excludes exhaust gas recirculation (EGR), which is considered an integral part of the engine.

3.1.10.   “full flow dilution method” means the process of mixing the total exhaust flow with diluent prior to separating a fraction of the diluted exhaust stream for analysis.

3.1.11.   “gaseous pollutants” means carbon monoxide, hydrocarbons and/or non-methane hydrocarbons (assuming a ratio of CH1,85 for diesel, CH2,525 for LPG and CH2.93 for NG, and an assumed molecule CH3O0,5 for ethanol fuelled diesel engines), methane (assuming a ratio of CH4 for NG) and oxides of nitrogen (expressed in nitrogen dioxide (NO2) equivalent).

3.1.12.   “high speed (n hi)” means the highest engine speed where 70 per cent of the declared maximum power occurs.

3.1.13.   “low speed (n lo)” means the lowest engine speed where 55 per cent of the declared maximum power occurs.

3.1.14.   “maximum power (Pmax)” means the maximum power in kW as specified by the manufacturer.

3.1.15.   “maximum torque speed” means the engine speed at which the maximum torque is obtained from the engine, as specified by the manufacturer.

3.1.16.   “normalised torque” means engine torque in per cent normalised to the maximum available torque at an engine speed.

3.1.17.   “operator demand” means an engine operator's input to control engine output. The operator may be a person (i.e. manual), or a governor (i.e. automatic) that mechanically or electronically signals an input that demands engine output. Input may be from an accelerator pedal or signal, a throttle-control lever or signal, a fuel lever or signal, a speed lever or signal, or a governor setpoint or signal.

3.1.18.   “parent engine” means an engine selected from an engine family in such a way that its emissions characteristics are representative for that engine family.

3.1.19.   “particulate after-treatment device” means an exhaust after-treatment system designed to reduce emissions of particulate pollutants (PM) through a mechanical, aerodynamic, diffusional or inertial separation.

3.1.20.   “partial flow dilution method” means the process of separating a part from the total exhaust flow, then mixing it with an appropriate amount of diluent prior to the particulate sampling filter.

3.1.21.   “particulate matter (PM)” means any material collected on a specified filter medium after diluting exhaust with a clean filtered diluent to a temperature between 315 K (42 °C) and 325 K (52 °C); this is primarily carbon, condensed hydrocarbons, and sulphates with associated water.

3.1.22.   “periodic regeneration” means the regeneration process of an exhaust after-treatment system that occurs periodically in typically less than 100 hours of normal engine operation. During cycles where regeneration occurs, emission standards may be exceeded.

3.1.23.   “ramped steady state test cycle” means a test cycle with a sequence of steady state engine test modes with defined speed and torque criteria at each mode and defined ramps between these modes (WHSC).

3.1.24.   “rated speed” means the maximum full load speed allowed by the governor as specified by the manufacturer in his sales and service literature, or, if such a governor is not present, the speed at which the maximum power is obtained from the engine, as specified by the manufacturer in his sales and service literature.

3.1.25.   “response time” means the difference in time between the change of the component to be measured at the reference point and a system response of 90 per cent of the final reading (t90) with the sampling probe being defined as the reference point, whereby the change of the measured component is at least 60 per cent full scale (FS) and takes place in less than 0,1 second. The system response time consists of the delay time to the system and of the rise time of the system.

3.1.26.   “rise time” means the difference in time between the 10 per cent and 90 per cent response of the final reading (t 90 – t 10).

3.1.27.   “span response” means the mean response to a span gas during a 30 s time interval.

3.1.28.   “specific emissions” means the mass emissions expressed in g/kWh.

3.1.29.   “test cycle” means a sequence of test points each with a defined speed and torque to be followed by the engine under steady state (WHSC) or transient operating conditions (WHTC).

3.1.30.   “transformation time” means the difference in time between the change of the component to be measured at the reference point and a system response of 50 per cent of the final reading (t 50) with the sampling probe being defined as the reference point. The transformation time is used for the signal alignment of different measurement instruments.

3.1.31.   “transient test cycle” means a test cycle with a sequence of normalised speed and torque values that vary relatively quickly with time (WHTC).

3.1.32.   “useful life” means the relevant period of distance and/or time over which compliance with the relevant gaseous and particulate emission limits has to be assured.

3.1.33.   “zero response” means the mean response to a zero gas during a 30 s time interval.

3.2.   General symbols

Symbol	Unit	Term
a 1	—	Slope of the regression
a 0	—	y intercept of the regression
A/F st	—	Stoichiometric air to fuel ratio
c	ppm/Vol per cent	Concentration
c d	ppm/Vol per cent	Concentration on dry basis
c w	ppm/Vol per cent	Concentration on wet basis
c b	ppm/Vol per cent	Background concentration
C d	—	Discharge coefficient of SSV
c gas	ppm/Vol per cent	Concentration on the gaseous components
d	m	Diameter
d V	m	Throat diameter of venturi
D 0	m3/s	PDP calibration intercept
D	—	Dilution factor
Δt	s	Time interval
e gas	g/kWh	Specific emission of gaseous components
e PM	g/kWh	Specific emission of particulates
e r	g/kWh	Specific emission during regeneration
e w	g/kWh	Weighted specific emission
E CO2	per cent	CO2 quench of NOx analyser
E E	per cent	Ethane efficiency
E H2O	per cent	Water quench of NOx analyser
E M	per cent	Methane efficiency
E NOx	per cent	Efficiency of NOx converter
f	Hz	Data sampling rate
f a	—	Laboratory atmospheric factor
F s	—	Stoichiometric factor
H a	g/kg	Absolute humidity of the intake air
H d	g/kg	Absolute humidity of the diluent
i	—	Subscript denoting an instantaneous measurement (e.g. 1 Hz)
k c	—	Carbon specific factor
k f,d	m3/kg fuel	Combustion additional volume of dry exhaust
k f,w	m3/kg fuel	Combustion additional volume of wet exhaust
k h,D	—	Humidity correction factor for NOx for CI engines
k h,G	—	Humidity correction factor for NOx for PI engines
k r,u	—	Upward regeneration adjustment factor
k r,d	—	Downward regeneration adjustment factor
k w,a	—	Dry to wet correction factor for the intake air
k w,d	—	Dry to wet correction factor for the diluent
k w,e	—	Dry to wet correction factor for the diluted exhaust gas
k w,r	—	Dry to wet correction factor for the raw exhaust gas
K V	—	CFV calibration function
λ	—	Excess air ratio
m b	mg	Particulate sample mass of the diluent collected
m d	kg	Mass of the diluent sample passed through the particulate sampling filters
m ed	kg	Total diluted exhaust mass over the cycle
m edf	kg	Mass of equivalent diluted exhaust gas over the test cycle
m ew	kg	Total exhaust mass over the cycle
m gas	g	Mass of gaseous emissions over the test cycle
m f	mg	Particulate sampling filter mass
m p	mg	Particulate sample mass collected
m PM	g	Mass of particulate emissions over the test cycle
m se	kg	Exhaust sample mass over the test cycle
m sed	kg	Mass of diluted exhaust gas passing the dilution tunnel
m sep	kg	Mass of diluted exhaust gas passing the particulate collection filters
m ssd	kg	Mass of secondary diluent
M	Nm	Torque
M a	g/mol	Molar mass of the intake air
M d	g/mol	Molar mass of the diluent
M e	g/mol	Molar mass of the exhaust
M f	Nm	Torque absorbed by auxiliaries/equipment to be fitted
M gas	g/mol	Molar mass of gaseous components
M r	Nm	Torque absorbed by auxiliaries/equipment to be removed
n	—	Number of measurements
nr	—	Number of measurements with regeneration
n	min–1	Engine rotational speed
n hi	min–1	High engine speed
n lo	min–1	Low engine speed
n pref	min–1	Preferred engine speed
n p	r/s	PDP pump speed
p a	kPa	Saturation vapour pressure of engine intake air
p b	kPa	Total atmospheric pressure
p d	kPa	Saturation vapour pressure of the diluent
P f	kW	Power absorbed by auxiliaries/equipment to be fitted
p p	kPa	Absolute pressure
p r	kW	Water vapour pressure after cooling bath
p s	kPa	Dry atmospheric pressure
P	kW	Power
P r	kW	Power absorbed by auxiliaries/equipment to be removed
q mad	kg/s	Intake air mass flow rate on dry basis
q maw	kg/s	Intake air mass flow rate on wet basis
q mCe	kg/s	Carbon mass flow rate in the raw exhaust gas
q mCf	kg/s	Carbon mass flow rate into the engine
q mCp	kg/s	Carbon mass flow rate in the partial flow dilution system
q mdew	kg/s	Diluted exhaust gas mass flow rate on wet basis
q mdw	kg/s	Diluent mass flow rate on wet basis
q medf	kg/s	Equivalent diluted exhaust gas mass flow rate on wet basis
q mew	kg/s	Exhaust gas mass flow rate on wet basis
q mex	kg/s	Sample mass flow rate extracted from dilution tunnel
q mf	kg/s	Fuel mass flow rate
q mp	kg/s	Sample flow of exhaust gas into partial flow dilution system
q vCVS	m3/s	CVS volume rate
q vs	dm3/min	System flow rate of exhaust analyser system
q vt	cm3/min	Tracer gas flow rate
r2	—	Coefficient of determination
r d	—	Dilution ratio
r D	—	Diameter ratio of SSV
r h	—	Hydrocarbon response factor of the FID
r m	—	Methanol response factor of the FID
r p	—	Pressure ratio of SSV
r s	—	Average sample ratio
ρ	kg/m3	Density
ρ e	kg/m3	Exhaust gas density
σ	—	Standard deviation
s		Standard deviation
T	K	Absolute temperature
T a	K	Absolute temperature of the intake air
t	s	Time
t 10	s	Time between step input and 10 per cent of final reading
t 50	s	Time between step input and 50 per cent of final reading
t 90	s	Time between step input and 90 per cent of final reading
u	—	Ratio between the densities (or molar masses) of the gas components and the exhaust gas divided by 1 000
V 0	m3/r	PDP gas volume pumped per revolution
V s	dm3	System volume of exhaust analyser bench
W act	kWh	Actual cycle work of the test cycle
W ref	kWh	Reference cycle work of the test cycle
X 0	m3/r	PDP calibration function

3.3.   Symbols and abbreviations for the fuel composition

w ALF	hydrogen content of fuel, per cent mass
w BET	carbon content of fuel, per cent mass
w GAM	sulphur content of fuel, per cent mass
w DEL	nitrogen content of fuel, per cent mass
w EPS	oxygen content of fuel, per cent mass
α	molar hydrogen ratio (H/C)
γ	molar sulphur ratio (S/C)
δ	molar nitrogen ratio (N/C)
ε	molar oxygen ratio (O/C)

referring to a fuel CH α O ε N δ S γ

3.4.   Symbols and abbreviations for the chemical components

C1	Carbon 1 equivalent hydrocarbon
CH4	Methane
C2H6	Ethane
C3H8	Propane
CO	Carbon monoxide
CO2	Carbon dioxide
DOP	Di-octylphtalate
HC	Hydrocarbons
H2O	Water
NMHC	Non-methane hydrocarbons
NOx	Oxides of nitrogen
NO	Nitric oxide
NO2	Nitrogen dioxide
PM	Particulate matter

3.5.   Abbreviations

CFV	Critical Flow Venturi
CLD	Chemiluminescent Detector
CVS	Constant Volume Sampling
deNOx	NOx after-treatment system
EGR	Exhaust gas recirculation
FID	Flame Ionisation Detector
GC	Gas Chromatograph
HCLD	Heated Chemiluminescent Detector
HFID	Heated Flame Ionisation Detector
LPG	Liquefied Petroleum Gas
NDIR	Non-Dispersive Infrared (Analyser)
NG	Natural Gas
NMC	Non-Methane Cutter
PDP	Positive Displacement Pump
Per cent FS	Per cent of full scale
PFS	Partial Flow System
SSV	Subsonic Venturi
VGT	Variable Geometry Turbine

4.   GENERAL REQUIREMENTS

The engine system shall be so designed, constructed and assembled as to enable the engine in normal use to comply with the provisions of this Annex during its useful life, as defined in this Regulation, including when installed in the vehicle.

5.   PERFORMANCE REQUIREMENTS

5.1.   Emission of gaseous and particulate pollutants

The emissions of gaseous and particulate pollutants by the engine shall be determined on the WHTC and WHSC test cycles, as described in paragraph 7. The measurement systems shall meet the linearity requirements in paragraph 9.2 and the specifications in paragraph 9.3 (gaseous emissions measurement), paragraph 9.4 (particulate measurement) and in Appendix 3.

Other systems or analysers may be approved by the type approval authority, if it is found that they yield equivalent results in accordance with paragraph 5.1.1.

5.1.1.   Equivalency

The determination of system equivalency shall be based on a seven-sample pair (or larger) correlation study between the system under consideration and one of the systems of this Annex.

“Results” refer to the specific cycle weighted emissions value. The correlation testing is to be performed at the same laboratory, test cell, and on the same engine, and is preferred to be run concurrently. The equivalency of the sample pair averages shall be determined by F-test and t-test statistics as described in Appendix 4, paragraph A.4.3, obtained under the laboratory test cell and the engine conditions described above. Outliers shall be determined in accordance with ISO 5725 and excluded from the database. The systems to be used for correlation testing shall be subject to the approval by the type approval authority.

5.2.   Engine family

5.2.1.   General

An engine family is characterised by design parameters. These shall be common to all engines within the family. The engine manufacturer may decide which engines belong to an engine family, as long as the membership criteria listed in paragraph 5.2.3 are respected. The engine family shall be approved by the type approval authority. The manufacturer shall provide to the type approval authority the appropriate information relating to the emission levels of the members of the engine family.

5.2.2.   Special cases

In some cases there may be interaction between parameters. This shall be taken into consideration to ensure that only engines with similar exhaust emission characteristics are included within the same engine family. These cases shall be identified by the manufacturer and notified to the type approval authority. It shall then be taken into account as a criterion for creating a new engine family.

In case of devices or features, which are not listed in paragraph 5.2.3 and which have a strong influence on the level of emissions, this equipment shall be identified by the manufacturer on the basis of good engineering practice, and shall be notified to the type approval authority. It shall then be taken into account as a criterion for creating a new engine family.

In addition to the parameters listed in paragraph 5.2.3, the manufacturer may introduce additional criteria allowing the definition of families of more restricted size. These parameters are not necessarily parameters that have an influence on the level of emissions.

5.2.3.   Parameters defining the engine family

5.2.3.1.   Combustion cycle

(a)	2-stroke cycle

(b)	4-stroke cycle

(c)	Rotary engine

(d)

Others

5.2.3.2.   Configuration of the cylinders

5.2.3.2.1.   Position of the cylinders in the block

(a)

V

(b)

In line

(c)

Radial

(d)	Others (F, W, etc.)

5.2.3.2.2.   Relative position of the cylinders

Engines with the same block may belong to the same family as long as their bore centre-to-centre dimensions are the same.

5.2.3.3.   Main cooling medium

(a)

air

(b)

water

(c)

oil

5.2.3.4.   Individual cylinder displacement

5.2.3.4.1.   Engine with a unit cylinder displacement ≥ 0,75 dm3

In order for engines with a unit cylinder displacement of ≥ 0,75 dm3 to be considered to belong to the same engine family, the spread of their individual cylinder displacements shall not exceed 15 per cent of the largest individual cylinder displacement within the family.

5.2.3.4.2.   Engine with a unit cylinder displacement < 0,75 dm3

In order for engines with a unit cylinder displacement of < 0,75 dm3 to be considered to belong to the same engine family, the spread of their individual cylinder displacements shall not exceed 30 per cent of the largest individual cylinder displacement within the family.

5.2.3.4.3.   Engine with other unit cylinder displacement limits

Engines with an individual cylinder displacement that exceeds the limits defined in paragraphs 5.2.3.4.1 and 5.2.3.4.2 may be considered to belong to the same family with the approval of the type approval authority. The approval shall be based on technical elements (calculations, simulations, experimental results etc.) showing that exceeding the limits does not have a significant influence on the exhaust emissions.

5.2.3.5.   Method of air aspiration

(a)	Naturally aspirated

(b)	Pressure charged

(c)	Pressure charged with charge cooler

5.2.3.6.   Fuel type

(a)

Diesel

(b)	Natural gas (NG)

(c)	Liquefied petroleum gas (LPG)

(d)

Ethanol

5.2.3.7.   Combustion chamber type

(a)	Open chamber

(b)	Divided chamber

(c)	Other types

5.2.3.8.   Ignition Type

(a)	Positive ignition

(b)	Compression ignition

5.2.3.9.   Valves and porting

(a)	Configuration

(b)	Number of valves per cylinder

5.2.3.10.   Fuel supply type

(a)	Liquid fuel supply type (i) Pump and (high pressure) line and injector (ii) In-line or distributor pump (iii) Unit pump or unit injector (iv) Common rail (v) Carburettor(s) (vi) Others

(b)	Gas fuel supply type (i) Gaseous (ii) Liquid (iii) Mixing units (iv) Others

(c)	Other types

5.2.3.11.   Miscellaneous devices

(a)	Exhaust gas recirculation (EGR)

(b)	Water injection

(c)	Air injection

(d)

Others

5.2.3.12.   Electronic control strategy

The presence or absence of an electronic control unit (ECU) on the engine is regarded as a basic parameter of the family.

In the case of electronically controlled engines, the manufacturer shall present the technical elements explaining the grouping of these engines in the same family, i.e. the reasons why these engines can be expected to satisfy the same emission requirements.

These elements can be calculations, simulations, estimations, description of injection parameters, experimental results, etc.

Examples of controlled features are:

(a)

Timing

(b)	Injection pressure

(c)	Multiple injections

(d)	Boost pressure

(e)

VGT

(f)

EGR

5.2.3.13.   Exhaust after-treatment systems

The function and combination of the following devices are regarded as membership criteria for an engine family:

(a)	Oxidation catalyst

(b)	Three-way catalyst

(c)	DeNOx system with selective reduction of NOx (addition of reducing agent)

(d)	Other DeNOx systems

(e)	Particulate trap with passive regeneration

(f)	Particulate trap with active regeneration

(g)	Other particulate traps

(h)	Other devices

When an engine has been certified without after-treatment system, whether as parent engine or as member of the family, then this engine, when equipped with an oxidation catalyst, may be included in the same engine family, if it does not require different fuel characteristics.

