Measurement of emissions at project start Date: 12. May 2000 File number: 270 0 0088 Authors: Ken Friis Hansen Michael Grouleff Jensen Contribution from: Peter Wåhlin, DMU
Background Energy division Engine Technique Aarhus In connection with the Danish Road Transport and Safety Agency's Large Scale Particulate Filter Demonstration Project in Odense, the Danish Technological Institute, Engine Technique are in charge of a follow-up project with the purpose of measuring, collecting, processing and reporting data from Odense. One of the tasks in the project is measurement of emissions from selected vehicles at the project start and after 1 and 2 years of operation. Emission measurements The emission measurements are carried out as 5-mode tests on a chassis dynamometer. De legislative test for emissions from heavy-duty engines in Europe is 88/77/EEC (the 13-mode test), in which the engine is mounted in a test stand and measured in 13 stationary modes. The emission (g/s) and power (kw) measured at each mode is given a predetermined weight to produce a single value. The emission limits, the so-called EURO-standards, are thus expressed in g/kwh. Engine Technique have previously shown that a similar result can be obtained by measuring the vehicle on a chassis dynamometer in 5 different modes selected from the 13-mode, but with different weight factors. This "5-mode test" has been used when measuring the buses. The weight factors and modes are shown in the table below: Mode no. rpm load % 1 idle 0 25 3 max. torque 25 16 4 max. torque 50 16 6 max. torque 100 25 8 max. power 100 18 Table 1: 5-mode load- and weight factors. weight % 2
The following parameters are measured: NO x, NO 2, HC, CO and particulates, and exhaust gas temperature, backpressure, engine power and a number of other parameters for conversion and control of emissions. I addition to these measurements the National Environmental Research Institute, NERI measured the particulate size distribution. A custom built Hauke-type DMA (Differential Mobility Analyser, Winklmayr et al., 1991) was used in a closed-loop flow arrangement (Jokinen and Mäkelä, 1997) in connection with a TSI Model 3010 Condensation Particle Counter and a 12 kv adjustable generator. The scanning voltages were used to define 29 adjacent electrical mobility channels corresponding to the size range 6-700 nm of spherical particles with one electron charge. A PC was used to control the scanning process. Each scanning had a duration of approximately 3 minutes. Corrections for 50% average channel efficiency, and for zero and multiple electron charging, were calculated after each scanning using the same PC (Wiedensohler, 1988). No corrections were made for particle counter efficiency, which drops below 50% for particle sizes <10 nm. The equipment and software was borrowed from JRC, Ispra, Italy (Rita van Dingenen, European Commission Environment Institute). The DMA measurements were performed in diluted exhaust from a mini dilution tunnel. The dilution ratio was between 14 and 60, highest when measuring on the bus without particulate filter due to the capacity of the particle counter. The highest particle concentrations, converted to raw exhaust, where measured during mode 3 (1400 rpm, 25% load). Bus- and filter types The following bus and filter types were measured: Bus 8 Bus 8 Bus 16 Bus 5 Bus 139 Year 1996 1996 1999 1996 1987 Engine Volvo DH10A Volvo DH10A Volvo DH10A Volvo THD104KF Volvo THD101GC Power kw 285 285 245 245 242 Filter - none Silentor Volvo CRT Ferrita Eminox CRT Installed at date/km - 20-01-00 239.100 31-08-99 44.098 16-11-99 281.247 30-12-00 1.036.429 Tested at date/km 20-01-00 238.980 21-01-00 239.100 25-01-00 76.960 27-01-00 293.880 01-02-00 1.042.900 Table 2: Overview of bus- and filter types. Bus no. 5, 8 and 16 are newer buses of Euro-2 standard, while bus 139 is an older bus of Euro-0 standard. The filters were installed at different dates. Bus 16 had the filter factory-installed, while the others where retrofitted in the Odense project; the Silentor filter as late as 2
the day before the test. This is assessed to have only marginal influence on the measurements and results. Bus no. 8 was measured both with and without the particulate filter, primarily to give NERI the opportunity to measure particle size distributions both with and without the particulate filter. 3
