20076559 (JSAE) 2007-32-0059 (SAE)

The Feasibility of Meeting CARB / EPA 3 Emission Regulations for Small Engines Roy Douglas1)

Stephen Glover2)

Copyright © 2007 Society of Automotive Engineers of Japan, Inc. and Copyright © 2007 SAE International With annual worldwide production of over 100 million units, small off-road engines (SORE) have been recognised as a major source of air pollution. It is estimated that non handheld SORE products in circulation annually produce over 1 million tonnes of HC+NOx and over 50 million tonnes of CO2. These SORE did not have to meet any emissions control legislation until its introduction in the USA in 1995. Since then the gradual implementation of several stages of increasingly more severe legislation has resulted in a decade of intensive emissions control development for utility engines. New carburetted stratified charge 2-stroke engines and catalytic after-treatment are being developed for the handheld products where weight and multi-orientation operation are key requirements. For the non-handheld 4-stroke dominated market, manufacturers are looking at improved fuel system design, improved engine design and the use of after-treatment to meet current and future legislative requirements. The bulk of the total 4-stroke SORE engine market, at around 65%, is taken up by single cylinder 4-stroke gasoline engines of under 6hp or under around 160cc. In this, the largest segment of the market, manufacturers must look to more conventional or new technologies that can be applied without significant add on cost or more preferable with a cost reduction. This paper presents the rational for the proposed EPA / CARB approach to meeting Phase 3 legislation with SORE and considers several possible alternative strategies. Key Words: Utility Engine, Intake System, Exhaust Emissions stroke SORE market (handheld and non handheld) at around 65% is made up of single cylinder air cooled 4stroke engines of under 6hp or under around 160cc with both horizontal and vertical crankshafts. Figure 1 show the results of an EPA population report (4) of 4-stroke gasoline SORE engines using 1996/98 base year population figures. The under 6 hp single cylinder engines generally now have overhead valves and are lubricated by splash lubrication. Larger capacity, more expensive, twin cylinder engines are still generally air cooled but have pressurised lubrication circuits.

Introduction Small Off-Road Engines (SORE) are found in a very wide and diverse range of products that can be separated into two main product groups, handheld and non handheld. These products are generally classified as utility engines with a power output of less than or equal to 19 kW. Common examples of handheld products are chainsaws, trimmers, concrete saws etc. and common examples of non handheld products are lawnmowers, pumps, generators, small tractors, etc, as shown in Figure 1. The handheld market is around 30 million units per annum and is almost completely dominated by the 2stroke engine. New carburetted, stratified-charge 2stroke engines and catalytic after-treatment are being developed for the handheld products where weight and multi-orientation operation are key requirements. There have been many SETC technical papers (8,9,10,11) presented on these new stratified charge 2-stroke technologies by companies such as Stihl, Electrolux and Komatsu. The 2-stroke engine is currently not the focus of the fuel system technology discussed in this paper although the technology is applicable to 2-stroke engines. An application programme has recently commenced on a stratified charge 2-stroke and it is hoped that a future paper will present the results of this application. The non-handheld SORE market is estimated at around 70 million units per annum and is dominated by the 4stroke engine. By far the largest segment of total 4-

Figure 1: Examples of Handheld and Non Handheld SORE 1) Queen’s University, Belfast 2) Fjölblendir ehf 1

20076559 (JSAE) 2007-32-0059 (SAE) 80.0

Number of Units (millions)

70.0 60.0

Approximate Capacity Range 0 to 160 cc

50.0 40.0 65 % 30.0 20.0 160 to 300cc

300 to 450 cc

450 to 700 cc

6 to 11 hp

11 to 16 hp

16 to 25 hp

10.0 0.0 Below 6 hp

Total below 25 hp

Horse Power Categories

Figure 2 1996/98 US Population of 4-Stroke Gasoline SORE (Handheld and Non Handheld) Below 25 hp (EPA Report EPA420-R-05-022) 2.5 Number of Automobiles

The non-handheld SORE market is estimated at around 70 million units per annum and is dominated by the 4-stroke engine. By far the largest segment of total 4-stroke SORE market (handheld and non handheld) at around 65% is made up of single cylinder air cooled 4-stroke engines of under 6hp or under around 160cc with both horizontal and vertical crankshafts. Figure 2 shows the results of an EPA population report (4) of 4-stroke gasoline SORE engines using 1996/98 base year population figures. The under 6 hp single cylinder engines generally now have overhead valves and are lubricated by splash lubrication. Larger capacity, more expensive, twin cylinder engines are still generally air cooled but have pressurised lubrication circuits.