If it requires specific fuel characteristics (e.g. particulate traps requiring special additives in the fuel to ensure the regeneration process), the decision to include it in the same family shall be based on technical elements provided by the manufacturer. These elements shall indicate that the expected emission level of the equipped engine complies with the same limit value as the non-equipped engine.

When an engine has been certified with after-treatment system, whether as parent engine or as member of a family, whose parent engine is equipped with the same after-treatment system, then this engine, when equipped without after-treatment system, shall not be added to the same engine family.

5.2.4.   Choice of the parent engine

5.2.4.1.   Compression ignition engines

Once the engine family has been agreed by the type approval authority, the parent engine of the family shall be selected using the primary criterion of the highest fuel delivery per stroke at the declared maximum torque speed. In the event that two or more engines share this primary criterion, the parent engine shall be selected using the secondary criterion of highest fuel delivery per stroke at rated speed.

5.2.4.2.   Positive ignition engines

Once the engine family has been agreed by the type approval authority, the parent engine of the family shall be selected using the primary criterion of the largest displacement. In the event that two or more engines share this primary criterion, the parent engine shall be selected using the secondary criterion in the following order of priority:

(a)	the highest fuel delivery per stroke at the speed of declared rated power;

(b)	the most advanced spark timing;

(c)	the lowest EGR rate.

5.2.4.3.   Remarks on the choice of the parent engine

The type approval authority may conclude that the worst-case emission of the family can best be characterised by testing additional engines. In this case, the engine manufacturer shall submit the appropriate information to determine the engines within the family likely to have the highest emissions level.

If engines within the family incorporate other features which may be considered to affect exhaust emissions, these features shall also be identified and taken into account in the selection of the parent engine.

If engines within the family meet the same emission values over different useful life periods, this shall be taken into account in the selection of the parent engine.

6.   TEST CONDITIONS

6.1.   Laboratory test conditions

The absolute temperature (T a) of the engine intake air expressed in Kelvin, and the dry atmospheric pressure (p s), expressed in kPa shall be measured and the parameter f a shall be determined according to the following provisions. In multi-cylinder engines having distinct groups of intake manifolds, such as in a “Vee” engine configuration, the average temperature of the distinct groups shall be taken. The parameter f a shall be reported with the test results. For better repeatability and reproducibility of the test results, it is recommended that the parameter f a be such that: 0,93 ≤ f a ≤ 1,07.

(a)	Compression-ignition engines: Naturally aspirated and mechanically supercharged engines: (1) Turbocharged engines with or without cooling of the intake air: (2)

(b)	Positive ignition engines: (3)

6.2.   Engines with charge air-cooling

The charge air temperature shall be recorded and shall be, at the rated speed and full load, within ± 5 K of the maximum charge air temperature specified by the manufacturer. The temperature of the cooling medium shall be at least 293 K (20 °C).

If a test laboratory system or external blower is used, the coolant flow rate shall be set to achieve a charge air temperature within ± 5 K of the maximum charge air temperature specified by the manufacturer at the rated speed and full load. Coolant temperature and coolant flow rate of the charge air cooler at the above set point shall not be changed for the whole test cycle, unless this results in unrepresentative overcooling of the charge air. The charge air cooler volume shall be based upon good engineering practice and shall be representative of the production engine's in-use installation. The laboratory system shall be designed to minimise accumulation of condensate. Any accumulated condensate shall be drained and all drains shall be completely closed before emission testing.

If the engine manufacturer specifies pressure-drop limits across the charge-air cooling system, it shall be ensured that the pressure drop across the charge-air cooling system at engine conditions specified by the manufacturer is within the manufacturer's specified limit(s). The pressure drop shall be measured at the manufacturer's specified locations.

6.3.   Engine power

The basis of specific emissions measurement is engine power and cycle work as determined in accordance with paragraphs 6.3.1 to 6.3.5.

6.3.1.   General engine installation

The engine shall be tested with the auxiliaries/equipment listed in Appendix 7.

If auxiliaries/equipment are not installed as required, their power shall be taken into account in accordance with paragraphs 6.3.2 to 6.3.5.

6.3.2.   Auxiliaries/equipment to be fitted for the emissions test

If it is inappropriate to install the auxiliaries/equipment required according to Appendix 7 on the test bench, the power absorbed by them shall be determined and subtracted from the measured engine power (reference and actual) over the whole engine speed range of the WHTC and over the test speeds of the WHSC.

6.3.3.   Auxiliaries/equipment to be removed for the test

Where the auxiliaries/equipment not required according to Appendix 7 cannot be removed, the power absorbed by them may be determined and added to the measured engine power (reference and actual) over the whole engine speed range of the WHTC and over the test speeds of the WHSC. If this value is greater than 3 per cent of the maximum power at the test speed it shall be demonstrated to the type approval authority.

6.3.4.   Determination of auxiliary power

The power absorbed by the auxiliaries/equipment needs only be determined, if

(a)	auxiliaries/equipment required according to Appendix 7, are not fitted to the engine and/or

(b)	auxiliaries/equipment not required according to Appendix 7, are fitted to the engine.

The values of auxiliary power and the measurement/calculation method for determining auxiliary power shall be submitted by the engine manufacturer for the whole operating area of the test cycles, and approved by the type approval authority.

6.3.5.   Engine cycle work

The calculation of reference and actual cycle work (see paragraphs 7.4.8 and 7.8.6) shall be based upon engine power according to paragraph 6.3.1. In this case, P f and P r of equation 4 are zero, and P equals P m.

If auxiliaries/equipment are installed according to paragraphs 6.3.2 and/or 6.3.3, the power absorbed by them shall be used to correct each instantaneous cycle power value P m,i, as follows:

(4)

where:

P m,i	is the measured engine power, kW
P f,i	is the power absorbed by auxiliaries/equipment to be fitted, kW
P r,i	is the power absorbed by auxiliaries/equipment to be removed, kW

6.4.   Engine air intake system

An engine air intake system or a test laboratory system shall be used presenting an air intake restriction within ± 300 Pa of the maximum value specified by the manufacturer for a clean air cleaner at the rated speed and full load. The static differential pressure of the restriction shall be measured at the location specified by the manufacturer.

6.5.   Engine exhaust system

An engine exhaust system or a test laboratory system shall be used presenting an exhaust backpressure within 80 to 100 per cent of the maximum value specified by the manufacturer at the rated speed and full load. If the maximum restriction is 5 kPa or less, the set point shall be no less than 1,0 kPa from the maximum. The exhaust system shall conform to the requirements for exhaust gas sampling, as set out in paragraphs 9.3.10 and 9.3.11.

6.6.   Engine with exhaust after-treatment system

If the engine is equipped with an exhaust after-treatment system, the exhaust pipe shall have the same diameter as found in-use, or as specified by the manufacturer, for at least four pipe diameters upstream of the expansion section containing the after-treatment device. The distance from the exhaust manifold flange or turbocharger outlet to the exhaust after-treatment system shall be the same as in the vehicle configuration or within the distance specifications of the manufacturer. The exhaust backpressure or restriction shall follow the same criteria as above, and may be set with a valve. For variable-restriction after-treatment devices, the maximum exhaust restriction is defined at the after-treatment condition (de-greening/aging and regeneration/loading level) specified by the manufacturer. If the maximum restriction is 5 kPa or less, the set point shall be no less than 1,0 kPa from the maximum. The after-treatment container may be removed during dummy tests and during engine mapping, and replaced with an equivalent container having an inactive catalyst support.

The emissions measured on the test cycle shall be representative of the emissions in the field. In the case of an engine equipped with a exhaust after-treatment system that requires the consumption of a reagent, the reagent used for all tests shall be declared by the manufacturer.

Engines equipped with exhaust after-treatment systems with continuous regeneration do not require a special test procedure, but the regeneration process needs to be demonstrated according to paragraph 6.6.1.

For engines equipped with exhaust after-treatment systems that are regenerated on a periodic basis, as described in paragraph 6.6.2, emission results shall be adjusted to account for regeneration events. In this case, the average emission depends on the frequency of the regeneration event in terms of fraction of tests during which the regeneration occurs.

6.6.1.   Continuous regeneration

The emissions shall be measured on an after-treatment system that has been stabilised so as to result in repeatable emissions behaviour. The regeneration process shall occur at least once during the WHTC hot start test and the manufacturer shall declare the normal conditions under which regeneration occurs (soot load, temperature, exhaust back-pressure, etc.).

In order to demonstrate that the regeneration process is continuous, at least three WHTC hot start tests shall be conducted. For the purpose of this demonstration, the engine shall be warmed up in accordance with paragraph 7.4.1, the engine be soaked according to paragraph 7.6.3 and the first WHTC hot start test be run. The subsequent hot start tests shall be started after soaking according to paragraph 7.6.3. During the tests, exhaust temperatures and pressures shall be recorded (temperature before and after the after-treatment system, exhaust back pressure, etc.).

If the conditions declared by the manufacturer occur during the tests and the results of the three (or more) WHTC hot start tests do not scatter by more than ± 25 per cent or 0,005 g/kWh, whichever is greater, the after-treatment system is considered to be of the continuous regeneration type, and the general test provisions of paragraph 7.6 (WHTC) and paragraph 7.7 (WHSC) apply.

If the exhaust after-treatment system has a security mode that shifts to a periodic regeneration mode, it shall be checked according to paragraph 6.6.2. For that specific case, the applicable emission limits may be exceeded and would not be weighted.

6.6.2.   Periodic regeneration

For an exhaust after-treatment based on a periodic regeneration process, the emissions shall be measured on at least three WHTC hot start tests, one with and two without a regeneration event on a stabilised after-treatment system, and the results be weighted in accordance with equation 5.

The regeneration process shall occur at least once during the WHTC hot start test. The engine may be equipped with a switch capable of preventing or permitting the regeneration process provided this operation has no effect on the original engine calibration.

The manufacturer shall declare the normal parameter conditions under which the regeneration process occurs (soot load, temperature, exhaust back-pressure, etc.) and its duration. The manufacturer shall also provide the frequency of the regeneration event in terms of number of tests during which the regeneration occurs compared to number of tests without regeneration. The exact procedure to determine this frequency shall be based upon in use data using good engineering judgement, and shall be agreed by the type approval or certification authority.

The manufacturer shall provide an after-treatment system that has been loaded in order to achieve regeneration during a WHTC test. For the purpose of this testing, the engine shall be warmed up in accordance with paragraph 7.4.1, the engine be soaked according to paragraph 7.6.3 and the WHTC hot start test be started. Regeneration shall not occur during the engine warm-up.

Average specific emissions between regeneration phases shall be determined from the arithmetic mean of several approximately equidistant WHTC hot start test results (g/kWh). As a minimum, at least one WHTC hot start test as close as possible prior to a regeneration test and one WHTC hot start test immediately after a regeneration test shall be conducted. As an alternative, the manufacturer may provide data to show that the emissions remain constant (±25 per cent or 0,005 g/kWh, whichever is greater) between regeneration phases. In this case, the emissions of only one WHTC hot start test may be used.

During the regeneration test, all the data needed to detect regeneration shall be recorded (CO or NOx emissions, temperature before and after the after-treatment system, exhaust back pressure, etc.).

During the regeneration test, the applicable emission limits may be exceeded.

The test procedure is schematically shown in figure 2.

The WHTC hot start emissions shall be weighted as follows:

(5)

where:

n	is the number of WHTC hot start tests without regeneration
nr	is the number of WHTC hot start tests with regeneration (minimum one test)
	is the average specific emission without regeneration, g/kWh
	is the average specific emission with regeneration, g/kWh

For the determination of , the following provisions apply:

(a)	If regeneration takes more than one hot start WHTC, consecutive full hot start WHTC tests shall be conducted and emissions continued to be measured without soaking and without shutting the engine off, until regeneration is completed, and the average of the hot start WHTC tests be calculated.

(b)	If regeneration is completed during any hot start WHTC, the test shall be continued over its entire length.

In agreement with the type approval authority, the regeneration adjustment factors may be applied either multiplicative (c) or additive (d) based upon good engineering analysis.

(c)	The multiplicative adjustment factors shall be calculated as follows: (upward) (6) (downward) (6a)

(d)	The additive adjustment factors shall be calculated as follows: (upward) (7) (downward) (8)

With reference to the specific emission calculations in paragraph 8.6.3, the regeneration adjustment factors shall be applied, as follows:

(e)	for a test without regeneration, k r,u shall be multiplied with or be added to, respectively, the specific emission e in equations 69 or 70,

(f)	for a test with regeneration, k r,d shall be multiplied with or be subtracted from, respectively, the specific emission e in equations 69 or 70.

At the request of the manufacturer, the regeneration adjustment factors

(g)	may be extended to other members of the same engine family,

(h)	may be extended to other engine families using the same after-treatment system with the prior approval of the type approval or certification authority based on technical evidence to be supplied by the manufacturer, that the emissions are similar.

6.7.   Cooling system

An engine cooling system with sufficient capacity to maintain the engine at normal operating temperatures prescribed by the manufacturer shall be used.

6.8.   Lubricating oil

The lubricating oil shall be specified by the manufacturer and be representative of lubricating oil available on the market; the specifications of the lubricating oil used for the test shall be recorded and presented with the results of the test.

6.9.   Specification of the reference fuel

The reference fuel is specified in Appendix 2 of this Annex for CI engines and in Annexes 6 and 7 for CNG and LPG fuelled engines.

The fuel temperature shall be in accordance with the manufacturer’s recommendations.

6.10.   Crankcase emissions

No crankcase emissions shall be discharged directly into the ambient atmosphere, with the following exception: engines equipped with turbochargers, pumps, blowers, or superchargers for air induction may discharge crankcase emissions to the ambient atmosphere if the emissions are added to the exhaust emissions (either physically or mathematically) during all emission testing. Manufacturers taking advantage of this exception shall install the engines so that all crankcase emission can be routed into the emissions sampling system.

For the purpose of this paragraph, crankcase emissions that are routed into the exhaust upstream of exhaust after-treatment during all operation are not considered to be discharged directly into the ambient atmosphere.

Open crankcase emissions shall be routed into the exhaust system for emission measurement, as follows:

(a)	The tubing materials shall be smooth-walled, electrically conductive, and not reactive with crankcase emissions. Tube lengths shall be minimised as far as possible.

(b)	The number of bends in the laboratory crankcase tubing shall be minimised, and the radius of any unavoidable bend shall be maximised.

(c)	The laboratory crankcase exhaust tubing shall be heated, thin-walled or insulated and shall meet the engine manufacturer’s specifications for crankcase back pressure.

(d)

The crankcase exhaust tubing shall connect into the raw exhaust downstream of any after-treatment system, downstream of any installed exhaust restriction, and sufficiently upstream of any sample probes to ensure complete mixing with the engine’s exhaust before sampling. The crankcase exhaust tube shall extend into the free stream of exhaust to avoid boundary-layer effects and to promote mixing. The crankcase exhaust tube’s outlet may orient in any direction relative to the raw exhaust flow.

7.   TEST PROCEDURES

7.1.   Principles of emissions measurement

To measure the specific emissions, the engine shall be operated over the test cycles defined in paragraphs 7.2.1 and 7.2.2. The measurement of specific emissions requires the determination of the mass of components in the exhaust and the corresponding engine cycle work. The components are determined by the sampling methods described in paragraphs 7.1.1 and 7.1.2.

7.1.1.   Continuous sampling

In continuous sampling, the component’s concentration is measured continuously from raw or dilute exhaust. This concentration is multiplied by the continuous (raw or dilute) exhaust flow rate at the emission sampling location to determine the component’s mass flow rate. The component’s emission is continuously summed over the test cycle. This sum is the total mass of the emitted component.

7.1.2.   Batch sampling

In batch sampling, a sample of raw or dilute exhaust is continuously extracted and stored for later measurement. The extracted sample shall be proportional to the raw or dilute exhaust flow rate. Examples of batch sampling are collecting diluted gaseous components in a bag and collecting particulate matter (PM) on a filter. The batch sampled concentrations are multiplied by the total exhaust mass or mass flow (raw or dilute) from which it was extracted during the test cycle. This product is the total mass or mass flow of the emitted component. To calculate the PM concentration, the PM deposited onto a filter from proportionally extracted exhaust shall be divided by the amount of filtered exhaust.

7.1.3.   Measurement procedures

This annex applies two measurement procedures that are functionally equivalent. Both procedures may be used for both the WHTC and the WHSC test cycle:

(a)	the gaseous components are sampled continuously in the raw exhaust gas, and the particulates are determined using a partial flow dilution system;

(b)	the gaseous components and the particulates are determined using a full flow dilution system (CVS system).

Any combination of the two principles (e.g. raw gaseous measurement and full flow particulate measurement) is permitted.

7.2.   Test cycles

7.2.1.   Transient test cycle WHTC

The transient test cycle WHTC is listed in Appendix 1 as a second-by-second sequence of normalised speed and torque values. In order to perform the test on an engine test cell, the normalised values shall be converted to the actual values for the individual engine under test based on the engine-mapping curve. The conversion is referred to as denormalisation, and the test cycle so developed as the reference cycle of the engine to be tested. With those reference speed and torque values, the cycle shall be run on the test cell, and the actual speed, torque and power values shall be recorded. In order to validate the test run, a regression analysis between reference and actual speed, torque and power values shall be conducted upon completion of the test.

For calculation of the brake specific emissions, the actual cycle work shall be calculated by integrating actual engine power over the cycle. For cycle validation, the actual cycle work shall be within prescribed limits of the reference cycle work.

For the gaseous pollutants, continuous sampling (raw or dilute exhaust gas) or batch sampling (dilute exhaust gas) may be used. The particulate sample shall be diluted with a conditioned diluent (such as ambient air), and collected on a single suitable filter. The WHTC is shown schematically in figure 3.

7.2.2.   Ramped steady state test cycle WHSC

The ramped steady state test cycle WHSC consists of a number of normalised speed and load modes which shall be converted to the reference values for the individual engine under test based on the engine-mapping curve. The engine shall be operated for the prescribed time in each mode, whereby engine speed and load shall be changed linearly within 20 ± 1 seconds. In order to validate the test run, a regression analysis between reference and actual speed, torque and power values shall be conducted upon completion of the test.

