THIRD DRAFT 270404

BWG-P-04-006

Emissions performance of buses powered by engines conforming to Euro 3 emissions legislation

The Low Carbon Vehicle Partnership (LowCVP) set up a Bus Working Group to investigate how the shift to low carbon buses could be achieved. The Group's Interim Report indicated that low carbon buses could be technically feasible at a whole-life cost comparable to present buses once in full-scale production. It made recommendations on the definition, testing and certification of a low carbon bus. It recommended that initial support from Government was needed to enable the widespread early use and adoption of low carbon buses.

The UK Government’s Powering Future Vehicles Strategy includes a target that by 2012 at least one in five new buses will be low carbon, with greenhouse gas emissions at least 30% below today’s levels. The Bus Working Group defined the principal Greenhouse Gases (GHG) of interest to be Carbon Dioxide (CO2), Methane (CH4) and Nitrous Oxide (N2O). The relative “global warming potentials” for the 3 gases are 1:21:310 respectively. For current diesel bus technology, Carbon Dioxide is the main GHG constituent of vehicle exhaust, while some methane may be found in lean-burn natural gas buses.

The objectives of the Low Carbon Bus Project

The primary purpose of the Low Carbon Bus Project is to provide PowerShift grants to support fleets of low carbon buses to demonstrate their performance, reliability, operating costs and effectiveness across a range of routes. In doing so the intention is to provide the bus industry with experience and authoritative data from the real time operation of low carbon buses, generating the confidence for bus operators to take up their use, and bus manufacturers to put them into full-scale production.

Performance specification of vehicles

The definition of a Low Carbon Bus against which applications for the project will be assessed is as follows:

  • A Low Carbon Bus shall produce at least 30% fewer Greenhouse Gas (GHG) emissions than a current Euro 3 equivalent diesel bus of the same total passenger capacity. GHG emissions shall be expressed in grams of carbon dioxide equivalent and will cover “Well to Wheel” performance, thereby taking into account the production of the fuel as well as its performance from fuel tank to wheel.
  • The Well to Tank (WTT) and Tank-To-Wheel (TTW) Performances shall be assessed using the reference procedures described in the Interim Report from the Low Carbon Bus Working Group. The test cycle shall be the London Buses Route 159 test cycle.

A capability to run with zero tailpipe emissions would be advantageous, particularly in urban air quality problem areas. Applications should indicate the capacity for zero tailpipe emission running where this is feasible.

Current Euro 3 diesel bus performance

A precursor to the assessment of the performance of Low Carbon Buses is the establishment of a baseline for GHG emissions generated by buses powered by Euro 3 engines.

European legislation currently only requires the determination of air quality emissions which does not include CO2, but would include methane if the engine were fuelled by Natural Gas. Euro 3 engine emissions are measured over the European Stationary Cycle (ESC) and European Load Response (ELR) test cycles for diesel fuelled engines. Tests are also carried out over the European Transient Cycle (ETC) for gaseous fuelled engines.

None of the above test cycles factor in vehicle parameters such as mass, inertia and aerodynamic drag and only the ETC test includes acceleration and deceleration of the engine as might be experienced in “real world” operation. The ETC test is, however, based upon the FiGE cycle which, whilst based upon the operation of a sample of vehicles including buses and heavy-duty trucks is considered not to be strictly suited to suburban bus application, particularly as it contains a significant portion of simulated motorway driving.

A request for data was made to members of the Low Carbon Bus Working group for data from Euro 3 buses tested at Millbrook on the Route 159 cycle. The major contributions were from work sponsored by TfL and TransportEnergy although it should be acknowledged that some legislative engine dynamometer data has been submitted by some OEMs.

With respect to greenhouse gases, only CO2 is represented within the data. Therefore further work in terms of determining levels of other GHGs would need to be commissioned for total GHGs to be represented. Experience has shown, however, that for diesel powered vehicles CO2 is representative of total GHG as leveles of methane and N2O are very low or zero.

Table 1 details the types of bus and their emissions levels as tested over the Route 159 test cycle. This data includes both regulated emissions and CO2 and also represents buses fitted with exhaust aftertreatment and pre-Euro 3 buses repowered with Euro 3 engines.

