NASA looks beyond batteries for space exploration
Syngenta in cellulosic fuel collaboration to increase ethanol plant productivity

ExxonMobil, Corning and Toyota develop onboard membrane system to separate gasoline into octane fractions to optimize engine efficiency and performance

ExxonMobil, Corning and Toyota have collaborated to develop an Onboard Separation System (OBS) to optimize gasoline engine efficiency and performance. OBS is a membrane-based process that separates gasoline into higher and lower octane fractions—essentially creating a dual fuel system from a single base fuel—allowing optimal use of fuel components based on engine requirements. The system, say the researchers, offers the potential to exploit most of the benefits of operating on premium high octane fuel while using less expensive regular grade more effectively.

In a paper on the OBS presented at the SAE 2014 World Congress, the researchers suggested that potential applications include downsizing to increase fuel economy by ∼10% while maintaining performance, and or using OBS with turbocharging to improve performance and knock resistance.

The novel polymer-ceramic composite monolith membrane has been demonstrated to be stable to E10 gasoline, and typically provides 20% yield of ∼100 RON product when using regular unleaded (RUL) 92 RON gasoline as the base fuel. The OBS system uses waste exhaust energy to effect the fuel separation and provides a simple and reliable means for managing the separated fuels. OBS has been demonstrated using several generations of dual-fuel test vehicles.

Background. Earlier work by ExxonMobil and Toyota on the effect of fuel octane number and compression on engine performance in a 2.0-liter high compression ratio (13:1), spark-ignition direct-injection (SIDI) engine found that under low-load, low-speed stratified conditions, a very low-octane (84 RON), aliphatic fuel resulted in higher efficiency and lower HC than regular 92 RON gasoline. They found spark-induced compression ignition (SICI) occurring with the low RON fuel.

Studies at wide open throttle condition, on the other hand, found that a higher octane, highly aromatic fuel (RON 103, 60% toluene) provided significant torque benefits compared to pure isooctane (RON=100).

With data points gathered from further testing at several load points under both stratified, lean-burn conditions and homogeneous stoichiometric conditions using a range of fuels from RON 84 to 103, the researchers developed an “Octane Requirements Map.” General conclusions were that:

  • At low loads in stratified operation, lower RON fuels give higher brake efficiency.

  • At intermediate loads and stoichiometric operation, intermediate RON fuels delivered maximum efficiency.

  • At higher loads and stoichiometric operation, higher RON fuels gave the best efficiency and were required to avoid knock.

Applying this to the LA-4 urban drive cycle, they found that optimal engine efficiency required only modest amounts of >96 RON fuel, even with the 13:1 compression ratio. Most of the drive cycle required <92 RON for maximum efficiency. The results suggested that under some conditions <84 RON fuel might be acceptable, but 84 RON was the lowest ExxonMobil and Toyota tested.

Applying optimal fuel octane throughout the LA-4 drive cycle would result in an overall potential 15.5% increase in fuel economy, according to their work; 8.5% from using optimal octane in the base engine, and another 7% potential improvement due to downsizing enabled by the increase in torque available.

These results were intriguing. This “Octane Requirement Map” indicated that nearly 80% of the fuel requirement using a very high compression engine might be met with conventional regular gasoline having RON 92 or less. When under load, about 20% premium gasoline of RON 97 or greater would be required for optimal performance. Notably, even with premium fuel it would not be possible to reach MBT [maximum brake torque] under all conditions at 13 compression ratio. However, concentrating the higher octane components in gasoline, such as aromatics and/or ethanol (in some markets), might provide the higher RON and ignition characteristics required.

This suggested potential for a dual (or multi) fuel strategy to realize significant efficiency and emissions benefits. Recognizing the difficulties of providing multiple fuels to a vehicle, and the desirability of providing a higher RON fuel than commonly available, the thought of separating gasoline onboard was raised.

—Partridge et al.