The concentration of each gaseous pollutant, exhaust flow and power output shall be determined over the test cycle. The gaseous pollutants may be recorded continuously or sampled into a sampling bag. The particulate sample shall be diluted with a conditioned diluent (such as ambient air). One sample over the complete test procedure shall be taken, and collected on a single suitable filter.

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

The WHSC is shown in table 1. Except for mode 1, the start of each mode is defined as the beginning of the ramp from the previous mode.

Table 1

WHSC test cycle

Mode	Normalised speed (per cent)	Normalised torque (per cent)	Mode length(s) incl. 20 s ramp
1	0	0	210
2	55	100	50
3	55	25	250
4	55	70	75
5	35	100	50
6	25	25	200
7	45	70	75
8	45	25	150
9	55	50	125
10	75	100	50
11	35	50	200
12	35	25	250
13	0	0	210
Sum			1 895

7.3.   General test sequence

The following flow chart outlines the general guidance that should be followed during testing. The details of each step are described in the relevant paragraphs. Deviations from the guidance are permitted where appropriate, but the specific requirements of the relevant paragraphs are mandatory.

For the WHTC, the test procedure consists of a cold start test following either natural or forced cool-down of the engine, a hot soak period and a hot start test.

For the WHSC, the test procedure consists of a hot start test following engine preconditioning at WHSC mode 9.

7.4.   Engine mapping and reference cycle

Pre-test engine measurements, pre-test engine performance checks and pre-test system calibrations shall be made prior to the engine mapping procedure in line with the general test sequence shown in paragraph 7.3.

As basis for WHTC and WHSC reference cycle generation, the engine shall be mapped under full load operation for determining the speed vs. maximum torque and speed vs. maximum power curves. The mapping curve shall be used for denormalising engine speed (paragraph 7.4.6) and engine torque (paragraph 7.4.7).

7.4.1.   Engine warm-up

The engine shall be warmed up between 75 per cent and 100 per cent of its maximum power or according to the recommendation of the manufacturer and good engineering judgment. Towards the end of the warm up it shall be operated in order to stabilise the engine coolant and lube oil temperatures to within ± 2 per cent of its mean values for at least 2 minutes or until the engine thermostat controls engine temperature.

7.4.2.   Determination of the mapping speed range

The minimum and maximum mapping speeds are defined as follows:

Minimum mapping speed	=	idle speed
Maximum mapping speed	=	n hi × 1,02 or speed where full load torque drops off to zero, whichever is smaller.

7.4.3.   Engine mapping curve

When the engine is stabilised according to paragraph 7.4.1, the engine mapping shall be performed according to the following procedure.

(a)	The engine shall be unloaded and operated at idle speed.

(b)	The engine shall be operated with maximum operator demand at minimum mapping speed.

(c)

The engine speed shall be increased at an average rate of 8 ± 1 min–1/s from minimum to maximum mapping speed, or at a constant rate such that it takes 4 to 6 min to sweep from minimum to maximum mapping speed. Engine speed and torque points shall be recorded at a sample rate of at least one point per second.

When selecting option (b) in paragraph 7.4.7 for determining negative reference torque, the mapping curve may directly continue with minimum operator demand from maximum to minimum mapping speed.

7.4.4.   Alternate mapping

If a manufacturer believes that the above mapping techniques are unsafe or unrepresentative for any given engine, alternate mapping techniques may be used. These alternate techniques shall satisfy the intent of the specified mapping procedures to determine the maximum available torque at all engine speeds achieved during the test cycles. Deviations from the mapping techniques specified in this paragraph for reasons of safety or representativeness shall be approved by the type approval authority along with the justification for their use. In no case, however, the torque curve shall be run by descending engine speeds for governed or turbocharged engines.

7.4.5.   Replicate tests

An engine need not be mapped before each and every test cycle. An engine shall be remapped prior to a test cycle if:

(a)	an unreasonable amount of time has transpired since the last map, as determined by engineering judgement, or

(b)	physical changes or recalibrations have been made to the engine which potentially affect engine performance.

7.4.6.   Denormalisation of engine speed

For generating the reference cycles, the normalised speeds of Appendix 1 (WHTC) and table 1 (WHSC) shall be denormalised using the following equation:

(9)

For determination of n pref, the integral of the maximum torque shall be calculated from n idle to n 95h from the engine mapping curve, as determined in accordance with paragraph 7.4.3.

The engine speeds in figures 4 and 5 are defined, as follows:

n lo	is the lowest speed where the power is 55 per cent of maximum power
n pref	is the engine speed where the integral of max. mapped torque is 51 per cent of the whole integral between n idle and n 95h
n hi	is the highest speed where the power is 70 per cent of maximum power
n idle	is the idle speed
n 95h	is the highest speed where the power is 95 per cent of maximum power

For engines (mainly positive ignition engines) with a steep governor droop curve, where fuel cut off does not permit to operate the engine up to n hi or n 95h, the following provisions apply:

n hi	in equation 9 is replaced with n Pmax × 1,02
n 95h	is replaced with n Pmax × 1,02

7.4.7.   Denormalisation of engine torque

The torque values in the engine dynamometer schedule of Appendix 1 (WHTC) and in table 1 (WHSC) are normalised to the maximum torque at the respective speed. For generating the reference cycles, the torque values for each individual reference speed value as determined in paragraph 7.4.6 shall be denormalised, using the mapping curve determined according to paragraph 7.4.3, as follows:

(10)

where:

M norm,i	is the normalised torque, per cent
M max,i	is the maximum torque from the mapping curve, Nm
M f,i	is the torque absorbed by auxiliaries/equipment to be fitted, Nm
M r,i	is the torque absorbed by auxiliaries/equipment to be removed, Nm

If auxiliaries/equipment are fitted in accordance with paragraph 6.3.1 and Appendix 7, M f and M r are zero.

The negative torque values of the motoring points (m in Appendix 1) shall take on, for purposes of reference cycle generation, reference values determined in either of the following ways:

(a)	negative 40 per cent of the positive torque available at the associated speed point;

(b)	mapping of the negative torque required to motor the engine from maximum to minimum mapping speed;

(c)	determination of the negative torque required to motor the engine at idle and at n hi and linear interpolation between these two points.

7.4.8.   Calculation of reference cycle work

Reference cycle work shall be determined over the test cycle by synchronously calculating instantaneous values for engine power from reference speed and reference torque, as determined in paragraphs 7.4.6 and 7.4.7. Instantaneous engine power values shall be integrated over the test cycle to calculate the reference cycle work W ref (kWh). If auxiliaries are not fitted in accordance with paragraph 6.3.1, the instantaneous power values shall be corrected using equation (4) in paragraph 6.3.5.

The same methodology shall be used for integrating both reference and actual engine power. If values are to be determined between adjacent reference or adjacent measured values, linear interpolation shall be used. In integrating the actual cycle work, any negative torque values shall be set equal to zero and included. If integration is performed at a frequency of less than 5 Hz, and if, during a given time segment, the torque value changes from positive to negative or negative to positive, the negative portion shall be computed and set equal to zero. The positive portion shall be included in the integrated value.

7.5.   Pre-test procedures

7.5.1.   Installation of the measurement equipment

The instrumentation and sample probes shall be installed as required. The tailpipe shall be connected to the full flow dilution system, if used.

7.5.2.   Preparation of measurement equipment for sampling

The following steps shall be taken before emission sampling begins:

(a)	Leak checks shall be performed within 8 hours prior to emission sampling according to paragraph 9.3.4.

(b)	For batch sampling, clean storage media shall be connected, such as evacuated bags.

(c)	All measurement instruments shall be started according to the instrument manufacturer’s instructions and good engineering judgment.

(d)	Dilution systems, sample pumps, cooling fans, and the data-collection system shall be started.

(e)	The sample flow rates shall be adjusted to desired levels, using bypass flow, if desired.

(f)	Heat exchangers in the sampling system shall be pre-heated or pre-cooled to within their operating temperature ranges for a test.

(g)	Heated or cooled components such as sample lines, filters, coolers, and pumps shall be allowed to stabilise at their operating temperatures.

(h)	Exhaust dilution system flow shall be switched on at least 10 minutes before a test sequence.

(i)	Any electronic integrating devices shall be zeroed or re-zeroed, before the start of any test interval.

7.5.3.   Checking the gas analysers

Gas analyser ranges shall be selected. Emission analysers with automatic or manual range switching are permitted. During the test cycle, the range of the emission analysers shall not be switched. At the same time the gains of an analyser’s analogue operational amplifier(s) may not be switched during the test cycle.

Zero and span response shall be determined for all analysers using internationally-traceable gases that meet the specifications of paragraph 9.3.3. FID analysers shall be spanned on a carbon number basis of one (C1).

7.5.4.   Preparation of the particulate sampling filter

At least one hour before the test, the filter shall be placed in a petri dish, which is protected against dust contamination and allows air exchange, and placed in a weighing chamber for stabilisation. At the end of the stabilisation period, the filter shall be weighed and the tare weight shall be recorded. The filter shall then be stored in a closed petri dish or sealed filter holder until needed for testing. The filter shall be used within eight hours of its removal from the weighing chamber.

7.5.5.   Adjustment of the dilution system

The total diluted exhaust gas flow of a full flow dilution system or the diluted exhaust gas flow through a partial flow dilution system shall be set to eliminate water condensation in the system, and to obtain a filter face temperature between 315 K (42 °C) and 325 K (52 °C).

7.5.6.   Starting the particulate sampling system

The particulate sampling system shall be started and operated on by-pass. The particulate background level of the diluent may be determined by sampling the diluent prior to the entrance of the exhaust gas into the dilution tunnel. The measurement may be done prior to or after the test. If the measurement is done both at the beginning and at the end of the cycle, the values may be averaged. If a different sampling system is used for background measurement, the measurement shall be done in parallel to the test run.

7.6.   WHTC cycle run

7.6.1.   Engine cool-down

A natural or forced cool-down procedure may be applied. For forced cool-down, good engineering judgment shall be used to set up systems to send cooling air across the engine, to send cool oil through the engine lubrication system, to remove heat from the coolant through the engine cooling system, and to remove heat from an exhaust after-treatment system. In the case of a forced after-treatment system cool down, cooling air shall not be applied until the after-treatment system has cooled below its catalytic activation temperature. Any cooling procedure that results in unrepresentative emissions is not permitted.

7.6.2.   Cold start test

The cold-start test shall be started when the temperatures of the engine’s lubricant, coolant, and after-treatment systems are all between 293 and 303 K (20 and 30 °C). The engine shall be started using one of the following methods:

(a)	the engine shall be started as recommended in the owners manual using a production starter motor and adequately charged battery or a suitable power supply, or

(b)

the engine shall be started by using the dynamometer. The engine shall be motored within ± 25 per cent of its typical in-use cranking speed. Cranking shall be stopped within 1 second after the engine is running. If the engine does not start after 15 seconds of cranking, cranking shall be stopped and the reason for the failure to start determined, unless the owners manual or the service-repair manual describes the longer cranking time as normal.

7.6.3.   Hot soak period

Immediately upon completion of the cold start test, the engine shall be conditioned for the hot start test using a 10 ± 1 minutes hot soak period.

7.6.4.   Hot start test

The engine shall be started at the end of the hot soak period as defined in paragraph 7.6.3. using the starting methods given in paragraph 7.6.2.

7.6.5.   Test sequence

The test sequence of both cold start and hot start test shall commence at the start of the engine. After the engine is running, cycle control shall be initiated so that engine operation matches the first set point of the cycle.

The WHTC shall be performed according to the reference cycle as set out in paragraph 7.4. Engine speed and torque command set points shall be issued at 5 Hz (10 Hz recommended) or greater. The set points shall be calculated by linear interpolation between the 1 Hz set points of the reference cycle. Actual engine speed and torque shall be recorded at least once every second during the test cycle (1 Hz), and the signals may be electronically filtered.

7.6.6.   Collection of emission relevant data

At the start of the test sequence, the measuring equipment shall be started, simultaneously:

(a)	start collecting or analysing diluent air, if a full flow dilution system is used;

(b)	start collecting or analysing raw or diluted exhaust gas, depending on the method used;

(c)	start measuring the amount of diluted exhaust gas and the required temperatures and pressures;

(d)	start recording the exhaust gas mass flow rate, if raw exhaust gas analysis is used;

(e)	start recording the feedback data of speed and torque of the dynamometer.

If raw exhaust measurement is used, the emission concentrations ((NM)HC, CO and NOx) and the exhaust gas mass flow rate shall be measured continuously and stored with at least 2 Hz on a computer system. All other data may be recorded with a sample rate of at least 1 Hz. For analogue analysers the response shall be recorded, and the calibration data may be applied online or offline during the data evaluation.

If a full flow dilution system is used, HC and NOx shall be measured continuously in the dilution tunnel with a frequency of at least 2 Hz. The average concentrations shall be determined by integrating the analyser signals over the test cycle. The system response time shall be no greater than 20 s, and shall be coordinated with CVS flow fluctuations and sampling time/test cycle offsets, if necessary. CO, CO2, and NMHC may be determined by integration of continuous measurement signals or by analysing the concentrations in the sample bag, collected over the cycle. The concentrations of the gaseous pollutants in the diluent shall be determined prior to the point where the exhaust enters into the dilution tunnel by integration or by collecting into the background bag. All other parameters that need to be measured shall be recorded with a minimum of one measurement per second (1 Hz).

7.6.7.   Particulate sampling

At the start of the test sequence, the particulate sampling system shall be switched from by-pass to collecting particulates.

If a partial flow dilution system is used, the sample pump(s) shall be controlled, so that the flow rate through the particulate sample probe or transfer tube is maintained proportional to the exhaust mass flow rate as determined in accordance with paragraph 9.4.6.1.

If a full flow dilution system is used, the sample pump(s) shall be adjusted so that the flow rate through the particulate sample probe or transfer tube is maintained at a value within ±2,5 per cent of the set flow rate. If flow compensation (i.e. proportional control of sample flow) is used, it shall be demonstrated that the ratio of main tunnel flow to particulate sample flow does not change by more than ±2,5 per cent of its set value (except for the first 10 seconds of sampling). The average temperature and pressure at the gas meter(s) or flow instrumentation inlet shall be recorded. If the set flow rate cannot be maintained over the complete cycle within ±2,5 per cent because of high particulate loading on the filter, the test shall be voided. The test shall be rerun using a lower sample flow rate.

7.6.8.   Engine stalling and equipment malfunction

If the engine stalls anywhere during the cold start test, the test shall be voided. The engine shall be preconditioned and restarted according to the requirements of paragraph 7.6.2, and the test repeated.

If the engine stalls anywhere during the hot start test, the hot start test shall be voided. The engine shall be soaked according to paragraph 7.6.3, and the hot start test repeated. In this case, the cold start test need not be repeated.

If a malfunction occurs in any of the required test equipment during the test cycle, the test shall be voided and repeated in line with the above provisions.

7.7.   WHSC cycle run

7.7.1.   Preconditioning the dilution system and the engine

The dilution system and the engine shall be started and warmed up in accordance with paragraph 7.4.1. After warm-up, the engine and sampling system shall be preconditioned by operating the engine at mode 9 (see paragraph 7.2.2, table 1) for a minimum of 10 minutes while simultaneously operating the dilution system. Dummy particulate emissions samples may be collected. Those sample filters need not be stabilised or weighed, and may be discarded. Flow rates shall be set at the approximate flow rates selected for testing. The engine shall be shut off after preconditioning.

7.7.2.   Engine starting

5 ± 1 minutes after completion of preconditioning at mode 9 as described in paragraph 7.7.1, the engine shall be started according to the manufacturer’s recommended starting procedure in the owner’s manual, using either a production starter motor or the dynamometer in accordance with paragraph 7.6.2.

7.7.3.   Test sequence

The test sequence shall commence after the engine is running and within one minute after engine operation is controlled to match the first mode of the cycle (idle).

The WHSC shall be performed according to the order of test modes listed in table 1 of paragraph 7.2.2.

7.7.4.   Collection of emission relevant data

At the start of the test sequence, the measuring equipment shall be started, simultaneously:

(a)	start collecting or analysing diluent, if a full flow dilution system is used;

(b)	start collecting or analysing raw or diluted exhaust gas, depending on the method used;

(c)	start measuring the amount of diluted exhaust gas and the required temperatures and pressures;

(d)	start recording the exhaust gas mass flow rate, if raw exhaust gas analysis is used;

(e)	start recording the feedback data of speed and torque of the dynamometer.

If raw exhaust measurement is used, the emission concentrations ((NM)HC, CO and NOx) and the exhaust gas mass flow rate shall be measured continuously and stored with at least 2 Hz on a computer system. All other data may be recorded with a sample rate of at least 1 Hz. For analogue analysers the response shall be recorded, and the calibration data may be applied online or offline during the data evaluation.

If a full flow dilution system is used, HC and NOx shall be measured continuously in the dilution tunnel with a frequency of at least 2 Hz. The average concentrations shall be determined by integrating the analyser signals over the test cycle. The system response time shall be no greater than 20 s, and shall be coordinated with CVS flow fluctuations and sampling time/test cycle offsets, if necessary. CO, CO2, and NMHC may be determined by integration of continuous measurement signals or by analysing the concentrations in the sample bag, collected over the cycle. The concentrations of the gaseous pollutants in the diluent shall be determined prior to the point where the exhaust enters into the dilution tunnel by integration or by collecting into the background bag. All other parameters that need to be measured shall be recorded with a minimum of one measurement per second (1 Hz).

7.7.5.   Particulate sampling

At the start of the test sequence, the particulate sampling system shall be switched from by-pass to collecting particulates. If a partial flow dilution system is used, the sample pump(s) shall be controlled, so that the flow rate through the particulate sample probe or transfer tube is maintained proportional to the exhaust mass flow rate as determined in accordance with paragraph 9.4.6.1.