Vehicle / HC / CO / NOx / PM / CO2 / l/100km
Trident #1 DD Cummins without CRT / 0.402 / 5.49 / 16.10 / 0.270 / 1461.2 / 55.22
Trident #1 DD Cummins with CRT / 0.020 / 0.064 / 15.47 / 0.038 / 1456.0 / 54.66
Trident #2 DD Cummins without CRT 260 BHP / 0.515 / 3.342 / 20.29 / 0.153 / 1323.5 / 49.94
Trident #2 DD Cummins without CRT 225 BHP / 0.504 / 2.05 / 19.21 / 0.154 / 1308.4 / 49.34
Trident #2 DD Cummins 225 BHP with CRT / 0.073 / 0.046 / 18.44 / 0.010 / 1320.8 / 49.59
Transbus Dennis Trident #3 DD with CRT / 0.046 / 0.062 / 20.39 / 0.020 / 1372.8 / 51.79
Transbus Dennis Dart #1 SD Cummins no trap / 0.143 / 1.674 / 15.67 / 0.148 / 1102.0 / 41.47
Transbus Dennis Dart #2 SD with CRT / 0.011 / 0.064 / 12.35 / 0.010 / 930.3 / 35.73
Volvo B7TL DD with CRT / 0.034 / 0.532 / 12.13 / 0.024 / 1406.0 / 54.03
Mercedes Citaro 12m SD with HJS CRT / 0.006 / 0.254 / 12.60 / 0.000 / 1422.0 / 54.62
Mercedes Citaro Artic SD with HJS CRT / 0.000 / 0.281 / 13.61 / 0.092 / 1585.7 / 59.82
DAF DB250 DD with Filter / 0.133 / 0.066 / 16.70 / 0.008 / 1238.5 / 46.73
Leyland Olympian DD Cummins repower no trap / 0.072 / 0.069 / 16.06 / 0.090 / 1288.3 / 48.75
Leyland Olympian DD Iveco Repower Eminox CRT / 0.046 / 0.068 / 16.609 / 0.035 / 1351.9 / 51.00
Optare Excell Cummins ISB Re power, Std Exhaust / 0.346 / 1.841 / 10.357 / 0.4 / 1118.6 / 42.34
Optare Excell Cummins ISB Re power, ECS Purifilter / 0.031 / 0.103 / 10.566 / 0.015 / 1088.4 / 41.06
Optare Solo Mercedes OM 906 LA Std Exhaust / 0.517 / 1.009 / 6.025 / 0.148 / 829.06 / 31.39
Marshall Midi Bus Cummins ISB Repower. Std Exhaust / 0.16 / 1.08 / 13.34 / 0.09 / 931.12 / 35.20
Marshall Midi Bus Cummins ISB Repower. ECS Purifilter / 0.003 / 0.016 / 13.443 / 0.023 / 968.22 / 36.520
Transbus Dennis Dart Dinex SCR + Trap, Urea Injection / 3E-06 / 0.078 / 3.977 / 0.044 / 991.643 / 37.407
Transbus Dennis Dart Dinex SCR + Trap, Ammonia Injection / 0.000 / 0.331 / 1.485 / 0.005 / 998.099 / 37.666
Transbus Dennis Dart, Dinex Trap, / 0.002 / 0.023 / 11.490 / 0.021 / 996.123 / 37.573
Transbus Dennis Dart, OE Silencer / 0.158 / 1.204 / 12.175 / 0.097 / 1030.876 / 38.973

Table 1: Euro 3 emissions data – Route 159 test cycle
At the initial first draft stage of this report it was stated that some of the data as received included information such as GVW, test inertia, number of seats, standees and total passenger numbers. This information was required in order to provide emissions figures on a gms/passenger/kilometre basis. Appendix 1 provides details of this more “complete” data set. It should be noted that the majority of vehicles within this dataset have exhaust aftertreatment and some are re-powers. A number of buses were removed from the data set at this stage due to inertia details not being available. These included some Dennis Tridents and Darts. Comments received at the Low Carbon Bus Working Group on 4th December 2003 suggested that inclusion of vehicles repowered with Euro 3 engines might result in a data set not representative of new vehicles.

Therefore a data set was produced which represents OEM fitted Euro 3 engines and has been refined to include UVW, GVW, test inertia, number of seats and total passenger numbers. This is shown in Appendix 2.

Methodology

The Low Carbon Bus Working Group had mandated that CO2 emissions figures for Low Carbon Buses should be based upon gms per kilometre and passenger number. Two stages of analysis were therefore undertaken. The first involved plotting CO2 data from the data set against test inertia to examine if a relationship between CO2 and test inertia could be established. The second phase sought to establish the relationship between CO2 emissions and passenger numbers.