2

1.5

1

0.5

0

HC+NOx

CO

CO2

Emission

With annual worldwide production of around 100 million units per annum, non handheld and handheld SORE have been recognised as a major source of air pollution. Figure 3 shows an emissions equivalence comparison of non handheld and handheld products to typical automotive engines.

Figure 3: Estimated Comparison of a Handheld and Non Handheld SORE to a US Automobile (Estimate based on US Federal Tier 2) SORE did not have to meet any emissions control legislation until its implementation in the USA in 1995. Since then the introduction of several stages of steadily more severe legislation has resulted in a decade of intensive emissions control development for utility engines. With more severe Stage 3 EPA and European legislation on its way in the very near future, manufacturers of SORE need to meet legislative requirements but very importantly they need to do so in a cost effective manner.

In circulation population estimates of exact global numbers of SORE have not been found. However, using the EPA USA non-road population estimates (2) there were around 70 million 4-stroke gasoline SORE (under 19kW) in circulation in the USA in 1996/98. Taking this it seems reasonable to estimate that the global population could be circa 6 to 7 times this volume today. This would estimate the world population of non handheld 4-stroke SORE at 400 to 500 million units. With these estimated population volumes, non handheld products annually produce over 1 Million tonnes of HC+NOx and over 50 Million tonnes of CO2.

The whole SORE market is very cost sensitive and manufacturers must look to more conventional or new 2

20076559 (JSAE) 2007-32-0059 (SAE) Legislative Body CARB 2 EPA 2 Europe 2 CARB 3 EPA 3 Europe 3 CARB Blue Skies EPA Blue Skies Europe Blue Skies

The legislation in the USA and Europe divides the various SORE into two different groups, handheld and non handheld and then further divides them into classifications. Figure 6 depicts how the classifications are divided for CARB and EPA. Each classification is effectively a range of engine capacities and each classification has its own legislative emissions limits. The classifications are slightly different between CARB 2 and EPA 2, but will be more inline with CARB 3 and EPA 3. The test cycle will be covered in more detail in the next section of this paper, but there are effectively three test cycles, Cycle A and B for non hand held products and intermediate or rated speed applications respectively and Cycle C for handheld products. The test cycle is the J1088 test cycle, figure 7.

Status Implemented, effective from 2000 to 2005 and later Implemented, effective from 2001 to 2007 and later Implemented, effective from 2004 to 2008 and later Implemented, effective from 2005 to 2008 and later Being proposed, likely implementation 2008 and later Being proposed, likely implementation +2008 Voluntary, Implemented Being Proposed Unknown

Figure 4: General Status of US and European SORE Legislation (Source EPA, CARB data bases and 2002/88/EC Directive)

Products in each classification must also be certified to a specific useful life. Full details of the useful life, which allows for a deterioration factor, can be found from the appropriate legislative body but the useful life effectively separates products in hobby, semi professional and professional products. In the same way the useful life effectively separates products into three price levels, entry level, medium and top level products. Products at entry level must potentially look to more cost competitive emissions control technologies than top entry products.

technologies that can be applied without significant add-on cost or more preferably with a cost reduction. For the nonhandheld 4-stroke dominated market, manufactures are looking at improved fuel system design, improved engine design, internal/external EGR and the use of after-treatment to meet current and future legislative requirements.

Legislation In the USA the California Air Resources Board (CARB) has been leading the way with regard to SORE legislation. CARB has the most severe legislation with a third phase and lower voluntary limits already implemented. The environmental protection Agency (EPA) is seen as more or less following CARB with EPA 2 legislation already implemented and proposed EPA 3 limits that match CARB 3 likely to be introduced in the near future. Europe has followed the EPA with very similar Stage 2 legislation to that of EPA 2 and is thought to be considering a Stage 3 to match EPA 3. The actual compliance dates are quite complex when taking into account average, trading and banking (ATB’s) but in general the table in Figure 4 summarises the current status.

Products

Engine Capacity

Handheld Handheld Handheld Non Handheld Non Handheld Non Handheld Non Handheld Non Handheld Non Handheld Non Handheld

under 50cc over 50cc 50cc to 80cc under 66cc/65cc 66cc to 100cc 100cc to 225cc 65cc to 225cc over 225cc 80cc to 225cc Over 225 cc

Figure 5 gives a summary view of the legislation that has actually been implemented. The legislation shows a distinct trend in the reduction of HC+NOx emissions levels with the implementation of each successive stage but shows no reduction in CO emissions levels which remain at ten times the HC+NOx limits. Figures 6 shows the legislative trend in terms of HC+NOx, taking CARB legislation as the data source. CO emissions limits do not change and remain at around 550 g/kWh. The legislation is clearly focused on the reduction of HC+NOx emissions but apparently there is no move to reduce CO emissions or CO2 emissions as CO eventually degrades to CO2. CARB CARB BLUE SKY BLUE SKY

