European NRMM PEMS Pilot Program · EC JRC - EU PEMS NRMM – Brussels – January 23, 2013 Exclusion of averaged window data The remaining valid test data is used to conduct the

European NRMM PEMS Pilot Program - Status of technical developments -

23th January 2013 European Commission – Joint Research Centre A. Perujo – P. Bonnel

(Some) components of vehicle/engine emissions

EC JRC - EU PEMS NRMM – Brussels – January 23, 2013

• Standard test cycle and test conditions

• Emissions limits

• Average energy consumption / CO2 emissions

Vehicle or engine type approval

• Durability

• In-Service Conformity

Lifetime « warranty »

• Product labelling (Energy consumption)

• OBD (ECS malfunctions) User information

Actors of vehicle/engine emissions legislation…

1. Product manufacturer > product compliance upon introduction on the market and during a certain period of the product life (e.g. durability, in-service conformity)

2. Type approval authority > compliance verification and market surveillance

3. User > product operation (e.g. maintenance, fueling and reagents), malfunctions (e.g. OBD)


1. Conformity of the product with the type approval requirements during the product ‘normal life’ – Not all type approval requirements covered (mainly emissions, also evaporative emissions for passenger cars)

2. For engines, the EU ISC provisions also include requirements to check the ‘functionality’ of the emissions control systems.

In-Service Conformity (ISC)…


1. Cost-effectiveness (no engine extraction required)

2. Intermediate step towards the control of real emissions

Why PEMS for ISC of HD and NRMM engines?


1. Cost-effectiveness (no engine extraction required)

2. Intermediate step towards the control of real emissions

3. Appropriate test protocols and data evaluation required

PEMS for ISC NRMM engines?


Approach

1. ‘Baseline’ procedures (i.e. for both emissions testing and data processing as in the draft annex) applied on a larger sample of cases (machine, duty cycles).

2. Modifications of the EU baseline test protocol shall be evaluated to possibly improve the procedures on a number of selected items.

3. Comparison of EU / US data evaluation methods in terms of coverage of test conditions, ability to judge the engine conformity (ISC), …


Outline

1. Status of testing activities

2. Instrumentation and testing

3. Test cycle and test conditions

4. Data evaluation method

5. Next steps


Lessons learned – Participants input 1. Status of testing activities

• Stage III A machines: 6 (Joint tests JCB/CNH/JRC)

• Stage III B machines: • John Deere/ JRC: 1

• Volvo Construction: 2

• Deutz / TUV Hessen: 1

• Cummins: 1

• FPT: 3

• Stage IV machines (prototypes): • Caterpillar: 4

• Stage IV machines (simulation): • Deutz: 1


Lessons learned – Participants input 2. Test protocol, instrumentation & testing

“Shopping list” - What can be done in the protocol for:

• Alternative exhaust mass flow measurement methods?

• Noisy, missing data?

• Site or machine based restrictions on PEMS use testing of small machines?

• ECU (ECM) Signal Calibration, e.g. torque?

• Power to the PEMS from machines?

• Ambient temperature measurements?


Lessons learned – Participants input Alternative exhaust mass flow measurement methods?

• Direct measurement shall remain the basis

• In case the direct measurements are not possible: Indirect exhaust flow measurements (From fuel and intake air flows) requires to demonstrate the conformity of ECU signal. Calibration of fuel / air flow measurements during both in-field tests and type-approval tests.

• In case of air “blow in” in the tailpipe, emissions sampling shall be conducted upstream the point of air injection

EC JRC - EU PEMS NRMM – Brussels – September 19 , 2012

Lessons learned – Participants input Noisy, missing data?

• Elements to ensure a sufficient data quality:

• Calibration

• Pre and post-test instruments checks

• Zero and auto-zero of instruments

• Drift corrections

• “Noisy” cannot be defined

• “Missing data’ could be interpreted as “interrupted test”: could it be allowed under certain conditions?


Lessons learned – Participants input Site or machine based restrictions on PEMS use testing of small machines?

• Machine selection is a joint process engine manufacturer – type approval authority, submitted to the approval of the TAA

• “Suitable” machines should be selected to demonstrate the engine conformity, i.e. not “difficult” machines

• Should machine “restrictions” remain for safety, the possibility not to report THC should be introduced

• For small machines (Definition of “small” to be discussed), the possibility to extract the engine to check its conformity should remain


Lessons learned – Participants input ECU (ECM) Signal Calibration?

• Important for engine torque, fuel rate and/or intake air flow

• Demonstration of ALL signals “correctness” at type approval

• Problems in demonstrating the ECM torque compliance during in-field testing (full load curve check): use of CO2 based window calculations


Lessons learned – Participants input Power to the PEMS from machines?