If a full flow dilution system is used, the sample pump(s) shall be adjusted so that the flow rate through the particulate sample probe or transfer tube is maintained at a value within ±2,5 per cent of the set flow rate. If flow compensation (i.e. proportional control of sample flow) is used, it shall be demonstrated that the ratio of main tunnel flow to particulate sample flow does not change by more than ±2,5 per cent of its set value (except for the first 10 seconds of sampling). The average temperature and pressure at the gas meter(s) or flow instrumentation inlet shall be recorded. If the set flow rate cannot be maintained over the complete cycle within ±2,5 per cent because of high particulate loading on the filter, the test shall be voided. The test shall be rerun using a lower sample flow rate.

7.7.6.   Engine stalling and equipment malfunction

If the engine stalls anywhere during the cycle, the test shall be voided. The engine shall be preconditioned according to paragraph 7.7.1 and restarted according to paragraph 7.7.2, and the test repeated.

If a malfunction occurs in any of the required test equipment during the test cycle, the test shall be voided and repeated in line with the above provisions.

7.8.   Post-test procedures

7.8.1.   Operations after test

At the completion of the test, the measurement of the exhaust gas mass flow rate, the diluted exhaust gas volume, the gas flow into the collecting bags and the particulate sample pump shall be stopped. For an integrating analyser system, sampling shall continue until system response times have elapsed.

7.8.2.   Verification of proportional sampling

For any proportional batch sample, such as a bag sample or PM sample, it shall be verified that proportional sampling was maintained according to paragraphs 7.6.7 and 7.7.5. Any sample that does not fulfil the requirements shall be voided.

7.8.3.   PM conditioning and weighing

The particulate filter shall be placed into covered or sealed containers or the filter holders shall be closed, in order to protect the sample filters against ambient contamination. Thus protected, the filter shall be returned to the weighing chamber. The filter shall be conditioned for at least one hour, and then weighed according to paragraph 9.4.5. The gross weight of the filter shall be recorded.

7.8.4.   Drift verification

As soon as practical but no later than 30 minutes after the test cycle is complete or during the soak period, the zero and span responses of the gaseous analyser ranges used shall be determined. For the purpose of this paragraph, test cycle is defined as follows:

(a)	for the WHTC: the complete sequence cold — soak — hot;

(b)	for the WHTC hot start test (paragraph 6.6): the sequence soak — hot;

(c)	for the multiple regeneration WHTC hot start test (paragraph 6.6): the total number of hot start tests;

(d)	for the WHSC: the test cycle.

The following provisions apply for analyser drift:

(a)	the pre-test zero and span and post-test zero and span responses may be directly inserted into equation 66 of paragraph 8.6.1 without determining drift;

(b)	if the drift difference between the pre-test and post-test results is less than 1 per cent of full scale, the measured concentrations may be used uncorrected or may be corrected for drift according to paragraph 8.6.1;

(c)	if the drift difference between the pre-test and post-test results is equal to or greater than 1 per cent of full scale, the test shall be voided or the measured concentrations shall be corrected for drift according to paragraph 8.6.1.

7.8.5.   Analysis of gaseous bag sampling

As soon as practical, the following shall be performed:

(a)	gaseous bag samples shall be analysed no later than 30 minutes after the hot start test is complete or during the soak period for the cold start test;

(b)	background samples shall be analysed no later than 60 minutes after the hot start test is complete.

7.8.6.   Validation of cycle work

Before calculating actual cycle work, any points recorded during engine starting shall be omitted. Actual cycle work shall be determined over the test cycle by synchronously using actual speed and actual torque values to calculate instantaneous values for engine power. Instantaneous engine power values shall be integrated over the test cycle to calculate the actual cycle work W act (kWh). If auxiliaries/equipment are not fitted in accordance with paragraph 6.3.1, the instantaneous power values shall be corrected using equation (4) in paragraph 6.3.5.

The same methodology as described in paragraph 7.4.8 shall be used for integrating actual engine power.

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 paragraph 8.6.3).

W act shall be between 85 per cent and 105 per cent of W ref.

7.8.7.   Validation statistics of the test cycle

Linear regressions of the actual values (n act, M act, P act) on the reference values (n ref, M ref, P ref) shall be performed for both the WHTC and the WHSC.

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

The method of least squares shall be used, with the best-fit equation having the form:

(11)

where:

y	is the actual value of speed (min–1), torque (Nm), or power (kW)
a1	is the slope of the regression line
x	is the reference value of speed (min–1), torque (Nm), or power (kW)
a0	is the y intercept of the regression line

The standard error of estimate (SEE) of y on x and the coefficient of determination (r2) shall be calculated for each regression line.

It is recommended that this analysis be performed at 1 Hz. For a test to be considered valid, the criteria of table 2 (WHTC) or table 3 (WHSC) shall be met.

Table 2

Regression line tolerances for the WHTC

	Speed	Torque	Power
Standard error of estimate (SEE) of y on x	maximum 5 per cent of maximum test speed	maximum 10 per cent of maximum engine torque	maximum 10 per cent of maximum engine power
Slope of the regression line, a1	0,95 to 1,03	0,83 – 1,03	0,89 – 1,03
Coefficient of determination, r2	minimum 0,970	minimum 0,850	minimum 0,910
y intercept of the regression line, a0	maximum 10 per cent of idle speed	± 20 Nm or ± 2 per cent of maximum torque whichever is greater	± 4 kW or ± 2 per cent of maximum power whichever is greater

Table 3

Regression line tolerances for the WHSC

	Speed	Torque	Power
Standard error of estimate (SEE) of y on x	maximum 1 per cent of maximum test speed	maximum 2 per cent of maximum engine torque	maximum 2 per cent of maximum engine power
Slope of the regression line, a1	0,99 to 1,01	0,98 – 1,02	0,98 – 1,02
Coefficient of determination, r2	minimum 0,990	minimum 0,950	minimum 0,950
y intercept of the regression line, a0	maximum 1 per cent of maximum test speed	± 20 Nm or ± 2 per cent of maximum torque whichever is greater	± 4 kW or ± 2 per cent of maximum power whichever is greater

For regression purposes only, point omissions are permitted where noted in table 4 before doing the regression calculation. However, those points shall not be omitted for the calculation of cycle work and emissions. Point omission may be applied to the whole or to any part of the cycle.

Table 4

Permitted point omissions from regression analysis

Event	Conditions	Permitted point omissions
Minimum operator demand (idle point)	n ref = 0 per cent and M ref = 0 per cent and M act > (M ref – 0,02M max. mapped torque) and M act < (M ref + 0,02M max. mapped torque)	speed and power
Minimum operator demand (motoring point)	M ref < 0 per cent	power and torque
Minimum operator demand	n act ≤ 1,02 n ref and M act > M ref or n act > n ref and M act ≤ M ref' or n act > 1,02 n ref and M ref < M act ≤ (M ref + 0,02M max. mapped torque)	power and either torque or speed
Maximum operator demand	n act < n ref and M act ≥ M ref or n act ≥ 0,98 n ref and M act < M ref or n act < 0,98 n ref and M ref > M act ≥ (M ref – 0,02M max. mapped torque)	power and either torque or speed

8.   EMISSION CALCULATION

The final test result shall be rounded in one step to the number of places to the right of the decimal point indicated by the applicable emission standard plus one additional significant figure, in accordance with ASTM E 29-06B. No rounding of intermediate values leading to the final break-specific emission result is permitted.

Examples of the calculation procedures are given in Appendix 6.

Emissions calculation on a molar basis, in accordance with Annex 7 of gtr No [xx] concerning the exhaust emission test protocol for Non-Road Mobile Machinery (NRMM), is permitted with the prior agreement of the type approval authority.

8.1.   Dry/wet correction

If the emissions are measured on a dry basis, the measured concentration shall be converted to a wet basis according to the following equation:

(12)

where:

c d	is the dry concentration in ppm or per cent volume
kw	is the dry/wet correction factor (k w,a, k w,e, or k w,d depending on respective equation used)

8.1.1.   Raw exhaust gas

(13)

or

(14)

or

(15)

with

(16)

and

(17)

where:

H a	is the intake air humidity, g water per kg dry air
w ALF	is the hydrogen content of the fuel, per cent mass
q mf,i	is the instantaneous fuel mass flow rate, kg/s
q mad,I	is the instantaneous dry intake air mass flow rate, kg/s
p r	is the water vapour pressure after cooling bath, kPa
p b	is the total atmospheric pressure, kPa
w DEL	is the nitrogen content of the fuel, per cent mass
w EPS	is the oxygen content of the fuel, per cent mass
α	is the molar hydrogen ratio of the fuel
c CO2	is the dry CO2 concentration, per cent
c CO	is the dry CO concentration, per cent

Equations (13) and (14) are principally identical with the factor 1,008 in equations (13) and (15) being an approximation for the more accurate denominator in equation (14).

8.1.2.   Diluted exhaust gas

(18)

or

(19)

with

(20)

where:

α	is the molar hydrogen ratio of the fuel
c CO2w	is the wet CO2 concentration, per cent
c CO2d	is the dry CO2 concentration, per cent
H d	is the diluent humidity, g water per kg dry air
H a	is the intake air humidity, g water per kg dry air
D	is the dilution factor (see paragraph 8.5.2.3.2)

8.1.3.   Diluent

(21)

with

(22)

where:

H d	is the diluent humidity, g water per kg dry air

8.2.   NOx correction for humidity

As the NOx emission depends on ambient air conditions, the NOx concentration shall be corrected for humidity with the factors given in paragraph 8.2.1 or 8.2.2. The intake air humidity Ha may be derived from relative humidity measurement, dew point measurement, vapour pressure measurement or dry/wet bulb measurement using generally accepted equations.

8.2.1.   Compression-ignition engines

(23)

where:

H a	is the intake air humidity, g water per kg dry air

8.2.2.   Positive ignition engines

(24)

where:

H a	is the intake air humidity, g water per kg dry air

8.3.   Particulate filter buoyancy correction

The sampling filter mass shall be corrected for its buoyancy in air. The buoyancy correction depends on sampling filter density, air density and the density of the balance calibration weight, and does not account for the buoyancy of the PM itself. The buoyancy correction shall be applied to both tare filter mass and gross filter mass.

If the density of the filter material is not known, the following densities shall be used:

(a)	teflon coated glass fiber filter: 2 300 kg/m3

(b)	teflon membrane filter: 2 144 kg/m3

(c)	teflon membrane filter with polymethylpentene support ring: 920 kg/m3

For stainless steel calibration weights, a density of 8 000 kg/m3 shall be used. If the material of the calibration weight is different, its density shall be known.

The following equation shall be used:

(25)

with

(26)

where:

m uncor	is the uncorrected particulate filter mass, mg
ρ a	is the density of the air, kg/m3
ρ w	is the density of balance calibration weight, kg/m3
ρ f	is the density of the particulate sampling filter, kg/m3
p b	is the total atmospheric pressure, kPa
T a	is the air temperature in the balance environment, K
28,836	is the molar mass of the air at reference humidity (282,5 K), g/mol
8,3144	is the molar gas constant

The particulate sample mass mp used in paragraphs 8.4.3 and 8.5.3 shall be calculated as follows:

(27)

where:

m f,G	is the buoyancy corrected gross particulate filter mass, mg
m f,T	is the buoyancy corrected tare particulate filter mass, mg

8.4.   Partial flow dilution (PFS) and raw gaseous measurement

The instantaneous concentration signals of the gaseous components are used for the calculation of the mass emissions by multiplication with the instantaneous exhaust mass flow rate. The exhaust mass flow rate may be measured directly, or calculated using the methods of intake air and fuel flow measurement, tracer method or intake air and air/fuel ratio measurement. Special attention shall be paid to the response times of the different instruments. These differences shall be accounted for by time aligning the signals. For particulates, the exhaust mass flow rate signals are used for controlling the partial flow dilution system to take a sample proportional to the exhaust mass flow rate. The quality of proportionality shall be checked by applying a regression analysis between sample and exhaust flow in accordance with paragraph 9.4.6.1. The complete test set up is schematically shown in figure 6.

8.4.1.   Determination of exhaust gas mass flow

8.4.1.1.   Introduction

For calculation of the emissions in the raw exhaust gas and for controlling of a partial flow dilution system, it is necessary to know the exhaust gas mass flow rate. For the determination of the exhaust mass flow rate, either of the methods described in paragraphs 8.4.1.3 to 8.4.1.7 may be used.

8.4.1.2.   Response time

For the purpose of emissions calculation, the response time of either method described in paragraphs 8.4.1.3 to 8.4.1.7 shall be equal to or less than the analyser response time of ≤ 10 s, as required in paragraph 9.3.5.

For the purpose of controlling of a partial flow dilution system, a faster response is required. For partial flow dilution systems with online control, the response time shall be ≤ 0,3 s. For partial flow dilution systems with look-ahead control based on a pre-recorded test run, the response time of the exhaust flow measurement system shall be ≤ 5 s with a rise time of ≤ 1 s. The system response time shall be specified by the instrument manufacturer. The combined response time requirements for the exhaust gas flow and partial flow dilution system are indicated in paragraph 9.4.6.1.

8.4.1.3.   Direct measurement method

Direct measurement of the instantaneous exhaust flow shall be done by systems, such as:

(a)	pressure differential devices, like flow nozzle, (details see ISO 5167);

(b)	ultrasonic flowmeter;

(c)	vortex flowmeter.

Precautions shall be taken to avoid measurement errors which will impact emission value errors. Such precautions include the careful installation of the device in the engine exhaust system according to the instrument manufacturers’ recommendations and to good engineering practice. Especially, engine performance and emissions shall not be affected by the installation of the device.

The flowmeters shall meet the linearity requirements of paragraph 9.2.

8.4.1.4.   Air and fuel measurement method

This involves measurement of the airflow and the fuel flow with suitable flowmeters. The calculation of the instantaneous exhaust gas flow shall be as follows:

(28)

where:

q mew,i	is the instantaneous exhaust mass flow rate, kg/s
q maw,i	is the instantaneous intake air mass flow rate, kg/s
q mf,i	is the instantaneous fuel mass flow rate, kg/s

The flowmeters shall meet the linearity requirements of paragraph 9.2, but shall be accurate enough to also meet the linearity requirements for the exhaust gas flow.

8.4.1.5.   Tracer measurement method

This involves measurement of the concentration of a tracer gas in the exhaust.

A known amount of an inert gas (e.g. pure helium) shall be injected into the exhaust gas flow as a tracer. The gas is mixed and diluted by the exhaust gas, but shall not react in the exhaust pipe. The concentration of the gas shall then be measured in the exhaust gas sample.

In order to ensure complete mixing of the tracer gas, the exhaust gas sampling probe shall be located at least 1 m or 30 times the diameter of the exhaust pipe, whichever is larger, downstream of the tracer gas injection point. The sampling probe may be located closer to the injection point if complete mixing is verified by comparing the tracer gas concentration with the reference concentration when the tracer gas is injected upstream of the engine.

The tracer gas flow rate shall be set so that the tracer gas concentration at engine idle speed after mixing becomes lower than the full scale of the trace gas analyser.

The calculation of the exhaust gas flow shall be as follows:

(29)

where:

q mew,i	is the instantaneous exhaust mass flow rate, kg/s
q vt	is tracer gas flow rate, cm3/min
c mix,i	is the instantaneous concentration of the tracer gas after mixing, ppm
ρ e	is the density of the exhaust gas, kg/m3 (cf. table 4)
c b	is the background concentration of the tracer gas in the intake air, ppm

The background concentration of the tracer gas (c b) may be determined by averaging the background concentration measured immediately before the test run and after the test run.

When the background concentration is less than 1 per cent of the concentration of the tracer gas after mixing (c mix.i) at maximum exhaust flow, the background concentration may be neglected.

The total system shall meet the linearity requirements for the exhaust gas flow of paragraph 9.2.

8.4.1.6.   Airflow and air to fuel ratio measurement method

This involves exhaust mass calculation from the air flow and the air to fuel ratio. The calculation of the instantaneous exhaust gas mass flow is as follows:

(30)

with

(31)

(32)

where:

q maw,i	is the instantaneous intake air mass flow rate, kg/s
A/F st	is the stoichiometric air to fuel ratio, kg/kg
λ i	is the instantaneous excess air ratio
c CO2d	is the dry CO2 concentration, per cent
c COd	is the dry CO concentration, ppm
c HCw	is the wet HC concentration, ppm

Airflowmeter and analysers shall meet the linearity requirements of paragraph 9.2, and the total system shall meet the linearity requirements for the exhaust gas flow of paragraph 9.2.

If an air to fuel ratio measurement equipment such as a zirconia-type sensor is used for the measurement of the excess air ratio, it shall meet the specifications of paragraph 9.3.2.7.

8.4.1.7.   Carbon balance method

This involves exhaust mass calculation from the fuel flow and the gaseous exhaust components that include carbon. The calculation of the instantaneous exhaust gas mass flow is as follows:

(33)

with

(34)

and

(35)

where:

q mf,i	is the instantaneous fuel mass flow rate, kg/s
H a	is the intake air humidity, g water per kg dry air
w BET	is the carbon content of the fuel, per cent mass
w ALF	is the hydrogen content of the fuel, per cent mass
w DEL	is the nitrogen content of the fuel, per cent mass
w EPS	is the oxygen content of the fuel, per cent mass
c CO2d	is the dry CO2 concentration, per cent
c CO2d,a	is the dry CO2 concentration of the intake air, per cent
c CO	is the dry CO concentration, ppm
c HCw	is the wet HC concentration, ppm

8.4.2.   Determination of the gaseous components

8.4.2.1.   Introduction

The gaseous components in the raw exhaust gas emitted by the engine submitted for testing shall be measured with the measurement and sampling systems described in paragraph 9.3 and Appendix 3. The data evaluation is described in paragraph 8.4.2.2.

Two calculation procedures are described in paragraphs 8.4.2.3 and 8.4.2.4, which are equivalent for the reference fuel of Appendix 2. The procedure in paragraph 8.4.2.3 is more straightforward, since it uses tabulated u values for the ratio between component and exhaust gas density. The procedure in paragraph 8.4.2.4 is more accurate for fuel qualities that deviate from the specifications in Appendix 2, but requires elementary analysis of the fuel composition.

8.4.2.2.   Data evaluation

The emission relevant data shall be recorded and stored in accordance with paragraph 7.6.6.