Effect of test Inertia

The data as received were from buses which, whilst tested over the same drive-cycle (Route 159), had been loaded under different inertia setting regimes. A large proportion of buses had been tested to an inertia defined as 50% of seated passenger load, a passenger being defined as weighing 63kg. A number of the buses were tested at ULW plus 50% payload at GVW, whilst one only was tested at the recommendations of LCBWG, i.e., 50% total passenger load. The weight of a passenger under the new bus and coach directive is taken as 68kg for Class 1 and A (city buses with provision for standing passengers), and 71kg – which includes 3kg of hand luggage for Class 2 and 3 (express buses and coaches).

It was therefore required to normalise the effect of these load conditions before a Euro 3 CO2 emission baseline could be determined.

By plotting the levels of CO2 (TTW) against the test inertia of the vehicles it is possible to develop a straight-line relationship between test inertia and CO2

The plot below shows this relationship.

Using the refined Euro 3 dataset (100% diesel powered vehicles only) the developed y = Mx + C function is:

y = 0.0637x + 461.03

With the above function a CO2 level based upon a defined operational mass i.e. ULW + payload of the vehicle can be developed where:

CO2 (TTW) = 0.0637 ((number of passengers x 68Kg) + ULW) + 461.03

Unladen bus weight (ULW) is listed in both EEC and ECE regulations as the ‘Mass of vehicle in running order’ which is defined as:

“The mass of an unladen vehicle with bodywork, and with coupling device in the case of a towing vehicle, or the mass of the chassis with cab if the manufacturer does not fit the bodywork and/or coupling device, including coolants, oils, 90% of fuel, 100% of other liquids except used waters, tools, spare wheel, driver (75kg) if there is a crew seat in the vehicle”.

Determination of the baseline

Recognising that the available data is limited, particularly for vehicles having fewer than around 40 seats it is felt that having established an essentially linear relationship between vehicle test inertia and CO2 emissions, it is now possible to develop the relationship to enable the CO2 emissions to be predicted at higher test inertias representing the higher passenger loadings defined by the Low Carbon Bus Working Group i.e. 50% total passengers. This is shown in Appendix 3 which also shows values for Well-to-Tank CO2 emissions (based on the LBST study) and the 30% reduction baseline for a Low Carbon Bus.

The graph below shows CO2 (TTW) in gms/km at 50% passenger loading plus ULW

As previously discussed data is limited for vehicles with passenger capacity of less than around 40. The Low Carbon Bus Target setting group originally proposed that smaller vehicles e.g. midi- and mini-buses may produce lower CO2 emissions than larger buses. This is possible, especially if fitted with manual gearbox which could provide a significant benefit over auto transmission variants. However, it is also possible that the ratio between GVW and ULW may be lower than for larger buses thus negating any such benefits. Therefore until sufficient data is made available this can neither be confirmed nor refuted.

Applying the methodology for the determination of Well-to-Tank emissions as defined in the LBST study produces the relationship shown in the figure below for Well-to Wheel CO2 emissions versus passenger capacity.

The bottom trend line is the result of applying a 30% reduction to this relationship as the recommended target reduction for Low Carbon Buses.

Discussion

The results presented vary slightly from the target figures initially proposed by the Low Carbon Bus Target Setting Group. These are attached as

Appendix 4.

Within the data set only one bus has been tested to the LCB inertia loading condition. This is a Dennis Dart with a Cummins ISB engine. This vehicle has 31 seats with 29 standees making a total of 60 passengers.

On the Route 159 test, 1192 gms/km of Carbon Dioxide were measured, with zero methane and zero nitrous oxide. Tank-to-Wheel performance is therefore 1192 gms/km

45.06 l/100 km fuel consumption was measured on test. This translates into 16.18 MJ/km energy input of diesel, and the LBST study shows GHG for diesel of 10.4 gms per MJ.

Well-to-Tank GHG is therefore 168 gms/km.

Overall Well-to-Wheel is 1360 gms/km

From the Target table (Appendix 4) at 60 passengers:

Target CO2 emissions from the original Low Carbon curve is 899.8 gms/km

This represents a reduction from a measured Euro 3 powered bus of 34%.

The predictive Low Carbon Bus curve, above, suggests that for a maximum loading of 60 passengers, a Low Carbon Bus should produce 873 gms CO2 per kilometre, a 36 % reduction from the Euro 3 figure for the Dart as actually tested to the recommended test loading of 50% total passengers.