Results and discussion The results of the emission measurements are given in the table below. Detailed measurement results can be found in appendix A. Bus 8 Bus 8 Bus 16 Bus 5 Bus 139 Filter - none Silentor Volvo CRT Ferrita Eminox CRT NO x g/kwh 11.8 10.7 10.9 8.69 13.5 NO 2 g/kwh 0.44 0.43 1.61 1.56 1.77 HC g/kwh 0.22 0.23 0.07 0.05 0.04 CO g/kwh 0.62 0.57 0.08 0.01 0.18 PM g/kwh 0.33 0.04 0.18 0.03 0.13 Table 3: Measured emissions. The most remarkable results are, without a doubt, the high particulate emission from the two CRT-filters. Both results are far above the expected level, and are a consequence of very high particulate emission at max. torque (mode 6), with a weight factor of 25%. The emissions are also high at max. power (mode 8), with a weight factor of 16%, although not as pronounced. The results have been presented to the producer of the filters, Johnson Matthey, who have given an explanation for the phenomena. In brief the explanation is that sulphate, formed by sulphur in the fuel 1, is stored in the wash-coat in the CRT-system pre-catalyst at temperatures below 380 C and released at temperatures above. As the exhaust gas temperature at mode 6 and 8 is around 380 C, this explanation seems likely. For those interested in the details Johnson Matthey's explanation is given in full in appendix. Please note that the arguments in the appendix are those of Johnson Matthey, and are not necessarily shared by neither the Danish Technological Institute nor the Danish Road Transport and Safety Agency. Another result worth noting is the high particulate emissions from bus 8 without filter. As new the bus should be below the Euro2 norm of 0.15 g/kwh. Normally Euro2 buses in use have particulate emissions around 0.22-0.26 g/kwh, but bus 8 is even higher than normal. The results for NO x and NO 2 (which is a subset of NO x ) clearly shows that while NO 2 is only 4% of the NO x -emission on the bus without filter, and on the bus with the Silentor filter, this share have risen to about 14% with the two CRT filters and about 18% with the Ferrita filter. This is because the CRT and Ferrita filters uses NO 2 to regenerate the filter (burn the collected particulates), and this increase was not unexpected, but unwelcome, as NO 2 is toxic. The filters with oxidation catalysts reduce the mission of HC and CO, as one would expect. 1 The sulphur content of the fuel is this context is of less importance. The sulphate storage will take place even with 1 ppm sulphur in the fuel (present content is below 50 ppm) it would just take more time before the wash-coat was saturated 4
The DMA measurements gave the following results: De highest particle concentrations, converted to raw exhaust, where measured at mode 3 (1400 rpm, 25% load). These size distributions are shown in the figures below. 4.0E+07 Average 3.0E+07 dn/dlogd (cm-3) 2.0E+07 1.0E+07 0.0E+00 1 10 100 1000 d (nm) Figure 1: Volvo DH10A 285, Odense Bytrafik no. 8, Mode 3 (1400 rpm, 25%) without filter. Particle concentrations measured in diluted exhaust, converted to raw exhaust. 4.0E+07 Average 3.0E+07 dn/dlogd (cm-3) 2.0E+07 1.0E+07 0.0E+00 1 10 100 1000 d (nm) Figure 2: Volvo DH10A 285, Odense Bytrafik no. 8, Mode 3 (1400 rpm, 25%) with Silentor-filter. Particle concentrations measured in diluted exhaust, converted to raw exhaust. The two first figures show that the Silentor filter gives a marked reduction in particle concentrations. The Silentor filter is especially efficient towards the smallest particulate sizes (the particulate number is reduced to 9%, the particulate volume to 20%). 5
No measurements where performed on bus 5 without filter, so no conclusion about the Feritta filters efficiency can be made. But the two next figure shows that the total particle concentration from the two buses with the filters mounted are similar. Please note that figure 3 is an enlargement of figure 2. 4.0E+06 Average 3.0E+06 dn/dlogd (cm-3) 2.0E+06 1.0E+06 0.0E+00 1 10 100 1000 d (nm) Figure 3: Volvo DH10A 285, Odense Bytrafik no. 8, Mode 3 (1400 rpm, 25%) with Silentor-filter. Particle concentrations measured in diluted exhaust, converted to raw exhaust. 4.0E+06 Average 3.0E+06 dn/dlogd (cm-3) 2.0E+06 1.0E+06 0.0E+00 1 10 100 1000 d (nm) Figure 4: Volvo THD104KF, Odense Bytrafik no. 5, Mode 3 (1400 rpm, 25 %) with Ferrita-filter. Particle concentrations measured in diluted exhaust, converted to raw exhaust. Presuming that the raw exhaust from the two buses have similar particle size distribution, it appears as if the Ferrita filter have a more even efficiency over the entire particle spectrum than the Silentor filter. 6