EPA P2

EPA P2

CARB 2

CARB 2

CARB P3

CARB P3

HC+Nox (g/kWhr)

CO (g/kWhr)

HC+Nox (g/kWhr)

CO (g/kWhr)

HC+Nox (g/kWhr)

CO (g/kWhr)

HC+Nox (g/kWhr)

CO (g/kWhr)

50 72

805 603

72 72

536 536

50

536

25

536

72

536

36

536

50 40 16.1

610 610 610

72

536

12.1

610

16.1 12.1

549 549 10 8

549 549

5 5

549 549

Figure 5: Summary of US Legislative Limits for Handheld and Non–Handheld products (Source EPA / CARB) 3

20076559 (JSAE) 2007-32-0059 (SAE)

Summary Bar Chart of US HC+NOx legislation Emissions limits ‰

NOTE Capacity catagories are not accurate and are a generalisation to facilitate plotting Summary Chart HC+ NOx limits Note, capacity categories areof grouped to facilitate plotting (Note capacity classifications are not accurate due to difference between governing bodies and phases)

70

Under 50cc 65cc, 80cc, 100cc to 225cc Over 225 cc

30%

HC+NOx (g/kWhr)

60

Handheld

EPA 2 2001 - 2007 (2010) EPA 3 2008 ? EPA Blue Skies 2008 ?

50

40

Non Handheld

Numbers are % Reduction from Phase 2

65%

30

20

70%

37%

66%

34% 10

0

CARB 2

Legislation

2000 - 2005

CARB 3

BLUE SKY

2005 - 2008

2005 - 2008

Figure 6 : Trend in CARB HC+NOx Legislative Limits (capacity categories are not accurate to facilitate plotting of the source data) Mode

1 2 3 4 5 6 7 8 9 10 11

Engine

Cycle A Cycle A Cycle B Cycle B Cycle C Cycle C

Speed

Load % Weight Load % Weight Load % Weight

Rated Rated Rated Rated Rated 85% Rated 85% Rated 85% Rated 85% Rated 85% Rated Idle

100 75 50 25 10 0

9 20 29 30 7 5

100 75 50 25 10

9 20 29 30 7

100

0

5

0

therefore HC+NOx emissions. However, for CARB 3 and subsequent legislation, this strategy alone will not be sufficient. EPA have stated (1, 2 and 5) that Phase 3 limits can be met using :

85

• • • 15

Improved production control Improved fuel system calibration Use of catalytic after treatment and passive secondary air injection

There is however considerable interest in the industry to reduce the costs associated with emissions control technology and political pressure to reduce fuel consumption to lower green house gas emissions.

Figure 7: J1088 Emissions Test Cycle for SORE (Source EPA/CARB) Setting of High CO Limits

The Test Cycle and its Importance

In response to the authors, EPA has stated that it has generally focused its efforts on regulating HC emissions and oxides of nitrogen (NOx) emissions. Regarding CO, very few areas (less than 10 counties) of the country exceed the NAAQS for CO. Therefore, EPA has not focused much effort on reducing CO emissions from non-road engines.

SORE engines are legislated in the USA and Europe on the J1088 test cycle. Figure 7 shows the test cycles for handheld and non handheld products. The test cycle is a steady state test cycle with no transient conditions. Each mode point has a weighting factor that effectively mimics the importance of each mode to real life usage. Cycle A and B for non handheld products have part load test points whereas Cycle C for handheld products is full throttle and idle only.

Other reasons for the CO limits remaining unchanged are related to the fuel system and engine functionality. SORE engines are generally quite thermally limited and use fuel to cool the engine at full throttle. The current strategy is to operate rich throughout the load range to minimise NOx and 4

20076559 (JSAE) 2007-32-0059 (SAE) 30.0

25 20% production variation on limits

Briggs Fuji Honda

20 HC+NOx (g/kWhr)

% Contribution

25.0 20.0 15.0 10.0 5.0

EPA 2 15

Average Bl 26

CARB Blue Skies

5

0.0

31

CARB / EPA 3

10

66

63

36 68

0

100% Load

75% Load

50% Load

25% Load

10% Load

Idle

0

100

200

300 CO (g/kWhr)

400

500

Cycle A Mode Point

Figure 9: Selection of Certified Non Catalyst Class 1 Phase 2 Engines Cycle Emissions

Figure 8: J1088 Modal Analysis of Percentage Contribution to Total Cycle Emissions for a Cycle A non Handheld Product

Position of Current Production Engines with Regard to Legislation

In order to meet legislation it is clearly important to look at the test cycle and match appropriate emissions control technology to reduce emissions at all or the most important mode points. For hand held products which are dominated by the 2-stroke engine, the technology must be effective at wide open throttle with a weighting factor of 85% and a high mass emissions rate but can potentially be less effective at idle. Stratified scavenging 2-stroke engines that are now in the market are effective at full load by increasing fuel trapping efficiency relative to air trapping efficiency, but do not reduce emissions and idle where the scavenging process is homogeneous and the main problem is incomplete combustion and misfire.