Within the dataset no bus at the higher end of passenger capacity has been tested at the recommended test loading therefore data is not available to provide a validation check. However, from the predictive charts above a bus with a capacity of 100 passengers when tested at 50% passenger loading over the 159 cycle, might be expected to produce well to wheel CO2 emissions in the region of 1695 gms/km. The Low Carbon Target for a bus of 100 passengers as originally defined by the Bus Working Group is 1185.6 gms/km. The corresponding value derived from the predictive curves is 1186.2 gms/km, essentially the same value.

The chart below shows the comparison between the original LCBWG target curve and the linear target curve predicted from the test data (y = 7.8322x + 403.1).

There is reasonably close agreement between the predictive curve and the original LCBWG curve at the middle to high range of passenger capacity, i.e. 40 – 110 passengers, but there is some diversion below around 40 and above 110 total passengers. It is recognised that the original LCBWG curve may have been unrepresentative in form. Therefore superimposed is a linear fit curve derived from the LCBWG curve (y = 8.005x + 352.65). It can be seen that the linear fit applied to the LCBWG curve is of a similar (slightly steeper) gradient to the predicted curve but with a lower intercept.

Coupled with the lack of available data at the low end of the passenger range, and the fact that the only bus to have been tested at the 50% total passenger test weight presented significantly higher CO2 emissions than might be predicted, it is suggested that at this stage a revised linear curve is adopted which is similar to the predictive curve but lifted at the lower end of the passenger range, whilst keeping the top end fixed.

The revised curve is shown below compared with the WTW Euro 3 baseline and the predictive 30% Low Carbon Target

Air Quality Emissions

Whilst the focus of the Low Carbon Bus Project is, as previously stated, Green House Gas emissions, it is also a requirement that Low Carbon buses should perform at least as well as conventional Euro 3 powered buses with respect to regulated “air quality” emissions, namely NOx, PM, HC and CO. Key amongst these pollutants, especially in respect of diesel powered vehicles operating in urban areas are NOx and PM.

The dataset used to determine the baseline CO2 curve includes some vehicles fitted with OEM exhaust systems and some, the majority, including exhaust after-treatment for the abatement of PM, NOx and/or both. The data available for the analysis of baseline air quality emission performance is therefore further reduced due to the effect of after-treatment systems on specific pollutants.

Oxides of Nitrogen (NOx)

The limited dataset provided a number of data points to enable analysis of NOx emissions against test inertia. The vehicles chosen had either no exhaust after-treatment or were fitted withparticulate traps. This “relationship” between NOx and test inertia is shown in the figure below

The data points are widely scattered. It is felt that the introduction of electronic engine management systems may have contributed to this. It is possible to fit a linear trend line through these points, resulting in a trend both broadly as expected from combustion theory and also of similar form to that seen in other analyses - an analysis carried out by the author on Euro 2 heavy-goods vehicles which, it is presumed, did not have such sophisticated control resulted in a linear fit with an R2 value of 0.8752.

The R2 value from the Euro 3 curve is low at 0.252 and does not provide confidence in the use of this line as the basis of an emissions target. However, applying the same methodology as adopted for CO2 results in the chart below

Comparison of the specific vehicles with this trend line suggests that it is perhaps inappropriate in its current form as a performance baseline for NOx. The Dennis Dart with 60 passengers tested at LCBWG inertia loadings produced 14.73 gms/km NOx, 21% higher than the 12.18 gms/km predicted. Whilst a trend is apparent, more debate is required within LCBWG, and more research work is required before an acceptable and defensible NOx performance target can be established. A key question is whether such a “real world” target for air quality emissions could or should be applied to Low Carbon buses, or whether accreditation of such vehicles should rely purely on the correct application of the appropriate engine homologation standard. This also raises the issue of how hybrid drive trains for heavy-duty applications should be homologated. Whilst the current requirements of ESC and ELR engine dynamometer tests for Euro 3 supplemented by the ETC test for Euro 4, provide procedures and process within the established legislative framework, these cannot be considered as representative of hybrid operation.

Particulate Matter (PM)

Again, the available data was severely reduced due to the need to exclude any vehicles with particulate traps and SCR systems. The resulting scatter plot is shown below. It is possible to obtain a linear fit to these data points but this would result in a negative PM intercept at zero test inertia. A second order polynomial presents a reasonable fit but would result in very high predictive PM levels at high, but credible, test inertias.