Conclusion The measurements on the Silentor filter shows that the minimum 80% efficiency, specified in the technical requirements, is obtained. The Ferrita filter should also satisfy the requirement although no measurements where made on this bus without the filter. The two CRT filters however do not fulfil the requirements. The DMA measurements done by NERI shows that the filters withhold more than 90% of the particles including the very smallest. The measurements also demonstrate that the CRT and Ferrita filters raise the NO 2 share of NO x by a factor of 3-4, while reducing CO and HC. As mentioned the measurements will be repeated after one and two years of service, and then it will be possible to make an assessment of the development of the filters efficiency and perhaps also about their lifetime. Aarhus, May 12th 2000 Danish Technological Institute, Energy Engine Technique References Jokeren, V. and Mäkelä, J.M. (1997) Closed-loop arrangement with critical orifice for DMA sheath/excess flow system. J. Aerosol Sci. 28, 643-648 Wiedensohler A. (1988) An approximation of the bipolar charge distribution for particles in the submicron size range. J. Aerosol Sci., 19, 387-389 Winklmayr, W., Reischl, G.P., Linde, A.O. and Berner, A. (1991) A new electromobility spectrometer for the measurements of aerosol size distributions in the size range from 1 to 1000 nm. J. Aerosol Sci. 22, 289-296 7
Johnson Matthey: Sulphate Storage and Release The CRT TM system comprises a Pt-based catalyst in front of a Diesel Particulate Filter (DPF). The catalyst is used to enhance the low temperature activity of the system such that the system gives better low temperature performance than alternative PM control devices [1]. The catalyst comprises active components supported on a washcoat; the washcoat greatly increases the durability of the catalyst, and ensures that the catalyst maintains very high performance even after high mileage accumulation [2]. All diesel fuel contains sulphur species, and these species are combusted within the engine to form sulphur oxides. These sulphur oxides can stick to the washcoat of the catalyst, where they are stored as sulphate species. If the PM control device does not contain a washcoated catalyst (e.g. alternative PM control devices), then this storage will be minimised (or may not be seen) with the result that these gaseous sulphates will be released straight into the atmosphere. Over the CRT TM catalyst the sulphate storage is most evident during low temperature (low speed) driving. Therefore, when the CRT TM system is used in an urban environment, a significant proportion of the sulphate will be stored on the washcoat of the catalyst. Above a certain temperature, this sulphate will be released from the washcoat, and, during legislative tests, this gaseous sulphate is condensed and collected on the filter papers and contributes to the Particulate Matter (PM) emissions of the system. Therefore, if the CRT TM system is operated for long periods of time under low temperature conditions, a large amount of sulphate may be collected on the washcoat. If the system is then tested under a legislative cycle which contains high temperature points (eg the R49 or ESC cycles) then this sulphate will be released and it will appear as PM during the measurement. It can, therefore, seem as though the CRT TM is emitting large quantities of PM, and not, therefore, filtering with high efficiency. However, the soot emissions of the system will actually be extremely low, as proved by the fact that the filter papers used in the CRT TM tests were still white/yellow, rather than the black colour which would be expected if the system was emitting soot. In addition, it has been shown that if the CRT TM is re-tested following the first test, then the PM emissions of the system are greatly reduced, because a large proportion of the stored sulphate is released during the first test. After three or four tests the PM level reaches a very low and stable level, because all of the sulphate has been released from the washcoat. It is important to note that the overall emissions of gaseous sulphate from the CRT TM and alternative PM control devices will, of course, be the same. The only difference will be in the location of the release. Non-washcoated systems will release the sulphates all the time. However, the storage and release behaviour of the CRT TM means that it will minimise the release of sulphate in urban areas (where the system will generally be at low temperatures due to the low speed driving cycle), and will then release the sulphate during high speed highway driving, away from population centres. 8
[1] DECSE Program, Final Report (Jan 2000), sponsored by US DOE, EMA and MECA. [2] R Allansson, BJ Cooper, JE Thoss, A Uusimaki, AP Walker and JP Warren, SAE 2000-01-0480, European experience of high mileage durability of continuously regenerating filter technology. 9