Study finds highly-turbocharged alcohol-fueled DISI engines could be more efficient than diesels
01 November 2010
Highly turbocharged alcohol-fueled direct-injection spark-ignition (DISI) engines operated at a high compression ratio could be as or more efficient than diesel engines while also providing advantages of lower vehicle cost, lower emissions and higher power, according to a recent modeling study by Leslie Bromberg and Daniel Cohn at MIT.
In their paper, presented at the SAE 2010 Powertrains Fuels & Lubricants Meeting last week, the two investigated the potential of ethanol and methanol in heavy duty (HD), long-haul applications, examining the issues of infrastructure, engine performance, and briefly, emissions and cost implications.
Alcohol-based transportation fuels from a wide range of biomass feedstocks including agricultural, forestry, municipal and industrial waste and specially grown crops and trees can potentially offer a substantial substitute for oil derived fuel. Alcohol fuels can also be produced from natural gas or coal. Use of ethanol, methanol or alcohol mixtures could be particularly attractive for heavy duty vehicles because of the need for cleaner engines than present diesel engines and a less demanding infrastructure requirement than that for light duty vehicles.
Small, very high power density, spark ignition engines which are fueled with ethanol methanol or mixed alcohols can be used as a substitute of heavy duty diesel engines, with higher engine thermal efficiency and much reduced size and weight. In this manner a 3.6 liter engine could potentially be used to replace a diesel engine with a displacement as high as 11 liters. These extreme downsizing indicates the potential of a SI, knock-free engine, operating at the same peak pressure as the diesel engine and higher engine speeds. However this aggressive downsizing may not be practical because other constraints (durability, exhaust temperatures). More modest downsizing up to 5 liters could offer a practical solution.
—Bromberg and Cohn
Diesel engines have been substantially more efficient than gasoline (spark ignition) engines, due in part to the problem of knock in the SI engines, which limits the compression ratio and the amount of pressure boosting.
With the use of alcohol-based fuels it is possible to reduce much further the tendency of knocking in spark ignited engines...By eliminating the knock constraint, much higher compression ratios can be used. Similarly, turbocharging allows for substantial engine downsizing.
The high knock-free pressure resulting from the use of direct injection in combination with stoichiometric operation with no EGR at high torque and use of high rpm operation makes possible an engine power density and a torque which are each up to 3 times those of a diesel engine having the same piston displacement. A representative number for the torque/liter capability of direct injection alcohol is 200 lb-ft/liter and that for potential engine power density is 200 hp/liter.
In their study, they found that a direct injection alcohol engine downsized by around a factor of two relative to a diesel engine could have an efficiency advantage over the diesel of about 4% for prolonged high torque operation in long haul trucks. Reforming about half the methanol or ethanol would result in capture of about an additional 5% from the exhaust, as a bottoming cycle. The lean operation could result in an additional 5% improvement in efficiency, they said. There could also be a small increase in efficiency with up-speeding and further downsizing.
They notes that their calculations did include the fuel penalty expense needed for the diesel engine aftertreatment, which when included would further increase the comparative advantage of the spark ignited engine by a few percentage points. Their calculations suggest that for long haul trucks, a 5-liter direct injection ethanol engine could be used to replace a diesel engine of 11-liter displacement and a 7-liter direct injection ethanol engine could replace a diesel engine of 15-liters displacement.
The high efficiency downsized alcohol engine approach described in this document could also be used in light duty vehicles.
—Bromberg and Cohn
Fuel consumption and range. Bromberg and Cohn note that M85 (around 85% methanol, 15% gasoline by volume) has roughly 60% of the range of diesel fuel for a given fuel tank size. However, the weight reduction from the highly downsized engines could essentially compensate for the increased alcohol tank and fuel weight needed to obtain the same range as a diesel vehicle, they suggested.
They have not yet evaluated the potential use of higher alcohols, which are more energy dense than methanol or ethanol.
Sandia researchers, for example, report (also at the SAE PFL meeting) that isopentanol—a five-carbon, long-chain alcohol—has superior physiochemical properties compared to ethanol and very similar HCCI combustion properties to gasoline. (Earlier post.)
