GM showcases Beat EV in demo program in New Delhi
US DOE, IEA member countries releasing a total of 60 million barrels of oil from reserves

Ricardo and Univ. of Brighton working on advanced combustion system for heavy-duty vehicles; CoolR features recuperated split-cycle with isothermal compression using cryogenic injection

The concept of the Ricardo Split-Cycle engine. The recuperated engine uses isothermal compression via cryogenic injection to enable significant exhaust to compressed gas heat transfer. Source: Neville Jackson. Click to enlarge.

Ricardo and the University of Brighton will model and evaluate an advanced split-cycle combustion system aimed at substantially reducing the carbon emissions of heavy-duty vehicles. The feasibility study is part-funded by the UK Technology Strategy Board as one of the winning submissions for the recent “Disruptive technologies in low carbon vehicles” competition.

Unlike many previous research projects that have focused on refining existing four-stroke engine technology to reduce fuel consumption and emissions from heavy-duty vehicles, the CoolR project will examine a fundamentally new split cycle combustion concept based on a recuperated split-cycle with isothermal compression using cryogenic injection. (In an isothermal process, the temperature is constant.) Professor Neville Jackson, Ricardo’s Chief Technology & Innovation Officer, had presented an overview of the basics of such an approach at the SAE 2011 High Efficiency IC Engines Symposium in April.

The Ricardo Split-Cycle engine concept incorporates the following:

  • Liquid Nitrogen (LN2) injection during compression to control temperature rise and increase mass.
  • LN2 produced using the engine (efficiency of generation is a key parameter).
  • Recuperator to transfer heat from exhaust gas to compressed air.

Turbine (Brayton cycle) efficiency can be substantially improved through recuperation, Jackson noted. Recuperated Brayton cycle efficiency improvement is a function of exhaust and end of compression temperatures, as well as of compression ratio. However, reciprocating engines with high compression ratios offer less opportunity for recuperation.

Lower compression ratios would provide more opportunities for recuperation, he said, but with lower overall efficiencies. High compression ratios reduce this temperature difference with less scope for recuperation. The key challenge is to maintain high pressure ratios for good simple cycle efficiency and to increase the capacity for recuperation.

Example of a recuperated diesel cycle with high compression ratio and isothermal compression. Source: Neville Jackson. Click to enlarge.

The key to successful recuperation with high pressure ratios is isothermal compression, Jackson said. Further, isothermal compression reduces compressor work; test show about a 17% reduction in energy requirement. However, implementation of practical isothermal compression is challenging.

A recuperated combustion engine transfers exhaust heat to the working gas at the end of compression and at constant volume. It requires a separate compressor and expander—e.g. cross-linked reciprocating piston and cylinders.

The IsoEngine concept. Click to enlarge.

Ricardo has already successfully demonstrated a spilt-cycle isothermal compression engine in static form for power generation purposes. The “IsoEngine” prototype demonstrated by the company for energy utility Innogy in the 1990s used water injection to achieve a thermal efficiency in excess of 60% in comparison with around 43% for a current state-of-the-art on-highway heavy duty diesel engine.

The IsoEngine separated compression and combustion/expansion processes in different cylinders. The charge air was compressed isothermally by spraying a large volume of water into the cylinder during compression to 100 bar. The charge air was heated to 750 °C by the low pressure exhaust gases in the recuperator.

Combustion occurred in the separate cylinder at constant pressure by staged injection of the fuel; work was extracted through expansion of the combustion products and recovered through the crank.

Ricardo has patented an approach to a split-cycle isothermal compression engine with cryogenic injection. Click to enlarge.

While water injection would not be practical for a vehicular implementation, the CoolR concept aims to achieve the same thermodynamic benefits using liquid cryogen injection. Allowing for the energy costs of cryogen production, this would result in a thermal efficiency improvement of around 40%. This is significantly better than that of other promising technologies also currently being researched such as exhaust heat recovery concepts based on thermo-electric generation or the Organic Rankine Cycle which offer improvements of around 10-15%.