For calculation of the mass emission of the gaseous components, the traces of the recorded concentrations and the trace of the exhaust gas mass flow rate shall be time aligned by the transformation time as defined in paragraph 3.1.30. Therefore, the response time of each gaseous emissions analyser and of the exhaust gas mass flow system shall be determined according to paragraphs 8.4.1.2 and 9.3.5, respectively, and recorded.

8.4.2.3.   Calculation of mass emission based on tabulated values

The mass of the pollutants (g/test) shall be determined by calculating the instantaneous mass emissions from the raw concentrations of the pollutants and the exhaust gas mass flow, aligned for the transformation time as determined in accordance with paragraph 8.4.2.2, integrating the instantaneous values over the cycle, and multiplying the integrated values with the u values from table 5. If measured on a dry basis, the dry/wet correction according to paragraph 8.1 shall be applied to the instantaneous concentration values before any further calculation is done.

For the calculation of NOx, the mass emission shall be multiplied, where applicable, with the humidity correction factor k h,D, or k h,G, as determined according to paragraph 8.2.

The following equation shall be applied:

(in g/test) (36)

where:

u gas	is the respective value of the exhaust component from table 5
c gas,i	is the instantaneous concentration of the component in the exhaust gas, ppm
q mew,i	is the instantaneous exhaust mass flow, kg/s
f	is the data sampling rate, Hz
n	is the number of measurements

Table 5

Raw exhaust gas u values and component densities

Fuel	ρ e	Gas
		NOx	CO	HC	CO2	O2	CH4
		ρ gas [kg/m3]
		2,053	1,250	(2)	1,9636	1,4277	0,716
		u gas  (3)
Diesel	1,2943	0,001586	0,000966	0,000479	0,001517	0,001103	0,000553
Ethanol	1,2757	0,001609	0,000980	0,000805	0,001539	0,001119	0,000561
CNG (4)	1,2661	0,001621	0,000987	0,000528 (5)	0,001551	0,001128	0,000565
Propane	1,2805	0,001603	0,000976	0,000512	0,001533	0,001115	0,000559
Butane	1,2832	0,001600	0,000974	0,000505	0,001530	0,001113	0,000558
LPG (6)	1,2811	0,001602	0,000976	0,000510	0,001533	0,001115	0,000559

8.4.2.4.   Calculation of mass emission based on exact equations

The mass of the pollutants (g/test) shall be determined by calculating the instantaneous mass emissions from the raw concentrations of the pollutants, the u values and the exhaust gas mass flow, aligned for the transformation time as determined in accordance with paragraph 8.4.2.2 and integrating the instantaneous values over the cycle. If measured on a dry basis, the dry/wet correction according to paragraph 8.1 shall be applied to the instantaneous concentration values before any further calculation is done.

For the calculation of NOx, the mass emission shall be multiplied with the humidity correction factor k h,D, or k h,G, as determined according to paragraph 8.2.

The following equation shall be applied:

(in g/test) (37)

where:

u gas,i	is calculated from equation 38 or 39
c gas,i	is the instantaneous concentration of the component in the exhaust gas, ppm
q mew,i	is the instantaneous exhaust mass flow, kg/s
f	is the data sampling rate, Hz
n	is the number of measurements

The instantaneous u values shall be calculated as follows:

(38)

or

(39)

with

(40)

where:

M gas	is the molar mass of the gas component, g/mol (cf. Appendix 6)
M e,i	is the instantaneous molar mass of the exhaust gas, g/mol
ρ gas	is the density of the gas component, kg/m3
ρ e,i	is the instantaneous density of the exhaust gas, kg/m3

The molar mass of the exhaust, M e, shall be derived for a general fuel composition CH α O ε N δ S γ under the assumption of complete combustion, as follows:

(41)

where:

q maw,i	is the instantaneous intake air mass flow rate on wet basis, kg/s
q mf,i	is the instantaneous fuel mass flow rate, kg/s
H a	is the intake air humidity, g water per kg dry air
M a	is the molar mass of the dry intake air = 28,965 g/mol

The exhaust density ρ e shall be derived, as follows:

(42)

where:

q mad,i	is the instantaneous intake air mass flow rate on dry basis, kg/s
q mf,i	is the instantaneous fuel mass flow rate, kg/s
H a	is the intake air humidity, g water per kg dry air
k fw	is the fuel specific factor of wet exhaust (equation 16) in paragraph 8.1.1.

8.4.3.   Particulate determination

8.4.3.1.   Data evaluation

The particulate mass shall be calculated according to equation 27 of paragraph 8.3. For the evaluation of the particulate concentration, the total sample mass (m sep) through the filter over the test cycle shall be recorded.

With the prior approval of the type approval authority, the particulate mass may be corrected for the particulate level of the diluent, as determined in paragraph 7.5.6, in line with good engineering practice and the specific design features of the particulate measurement system used.

8.4.3.2.   Calculation of mass emission

Depending on system design, the mass of particulates (g/test) shall be calculated by either of the methods in paragraphs 8.4.3.2.1 or 8.4.3.2.2 after buoyancy correction of the particulate sample filter according to paragraph 8.3.

8.4.3.2.1.   Calculation based on sample ratio

(43)

where:

m p	is the particulate mass sampled over the cycle, mg
r s	is the average sample ratio over the test cycle

with

(44)

where:

m se	is the sample mass over the cycle, kg
m ew	is the total exhaust mass flow over the cycle, kg
m sep	is the mass of diluted exhaust gas passing the particulate collection filters, kg
m sed	is the mass of diluted exhaust gas passing the dilution tunnel, kg

In case of the total sampling type system, m sep and m sed are identical.

8.4.3.2.2.   Calculation based on dilution ratio

(45)

where:

m p	is the particulate mass sampled over the cycle, mg
m sep	is the mass of diluted exhaust gas passing the particulate collection filters, kg
m edf	is the mass of equivalent diluted exhaust gas over the cycle, kg

The total mass of equivalent diluted exhaust gas mass over the cycle shall be determined as follows:

(46)

(47)

(48)

where:

q medf,i	is the instantaneous equivalent diluted exhaust mass flow rate, kg/s
q mew,i	is the instantaneous exhaust mass flow rate, kg/s
r d,i	is the instantaneous dilution ratio
q mdew,i	is the instantaneous diluted exhaust mass flow rate, kg/s
q mdw,i	is the instantaneous diluent mass flow rate, kg/s
f	is the data sampling rate, Hz
n	is the number of measurements

8.5.   Full flow dilution measurement (CVS)

The concentration signals, either by integration over the cycle or by bag sampling, of the gaseous components shall be used for the calculation of the mass emissions by multiplication with the diluted exhaust mass flow rate. The exhaust mass flow rate shall be measured with a constant volume sampling (CVS) system, which may use a positive displacement pump (PDP), a critical flow venturi (CFV) or a subsonic venturi (SSV) with or without flow compensation.

For bag sampling and particulate sampling, a proportional sample shall be taken from the diluted exhaust gas of the CVS system. For a system without flow compensation, the ratio of sample flow to CVS flow shall not vary by more than ±2,5 per cent from the set point of the test. For a system with flow compensation, each individual flow rate shall be constant within ±2,5 per cent of its respective target flow rate.

The complete test set up is schematically shown in figure 7.

8.5.1.   Determination of the diluted exhaust gas flow

8.5.1.1.   Introduction

For calculation of the emissions in the diluted exhaust gas, it is necessary to know the diluted exhaust gas mass flow rate. The total diluted exhaust gas flow over the cycle (kg/test) shall be calculated from the measurement values over the cycle and the corresponding calibration data of the flow measurement device (V 0 for PDP, K V for CFV, C d for SSV) by either of the methods described in paragraphs 8.5.1.2 to 8.5.1.4. If the total sample flow of particulates (m sep) exceeds 0,5 per cent of the total CVS flow (m ed), the CVS flow shall be corrected for m sep or the particulate sample flow shall be returned to the CVS prior to the flow measuring device.

8.5.1.2.   PDP-CVS system

The calculation of the mass flow over the cycle is as follows, if the temperature of the diluted exhaust is kept within ± 6 K over the cycle by using a heat exchanger:

(49)

where:

V 0	is the volume of gas pumped per revolution under test conditions, m3/rev
n P	is the total revolutions of pump per test
p p	is the absolute pressure at pump inlet, kPa
T	is the average temperature of the diluted exhaust gas at pump inlet, K

If a system with flow compensation is used (i.e. without heat exchanger), the instantaneous mass emissions shall be calculated and integrated over the cycle. In this case, the instantaneous mass of the diluted exhaust gas shall be calculated as follows:

(50)

where:

n P,i

is the total revolutions of pump per time interval

8.5.1.3.   CFV-CVS system

The calculation of the mass flow over the cycle is as follows, if the temperature of the diluted exhaust is kept within ± 11 K over the cycle by using a heat exchanger:

(51)

where:

t	is the cycle time, s
K V	is the calibration coefficient of the critical flow venturi for standard conditions,
p p	is the absolute pressure at venturi inlet, kPa
T	is the absolute temperature at venturi inlet, K

If a system with flow compensation is used (i.e. without heat exchanger), the instantaneous mass emissions shall be calculated and integrated over the cycle. In this case, the instantaneous mass of the diluted exhaust gas shall be calculated as follows:

(52)

where:

Δti	is the time interval, s

8.5.1.4.   SSV-CVS system

The calculation of the mass flow over the cycle shall be as follows, if the temperature of the diluted exhaust is kept within ± 11 K over the cycle by using a heat exchanger:

(53)

with

(54)

where:

A 0	is 0,006111 in SI units of
d V	is the diameter of the SSV throat, m
C d	is the discharge coefficient of the SSV
p p	is the absolute pressure at venturi inlet, kPa
T	is the temperature at the venturi inlet, K
r p	is the ratio of the SSV throat to inlet absolute static pressure,
r D	is the ratio of the SSV throat diameter, d, to the inlet pipe inner diameter D

If a system with flow compensation is used (i.e. without heat exchanger), the instantaneous mass emissions shall be calculated and integrated over the cycle. In this case, the instantaneous mass of the diluted exhaust gas shall be calculated as follows:

(55)

where:

Δt i	is the time interval, s

The real time calculation shall be initialised with either a reasonable value for C d, such as 0,98, or a reasonable value of Q ssv. If the calculation is initialised with Q ssv, the initial value of Q ssv shall be used to evaluate the Reynolds number.

During all emissions tests, the Reynolds number at the SSV throat shall be in the range of Reynolds numbers used to derive the calibration curve developed in paragraph 9.5.4.

8.5.2.   Determination of the gaseous components

8.5.2.1.   Introduction

The gaseous components in the diluted exhaust gas emitted by the engine submitted for testing shall be measured by the methods described in Appendix 3. Dilution of the exhaust shall be done with filtered ambient air, synthetic air or nitrogen. The flow capacity of the full flow system shall be large enough to completely eliminate water condensation in the dilution and sampling systems. Data evaluation and calculation procedures are described in paragraphs 8.5.2.2 and 8.5.2.3.

8.5.2.2.   Data evaluation

The emission relevant data shall be recorded and stored in accordance with paragraph 7.6.6.

8.5.2.3.   Calculation of mass emission

8.5.2.3.1.   Systems with constant mass flow

For systems with heat exchanger, the mass of the pollutants shall be determined from the following equation:

(in g/test) (56)

where:

u gas	is the respective value of the exhaust component from table 6
c gas	is the average background corrected concentration of the component, ppm
m ed	is the total diluted exhaust mass over the cycle, kg

If measured on a dry basis, the dry/wet correction according to paragraph 8.1 shall be applied.

For the calculation of NOx, the mass emission shall be multiplied, if applicable, with the humidity correction factor k h,D, or k h,G, as determined according to paragraph 8.2.

The u values are given in table 6. For calculating the u gas values, the density of the diluted exhaust gas has been assumed to be equal to air density. Therefore, the u gas values are identical for single gas components, but different for HC.

Table 6

Diluted exhaust gas u values and component densities

Fuel	ρ de	Gas
		NOx	CO	HC	CO2	O2	CH4
		ρ gas [kg/m3]
		2,053	1,250	(7)	1,9636	1,4277	0,716
		u gas  (8)
Diesel	1,293	0,001588	0,000967	0,000480	0,001519	0,001104	0,000553
Ethanol	1,293	0,001588	0,000967	0,000795	0,001519	0,001104	0,000553
CNG (9)	1,293	0,001588	0,000967	0,000517 (10)	0,001519	0,001104	0,000553
Propane	1,293	0,001588	0,000967	0,000507	0,001519	0,001104	0,000553
Butane	1,293	0,001588	0,000967	0,000501	0,001519	0,001104	0,000553
LPG (11)	1,293	0,001588	0,000967	0,000505	0,001519	0,001104	0,000553

Alternatively, the u values may be calculated using the exact calculation method generally described in paragraph 8.4.2.4, as follows:

(57)

where:

M gas	is the molar mass of the gas component, g/mol (cf. Appendix 6)
M e	is the molar mass of the exhaust gas, g/mol
M d	is the molar mass of the diluent = 28,965 g/mol
D	is the dilution factor (see paragraph 8.5.2.3.2)

8.5.2.3.2.   Determination of the background corrected concentrations

The average background concentration of the gaseous pollutants in the diluent shall be subtracted from the measured concentrations to get the net concentrations of the pollutants. The average values of the background concentrations can be determined by the sample bag method or by continuous measurement with integration. The following equation shall be used:

(58)

where:

c gas,e	is the concentration of the component measured in the diluted exhaust gas, ppm
c d	is the concentration of the component measured in the diluent, ppm
D	is the dilution factor

The dilution factor shall be calculated as follows:

(a)	for diesel and LPG fuelled gas engines (59)

(b)	for NG fuelled gas engines (60)

where:

c CO2,e	is the wet concentration of CO2 in the diluted exhaust gas, per cent vol
c HC,e	is the wet concentration of HC in the diluted exhaust gas, ppm C1
c NMHC,e	is the wet concentration of NMHC in the diluted exhaust gas, ppm C1
c CO,e	is the wet concentration of CO in the diluted exhaust gas, ppm
F S	is the stoichiometric factor

The stoichiometric factor shall be calculated as follows:

(61)

where:

α	is the molar hydrogen ratio of the fuel (H/C)

Alternatively, if the fuel composition is not known, the following stoichiometric factors may be used:

F S (diesel)	=	13,4
F S (LPG)	=	11,6
F S (NG)	=	9,5

8.5.2.3.3.   Systems with flow compensation

For systems without heat exchanger, the mass of the pollutants (g/test) shall be determined by calculating the instantaneous mass emissions and integrating the instantaneous values over the cycle. Also, the background correction shall be applied directly to the instantaneous concentration value. The following equation shall be applied:

(62)

where:

c gas,e	is the concentration of the component measured in the diluted exhaust gas, ppm
c d	is the concentration of the component measured in the diluent, ppm
m ed,i	is the instantaneous mass of the diluted exhaust gas, kg
m ed	is the total mass of diluted exhaust gas over the cycle, kg
u gas	is the tabulated value from table 6
D	is the dilution factor

8.5.3.   Particulate determination

8.5.3.1.   Calculation of mass emission

The particulate mass (g/test) shall be calculated after buoyancy correction of the particulate sample filter according to paragraph 8.3, as follows:

(63)

where:

m p	is the particulate mass sampled over the cycle, mg
m sep	is the mass of diluted exhaust gas passing the particulate collection filters, kg
m ed	is the mass of diluted exhaust gas over the cycle, kg

with

(64)

where:

m set	is the mass of double diluted exhaust gas through particulate filter, kg
m ssd	is the mass of secondary diluent, kg

If the particulate background level of the diluent is determined in accordance with paragraph 7.5.6, the particulate mass may be background corrected. In this case, the particulate mass (g/test) shall be calculated as follows:

(65)

where:

m sep	is the mass of diluted exhaust gas passing the particulate collection filters, kg
m ed	is the mass of diluted exhaust gas over the cycle, kg
m sd	is the mass of diluent sampled by background particulate sampler, kg
m b	is the mass of the collected background particulates of the diluent, mg
D	is the dilution factor as determined in paragraph 8.5.2.3.2.

8.6.   General calculations

8.6.1.   Drift correction

With respect to drift verification in paragraph 7.8.4, the corrected concentration value shall be calculated as follows:

(66)

where:

c ref,z	is the reference concentration of the zero gas (usually zero), ppm
c ref,s	is the reference concentration of the span gas, ppm
c pre,z	is the pre-test analyser concentration of the zero gas, ppm
c pre,s	is the pre-test analyser concentration of the span gas, ppm,
c post,z	is the post-test analyser concentration of the zero gas, ppm,
c post,s	is the post-test analyser concentration of the span gas, ppm
cgas	is the sample gas concentration, ppm

Two sets of specific emission results shall be calculated for each component in accordance with paragraph 8.6.3, after any other corrections have been applied. One set shall be calculated using uncorrected concentrations and another set shall be calculated using the concentrations corrected for drift according to equation 66.

Depending on the measurement system and calculation method used, the uncorrected emissions results shall be calculated with equations 36, 37 56, 57 or 62, respectively. For calculation of the corrected emissions, c gas in equations 36, 37 56, 57 or 62, respectively, shall be replaced with c cor of equation 66. If instantaneous concentration values c gas,i are used in the respective equation, the corrected value shall also be applied as instantaneous value c cor,i. In equation 57, the correction shall be applied to both the measured and the background concentration.

The comparison shall be made as a percentage of the uncorrected results. The difference between the uncorrected and the corrected brake-specific emission values shall be within ± 4 per cent of the uncorrected brake-specific emission values or within ± 4 per cent of the respective limit value, whichever is greater. If the drift is greater than 4 per cent, the test shall be voided.

If drift correction is applied, only the drift-corrected emission results shall be used when reporting emissions.

8.6.2.   Calculation of NMHC and CH4

The calculation of NMHC and CH4 depends on the calibration method used. The FID for the measurement without NMC (lower path of Appendix 3, figure 11), shall be calibrated with propane. For the calibration of the FID in series with NMC (upper path of Appendix 3, figure 11), the following methods are permitted.