The OBS system. With aromatics among the highest octane components present in all gasoline, the team developed the concept of separating the gasoline into high and low octane fractions by using a pervaporation process—a selective membrane process that separates a binary or multicomponent liquid mixture by partial vaporization through a dense non-porous membrane.

The OBS system comprises a membrane module; heat pipe (exhaust to fuel heat exchanger); integrated heat exchanger (fuel coolers); a modified fuel tank; and a dual fuel engine.

The researchers have gone through several polymer formulations for the membrane, ultimately developing cross-linked polyether-amine/epoxy polymer formulations that are stable to ethanol in gasoline while separating ethanol and aromatics from the lower octane aliphatic components of gasoline. The monolith structure is similar to an automotive emissions control substrate.

Under the pervaporation process, high octane aromatics and ethanol are preferentially absorbed by the membrane polymer. A vacuum is applied to the opposite side, pulling the concentrated aromatics and ethanol as vapor through the porous membrane support. Vapors are condensed by cooling, and the HiRON fuel is stored until needed.

Dual fuel engines were Toyota D-4(S) 1AZ-FSE 2-liter engines, with a compression ratio boosted from 9.8 to 13:1. Direct injection is maintained for the lower octane fuels; port fuel injection (PFI) is added for the high octane fraction. The original pistons were replaced with shallow cavity pistons with an aim for lower HC; variable valve timing was not used.

The separation process requires heating and partially vaporizing about 0.5 to 3 g/s gasoline to about 140-160 ˚C (using exhaust heat) at 400 kPa—about the same pressure as the fuel rail. Fuel rate to the membrane is similar to the average fuel rate, about 1 g/s. A variable fuel rate is used to control temperature in response to exhaust energy.

The resulting vapor-liquid fuel mixture is separated into a high octane (HiRON) permeate and lower octane (LoRON) retentate by the membrane. LoRON is provided to the direct injection system; excess LoRON returns to the tank. HiRON product is stored in a small 2- to 4-liter HiRON tank located within the main fuel tank. With a shared vapor space, no additional fuel vapor management is required.

The team used a Toyota RAV-4 as the primary OBS test vehicle, using several versions of dual fuel engines; as the program proceeded, the system was greatly simplified and miniaturized.

In the most recent tests, the team also used a Camry, with an under-floor mounted OBS system; ExxonMobil Research-Corning polymer ceramic composite membrane, heat pipe and integrated fuel coolers, working with a dual-fuel stoichiometric 2.4-liter engine.

Bench and in-vehicle testing showed the potential of increasing fuel economy by about 5%, with a torque increase of 8-10%, or potentially 8% fuel economy improvement at constant performance.

Issues. There are a variety of issues that need to be considered, the team pointed out in their study.

  • The efficiency and performance gains are a direct function of the octane number and fraction of the HiRON produced; these is turn depend on the composition of the regular gasoline fuel used, which can vary widely in different markets.

  • Under high load driving, the consumption of HiRON would exceed the present ability of the OBS system to supply it. However, they note, OBS could be applied to premium gasoline blends as well.

  • The ethanol blend level in the US could open up additional opportunities for OBS. Under optimal conditions, OBS has produced fuels with more than 35% ethanol content and 102 RON from E10. The cooling effects of ethanol would enable higher boost pressures without knock and hence greater efficiencies.

  • OBS could also be applied to flex fuel vehicles.

  • Lower startup emissions are also possible with the higher volatility LoRON fuels.

In the face of more stringent fuel economy standards in the future, concepts that build on existing fuel and vehicle platforms, such as the OBS system, will likely garner increased interest from auto manufacturers. The OBS system will, of course, require further development to be ready for commercial consideration.

—Partridge et al.

The researchers and their companies have filed for patents on the work.




Clever. Hopefully this doesn't add too much in terms of cost (eyeballing the components, it shouldn't add more than $500).


Looks good, im interested that they continue to experiment that on the road. If ever they need a driver to make the experiments I put my name on the list. Im tired of always driving my old boring car.