The EPA and CARB (1, 2) data bases are a very good source of engine test data from certified engines or products. The data bases also list the general technical specification of the engines and the emissions control technology used. This paper focuses on the Class 1 EPA Phase 2 engines as the data presented in this paper is from a Class 1 engine. Figure 9 shows cycle emissions data from a range of different Class 1 Phase 2 engines from different manufacturers. None of the engines are fitted with catalytic after-treatment. Cleary there may be examples of products with higher and lower emissions than those presented. This data in is not meant to be a complete analysis of all the engine data in the data bases but merely a “snap shot” to highlight the general picture.

In the same way, an analysis should be carried out for the non handheld products to determine which of the mode points are key to reducing emissions. Once this analysis has been carried out, various technologies can be assessed in terms of how good they are at reducing the total cycle emissions. Figure 8 shows the importance of the test cycle for a typical nonhandheld products in terms of the contribution of each mode point to total cycle HC+NOx emissions for a non handheld 4stroke product. A similar analysis can obviously be carried out for individual specific emissions or groups of emissions.

All the engines shown meet EPA and CARB 2 limits, most with a production margin of 20% on HC+NOx emissions. All the engines meet the Phase 2, 3 and Blue Skies CO limits. A few of the engines just meet the CARB 3 / EPA 3 HC+NOx limits. In general to meet CARB 3 / EPA 3, HC+NOx limits with a 20% production margin, the engines need a further 30% reduction in HC + NOx emissions. To meet Blue Skies with a 20% margin a 65% reduction is required. In general, it can be stated that Class 1 Phase 2 SORE appear to be able to meet Phase 2 limits at the required useful life without a catalyst, but that further emissions control technology is required to meet phase 3 limits, especially with sufficient production margin. It must be stressed that this statement is a general statement and that undoubtedly there are engines in production that that do not meet the Phase 2 limits either as new or at their designated useful lives.

The modal analysis for this typical, stock-rich, phase 2 compliant product clearly shows that 50%, 25% and 75% load are the three biggest contributors, in order of importance, to the total emissions. Any technology applied to reduce emissions needs to at least be effective at all or some of these test conditions. Although it is clearly advantageous to have a technology that is effective across the board, this is sometimes cost prohibitive and it may be necessary to select a technology that is less effective at mode points that are not large contributors to total cycle emissions and more effective at those mode points that do contribute more to total cycle emissions.

Production variation is a key issue facing manufacturers. The greater the production variation the more effective, and potentially the more costly, the emissions control technology needs to be to maximise production compliance. For the more 5

20076559 (JSAE) 2007-32-0059 (SAE)

Figure 10 : Production Variation in HC+NOx Emissions from Non Catalyst Class 1 Phase 2 Engines at 10 and 125 Hours Usage (source EPA/CARB)

Tecumseh Toro eng 1+ Cat

18 EPA 2 HC+Nox 16.1, 16 HC+NOx (g/kWhr)

variation of 25 to 50% compared to the 16.1 g/kWhr limit. Figure 10 also shows the deterioration factors over the 125 hour useful life for each of the engines. The engines emissions clearly increase with usage but all still meet the Phase 2 limits at 125 hours.

Tecumseh Toro eng 1

20

Honda GCV160 eng1 Tecumseh No Cat

Honda GCV160 eng 1 + Cat

14 12 10

Figure 11 shows the HC+NOx and CO emissions production variation of a selection of new Briggs and Stratton Quantum and Snapper engines. The exact engine usage in hours is not known but believed to be at zero or 10 hours. The Snapper engine apparently has much greater CO variation and the Quantum much greater HC+NOx variation. This suggests that the base Quantum engine has greater production variation than the Snapper and that the snapper fuel system has greater production variation than the Quantum fuel system.

Honda No cat

EPA 3 HC+Nox 10, CO 20% below limit

8 6

Tecumseh + Cat

CARB Blue Skies Honda + Cat

4

CO limit at 549

2 0 0

50

100

150

200

250

300

350

400

450

CO g/kWhr

The presented EPA / CARB data demonstrates that in general the current, phase 2, production engines are unable to meet the phase 3 limits when new or after usage at the certified life standard period. It can also be stated that a 20% HC+NOx production margin or perhaps even a 40% HC+NOx production margin will be needed to account for production variation and product deterioration at the certified useful life standard.