(a)	calibration gas — propane; propane bypasses NMC,

(b)	calibration gas — methane; methane passes through NMC

The concentration of NMHC and CH4 shall be calculated as follows for (a):

(67)

(68)

The concentration of NMHC and CH4 shall be calculated as follows for (b):

(67a)

(68a)

where:

c HC(w/NMC)	is the HC concentration with sample gas flowing through the NMC, ppm
c HC(w/oNMC)	is the HC concentration with sample gas bypassing the NMC, ppm
r h	is the methane response factor as determined per paragraph 9.3.7.2.
E M	is the methane efficiency as determined per paragraph 9.3.8.1.
E E	is the ethane efficiency as determined per paragraph 9.3.8.2.

If r h < 1,05, it may be omitted in equations 67, 67a and 68a.

8.6.3.   Calculation of the specific emissions

The specific emissions e gas or e PM (g/kWh) shall be calculated for each individual component in the following ways depending on the type of test cycle.

For the WHSC, hot WHTC, or cold WHTC, the following equation shall be applied:

(69)

where:

m	is the mass emission of the component, g/test
W act	is the actual cycle work as determined according to paragraph 7.8.6, kWh

For the WHTC, the final test result shall be a weighted average from cold start test and hot start test according to the following equation:

(70)

where:

m cold	is the mass emission of the component on the cold start test, g/test
m hot	is the mass emission of the component on the hot start test, g/test
W act,cold	is the actual cycle work on the cold start test, kWh
W act,hot	is the actual cycle work on the hot start test, kWh

If periodic regeneration in accordance with paragraph 6.6.2 applies, the regeneration adjustment factors k r,u or k r,d shall be multiplied with or be added to, respectively, the specific emissions result e as determined in equations 69 and 70.

9.   EQUIPMENT SPECIFICATION AND VERIFICATION

This annex does not contain details of flow, pressure, and temperature measuring equipment or systems. Instead, only the linearity requirements of such equipment or systems necessary for conducting an emissions test are given in paragraph 9.2.

9.1.   Dynamometer specification

An engine dynamometer with adequate characteristics to perform the appropriate test cycle described in paragraphs 7.2.1 and 7.2.2 shall be used.

The instrumentation for torque and speed measurement shall allow the measurement accuracy of the shaft power as needed to comply with the cycle validation criteria. Additional calculations may be necessary. The accuracy of the measuring equipment shall be such that the linearity requirements given in paragraph 9.2, table 7 are not exceeded.

9.2.   Linearity requirements

The calibration of all measuring instruments and systems shall be traceable to national (international) standards. The measuring instruments and systems shall comply with the linearity requirements given in table 7. The linearity verification according to paragraph 9.2.1 shall be performed for the gas analysers at least every 3 months or whenever a system repair or change is made that could influence calibration. For the other instruments and systems, the linearity verification shall be done as required by internal audit procedures, by the instrument manufacturer or in accordance with ISO 9000 requirements.

Table 7

Linearity requirements of instruments and measurement systems

Measurement system		Slope a1	Standard error SEE	Coefficient of determination r2
Engine speed	≤ 0,05 % max	0,98 – 1,02	≤ 2 % max	≥ 0,990
Engine torque	≤ 1 % max	0,98 – 1,02	≤ 2 % max	≥ 0,990
Fuel flow	≤ 1 % max	0,98 – 1,02	≤ 2 % max	≥ 0,990
Airflow	≤ 1 % max	0,98 – 1,02	≤ 2 % max	≥ 0,990
Exhaust gas flow	≤ 1 % max	0,98 – 1,02	≤ 2 % max	≥ 0,990
Diluent flow	≤ 1 % max	0,98 – 1,02	≤ 2 % max	≥ 0,990
Diluted exhaust gas flow	≤ 1 % max	0,98 – 1,02	≤ 2 % max	≥ 0,990
Sample flow	≤ 1 % max	0,98 – 1,02	≤ 2 % max	≥ 0,990
Gas analysers	≤ 0,5 % max	0,99 – 1,01	≤ 1 % max	≥ 0,998
Gas dividers	≤ 0,5 % max	0,98 – 1,02	≤ 2 % max	≥ 0,990
Temperatures	≤ 1 % max	0,99 – 1,01	≤ 1 % max	≥ 0,998
Pressures	≤ 1 % max	0,99 – 1,01	≤ 1 % max	≥ 0,998
PM balance	≤ 1 % max	0,99 – 1,01	≤ 1 % max	≥ 0,998

9.2.1.   Linearity verification

9.2.1.1.   Introduction

A linearity verification shall be performed for each measurement system listed in table 7. At least 10 reference values, or as specified otherwise, shall be introduced to the measurement system, and the measured values shall be compared to the reference values by using a least squares linear regression in accordance with equation 11. The maximum limits in table 6 refer to the maximum values expected during testing.

9.2.1.2.   General requirements

The measurement systems shall be warmed up according to the recommendations of the instrument manufacturer. The measurement systems shall be operated at their specified temperatures, pressures and flows.

9.2.1.3.   Procedure

The linearity verification shall be run for each normally used operating range with the following steps:

(a)	The instrument shall be set at zero by introducing a zero signal. For gas analysers, purified synthetic air (or nitrogen) shall be introduced directly to the analyser port.

(b)	The instrument shall be spanned by introducing a span signal. For gas analysers, an appropriate span gas shall be introduced directly to the analyser port.

(c)	The zero procedure of (a) shall be repeated.

(d)

The verification shall be established by introducing at least 10 reference values (including zero) that are within the range from zero to the highest values expected during emission testing. For gas analysers, known gas concentrations in accordance with paragraph 9.3.3.2 shall be introduced directly to the analyser port.

(e)	At a recording frequency of at least 1 Hz, the reference values shall be measured and the measured values recorded for 30 s.

(f)	The arithmetic mean values over the 30 s period shall be used to calculate the least squares linear regression parameters according to equation 11 in paragraph 7.8.7.

(g)	The linear regression parameters shall meet the requirements of paragraph 9.2, table 7.

(h)	The zero setting shall be rechecked and the verification procedure repeated, if necessary.

9.3.   Gaseous emissions measurement and sampling system

9.3.1.   Analyser specifications

9.3.1.1.   General

The analysers shall have a measuring range and response time appropriate for the accuracy required to measure the concentrations of the exhaust gas components under transient and steady state conditions.

The electromagnetic compatibility (EMC) of the equipment shall be on a level as to minimise additional errors.

9.3.1.2.   Accuracy

The accuracy, defined as the deviation of the analyser reading from the reference value, shall not exceed ±2 per cent of the reading or ±0,3 per cent of full scale whichever is larger.

9.3.1.3.   Precision

The precision, defined as 2,5 times the standard deviation of 10 repetitive responses to a given calibration or span gas, shall be no greater than 1 per cent of full scale concentration for each range used above 155 ppm (or ppm C) or 2 per cent of each range used below 155 ppm (or ppm C).

9.3.1.4.   Noise

The analyser peak-to-peak response to zero and calibration or span gases over any 10 seconds period shall not exceed 2 per cent of full scale on all ranges used.

9.3.1.5.   Zero drift

The drift of the zero response shall be specified by the instrument manufacturer.

9.3.1.6.   Span drift

The drift of the span response shall be specified by the instrument manufacturer.

9.3.1.7.   Rise time

The rise time of the analyser installed in the measurement system shall not exceed 2,5 s.

9.3.1.8.   Gas drying

Exhaust gases may be measured wet or dry. A gas-drying device, if used, shall have a minimal effect on the composition of the measured gases. Chemical dryers are not an acceptable method of removing water from the sample.

9.3.2.   Gas analysers

9.3.2.1.   Introduction

Paragraphs 9.3.2.2 to 9.3.2.7 describe the measurement principles to be used. A detailed description of the measurement systems is given in Appendix 3. The gases to be measured shall be analysed with the following instruments. For non-linear analysers, the use of linearising circuits is permitted.

9.3.2.2.   Carbon monoxide (CO) analysis

The carbon monoxide analyser shall be of the non-dispersive infrared (NDIR) absorption type.

9.3.2.3.   Carbon dioxide (CO2) analysis

The carbon dioxide analyser shall be of the non-dispersive infrared (NDIR) absorption type.

9.3.2.4.   Hydrocarbon (HC) analysis

The hydrocarbon analyser shall be of the heated flame ionisation detector (HFID) type with detector, valves, pipework, etc. heated so as to maintain a gas temperature of 463 K ± 10 K (190 ± 10 °C). Optionally, for NG fuelled and PI engines, the hydrocarbon analyser may be of the non-heated flame ionisation detector (FID) type depending upon the method used (see Appendix 3, paragraph A.3.1.3).

9.3.2.5.   Methane (CH4) and non-methane hydrocarbon (NMHC) analysis

The determination of the methane and non-methane hydrocarbon fraction shall be performed with a heated non-methane cutter (NMC) and two FID’s as per Appendix 3, paragraph A.3.1.4 and paragraph A.3.1.5. The concentration of the components shall be determined as per paragraph 8.6.2.

9.3.2.6.   Oxides of nitrogen (NOx) analysis

Two measurement instruments are specified for NOx measurement and either instrument may be used provided it meets the criteria specified in paragraphs 9.3.2.6.1 or 9.3.2.6.2, respectively. For the determination of system equivalency of an alternate measurement procedure in accordance with paragraph 5.1.1, only the CLD is permitted.

9.3.2.6.1.   Chemiluminescent detector (CLD)

If measured on a dry basis, the oxides of nitrogen analyser shall be of the chemiluminescent detector (CLD) or heated chemiluminescent detector (HCLD) type with a NO2/NO converter. If measured on a wet basis, a HCLD with converter maintained above 328 K (55 °C) shall be used, provided the water quench check (see paragraph 9.3.9.2.2) is satisfied. For both CLD and HCLD, the sampling path shall be maintained at a wall temperature of 328 K to 473 K (55 °C to 200 °C) up to the converter for dry measurement and up to the analyser for wet measurement.

9.3.2.6.2.   Non-dispersive ultraviolet detector (NDUV)

A non-dispersive ultraviolet (NDUV) analyser shall be used to measure NOx concentration. If the NDUV analyser measures only NO, a NO2/NO converter shall be placed upstream of the NDUV analyser. The NDUV temperature shall be maintained to prevent aqueous condensation, unless a sample dryer is installed upstream of the NO2/NO converter, if used, or upstream of the analyser.

9.3.2.7.   Air to fuel measurement

The air to fuel measurement equipment used to determine the exhaust gas flow as specified in paragraph 8.4.1.6 shall be a wide range air to fuel ratio sensor or lambda sensor of Zirconia type. The sensor shall be mounted directly on the exhaust pipe where the exhaust gas temperature is high enough to eliminate water condensation.

The accuracy of the sensor with incorporated electronics shall be within:

± 3 per cent of reading	for	λ < 2
± 5 per cent of reading	for	2 ≤ λ < 5
± 10 per cent of reading	for	5 ≤ λ

To fulfill the accuracy specified above, the sensor shall be calibrated as specified by the instrument manufacturer.

9.3.3.   Gases

The shelf life of all gases shall be respected.

9.3.3.1.   Pure gases

The required purity of the gases is defined by the contamination limits given below. The following gases shall be available for operation:

(a)

For raw exhaust gas Purified nitrogen (Contamination ≤ 1 ppm C1, ≤ 1 ppm CO, ≤ 400 ppm CO2, ≤ 0,1 ppm NO) Purified oxygen (Purity > 99,5 per cent vol O2) Hydrogen-helium mixture (FID burner fuel) (40 ± 1 per cent hydrogen, balance helium) (Contamination ≤ 1 ppm C1, ≤ 400 ppm CO2) Purified synthetic air (Contamination ≤ 1 ppm C1, ≤ 1 ppm CO, ≤ 400 ppm CO2, ≤ 0,1 ppm NO) (Oxygen content between 18-21 per cent vol.)

(b)

For dilute exhaust gas (optionally for raw exhaust gas) Purified nitrogen (Contamination ≤ 0,05 ppm C1, ≤ 1 ppm CO, ≤ 10 ppm CO2, ≤ 0,02 ppm NO) Purified oxygen (Purity > 99,5 per cent vol O2) Hydrogen-helium mixture (FID burner fuel) (40 ± 1 per cent hydrogen, balance helium) (Contamination ≤ 0,05 ppm C1, ≤ 10 ppm CO2) Purified synthetic air (Contamination ≤ 0,05 ppm C1, ≤ 1 ppm CO, ≤ 10 ppm CO2, ≤ 0,02 ppm NO) (Oxygen content between 20,5 - 21,5 per cent vol.) If gas bottles are not available, a gas purifier may be used, if contamination levels can be demonstrated.

9.3.3.2.   Calibration and span gases

Mixtures of gases having the following chemical compositions shall be available, if applicable. Other gas combinations are allowed provided the gases do not react with one another. The expiration date of the calibration gases stated by the manufacturer shall be recorded.

C3H8 and purified synthetic air (see paragraph 9.3.3.1);

CO and purified nitrogen;

NO and purified nitrogen;

NO2 and purified synthetic air;

CO2 and purified nitrogen;

CH4 and purified synthetic air;

C2H6 and purified synthetic air

The true concentration of a calibration and span gas shall be within ± 1 per cent of the nominal value, and shall be traceable to national or international standards. All concentrations of calibration gas shall be given on a volume basis (volume percent or volume ppm).

9.3.3.3.   Gas dividers

The gases used for calibration and span may also be obtained by means of gas dividers (precision blending devices), diluting with purified N2 or with purified synthetic air. The accuracy of the gas divider shall be such that the concentration of the blended calibration gases is accurate to within ± 2 per cent. This accuracy implies that primary gases used for blending shall be known to an accuracy of at least ± 1 per cent, traceable to national or international gas standards. The verification shall be performed at between 15 and 50 per cent of full scale for each calibration incorporating a gas divider. An additional verification may be performed using another calibration gas, if the first verification has failed.

Optionally, the blending device may be checked with an instrument which by nature is linear, e.g. using NO gas with a CLD. The span value of the instrument shall be adjusted with the span gas directly connected to the instrument. The gas divider shall be checked at the settings used and the nominal value shall be compared to the measured concentration of the instrument. This difference shall in each point be within ± 1 per cent of the nominal value.

For conducting the linearity verification according to paragraph 9.2.1, the gas divider shall be accurate to within ± 1 per cent.

9.3.3.4.   Oxygen interference check gases

Oxygen interference check gases are a blend of propane, oxygen and nitrogen. They shall contain propane with 350 ppm C ± 75 ppm C hydrocarbon. The concentration value shall be determined to calibration gas tolerances by chromatographic analysis of total hydrocarbons plus impurities or by dynamic blending. The oxygen concentrations required for positive ignition and compression ignition engine testing are listed in table 8 with the remainder being purified nitrogen.

Table 8

Oxygen interference check gases

Type of engine	O2 concentration (per cent)
Compression ignition	21 (20 to 22)
Compression and positive ignition	10 (9 to 11)
Compression and positive ignition	5 (4 to 6)
Positive ignition	0 (0 to 1)

9.3.4.   Leak check

A system leak check shall be performed. The probe shall be disconnected from the exhaust system and the end plugged. The analyser pump shall be switched on. After an initial stabilisation period all flowmeters will read approximately zero in the absence of a leak. If not, the sampling lines shall be checked and the fault corrected.

The maximum allowable leakage rate on the vacuum side shall be 0,5 per cent of the in-use flow rate for the portion of the system being checked. The analyser flows and bypass flows may be used to estimate the in-use flow rates.

Alternatively, the system may be evacuated to a pressure of at least 20 kPa vacuum (80 kPa absolute). After an initial stabilisation period the pressure increase Δp (kPa/min) in the system shall not exceed:

(71)

where:

V s	is the system volume, l
q vs	is the system flow rate, l/min

Another method is the introduction of a concentration step change at the beginning of the sampling line by switching from zero to span gas. If for a correctly calibrated analyser after an adequate period of time the reading is ≤ 99 per cent compared to the introduced concentration, this points to a leakage problem that shall be corrected.

9.3.5.   Response time check of the analytical system

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 analysers 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 gas switching shall be done in less than 0,1 s. The gases used for the test shall cause a concentration change of at least 60 per cent full scale (FS).

The concentration trace of each single gas component shall be recorded. The response time is defined to be the difference in time between the gas switching and the appropriate change of the recorded concentration. The system response time (t 90) consists of the delay time to the measuring detector and the rise time of the detector. The delay time is defined as the time from the change (t 0) until the response is 10 per cent of the final reading (t 10). The rise time is defined as the time between 10 per cent and 90 per cent response of the final reading (t 90 – t 10).

For time alignment of the analyser and exhaust flow signals, the transformation time is defined as the time from the change (t 0) until the response is 50 per cent of the final reading (t 50).

The system response time shall be ≤ 10 s with a rise time of ≤ 2,5 s in accordance with paragraph 9.3.1.7 for all limited components (CO, NOx, HC or NMHC) and all ranges used. When using a NMC for the measurement of NMHC, the system response time may exceed 10 s.

9.3.6.   Efficiency test of NOx converter

The efficiency of the converter used for the conversion of NO2 into NO is tested as given in paragraphs 9.3.6.1 to 9.3.6.8 (see figure 8).

9.3.6.1.   Test setup

Using the test setup as schematically shown in figure 8 and the procedure below, the efficiency of the converter shall be tested by means of an ozonator.

9.3.6.2.   Calibration

The CLD and the HCLD shall be calibrated in the most common operating range following the manufacturer’s specifications using zero and span gas (the NO content of which shall amount to about 80 per cent of the operating range and the NO2 concentration of the gas mixture to less than 5 per cent of the NO concentration). The NOx analyser shall be in the NO mode so that the span gas does not pass through the converter. The indicated concentration has to be recorded.

9.3.6.3.   Calculation

The per cent efficiency of the converter shall be calculated as follows:

(72)

where:

a	is the NOx concentration according to paragraph 9.3.6.6.
b	is the NOx concentration according to paragraph 9.3.6.7.
c	is the NO concentration according to paragraph 9.3.6.4.
d	is the NO concentration according to paragraph 9.3.6.5.

9.3.6.4.   Adding of oxygen

Via a T-fitting, oxygen or zero air shall be added continuously to the gas flow until the concentration indicated is about 20 per cent less than the indicated calibration concentration given in paragraph 9.3.6.2 (the analyser is in the NO mode).

