Blauer Engel Fuel Economy Test Feedback from ATC

1
Blauer Engel Fuel Economy TestFeedback from
ATC

• Ian Field
• Chairman, ATC-PTS

2
Background

• ATC supports the proposal to introduce a Blauer
  Engel qualification
• Provided advice on draft regarding toxicological
  and testing issues
• ATC believes the focus on fuel economy is
  appropriate for the definition of an
  environmentally friendly oil
• The Blauer Engel label is of value if the best
  10-15 of lubricants can achieve the required
  standard

3
ATC Concerns about Current Proposal

• The test method has not been fully developed
• There is potential for considerable variability
  in the test method, leading to unreliable
  conclusions about the lubricant performance
• Much of the variability arises from the aging
  process
• The target of 5 fuel economy is unrealistic

4
Test method development

• Today the test only exists on paper, the full
  procedure has never been run
• Normal process of test development would include
  the following
• Definition of draft method, targeted at meeting
  objectives of test
• Establishment of correlation with the field
• Selection of appropriate reference fuel
• Establishment of reference oils to allow quality
  monitoring and evaluation of ability of the test
  to discriminate
• Testing to establish repeatability and
  discrimination of the test
• Adjustment of test method if necessary
• Testing of method by other laboratories to
  establish reproducibility
• Establishment of quality monitoring requirements

5
Potential for Test Variability

• Aging of reference and candidate can never take
  place under identical conditions
• If the same engine is used, in the second test
  the oil will run in an engine that is 15,000km
  older, and the weather conditions may be
  different
• If the test is run in different engines of the
  same age then engine-to-engine variability is
  important
• For both procedures the oil consumption may be
  different
• Either way the aging process will be different
• Experience in API Seq. VIB test is that the
  engine and results change as the block ages
• Errors are increased as we have to compare the
  candidate with the reference oil, both of which
  will have variability

6
Factors which affect aging

• End of aging viscosity is the key parameter
• Key engine variables
• Oil consumption very large influence on aged
  oil viscosity
• Amount of blow-by impact on oil degradation
• Amount of wear wear metals catalyse oxidation
• Oil temperature controls degree of oxidation
• Shearing of viscosity modifier
• Fuel dilution
• Soot loading not important for gasoline engines
• Unless these variables are well controlled in the
  aging stage, the viscosity of the aged oil will
  vary considerably

7
Analysis of the Proposed Procedure

• This analysis has been prepared by Chris Gray and
  Colin Lewis, Infineum and CEC SDG statisticians
• Ageing of the candidate and reference oils for
  15,000km in a passenger car vehicle on a
  dynamometer
• Evaluation of the FE by comparing candidate and
  reference in a bench engine test

8
Estimation of the Test Variability

• The variability in the test can be broken down
  into 2 parts, which may be assessed using data
  from industry databases
• Variability of the Oil after it has been aged
• Look at variability in viscosity
• Look at FE sensitivity to viscosity
• Variability of the Testing Protocol
• Look at the precision of the FE test

9
Variability of Viscosity in Aged Oils
Variability varies significantly by test. The
TU3MH is a gasoline engine, with no soot and low
levels of viscosity increase, consequently the
viscosity is relatively well controlled, with an
CoV of around 5 Other engines typically have CoV
up to 50
10
Coefficient of Variation varies with Viscosity
Change
Out of range value

• The Coefficient of Variation (CoV) is calculated
  as SD/Mean
• The CoV increases as the severity of aging
  increases

11
Measure FE and Variability

• The measured fuel economy at the end of test (F1)
  may be written as follows
• FE1 FE0 ß(V1 - V0) D
  e

Test Error
Non-Viscosity Effects
Viscosity Effect
Initial True FE
Final Measured FE
where V0 and V1 indicate the initial and final
viscosities and ß is the sensitivity of the FE
measurement to Viscosity.
Notes D captures any FE effects due to ageing
which are NOT due to the changes in the
viscometrics. We ignore the effect of D in the
calculations as we cannot quantify it. This may
cause us to under-estimate the variability.
12
Uncertainty in FE Measurement for an Oil after
Testing

• Possible values based on the previous analysis
• M111FE test has standard deviation of 0.2,
  proposed procedure is more complex and likely to
  be worse
• Estimated 95 confidence interval for determined

Best case
More realistic
13
Conclusions from Analysis of Variability

• Variability of the viscosity of the used oil has
  the biggest impact on overall variability
• Difficult to control reliably in the aging
  procedure
• Compounded by comparing the candidate and
  reference results
• Both results suffer form variability
• If the reference oil gives FE at the low end of
  the statistical range, candidate oils can appear
  to have unrealistically high FE benefit
• The overall variability could approach 3
  (absolute) in terms of the final fuel economy
  measurement
• Low confidence that Blauer Engel oils would
  really have the fuel economy expected

14
Experience in Other Fuel Economy Tests

• M111FE
• CEC test, and only method accepted by ACEA as
  valid for lubricant fuel economy claims
• Gasoline engine
• OEM Tests
• VW, Opel, BMW
• All gasoline engines
• Ford Puma
• Diesel test, developed by CEC but determined not
  to be useful
• Seq. VIB
• US engine, not relevant to European application

15
M111FE Test

• Responds mainly to viscosity (HTHS), but effects
  of friction modifiers also seen
• Repeatability of reference oil 0.2 (absolute)

16
M111FE Results

• Best result ever seen for FE is 4.15 (1645
  results)
• Very few results above 3

17
M111FE Results (2)

• Only oils that give gt3.6 are xW20 grades
• These grades are not accepted by European OEMs
  based on the risk when fuel dilution occurs
• Greater number of high results from 5Wx grades is
  simply based on the number of oils tested
• No other viscosity grades can give results gt3.0
• Extrapolation of reference data suggests that
  HTHS 1.5 mPa.s is required to give 5 FE
• This is much to thin for engine operation

18
OEM Tests

• VW, Opel and BMW have in-house fuel economy tests
  in their specifications
• VW test proposed for Blauer Engel
• Unable to present results due to commercial
  confidentiality, but believed to be similar to
  M111FE results

19
Ford Puma Test

• CEC developed the Ford Puma (Duratorq) fuel
  economy test
• Diesel engine
• Examined oil aging
• Conclusion was that fresh and aged oil fuel
  economy was accurately predicted by the HTHS
  viscosity of the oil, so test was not adopted
• Data presented in SAE 2004-01-2023
• Aged oils gave slightly better fuel economy
• Aging was insensitive to formulation effects
• Extrapolation of results for reference oils
  suggest HTHS 2.2 mPa.s would give 5 fuel
  economy
• This is much to thin for engine operation

20
Ford Puma Analysis

• Linear relationship between HTHS viscosity and
  fuel economy
• No formulation effects

21
Conclusions

• In order to obtain a precise aged oil fuel
  economy test, the oil must be aged very
  consistently in order to avoid variation in the
  viscosity
• Viscosity is the major factor affecting fuel
  economy
• The FE benefits from moving to low viscosity oils
  vary from engine to engine, and according to test
  conditions
• Any test developed should provide additional
  information about the oil that could not be
  obtained by measuring the fresh oil viscometric
  properties
• It is highly unlikely that a result gt5 fuel
  economy will be seen in this test, except through
  random statistical variation