Figure 11: Production Variation in HC+NOx and CO emissions from a Selection of Non Catalyst Briggs and Stratton Quantum and Snapper Engines (Source EPA / CARB) expensive larger capacity products with higher useful life standards, the higher cost may allow tighter production tolerancing and the use of better materials like cast liners for example. Cleary the key to improving production variation is to more tightly control the production process and reduce variation in critical areas such as ring-pack design, bore distortion, blow-by and fuel system AFR control.

Figures 12 shows further data from the EPA / CARB data bases regarding production variation and product deterioration for various non catalyst Class 1 Phase 2 products. The EPA and CARB have carried out several very in-depth studies (3, 4, 5) to determine if the proposed Phase 3 limits can be reasonable attained by SORE. Also, they recommend possible minor modifications to address production and durability issues. The two notable reports are that by South West Research (3) for CARB and the EPA Technical Study (4).

Figure 10 shows the production variation in HC+NOx emissions for a selection of new Honda and Briggs and Stratton Class1 Phase 2 engines after a 10 hour break-in period. The Honda engines show a production variation of around 10 to 30% compared to the 16.1 g/kWhr limit. The Briggs and Stratton engines show a slightly higher production 6

20076559 (JSAE) 2007-32-0059 (SAE)

Figure 12: Production Variation and Product Deterioration HC+NOx Emissions Data for a Range of Non Catalyst Class 1 Phase 2 Engines (Source EPA / CARB)

Figure 13: EPA Conclusions for Phase 3 limits (4) The South West Research Report (3) presents a significant amount of test data for several utility engines and concludes that Phase 3 emissions limits can be met with an HC+NOx reduction of 50 to 70% from the Phase 2 Class 1 and Class 2 engines by using catalytic after-treatment, as concluded in the EPA report.

Some of these studies and their released statements and conclusions also evaluated the health and safety risks in attaining Phase 3 limits. These reports do not really look into base engine design changes to give reductions in HC emissions but look more towards durability related HC emissions issues and tend to concentrate more on aftertreatment solutions.

The report also concludes that the catalytic after-treatment systems were durable and that stock “rich” carburettor calibration was suitable for purpose. In order to generate sufficient oxidation activity in the catalytic converter, a passive secondary air injection (SAI) was used. Both CARB and EPA (3, 4) discount the use of a reed valve pulse

Both the EPA and CARB have concluded (3,4) that the CARB Phase 3 limits and the proposed EPA Phase 3 limits can be reasonably met using rich operation in conjunction with catalyst based technology without incremental increased risk of fire or of skin burns. 7

20076559 (JSAE) 2007-32-0059 (SAE) air system due to back pressure in the exhaust. However Ricardo UK Ltd. has developed a design methodology (6) for such pulse air system specifically for motorcycles that could potentially work for SORE. The engine cooling circuits were redesigned and the muffler provided with improved shielding to maintain or limit muffler skin temperature. It should be noted as show in Figure 12 that not all the engines were able to meet the projected target emissions at the end of their useful lives. This report clearly shows that, with a durable engine, the Phase 3 limits are potentially achievable. The EPA reports (4, 5) backup the CARB (3) findings as is shown in Figure 13. The EPA (4) catalyst and non catalyst exhaust outlet temperature trends shown in Figure 14 also support the CARB findings in that for Phase 3 catalyst muffler temperatures can be controlled or limited with little or no temperature rise compared to the standard muffler.

Figure 14: EPA Exhaust Muffler Temperature Test Data (4) for Phase 3 Compliant Engine with a Catalyst Compared to Phase 2 Engines without a Catalyst

The findings of the CARB and EPA studies show that in general catalytic after-treatment with SAI applied to a “durable” engine and fuel system with a stock “rich” carburettor calibration can meet Phase 3 limits and that muffler skin temperature can be controlled or limited by muffler and cooling circuit redesign. Clearly with “rich” carburettor calibrations there are high CO emissions in the exhaust gas which are preferentially converted before HC Emissions. Given this fact, the amount of oxidation activity in the catalyst must be limited to control or limit muffler skin temperature. Operating an engine rich without after-treatment results in lower engine-out temperatures due to the effect of excess fuel cooling in the combustion process and in the muffler. However, with a catalyst present, operating rich clearly provides the catalyst with additional oxidation fuel that, if oxidised, will produce more heat than for an engine operating lean. There is a trade off for a target level of tail pipe emissions. When operating rich with added air, there are higher engine out emissions, higher oxidation activity and higher temperature rise in the catalyst but engine out temperatures are lower. When leaner, there are lower engine out emissions, lower oxidation activity and lower temperature rise in the catalyst but engine out temperatures are higher. Also, if the engine is operating rich of stoichiometric then NOx conversion can take place in the catalyst. However if SAI is used to provide oxygen for catalytic oxidation, then in these “lean” exhaust conditions little or no NOx conversion can take place. So increases in NOx with leaner but still rich operation with SAI are important for the cycle emissions of HC+NOx.