Resources
Leslie Bromberg and Daniel Cohn (2010) Alcohol Fueled Heavy Duty Vehicles Using Clean, High Efficiency Engines. (SAE 2010-01-2199)
Wow-would somebody please get to work on a prototype!! They fail to mention another factor, the inherent inefficiency of diesel because its a distillate fuel (ie produced by distillation which consumes energy) whereas gasoline is produced by catalytic cracking.
Posted by: nordic | 01 November 2010 at 11:59 AM
what about MTBF of the downsized engine? Long Haul Class 8 low/medium speed large displacement engines go for a million miles TBO. They run at 1500-2000 rpm at peak torque, these engines will be screaming at 4000+RPM at full torque.
Cylinder liner wear rates are directly proportional to peak torque and cylinder sweep velocity (rpm) operating ANY piston engine at near peak torque for most of the duty cycle will greatly increase ring and cylinder liner wear rates. This is physics I could go into a long spat about how rotational torque is only generated by side force of the piston skirt acting on the cylinder liner wall but anyone who has taken freshman physics should know this already and if not get a freshman physic text book.
In physics there is no free lunch, drastically downsizing(increasing torque produced per cubic inch displacement) and upspeeding increasing RPM since work is a function of force(torque)*distance which is rotatinal in this case RPM an engine will increase the wear rates at the log of the piston skirt sweep rate over torque density per square centimeter of piston skirt area this is unavoidable.
A perfect example is small petrol generators vs low speed diesel generators of similar size outputs a 10kw petrol motor genset is screaming at 3600 rpm and has a TBO of less than 2500 hrs a 10kw low speed diesel hums along at 900rpm and have TBO of 40000hrs+ again physics cannot be monkeyed with.
Posted by: TXGeologist | 01 November 2010 at 04:06 PM
This repeats the results of the MIT ultraturbocharged engine with direct ethanol injection, except that this study does not propose using conventional port-injected fuel for most operations.
Wear might be an issue for medium and heavy-duty applications, but most passenger vehicles spend so much of their time at very light loads that it's unlikely that engine wear would be an issue in the service life of most vehicles. If worse comes to worst, borrow the idea of the cross-head from steam engines to take the side loads.
Posted by: Engineer-Poet | 01 November 2010 at 05:34 PM
Please note that when these guys talk of greater efficiency they mean more power per engine size NOT more MPG: In fact they're counting on "the weight reduction from the highly downsized engines could essentially compensate for the increased alcohol tank and fuel weight needed to obtain the same range as a diesel vehicle..."
We need to use less liquid fuel not more.
Posted by: ai_vin | 01 November 2010 at 06:44 PM
"Small, very high power density, spark ignition engines which are fueled with ethanol methanol or mixed alcohols can be used as a substitute of heavy duty diesel engines, with higher engine thermal efficiency and much reduced size and weight."
They are talking about heavy duty applications which as TXGeologist points out will have MTBF problems. But it may be that the weight savings and lower materials costs - plus lower cost of domestic produced alcohol - will offset higher maintenance.
Interesting finding. And yes, let's put a prototype in the field.
Posted by: Reel$$ | 01 November 2010 at 06:51 PM
What
They are only finding that now. DISI has been commercial for years as has the knowledge that ethanol at higher compression could be more efficient. It is a no brainer to put the two together and say it could be more efficient.
Posted by: Donough Shanahan | 02 November 2010 at 01:34 AM
EP- Cross heads good point steam has enormous torque loads even from 0 rpm wow thats a great idea large low speed marine Diesels still use them for that very reason to off load the cylinder wall there is no reason a scaled down version could not be put in a medium speed heavy duty ICE the only sacrifice is overall height of the engine but for medium/heavy duty whats 20 cm or so in overall engine height. So lets see a working test engine anything that keeps jobs and money in this country rather than sending it to people who want to exterminate us is a great idea.