Roger Pham

I think that an alcohol storage tank and an additional alcohol-compatible fuel pump to supply additional alcohol to the fuel rail would be a lot cheaper than the OBS device.
So, high compression can be used to increase efficiency. 87-octane fuel is used during partial load while additional alcohol will be injected during high load. This is nothing new but was used in WWII aircraft piston engines for WEP (War Emergency Power) when water methanol injection was done to allowing higher manifold pressure.

If this catches on, perhaps the gas station can supply gasoline w/out alcohol in one pump, and another pump just for ethanol to allow for mixing-in-the-run later, depending on engine load.
This will save a lot of money and increase efficiency, because anhydrous ethanol is required before it can be mixed with the gasoline to produce E-10 gasoline, and producing anhydrous ethanol (removal of the last trace of water) is expensive and energy consuming, because ethanol is very hygroscopic (affinity for water). Instead, the ethanol used on a dedicate tank contains a small amount of water that cannot be easily separated, and thus, the ethanol can be produced more cheaply and efficiently.


With Open Fuel Standard (OFS) cars, they could run M85 and have variable super charge or turbo charge to get performance and mileage. Methanol can be made from many feed stocks, for that matter the ethanol in E85 can by synthesized as well.

If we can get efficient spark ignited engines running series/parallel EREV designs like the Volt, every mid sized car could get 50 MPG, pollute less and use less imported oil. It is a matter of will and not technology, it can be done it we want to do it.

Nick Lyons

Under high load driving, the consumption of HiRON would exceed the present ability of the OBS system to supply it.

It seems the car will need the ability to run its high-compression engine on low octane fuel under such conditions (towing trailer up a long grade, e.g.). I suppose with variable duration intake valves you could close them late (as in Atkinson cycle) to effectively reduce the compression ratio and live with the reduced power and torque.

Roger Pham

Good point, SJC.

Alternatively, the gas station can just sell gasoline w/out pre-mixed alcohol in order for the user to mix it later with either hydrous ethanol or hydrous methanol, on the run basis, depending on engine load and depending on the season. The use of hydrous alcohol instead of anhydrous ethanol as is the current practice in gasoline will save money and energy, since anhydrous ethanol takes extra energy and cost to produce. The saving in alcohol fuel cost will pay for the modest additional hardware and infrastructure costs of having to adapt to two different fuels. The gain in fuel economy as the result of running on high compression ratio will be pure bonus!

So, near-future cars may have two fuel tanks, one for straight gasoline and the other tank for hydrous alcohol of either ethanol or methanol.
Summer may require higher percentage of alcohol mixed in, due to higher smog potential, while other season may not, except for at high load, alcohol fuel will be injected together with gasoline in various proportions, depending on engine load and ambient condition such as temperature and humidity.

The beauty of having two separate fuel tanks is that the car is essentially a FFV (flexible fuel vehicle) that can operate on alcohol fuel alone, or largely alcohol, should there be a shortage of petroleum, or lack of availability of petroleum fuel in certain region of the world.


Blender pumps are a proven technology, it costs very little to upgrade existing pumps, just an extra tank for the mixed alcohols. They can use the tank for mid grade, no car manufacturer specifies 89 octane anyway.

GM and other car makers have said they would produce OFS vehicles at an extra cost of about $100 each. They have no problem doing this and said they would absorb the cost. It is politics and special interests that have delayed OFS for more than 6 years now.


Roger, you just reinvented the Ford-MIT engine of 2006:

And indeed, if we're going to make ethanol from corn, we should minimize the energy we spend on it.  180-proof "wet" ethanol in a direct-injection system would be a better knock suppressor than anhydrous.

Verify your Comment

Previewing your Comment

This is only a preview. Your comment has not yet been posted.

Your comment could not be posted. Error type:
Your comment has been posted. Post another comment

The letters and numbers you entered did not match the image. Please try again.

As a final step before posting your comment, enter the letters and numbers you see in the image below. This prevents automated programs from posting comments.

Having trouble reading this image? View an alternate.


Post a comment

Your Information

(Name is required. Email address will not be displayed with the comment.)