• Should be introduced, under certain conditions, to facilitate testing

• Should possibly follow the US rules/recommendations (i.e. the power consumption shall not represent more than a given percentage of the engine maximum power)


Lessons learned – Participants input Ambient temperature measurements?

• Option 1: Should be measured once [beginning of the test] (or twice, [also at the end of the test]) within a reasonable distance from the machine

• Option 2: Us the CAN signal for Intake Air temperature (Temperature experienced by the engine)


Lessons learned – Participants input 3. Test conditions

Not to be discussed, the applicable:

• Ambient conditions

• Maximum altitude (i.e. Min. barometric pressure)

• Engine conditioning (definition of a cold engine based on coolant temperature)


Lessons learned – Participants input 3. Test conditions

Main items being considered:

• PEMS test duration, expressed in engine CO2 emissions or work over the NRTC cycle, deciding whether the requirements apply to all data or only to valid data

• At the scale of the ‘averaging windows’: Applicable engine loads (e.g. power threshold)


Lessons learned – Participants input 4. Data evaluation methods

The ‘baseline’ method is the EURO VI ISC method

In May 2012, JRC proposed a first ‘modified’ method, to be assessed by program participants. Its main objective was to seek for solutions to deal with long idling sequences and low load operation.

In Sept. 2012 JRC presented results from the first iteration

New proposal NRMM PEMS-ISC evaluation:

Definition of “boundary conditions”

Subsequent evaluation rules


Definition the working and non-working event: Based on duration parameters D0, D1, D2, D3

Exclusion of test data points


D0 2 minutes D1 2 minutes D2 10 minutes D3 4 minutes

Using a power or idling criterion [engine power is lower than <10% non-working situation]

Exclusion of test data points: a. Engine system condition: Cold start (Based on coolant

temperature) b. (Potentially, but very unlikely) - Operational: Ambient

temperature c. Short (<D2 min) and Long “NON WORKING” (>D2 min)

events. d. For all non-working events, the first D1 minutes of the

event are valid. e. For long non-working events, the take-off phase following

the idling event may also be excluded until the exhaust gas temperature reaches 250C. If the exhaust gas temperature does not reach 250C within D3 minutes, the data analysis shall restart.



Exclusion of averaged window data The remaining valid test data is used to conduct the moving averaging window calculations. The engine performance will be assessed from the window averaged emissions using:

a. The windows whose average power exceeds the power threshold [20%] are considered as ‘valid’.

b. A percentile obtained from the distribution of valid window emissions [90% cumulative percentile]

Lessons learned – Participants input


“Machine Work” marking algorithm (1)

Step 1: Detection, data splitting into working and non-working events

• using a power or idling criterion [engine power is lower than

<10% non-working situation]

• Calculate the duration of “non-working events”

• Mark the non-working events shorter than D0 minutes as ‘working’

• Calculate the duration of the remaining non-working events


Exclusions non-working data at the end of Step 1



Step 2: Merging of short working events into non-working

• Working events shorter than D0 are merged with surrounding non-working events longer than D1


Exclusions non-working data at the end of Step 2


Step 3: Exclusion of post non-working (take off) data


• Exclusion of D3 minutes (take off) after long non-working events


Exclusions of non-working data at the end of Step 3


Step 4: Inclusion of post-working data • Inclusion of D1 minutes of non-working at the end of working

events



0.80

0.85

0.90

0.95

1.00

1.05

0

20

40

60

80

100

120

3000 3500 4000 4500 5000 5500 6000 6500 7000 7500 8000

Engine Power [%]

Valid[1]/Invalid[0]

End of Step4 - Final


Case 1 (IIIB) - Baseline

Examples (1)

0

20

40

60

80

100

120

0 2000 4000 6000 8000 10000 12000 14000

Engine Power [%]

WBW Power [%]

WBW emissions

Valid[1]/Invalid[0]

Examples (1)


Case 1 (IIIB) - Modified

0

20

40

60

80

100

120

0 2000 4000 6000 8000 10000 12000 14000

Engine Power [%]

WBW Power [%]

WBW emissions

Valid[1]/Invalid[0]


Case 1 (IIIB) – Baseline vs. Modified

0

10

20

30

40

50

60

70

80

90

100

0.0 0.5 1.0 1.5 2.0 2.5 3.0 3.5 4.0

CONFORMITY FACTOR

NOx_Cum. Freq [BASELINE]

NOx_Cum. Freq [MODIFIED]

Examples (1)