Posted by: TXGeologist | 02 November 2010 at 01:37 AM
I am not so sure the longevity is going to be a problem. Conventional diesel engines use cast iron cylinder liners. Today's best cylinders are Nickel/Silicon Carbide or Nikasil (NiCom). It's not ideal for diesels and engines that have sulfur in the fuel, however, it does provide about 10 times the wear life when compared to conventional iron cylinders. And at least 5 times the wear life of a nitrided steel cylinder.
Alcohol fueled engines do very well with Nikasil bores. They are commonly used in racing engines, including formula 1 engines. The very high specific output goes hand in hand with long life cylinder coatings/platings.
Posted by: cujet | 02 November 2010 at 05:50 AM
@EP and TXGeologist
IIRC, on a recent episode of Dirty Jobs they showed a rebuild of a train's diesel engine and it had a cross head. Only downsides would be increased reciprocating mass and packaging, as you noted TX.
In a passenger vehicle, such an engine could be made in a flat six configuration and mounted in the rear such that the engine block and cylinder head become structural components. Dreaming here... maybe for a super car?
Posted by: GreenPlease | 02 November 2010 at 08:14 AM
In particular I'd like to see how alcohol fuelled piston engines stack up against micro-turbine HEV range extenders
http://www.gizmag.com/gas-turbine-electric-vehicle-project-funding/13995/
Posted by: Aussie | 02 November 2010 at 02:32 PM
The piston engine will have much higher efficiency than the turbine at reasonable power ratings; the turbine can't get close to the same pressure ratio, and constant-pressure combustion has much greater entropy losses than constant-volume combustion.
Posted by: Engineer-Poet | 02 November 2010 at 03:13 PM
A downsized and turbo-charged piston engine basically retains the thermal efficiency of a piston engine while embracing the volumetric efficiency/power density of a turbine.
Posted by: GreenPlease | 02 November 2010 at 04:56 PM
This conversation reminds me of various vis Honda's assymetric piston designs. These include the removal of much skirt from the non thrust side to reduce friction and add support to the heavily loaded side. Offsetting of the gudgeon pin in combination with offsetting the cylinder bore center to provide better utilisation of the expanding gas characteristics.3
While the examples we have seen to date are probably fuel and application specific, it would seem to me that a similar approach could have a major affect on pistobn side loads.
We would like to see a further ability in the area of multi fuel versatility possibly via the turbulent prechamber spark, piezzo DI, variable compression etc.
I don't seen why heavier bearing areas, more robust conrods and possibly fewer but larger pistons would not be every bit as strong as any industrialised engine in production.
The weight penalty would be marginal and we can observe that the current production DI engines are very slow revving compared to their port injected counterparts.
As for X slides, that sounds a bit too much Casey Jones to me.
Posted by: Arnold | 04 November 2010 at 10:38 PM
Please don't forget, that according to the rigid and real-world (vs. here done studies and assumptions and projections) testing requirements of the Progressive X-Prize the most economic 4-door, 4 passenger vehicle's propulsion system is a 1-cylinder, 250ccm, turbocharged, direct-injected E85 engine - the Edison vehicle. So the authors are definitely on the right track. An inherent advantage in such a vehicle would be: The engine main hardware itself as a massproduced motorcycle format engine would be dirt cheap. So it would be totally economic to rebuilt or replace it many times, which I would prize at around 100-200$ for a liner refinish and oversize piston which could be possible 2-3 times and maybe at 500$ for a complete new cylinder with piston. The car industry with it's unsustainable business model of selling electronic-intensiv, super-expensive to maintain gas suckers should highly aim for such a model. If due to the necessary high-tech lightweight materials, I think they used aluminum tubular frame and carbonfibre body shell, the vehicle could cost 50000$ the potential profits per car would be huge. All C02 requirements would be fullfilled. Additional money could be made on add-on electronics, which would keep/ improve customer satisfaction on this car, while aging. That's the way light aircrafts are built and it's a sustainable business, with the products being built to high profits by the same entreprises since over 50 years and with models that old still flying and being upgradable today.
Posted by: Electronhead | 11 November 2010 at 04:39 AM