75% Load 50% Load 25% Load 10% Load

40

HC+NOx (g/kWhr)

35 30 25 20 15 10 5 0 10

12

14

16

18

20

AFR (Spindt)

Figure 15, the general trends in HC and NOx with AFR for a Honda GX160 SORE engine at each of the J1088 part load test points operating with the TCT fuel system HC+NOx emissions and the reduction of this forms the basic technical challenge for SORE in meeting phase 3 limits. If engines could be run leaner at part load, where the engines are not thermally limited, then with much lower engine out CO and HC emissions in the exhaust gas, the HC emissions could be much further reduced without potentially exceeding muffler skin temperature limits. In addition the fuel consumption would be reduced reducing global warming CO2 emissions. Currently, there are no regulations regarding muffler skin temperature limits but they are currently being looked into by the Outdoor Power equipment Institute and an ANSI standard is due soon. Also, with less CO emissions in the exhaust gas a smaller and therefore lower cost catalyst could be used.

It is important to note that HC emissions are mainly engine design related in terms of their source and that NOx emissions are combustion related. This is a key statement as improvements to the combustion process in terms of improving its efficiency generally increase NOx and improvements to the base engine generally reduce HC. Here is the paradox, more efficient better designed engines produce lower CO2 emissions and lower HC emissions but much higher NOx emissions. The end result is in general higher 8

20076559 (JSAE) 2007-32-0059 (SAE)

Ideal for Min Engine Out HC+NOx 18

Standard Carburettor

AFR

16

14

12

10

8 0%

25%

50%

75%

100%

Engine load / J1088 Load Mode Point

Figure 16 Optimised AFR for Minimium HC+NOx on J1088 Cycle Weighted J1088 Modal HC+NOx Emission in g/hr for Ideal AFR and Stock Carburettor AFR Settings (Honda GX160 test Data Crank vent Disconnected) 5 Stock Carburettor 4.5

Weighted HC+NOx Emissions (g/hr)

Figure 17 shows that if this particular, phase-2 engine is operated with the ideal AFR curve over the J1088 cycle then the weighted engine-out HC+NOx and CO emissions are both reduced by around 30%. This is a significant reduction in emissions and clearly shows the potential of a fuel system with good AFR control and improved combustion stability (12). Alternatively, with stoichiometric operation at the 25% and 10% mode points reduces the engine-out HC+NOx and CO emissions reductions over the stock carburettor by 15 and 28 % respectively

Honda GX 160 4-Stroke Generator 20

Ideal AFR for Minimum HC+NOx 4 Sum of weighted 100, 75, 50, 25 and 10% HC+NOx Emissions:

3.5

Stock Carburettor = 16.6 g/hr Ideal AFR = 11.9 g/hr Reduction = 27.8 %

3 2.5 2 1.5 1 0.5 0 100

75

50

25

10

J1088 Mode Point

Weighted J1088 Modal CO Emission in g/hr for Ideal AFR and Stock Carburettor AFR Settings (Honda GX160 test Data, Crank Vent Disconnected) 250 Stock Carburettor

Weighted CO Emissions (g/hr)

Operating leaner reduces HC emissions but produces higher NOx emissions and with legislation using the sum of the HC plus NOx emissions, this is an important point. The trend for HC and NOx emissions with AFR for air cooled SORE engines is very dependent on the operating load of the engine and therefore is clearly important for the J1088 cycle mode points. Figure 15 shows the general trends in HC and NOx with AFR for a Honda GX160 SORE engine at each of the J1088 part load test points operating with a new design of fuel system, as discussed by Glover et al. (12). As shown above but discussed in detail in (12), this system allows for stable engine operation at much leaner AFR. At the 75% and 50% J1088 mode points, leaner operation results in higher HC+NOx emissions due to the NOx increasing faster than HC decreases. The trend flattens with the lower loads until at 25% and 10% load, in the case of this particular engine, the trend actually reverses. Here HC+NOx decreases with leaner operation for this particular engine. The HC+NOx at these lows loads is effectively an HC trend in nature as HC dominates. Clearly once the lean limit of combustion is reached as can be seen for 10% load at around 16.5 to 1 AFR the HC+NOx emissions increase again due to incomplete combustion or misfire. This shows that, for this Honda GX160 engine, minimum emissions on the J1088 cycle will be achieved with optimised “rich” operation at 100%, 75 and 50% load combined with “leaner” operation at 25% and 10% load. Figure 16 depicts this “ideal” trend together with the stock or conventional carburettor trend.