The indicated concentration (c) shall be recorded. The ozonator is kept deactivated throughout the process.

9.3.6.5.   Activation of the ozonator

The ozonator shall be activated to generate enough ozone to bring the NO concentration down to about 20 per cent (minimum 10 per cent) of the calibration concentration given in paragraph 9.3.6.2. The indicated concentration (d) shall be recorded (the analyser is in the NO mode).

9.3.6.6.   NOx mode

The NO analyser shall be switched to the NOx mode so that the gas mixture (consisting of NO, NO2, O2 and N2) now passes through the converter. The indicated concentration (a) shall be recorded (the analyser is in the NOx mode).

9.3.6.7.   Deactivation of the ozonator

The ozonator is now deactivated. The mixture of gases described in paragraph 9.3.6.6 passes through the converter into the detector. The indicated concentration (b) shall be recorded (the analyser is in the NOx mode).

9.3.6.8.   NO mode

Switched to NO mode with the ozonator deactivated, the flow of oxygen or synthetic air shall be shut off. The NOx reading of the analyser shall not deviate by more than ± 5 per cent from the value measured according to paragraph 9.3.6.2 (the analyser is in the NO mode).

9.3.6.9.   Test interval

The efficiency of the converter shall be tested at least once per month.

9.3.6.10.   Efficiency requirement

The efficiency of the converter E NOx shall not be less than 95 per cent.

If, with the analyser in the most common range, the ozonator cannot give a reduction from 80 per cent to 20 per cent according to paragraph 9.3.6.5, the highest range which will give the reduction shall be used.

9.3.7.   Adjustment of the FID

9.3.7.1.   Optimisation of the detector response

The FID shall be adjusted as specified by the instrument manufacturer. A propane in air span gas shall be used to optimise the response on the most common operating range.

With the fuel and airflow rates set at the manufacturer’s recommendations, a 350 ± 75 ppm C span gas shall be introduced to the analyser. The response at a given fuel flow shall be determined from the difference between the span gas response and the zero gas response. The fuel flow shall be incrementally adjusted above and below the manufacturer’s specification. The span and zero response at these fuel flows shall be recorded. The difference between the span and zero response shall be plotted and the fuel flow adjusted to the rich side of the curve. This is the initial flow rate setting which may need further optimisation depending on the results of the hydrocarbon response factors and the oxygen interference check according to paragraphs 9.3.7.2 and 9.3.7.3. If the oxygen interference or the hydrocarbon response factors do not meet the following specifications, the airflow shall be incrementally adjusted above and below the manufacturer’s specifications, repeating paragraphs 9.3.7.2 and 9.3.7.3 for each flow.

The optimisation may optionally be conducted using the procedures outlined in SAE paper No 770141.

9.3.7.2.   Hydrocarbon response factors

A linearity verification of the analyser shall be performed using propane in air and purified synthetic air according to paragraph 9.2.1.3.

Response factors shall be determined when introducing an analyser into service and after major service intervals. The response factor (r h) for a particular hydrocarbon species is the ratio of the FID C1 reading to the gas concentration in the cylinder expressed by ppm C1.

The concentration of the test gas shall be at a level to give a response of approximately 80 per cent of full scale. The concentration shall be known to an accuracy of ± 2 per cent in reference to a gravimetric standard expressed in volume. In addition, the gas cylinder shall be preconditioned for 24 hours at a temperature of 298 K ± 5 K (25 °C ± 5 °C).

The test gases to be used and the relative response factor ranges are as follows:

(a)	methane and purified synthetic air	1,00 ≤ r h ≤ 1,15
(b)	propylene and purified synthetic air	0,90 ≤ r h ≤ 1,1
(c)	toluene and purified synthetic air	0,90 ≤ r h ≤ 1,1

These values are relative to a r h of 1 for propane and purified synthetic air.

9.3.7.3.   Oxygen interference check

For raw exhaust gas analysers only, the oxygen interference check shall be performed when introducing an analyser into service and after major service intervals.

A measuring range shall be chosen where the oxygen interference check gases will fall in the upper 50 per cent. The test shall be conducted with the oven temperature set as required. Oxygen interference check gas specifications are found in paragraph 9.3.3.4.

(a)	the analyser shall be set at zero;

(b)	the analyser shall be spanned with the 0 per cent oxygen blend for positive ignition engines. Compression ignition engine instruments shall be spanned with the 21 per cent oxygen blend;

(c)	the zero response shall be rechecked. If it has changed by more than 0,5 per cent of full scale, steps (a) and (b) of this paragraph shall be repeated;

(d)	the 5 per cent and 10 per cent oxygen interference check gases shall be introduced;

(e)	the zero response shall be rechecked. If it has changed by more than ± 1 per cent of full scale, the test shall be repeated;

(f)

the oxygen interference E O2 shall be calculated for each mixture in step (d) as follows: (73) with the analyser response being (74) where: c ref,b is the reference HC concentration in step (b), ppm C c ref,d is the reference HC concentration in step (d), ppm C c FS,b is the full scale HC concentration in step (b), ppm C c FS,d is the full scale HC concentration in step (d), ppm C c m,b is the measured HC concentration in step (b), ppm C c m,d is the measured HC concentration in step (d), ppm C

(g)	The oxygen interference E O2 shall be less than ±1,5 per cent for all required oxygen interference check gases prior to testing.

(h)	If the oxygen interference E O2 is greater than ±1,5 per cent, corrective action may be taken by incrementally adjusting the airflow above and below the manufacturer’s specifications, the fuel flow and the sample flow.

(i)	The oxygen interference shall be repeated for each new setting.

9.3.8.   Efficiency of the non-methane cutter (NMC)

The NMC is used for the removal of the non-methane hydrocarbons from the sample gas by oxidising all hydrocarbons except methane. Ideally, the conversion for methane is 0 per cent, and for the other hydrocarbons represented by ethane is 100 per cent. For the accurate measurement of NMHC, the two efficiencies shall be determined and used for the calculation of the NMHC emission mass flow rate (see paragraph 8.5.2).

9.3.8.1.   Methane efficiency

Methane calibration gas shall be flown through the FID with and without bypassing the NMC and the two concentrations recorded. The efficiency shall be determined as follows:

(75)

where:

c HC(w/NMC)	is the HC concentration with CH4 flowing through the NMC, ppm C
c HC(w/oNMC)	is the HC concentration with CH4 bypassing the NMC, ppm C

9.3.8.2.   Ethane efficiency

Ethane calibration gas shall be flown through the FID with and without bypassing the NMC and the two concentrations recorded. The efficiency shall be determined as follows:

(76)

where:

c HC(w/NMC)	is the HC concentration with C2H6 flowing through the NMC, ppm C
c HC(w/oNMC)	is the HC concentration with C2H6 bypassing the NMC, ppm C

9.3.9.   Interference effects

Other gases than the one being analysed can interfere with the reading in several ways. Positive interference occurs in NDIR instruments where the interfering gas gives the same effect as the gas being measured, but to a lesser degree. Negative interference occurs in NDIR instruments by the interfering gas broadening the absorption band of the measured gas, and in CLD instruments by the interfering gas quenching the reaction. The interference checks in paragraphs 9.3.9.1 and 9.3.9.3 shall be performed prior to an analyser’s initial use and after major service intervals.

9.3.9.1.   CO analyser interference check

Water and CO2 can interfere with the CO analyser performance. Therefore, a CO2 span gas having a concentration of 80 to 100 per cent of full scale of the maximum operating range used during testing shall be bubbled through water at room temperature and the analyser response recorded. The analyser response shall not be more than 2 per cent of the mean CO concentration expected during testing.

Interference procedures for CO2 and H2O may also be run separately. If the CO2 and H2O levels used are higher than the maximum levels expected during testing, each observed interference value shall be scaled down by multiplying the observed interference by the ratio of the maximum expected concentration value to the actual value used during this procedure. Separate interference procedures concentrations of H2O that are lower than the maximum levels expected during testing may be run, but the observed H2O interference shall be scaled up by multiplying the observed interference by the ratio of the maximum expected H2O concentration value to the actual value used during this procedure. The sum of the two scaled interference values shall meet the tolerance specified in this paragraph.

9.3.9.2.   NOx analyser quench checks for CLD analyser

The two gases of concern for CLD (and HCLD) analysers are CO2 and water vapour. Quench responses to these gases are proportional to their concentrations, and therefore require test techniques to determine the quench at the highest expected concentrations experienced during testing. If the CLD analyser uses quench compensation algorithms that utilise H2O and/or CO2 measurement instruments, quench shall be evaluated with these instruments active and with the compensation algorithms applied.

9.3.9.2.1.   CO2 quench check

A CO2 span gas having a concentration of 80 to 100 per cent of full scale of the maximum operating range shall be passed through the NDIR analyser and the CO2 value recorded as A. It shall then be diluted approximately 50 per cent with NO span gas and passed through the NDIR and CLD, with the CO2 and NO values recorded as B and C, respectively. The CO2 shall then be shut off and only the NO span gas be passed through the (H)CLD and the NO value recorded as D.

The per cent quench shall be calculated as follows:

(77)

where:

A	is the undiluted CO2 concentration measured with NDIR, per cent
B	is the diluted CO2 concentration measured with NDIR, per cent
C	is the diluted NO concentration measured with (H)CLD, ppm
D	is the undiluted NO concentration measured with (H)CLD, ppm

Alternative methods of diluting and quantifying of CO2 and NO span gas values such as dynamic mixing/blending are permitted with the approval of the type approval authority.

9.3.9.2.2.   Water quench check

This check applies to wet gas concentration measurements only. Calculation of water quench shall consider dilution of the NO span gas with water vapour and scaling of water vapour concentration of the mixture to that expected during testing.

A NO span gas having a concentration of 80 per cent to 100 per cent of full scale of the normal operating range shall be passed through the (H) CLD and the NO value recorded as D. The NO span gas shall then be bubbled through water at room temperature and passed through the (H) CLD and the NO value recorded as C. The water temperature shall be determined and recorded as F. The mixture’s saturation vapour pressure that corresponds to the bubbler water temperature (F) shall be determined and recorded as G.

The water vapour concentration (in per cent) of the mixture shall be calculated as follows:

(78)

and recorded as H. The expected diluted NO span gas (in water vapour) concentration shall be calculated as follows:

(79)

and recorded as D e. For diesel exhaust, the maximum exhaust water vapour concentration (in per cent) expected during testing shall be estimated, under the assumption of a fuel H/C ratio of 1,8/1, from the maximum CO2 concentration in the exhaust gas A as follows:

(80)

and recorded as H m

The per cent water quench shall be calculated as follows:

(81)

where:

D e	De is the expected diluted NO concentration, ppm
C	is the measured diluted NO concentration, ppm
H m	is the maximum water vapour concentration, per cent
H	is the actual water vapour concentration, per cent

9.3.9.2.3.   Maximum allowable quench

The combined CO2 and water quench shall not exceed 2 per cent of full scale.

9.3.9.3.   NOx analyser quench check for NDUV analyser

Hydrocarbons and H2O can positively interfere with a NDUV analyser by causing a response similar to NOx. If the NDUV analyser uses compensation algorithms that utilise measurements of other gases to meet this interference verification, simultaneously such measurements shall be conducted to test the algorithms during the analyser interference verification.

9.3.9.3.1.   Procedure

The NDUV analyser shall be started, operated, zeroed, and spanned according to the instrument manufacturer’s instructions. It is recommended to extract engine exhaust to perform this verification. A CLD shall be used to quantify NOx in the exhaust. The CLD response shall be used as the reference value. Also HC shall be measured in the exhaust with a FID analyser. The FID response shall be used as the reference hydrocarbon value.

Upstream of any sample dryer, if used during testing, the engine exhaust shall be introduced into the NDUV analyser. Time shall be allowed for the analyser response to stabilise. Stabilisation time may include time to purge the transfer line and to account for analyser response. While all analysers measure the sample’s concentration, 30 s of sampled data shall be recorded, and the arithmetic means for the three analysers calculated.

The CLD mean value shall be subtracted from the NDUV mean value. This difference shall be multiplied by the ratio of the expected mean HC concentration to the HC concentration measured during the verification, as follows:

(82)

where

c NOx,CLD	is the measured NOx concentration with CLD, ppm
c NOx,NDUV	is the measured NOx concentration with NDUV, ppm
c HC,e	is the expected max. HC concentration, ppm
c HC,e	is the measured HC concentration, ppm

9.3.9.3.2.   Maximum allowable quench

The combined HC and water quench shall not exceed 2 per cent of the NOx concentration expected during testing.

9.3.9.4.   Sample dryer

A sample dryer removes water, which can otherwise interfere with a NOx measurement.

9.3.9.4.1.   Sample dryer efficiency

For dry CLD analysers, it shall be demonstrated that for the highest expected water vapour concentration H m (see paragraph 9.3.9.2.2), the sample dryer maintains CLD humidity at ≤ 5 g water/kg dry air (or about 0,008 per cent H2O), which is 100 per cent relative humidity at 3,9 °C and 101,3 kPa. This humidity specification is also equivalent to about 25 per cent relative humidity at 25 °C and 101,3 kPa. This may be demonstrated by measuring the temperature at the outlet of a thermal dehumidifier, or by measuring humidity at a point just upstream of the CLD. Humidity of the CLD exhaust might also be measured as long as the only flow into the CLD is the flow from the dehumidifier.

9.3.9.4.2.   Sample dryer NO2 penetration

Liquid water remaining in an improperly designed sample dryer can remove NO2 from the sample. If a sample dryer is used in combination with an NDUV analyser without an NO2/NO converter upstream, it could therefore remove NO2 from the sample prior NOx measurement.

The sample dryer shall allow for measuring at least 95 per cent of the total NO2 at the maximum expected concentration of NO2.

9.3.10.   Sampling for raw gaseous emissions, if applicable

The gaseous emissions sampling probes shall be fitted at least 0,5 m or 3 times the diameter of the exhaust pipe — whichever is the larger — upstream of the exit of the exhaust gas system but sufficiently close to the engine as to ensure an exhaust gas temperature of at least 343 K (70 °C) at the probe.

In the case of a multi-cylinder engine with a branched exhaust manifold, the inlet of the probe shall be located sufficiently far downstream so as to ensure that the sample is representative of the average exhaust emissions from all cylinders. In multi-cylinder engines having distinct groups of manifolds, such as in a “Vee” engine configuration, it is recommended to combine the manifolds upstream of the sampling probe. If this is not practical, it is permissible to acquire a sample from the group with the highest CO2 emission. For exhaust emission calculation the total exhaust mass flow shall be used.

If the engine is equipped with an exhaust after-treatment system, the exhaust sample shall be taken downstream of the exhaust after-treatment system.

9.3.11.   Sampling for dilute gaseous emissions, if applicable

The exhaust pipe between the engine and the full flow dilution system shall conform to the requirements laid down in Appendix 3. The gaseous emissions sample probe(s) shall be installed in the dilution tunnel at a point where the diluent and exhaust gas are well mixed, and in close proximity to the particulates sampling probe.

Sampling can generally be done in two ways:

(a)	the emissions are sampled into a sampling bag over the cycle and measured after completion of the test; for HC, the sample bag shall be heated to 464 ± 11 K (191 ± 11 °C), for NOx, the sample bag temperature shall be above the dew point temperature;

(b)	the emissions are sampled continuously and integrated over the cycle.

The background concentration shall be determined upstream of the dilution tunnel according to (a) or (b), and shall be subtracted from the emissions concentration according to paragraph 8.5.2.3.2.

9.4.   Particulate measurement and sampling system

9.4.1.   General specifications

To determine the mass of the particulates, a particulate dilution and sampling system, a particulate sampling filter, a microgram balance, and a temperature and humidity controlled weighing chamber, are required. The particulate sampling system shall be designed to ensure a representative sample of the particulates proportional to the exhaust flow.

9.4.2.   General requirements of the dilution system

The determination of the particulates requires dilution of the sample with filtered ambient air, synthetic air or nitrogen (the diluent). The dilution system shall be set as follows:

(a)	completely eliminate water condensation in the dilution and sampling systems;

(b)	maintain the temperature of the diluted exhaust gas between 315 K (42 °C) and 325 K (52 °C) within 20 cm upstream or downstream of the filter holder(s);

(c)	the diluent temperature shall be between 293 K and 325 K (20 °C to 52 °C) in close proximity to the entrance into the dilution tunnel;

(d)	the minimum dilution ratio shall be within the range of 5:1 to 7:1 and at least 2:1 for the primary dilution stage based on the maximum engine exhaust flow rate;

(e)	for a partial flow dilution system, the residence time in the system from the point of diluent introduction to the filter holder(s) shall be between 0,5 and 5 seconds;

(f)

for a full flow dilution system, the overall residence time in the system from the point of diluent introduction to the filter holder(s) shall be between 1 and 5 seconds, and the residence time in the secondary dilution system, if used, from the point of secondary diluent introduction to the filter holder(s) shall be at least 0,5 seconds.

Dehumidifying the diluent before entering the dilution system is permitted, and especially useful if diluent humidity is high.

9.4.3.   Particulate sampling

9.4.3.1.   Partial flow dilution system

The particulate sampling probe shall be installed in close proximity to the gaseous emissions sampling probe, but sufficiently distant as to not cause interference. Therefore, the installation provisions of paragraph 9.3.10 also apply to particulate sampling. The sampling line shall conform to the requirements laid down in Appendix 3.

In the case of a multi-cylinder engine with a branched exhaust manifold, the inlet of the probe shall be located sufficiently far downstream so as to ensure that the sample is representative of the average exhaust emissions from all cylinders. In multi-cylinder engines having distinct groups of manifolds, such as in a “Vee” engine configuration, it is recommended to combine the manifolds upstream of the sampling probe. If this is not practical, it is permissible to acquire a sample from the group with the highest particulate emission. For exhaust emission calculation the total exhaust mass flow of the manifold shall be used.

9.4.3.2.   Full flow dilution system

The particulate sampling probe shall be installed in close proximity to the gaseous emissions sampling probe, but sufficiently distant as to not cause interference, in the dilution tunnel. Therefore, the installation provisions of paragraph 9.3.11 also apply to particulate sampling. The sampling line shall conform to the requirements laid down in Appendix 3.