Examples (2)

Case 2 (IV) - Baseline

0

20

40

60

80

100

120

0 2000 4000 6000 8000 10000 12000 14000

Engine Power [%]

WBW Power [%]

WBW emissions

Valid[1]/Invalid[0]


Examples (2)

Case 2 (IV) - Modified

0

20

40

60

80

100

120

0 2000 4000 6000 8000 10000 12000 14000

Engine Power [%]

WBW Power [%]

WBW emissions

Valid[1]/Invalid[0]


Examples (2)


Case 2 (IV) – Baseline vs. Modified

0

10

20

30

40

50

60

70

80

90

100

0.00 1.00 2.00 3.00 4.00 5.00 6.00 7.00

CONFORMITY FACTOR

NOx_Cum. Freq [BASELINE]

NOx_Cum. Freq [MODIFIED]


Examples (3)

Case 3 (IIIB) – Baseline (little or not non-

working events)


Examples (3)

Case 3 (IIIB) – Modified (little or not non-working

events)


Scenario Baseline Option 1 Option 2

Cold Start Exclusion Yes Yes Yes

Definition of

working / non-

working machine

None Yes Yes

Minimum duration

for non-working

events (1 a,b)

N/A D0 D0

Exclusion of non-

working events (2) Yes Yes

Inclusion of first

minutes from non-

working events

(2) Yes (D1) Yes (D1)

Minimum duration

for long non-

working events

(2) D2 N/A

Exclusion of “take-

off “after long-non-

working events

(2) Yes (D3) No

Power Threshold Yes - 20% Yes - 20% Yes - 20%

(1a) Events shorter than D0 are considered as ‘working’ events.

(1b) Working events shorter than D0 between 2 ‘non-working’ events (longer than D0) are merged with the preceding and the following events and therefore considered as ‘non-working’.

(2) Indirect and incomplete through the Power Threshold rule

Proposal

D0 2 minutes D1 2 minutes D2 10 minutes D3 4 minutes

Further … • The time criterion could be complemented by a temperature

criterion based on the exhaust temperature (e.g. for D3: the data can be excluded until the exhaust temperature reaches 250 degC. The maximum duration of excluded data shall not exceed 2 minutes)

• For the purpose of in- service conformity testing the operating conditions may be reproduced by an artificial load cycle. In this case a suitable load cycle shall be determined by the manufacturer in consultation with the type approval authority, in accordance with section 4.5 of this Annex.



Engine and equipment operating conditions Regardless of whether the test is performed using an artificial load cycle or by operating the machine in its usual work activity, to the maximum extent possible the load cycle should;

• comprise loaded work activities that the majority of the in-service population of the selected machine type could reasonably be expected to perform, and

• not include a disproportionate amount of non-working activity, and

• comprise sufficient loaded work activity that it can be reasonably expected that the minimum test duration defined in point 4.6.5 will be achieved during the test.

Operational requirements

• To the maximum extent possible, the machine operation shall be selected in such a way that the testing is uninterrupted and the data continuously sampled to reach the minimum test duration defined in point 4.6.5.

• When testing a machine that either; o does not complete the minimum test duration defined in

point 4.6.5, despite attempting to achieve this, or o would normally perform multiple different work activities,

• the test data may contain a composite of test data obtained from several operating sequences .

4.6.5 The minimum test duration shall be long enough to complete five times the work performed during the NRTC or produce five times the CO2 reference mass in kg/cycle from the NRTC as applicable .


Because the tests are sometimes performed under very harsh conditions then:


When several operation sequences are combine to reach the minimum test duration in point 4.6.5 the following precautions shall apply: • the different sequences are obtained using the same

machine and engine • the composite test data contains a maximum of [three]

sequences • each sequence in the composite test data contains a

minimum of [one] NRTC work • the sequences in the composite test data are obtained and

put together in a chronological order • The data analysis will applied to the complete composite test

data • The maximum period elapsed between sequences is [72]

hours

Sequences requirements:

Emissions evaluation with stitched data:

• Files must be stitched first

• Exclusions and evaluations are conducted on the stitched (cold start, non-working, DPF regeneration)

• Initial cold start emissions can be excluded for each sequence

Lessons learned – Participants input 5. Next Steps

• Evaluation of new proposals by JRC/participants and consolidated reporting by JRC based on the data available

• Use of the consolidated reporting for decision making

• Comparison EU / US data evaluation methods


Documents

European NRMM PEMS Pilot Program · EC JRC - EU PEMS NRMM – Brussels – January 23, 2013 Exclusion of averaged window data The remaining valid test data is used to conduct the