If we now consider the strategy of improving the base engine with, for example, improved air motion and leaner operation, the HC+NOx trend worsens on the rich side of stoichiometry. This is due to the fact that the NOx rises faster than the HC decreases as the mixture is made leaner. Figure 18, courtesy of Ricardo UK Ltd, is a mixture loop on a water cooled automotive 4 cylinder gasoline engine at 1500 rpm 1.5 bar BMEP with high and low air motion (tumble) and shows this trend. Motorcycles have faced more severe legislation than SORE over the past few years. If we consider a carburetted, Euro-2, SOHC, air-cooled 150cc motorcycle engine as being representative of what an “emissions developed” base SORE could become, then the HC+NOx trends would be as shown in Figure 19. These motorcycle engines are still low cost designs but will mostly likely have reduced engine-out HC emissions (from crevices, valve sealing, etc) and improved combustion stability from improved air motion. However, the trends in Figure 19 show that these also have high engine-out

Ideal AFR for Minimum HC+NOx 200 Sum of weighted 100, 75, 50, 25 and 10% CO Emissions: Stock Carburettor = 651 g/hr Ideal AFR = 453 g/hr Reduction = 30 %

150

100

50

0 100

75

50

25

10

J1088 Mode Point

Figure 17 Weighted J1088 Modal Emissions for Optimised AFR NOx at light load. The trend is no longer HC dominated at light load but is NOx dominated. This approach works well with advanced after-treatment systems, such as three-way catalysts, and more complex fuel systems, such as CV 9

20076559 (JSAE) 2007-32-0059 (SAE) 40 HC+NOx Emissions (g/kWhr)

35 H C +N O x

Suzuki 150

30

Low Tumble Engine High Tumble engine

30 25 20 15 12

14

16

18

20

Daelim 125

28

Yamaha 150 26

Suzuki 150 Daelim 125

24

Yamaha 150

22 20 18 16 14

1.5 bar BEMP

12

2.5 bar BMEP

10 12

AFR

13

14

15

16

17

AFR

Figure 19 Emissions trends with AFR for three single cylinder air cooled motorcycle engines at light load HC+NOx emissions (g/kWhr)

Figure 18, courtesy of Ricardo UK Ltd, of a mixture loop on a water cooled automotive 4 cylinder gasoline engine at 1500 rpm 1.5 bar BMEP with high and low air motion (tumble). carburettors but this may be too costly a strategy for SORE. Thus, by improving the efficiency of the engine and reducing fuel consumption and carbon emissions, the SORE engines could have worse HC+NOx emissions, especially at higher loads. For each SORE J1088 mode point there is an optimum AFR for minimum HC+NOx which will require AFR control with load. If we look to the historical development of automotive engines, then the trend shows generally leaner operation with EFI providing AFR control and the use of catalytic aftertreatment and EGR to reduce emissions. The majority of automotive engines today have very robust combustion systems facilitating operation at Lambda 1 with good response and using EGR to reduce engine out NOx and three way catalytic conversion to reduce HC, CO and NOx. With the recent development of lean NOX after-treatment, automotive engine may soon be operating leaner than stoichiometric with effective three way after-treatment emissions control.

30

Suzuki 150 Daelim 125 Yamaha 150

25 20 15 10 5 0 0

2

4

6

8

10

% EGR

Figure 20 : Emissions trends with EGR for three single cylinder air cooled motorcycle engines at light load Possible Strategies for Meeting EPA Phase 3

Clearly, with a more robust combustion system, SORE could use EGR (internal or external) to effectively reduce the NOx emissions without increasing the HC emissions. This would improve the HC+NOx trend giving a decreasing trend at leaner operation and high load. If SORE are to operate leaner for improved efficiency and more effective after-treatment then, for these air-cooled hot-running engines that produce high NOx, EGR is potentially a key emissions control strategy. This, together with better AFR control and more robust combustions systems, gives a better overall emissions strategy. Figure 20 shows the effect of EGR on HC+NOx emissions at a medium to light load point on several motorcycle engines which would be representative of a developed SORE.

1.

2.