9.4.4.   Particulate sampling filters

The diluted exhaust shall be sampled by a filter that meets the requirements of paragraphs 9.4.4.1 to 9.4.4.3 during the test sequence.

9.4.4.1.   Filter specification

All filter types shall have a 0,3 µm DOP (di-octylphthalate) collection efficiency of at least 99 per cent. The filter material shall be either:

(a)	fluorocarbon (PTFE) coated glass fibre, or

(b)	fluorocarbon (PTFE) membrane.

9.4.4.2.   Filter size

The filter shall be circular with a nominal diameter of 47 mm (tolerance of 46,50 ± 0,6 mm) and an exposed diameter (filter stain diameter) of at least 38 mm.

9.4.4.3.   Filter face velocity

The face velocity through the filter shall be between 0,90 and 1,00 m/s with less than 5 per cent of the recorded flow values exceeding this range. If the total PM mass on the filter exceeds 400 µg, the filter face velocity may be reduced to 0,50 m/s. The face velocity shall be calculated as the volumetric flow rate of the sample at the pressure upstream of the filter and temperature of the filter face, divided by the filter’s exposed area.

9.4.5.   Weighing chamber and analytical balance specifications

The chamber (or room) environment shall be free of any ambient contaminants (such as dust, aerosol, or semi-volatile material) that could contaminate the particulate filters. The weighing room shall meet the required specifications for at least 60 min before weighing filters.

9.4.5.1.   Weighing chamber conditions

The temperature of the chamber (or room) in which the particulate filters are conditioned and weighed shall be maintained to within 295 K ± 1 K (22 °C ± 1 °C) during all filter conditioning and weighing. The humidity shall be maintained to a dew point of 282,5 K ± 1 K (9,5 °C ± 1 °C).

If the stabilisation and weighing environments are separate, the temperature of the stabilisation environment shall be maintained at a tolerance of 295 K ± 3 K (22 °C ± 3 °C), but the dew point requirement remains at 282,5 K ± 1 K (9,5 °C ± 1 °C).

Humidity and ambient temperature shall be recorded.

9.4.5.2.   Reference filter weighing

At least two unused reference filters shall be weighed within 12 hours of, but preferably at the same time as the sample filter weighing. They shall be the same material as the sample filters. Buoyancy correction shall be applied to the weighings.

If the weight of any of the reference filters changes between sample filter weighings by more than 10 µg, all sample filters shall be discarded and the emissions test repeated.

The reference filters shall be periodically replaced based on good engineering judgement, but at least once per year.

9.4.5.3.   Analytical balance

The analytical balance used to determine the filter weight shall meet the linearity verification criterion of paragraph 9.2, table 7. This implies a precision (standard deviation) of at least 2 µg and a resolution of at least 1 µg (1 digit = 1 µg).

In order to ensure accurate filter weighing, it is recommended that the balance be installed as follows:

(a)	installed on a vibration-isolation platform to isolate it from external noise and vibration;

(b)	shielded from convective airflow with a static-dissipating draft shield that is electrically grounded.

9.4.5.4.   Elimination of static electricity effects

The filter shall be neutralised prior to weighing, e.g. by a Polonium neutraliser or a device of similar effect. If a PTFE membrane filter is used, the static electricity shall be measured and is recommended to be within ±2,0 V of neutral.

Static electric charge shall be minimised in the balance environment. Possible methods are as follows:

(a)	the balance shall be electrically grounded;

(b)	stainless steel tweezers shall be used if PM samples are handled manually;

(c)	tweezers shall be grounded with a grounding strap, or a grounding strap shall be provided for the operator such that the grounding strap shares a common ground with the balance. Grounding straps shall have an appropriate resistor to protect operators from accidental shock.

9.4.5.5.   Additional specifications

All parts of the dilution system and the sampling system from the exhaust pipe up to the filter holder, which are in contact with raw and diluted exhaust gas, shall be designed to minimise deposition or alteration of the particulates. All parts shall be made of electrically conductive materials that do not react with exhaust gas components, and shall be electrically grounded to prevent electrostatic effects.

9.4.5.6.   Calibration of the flow measurement instrumentation

Each flowmeter used in a particulate sampling and partial flow dilution system shall be subjected to the linearity verification, as described in paragraph 9.2.1, as often as necessary to fulfil the accuracy requirements of this gtr. For the flow reference values, an accurate flowmeter traceable to international and/or national standards shall be used. For differential flow measurement calibration see paragraph 9.4.6.2.

9.4.6.   Special requirements for the partial flow dilution system

The partial flow dilution system has to be designed to extract a proportional raw exhaust sample from the engine exhaust stream, thus responding to excursions in the exhaust stream flow rate. For this it is essential that the dilution ratio or the sampling ratio r d or r s be determined such that the accuracy requirements of paragraph 9.4.6.2 are fulfilled.

9.4.6.1.   System response time

For the control of a partial flow dilution system, a fast system response is required. The transformation time for the system shall be determined by the procedure in paragraph 9.4.6.6. If the combined transformation time of the exhaust flow measurement (see paragraph 8.3.1.2) and the partial flow system is ≤ 0,3 s, online control shall be used. If the transformation time exceeds 0,3 s, look ahead control based on a pre-recorded test run shall be used. In this case, the combined rise time shall be ≤ 1 s and the combined delay time ≤ 10 s.

The total system response shall be designed as to ensure a representative sample of the particulates, q mp,i, proportional to the exhaust mass flow. To determine the proportionality, a regression analysis of q mp,i versus q mew,i shall be conducted on a minimum 5 Hz data acquisition rate, and the following criteria shall be met:

(a)	the coefficient of determination r 2 of the linear regression between q mp,i and q mew,i shall not be less than 0,95;

(b)	the standard error of estimate of q mp,i on q mew,i shall not exceed 5 per cent of q mp maximum;

(c)	q mp intercept of the regression line shall not exceed ± 2 per cent of q mp maximum.

Look-ahead control is required if the combined transformation times of the particulate system, t 50,P and of the exhaust mass flow signal, t 50,F are > 0,3 s. In this case, a pre-test shall be run, and the exhaust mass flow signal of the pre-test be used for controlling the sample flow into the particulate system. A correct control of the partial dilution system is obtained, if the time trace of q mew,pre of the pre-test, which controls q mp, is shifted by a “look-ahead” time of t 50,P + t 50,F.

For establishing the correlation between q mp,i and q mew,i the data taken during the actual test shall be used, with q mew,i time aligned by t50,F relative to q mp,i (no contribution from t 50,P to the time alignment). That is, the time shift between q mew and q mp is the difference in their transformation times that were determined in paragraph 9.4.6.6.

9.4.6.2.   Specifications for differential flow measurement

For partial flow dilution systems, the accuracy of the sample flow q mp is of special concern, if not measured directly, but determined by differential flow measurement:

(83)

In this case, the maximum error of the difference shall be such that the accuracy of q mp is within ± 5 per cent when the dilution ratio is less than 15. It can be calculated by taking root-mean-square of the errors of each instrument.

Acceptable accuracies of q mp can be obtained by either of the following methods:

(a)	the absolute accuracies of q mdew and q mdw are ±0,2 per cent which guarantees an accuracy of q mp of ≤ 5 per cent at a dilution ratio of 15. However, greater errors will occur at higher dilution ratios;

(b)	calibration of q mdw relative to q mdew is carried out such that the same accuracies for q mp as in (a) are obtained. For details see paragraph 9.4.6.2;

(c)	the accuracy of q mp is determined indirectly from the accuracy of the dilution ratio as determined by a tracer gas, e.g. CO2. Accuracies equivalent to method (a) for q mp are required;

(d)	the absolute accuracy of q mdew and q mdw is within ± 2 per cent of full scale, the maximum error of the difference between q mdew and q mdw is within 0,2 per cent, and the linearity error is within ±0,2 per cent of the highest q mdew observed during the test.

9.4.6.3.   Calibration of differential flow measurement

The flowmeter or the flow measurement instrumentation shall be calibrated in one of the following procedures, such that the probe flow q mp into the tunnel shall fulfil the accuracy requirements of paragraph 9.4.6.2:

(a)

The flowmeter for q mdw shall be connected in series to the flowmeter for q mdew, the difference between the two flowmeters shall be calibrated for at least 5 set points with flow values equally spaced between the lowest q mdw value used during the test and the value of q mdew used during the test. The dilution tunnel may be bypassed.

(b)

A calibrated flow device shall be connected in series to the flowmeter for q mdew and the accuracy shall be checked for the value used for the test. The calibrated flow device shall be connected in series to the flowmeter for q mdw, and the accuracy shall be checked for at least 5 settings corresponding to dilution ratio between 3 and 50, relative to q mdew used during the test.

(c)

The transfer tube (TT) shall be disconnected from the exhaust, and a calibrated flow-measuring device with a suitable range to measure q mp shall be connected to the transfer tube. q mdew shall be set to the value used during the test, and q mdw shall be sequentially set to at least 5 values corresponding to dilution ratios between 3 and 50. Alternatively, a special calibration flow path may be provided, in which the tunnel is bypassed, but the total and dilution airflow through the corresponding meters as in the actual test.

(d)

A tracer gas shall be fed into the exhaust transfer tube TT. This tracer gas may be a component of the exhaust gas, like CO2 or NOx. After dilution in the tunnel the tracer gas component shall be measured. This shall be carried out for 5 dilution ratios between 3 and 50. The accuracy of the sample flow shall be determined from the dilution ratio r d: (84) The accuracies of the gas analysers shall be taken into account to guarantee the accuracy of q mp.

9.4.6.4.   Carbon flow check

A carbon flow check using actual exhaust is strongly recommended for detecting measurement and control problems and verifying the proper operation of the partial flow system. The carbon flow check should be run at least each time a new engine is installed, or something significant is changed in the test cell configuration.

The engine shall be operated at peak torque load and speed or any other steady state mode that produces 5 per cent or more of CO2. The partial flow sampling system shall be operated with a dilution factor of about 15 to 1.

If a carbon flow check is conducted, the procedure given in Appendix 5 shall be applied. The carbon flow rates shall be calculated according to equations 80 to 82 in Appendix 5. All carbon flow rates should agree to within 3 per cent.

9.4.6.5.   Pre-test check

A pre-test check shall be performed within 2 hours before the test run in the following way.

The accuracy of the flowmeters shall be checked by the same method as used for calibration (see paragraph 9.4.6.2) for at least two points, including flow values of q mdw that correspond to dilution ratios between 5 and 15 for the q mdew value used during the test.

If it can be demonstrated by records of the calibration procedure under paragraph 9.4.6.2 that the flowmeter calibration is stable over a longer period of time, the pre-test check may be omitted.

9.4.6.6.   Determination of the transformation time

The system settings for the transformation time evaluation shall be exactly the same as during measurement of the test run. The transformation time shall be determined by the following method.

An independent reference flowmeter with a measurement range appropriate for the probe flow shall be put in series with and closely coupled to the probe. This flowmeter shall have a transformation time of less than 100 ms for the flow step size used in the response time measurement, with flow restriction sufficiently low as to not affect the dynamic performance of the partial flow dilution system, and consistent with good engineering practice.

A step change shall be introduced to the exhaust flow (or airflow if exhaust flow is calculated) input of the partial flow dilution system, from a low flow to at least 90 per cent of maximum exhaust flow. The trigger for the step change shall be the same one used to start the look-ahead control in actual testing. The exhaust flow step stimulus and the flowmeter response shall be recorded at a sample rate of at least 10 Hz.

From this data, the transformation time shall be determined for the partial flow dilution system, which is the time from the initiation of the step stimulus to the 50 per cent point of the flowmeter response. In a similar manner, the transformation times of the q mp signal of the partial flow dilution system and of the q mew,i signal of the exhaust flowmeter shall be determined. These signals are used in the regression checks performed after each test (see paragraph 9.4.6.1)

The calculation shall be repeated for at least 5 rise and fall stimuli, and the results shall be averaged. The internal transformation time (< 100 ms) of the reference flowmeter shall be subtracted from this value. This is the “look-ahead” value of the partial flow dilution system, which shall be applied in accordance with paragraph 9.4.6.1.

9.5.   Calibration of the CVS system

9.5.1.   General

The CVS system shall be calibrated by using an accurate flowmeter and a restricting device. The flow through the system shall be measured at different restriction settings, and the control parameters of the system shall be measured and related to the flow.

Various types of flowmeters may be used, e.g. calibrated venturi, calibrated laminar flowmeter, calibrated turbine meter.

9.5.2.   Calibration of the positive displacement pump (PDP)

All the parameters related to the pump shall be simultaneously measured along with the parameters related to a calibration venturi which is connected in series with the pump. The calculated flow rate (in m3/s at pump inlet, absolute pressure and temperature) shall be plotted versus a correlation function which is the value of a specific combination of pump parameters. The linear equation which relates the pump flow and the correlation function shall be determined. If a CVS has a multiple speed drive, the calibration shall be performed for each range used.

Temperature stability shall be maintained during calibration.

Leaks in all the connections and ducting between the calibration venturi and the CVS pump shall be maintained lower than 0,3 per cent of the lowest flow point (highest restriction and lowest PDP speed point).

9.5.2.1.   Data analysis

The airflow rate (q vCVS) at each restriction setting (minimum 6 settings) shall be calculated in standard m3/s from the flowmeter data using the manufacturer’s prescribed method. The airflow rate shall then be converted to pump flow (V 0) in m3/rev at absolute pump inlet temperature and pressure as follows:

(85)

where:

q vCVS	is the airflow rate at standard conditions (101,3 kPa, 273 K), m3/s
T	is the temperature at pump inlet, K
p p	is the absolute pressure at pump inlet, kPa
n	is the pump speed, rev/s

To account for the interaction of pressure variations at the pump and the pump slip rate, the correlation function (X 0) between pump speed, pressure differential from pump inlet to pump outlet and absolute pump outlet pressure shall be calculated as follows:

(86)

where:

Δp p	is the pressure differential from pump inlet to pump outlet, kPa
p p	is the absolute outlet pressure at pump outlet, kPa

A linear least-square fit shall be performed to generate the calibration equation as follows:

(87)

D 0 and m are the intercept and slope, respectively, describing the regression lines.

For a CVS system with multiple speeds, the calibration curves generated for the different pump flow ranges shall be approximately parallel, and the intercept values (D 0) shall increase as the pump flow range decreases.

The calculated values from the equation shall be within ±0,5 per cent of the measured value of V 0. Values of m will vary from one pump to another. Particulate influx over time will cause the pump slip to decrease, as reflected by lower values for m. Therefore, calibration shall be performed at pump start-up, after major maintenance, and if the total system verification indicates a change of the slip rate.

9.5.3.   Calibration of the critical flow venturi (CFV)

Calibration of the CFV is based upon the flow equation for a critical venturi. Gas flow is a function of venturi inlet pressure and temperature.

To determine the range of critical flow, K v shall be plotted as a function of venturi inlet pressure. For critical (choked) flow, K v will have a relatively constant value. As pressure decreases (vacuum increases), the venturi becomes unchoked and K v decreases, which indicates that the CFV is operated outside the permissible range.

9.5.3.1.   Data analysis

The airflow rate (q vCVS) at each restriction setting (minimum 8 settings) shall be calculated in standard m3/s from the flowmeter data using the manufacturer’s prescribed method. The calibration coefficient shall be calculated from the calibration data for each setting as follows:

(88)

where:

q vCVS	is the airflow rate at standard conditions (101,3 kPa, 273 K), m3/s
T	is the temperature at the venturi inlet, K
p p	is the absolute pressure at venturi inlet, kPa

The average K V and the standard deviation shall be calculated. The standard deviation shall not exceed ±0,3 per cent of the average K V.

9.5.4.   Calibration of the subsonic venturi (SSV)

Calibration of the SSV is based upon the flow equation for a subsonic venturi. Gas flow is a function of inlet pressure and temperature, pressure drop between the SSV inlet and throat, as shown in equation 43 (see paragraph 8.5.1.4).

9.5.4.1.   Data analysis

The airflow rate (Q SSV) at each restriction setting (minimum 16 settings) shall be calculated in standard m3/s from the flowmeter data using the manufacturer’s prescribed method. The discharge coefficient shall be calculated from the calibration data for each setting as follows:

(89)

where:

Q SSV	is the airflow rate at standard conditions (101,3 kPa, 273 K), m3/s
T	is the temperature at the venturi inlet, K
d V	is the diameter of the SSV throat, m
r p	is the ratio of the SSV throat to inlet absolute static pressure =
r D	is the ratio of the SSV throat diameter, d V, to the inlet pipe inner diameter D

To determine the range of subsonic flow, C d shall be plotted as a function of Reynolds number Re, at the SSV throat. The Re at the SSV throat shall be calculated with the following equation:

(90)

with

(91)

where:

A1	is 25,55152 in SI units of
Q SSV	is the airflow rate at standard conditions (101,3 kPa, 273 K), m3/s
d V	is the diameter of the SSV throat, m
μ	is the absolute or dynamic viscosity of the gas, kg/ms
b	is 1,458 × 106 (empirical constant), kg/ms K0,5
S	is 110,4 (empirical constant), K

Because Q SSV is an input to the Re equation, the calculations shall be started with an initial guess for Q SSV or C d of the calibration venturi, and repeated until Q SSV converges. The convergence method shall be accurate to 0,1 per cent of point or better.

For a minimum of sixteen points in the region of subsonic flow, the calculated values of C d from the resulting calibration curve fit equation shall be within ±0,5 per cent of the measured C d for each calibration point.

9.5.5.   Total system verification

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 analysed, and the mass calculated according to paragraph 8.5.2.4 except in the case of propane where a u factor of 0,000472 is used in place of 0,000480 for HC. Either of the following two techniques shall be used.

9.5.5.1.   Metering with a critical flow orifice

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 emission test for about 5 to 10 minutes. A gas sample shall be analysed with the usual equipment (sampling bag or integrating method), and the mass of the gas calculated.

The mass so determined shall be within ± 3 per cent of the known mass of the gas injected.

9.5.5.2.   Metering by means of a gravimetric technique

The 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 emission 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 analysed with the usual equipment (sampling bag or integrating method), and the mass of the gas calculated.

The mass so determined shall be within ± 3 per cent of the known mass of the gas injected.