3. 10

CARB/EPA strategy uses a durable phase 2 engine with improved production tolerances, secondary air injection (SAI) and a limited oxidation catalyst. This does meet the limits but with the additional cost of aftertreatment. No further reductions are possible due to temperature limitations in the catalyst. Durable phase 2 engine as CARB/ EPA strategy but with optimised AFR control with load for minimum engineout HC+NOx but with no aftertreatment. This should come very close for some engine and so is a possibility for those engine but is not guaranteed. As strategy #2 but with SAI and limited oxidation catalyst. This meets the phase 3 limits and has the

20076559 (JSAE) 2007-32-0059 (SAE)

4.

5.

6.

7.

strategy. Clearly the cost of these alternatives is of key importance.

advantage of using a smaller catalyst with less conversion of emissions giving a more durable aftertreatment package. As strategy #2 but with lambda-1 operation at part load and a three-way catalyst. This has the potential to achieve “Blue Skies” limits without excessive cost. As strategy #2 but with reduced engine HC sources to give lower engine-out HC+NOx. This has the potential for 10% to 20% lower emissions that strategy #2. As strategy #5 but with SAI and limited oxidation catalyst. This has lower emissions than either strategy #5 or strategy #3 giving good potential to meet phase 3 with a reasonable to cover for production tolerance variations. As strategy #5 but with lambda-1 operation at part load and a three-way catalyst. This gives lower emissions that strategy #6.

References (1) CARB WEB Site : www.arb.ca.gov (2) EPA WEB Site : www.epa.gov (3) Durability of Low Emissions Small Off Road Engines SwRI 08.05734 (4) EPA Technical Study on the Safety of Emissions Controls for Nonroad Spark Ignition Engines less than 50hp – EPA420-R-06-006, March 2006 (5) EPA-OTAQ-ASD November 22, 2004 / Fall 2004 (6) Ricardo UK Ltd. System for Supplying Secondary Air in the Exhaust System of an Internal Combustion Engine, International PCT Patent number WO 02/064955 A1. (7) Ricardo UK Ltd. Fjolbelndir TCT Assessment Report Dec 06 (8) Moesner et al, Emissions and Performance Potential of a Small Stratified Charge Two-Stroke Engine using Reed Valves. SAE Paper No, 2006-32-0058, SETC 2006. (9) M Bergman and J Berneklev, A Novel Method of Tuning a Stratified Scavenged Two-Stroke Engine. SAE Paper No 2006-32-0055, SETC 2006. (10) R Gustafsson, A Practical Application to Reduce Exhaust Emissions on a Two-Stroke Engine with Tuned Exhaust Pipe. SAE Paper No 2006-32-0054, SETC 2006. (11) M Bergman and R Gustafsson, Emissions and Performance Evaluation of a 25cc Stratified Scavenged Two-Stroke Engine. SAE Paper No 2003-32-0047, SETC 2003. (12) S Glover, R Douglas and K Omarsson, The Potential of a New Type of Carburettor to Assist SORE in Meeting EPA/CARB Phase 3 Legislation. SAE Paper 2007-320015, SETC 2007.

All of the above strategies can be employed with other engine improvements such as increased air motion or EGR which would clearly effect engine-out NOx emissions. What is apparent is that to go beyond the current EPA/CARB strategy requires good/optimised AFR control matched to an effective after-treatment strategy. This mirrors the development of automotive and motorcycle engines where EFI is dominate in all areas except for the sub 150 cc motorcycles where the cost is an issue. Clearly cost is also of paramount importance for SORE, hence the requirement of a cost effective means of optimising AFR. Glover et al () discuss such a carburetted fuel system that not only provides AFR control but also gives improved combustion robustness and a lower cyclic variation in delivered AFR which is essential for more effective catalytic after-treatment, especially three way conversion.

Conclusions The question of why rich fuel system calibrations has been asked and has hopefully been answered in the above discussion. SORE engines operate rich because it tends to give lower HC+NOx emissions and they do not have AFR control with load. Also, the poor air-fuel mixture provided to the engine by the conventional fuel systems does not allow leaner operation due to poor combustion robustness. If we also look in terms of costs, it is potentially lower cost to run rich and produce less HC+NOx rather than have a more efficient engine but costlier design, producing higher engine out HC+NOx. This latter approach would require more expensive emissions control technology to reach lower HC+NOx limits. This unfortunate trend in HC+NOx emissions and perceived increased costs associated with more efficient engines and leaner operation leads to rich operation, high CO emissions, high fuel consumption and poor aftertreatment system efficiency. This is not very promising for the future with regard to global warming and the general efficiency in fuel consumption demands. Is there a cost effective alternative emissions control strategy for SORE? The authors believe that using the TCT fuel system technology provides effective technical alternatives to the EAP/CARB 11