Key requirements for such a mobile application would be the ability to handle transient loads and speeds, compact packaging, and fast start and load acceptance, Jackson said in his talk. There might also be consideration of the capability for kinetic energy storage and release.

In his talk, Jackson said that a 2-liter split-cycle engine operating at 50 kW and 40 bar Pmax could utilize more than 18 kW of exhaust heat and deliver indicated efficiency of 60%.

In this one year feasibility project, the partners will carry out a concept study aimed at addressing the fundamental questions that industry will face if such a radically new technology were to be adopted. In doing so, it is intended that a road map be developed to identify the necessary work required to bring the CoolR concept from feasibility to systems prototype and beyond.

The global imperative to reduce the carbon footprint of road transportation is now almost universally accepted. While electrification, hybridization and improvements of the existing internal combustion engine offer a pathway to sustainability for light vehicles, a major problem remains in the heavy duty sector. By fundamentally reviewing the underlying thermodynamics of the internal combustion engine in a manner unseen for many decades, we believe that the CoolR spilt cycle cryogenic injection combustion concept offers the prospect of very significant improvements in thermal efficiency and hence reduced carbon dioxide emissions in the economically crucial heavy vehicle sector.

—Nick Owen, project director for research and collaboration at Ricardo UK


  • Neville Jackson, “Future Reciprocating Combustion Engine Efficiency – How much further can we go?” (SAE 2011 High Efficiency IC Engines Symposium)



Nice but complicated, might work for heavy truck though


Six cycle recovers heat and is simpler.



Roger Pham

LN2 in a separate tank? This requires entirely new infrastructure. LN2 is expensive and requires expensive and energy-intensive refrigeration system.

In an unproven split-cycle engine layout with unproven valve mechanism? Too long R&D process.

Exhaust heat reuperation? More complication and plumbing!

How about this: "Lund Team Shows 57% Thermodynamic Efficiency in a Gasoline-Fueled Heavy-Duty Diesel Engine Using PPC" That's right, well-proven diesel engine design available right now! See:


I agree with you Roger. If efficiency was equal in both concepts, it would be an easy choice. With more efficient turbocharging, adding a turbocompound system and reducing friction, the Lund University engine would be pretty close to the 60% indicated efficiency target mentioned by Ricardo. The Lund University engine would still look like a diesel engine, while the Ricardo engine looks like a chemical factory with all the plumbing associated with a plant of this kind. However, with the reputation of Ricardo as an engine consultant company, the engine they envision is certainly not a joke…

I have not looked at Ricardo´s concept in detail but one feature catches my attention. The efficiency would be achieved at a maximum cylinder pressure of only 40 bar. State-of-art diesel engines today must be designed for pressures well above 200 bar. This illustrates a “wonderful” feature of the cryogenic system. Eventually, I also think that we will have to look at options to reduce cylinder pressure for conventional diesel engines (at a given specific power and torque) without getting negative impact on efficiency. The margin we get by doing this could be used to increase specific power and torque to enable further downsizing, while still further increasing efficiency. Of course, we should keep trying to increase the pressure capability but I am afraid that further progress beyond the level we could extrapolate to in the near future, say 250 bar, will be very difficult. So, on a loner time horizon, Ricardo’s concept could be interesting.

It is interesting to note that also Rudolf Diesel had some ideas about isothermal compression. This was when he tried to create the “ultimate” thermodynamic cycle, i.e. an engine using the Carnot process. As we now know, the Carnot cycle was not a very good idea due to very high cylinder pressures and low specific power but his later “compromise”, the diesel cycle, turned out to be quite successful. Another interesting note is that the first type of combustion system that Rudolf Diesel worked on had some similarities with the PPC from the Lund University. Eventually, Rudolf Diesel did not succeed with that either. If he had, we might not have had conventional otto engines today.

Thomas Pedersen


LIN (liquid nitrogen) is made onboard and let to a buffer/storage tank to smooth out transients.

For large-scale plants, making LIN takes about 0.6 kWh/kg, assuming that pure nitrogen is required.

The sketched process will make liquid air, because oxygen condenses at higher temperature than nitrogen. Besides, it makes no sense to me to prefer nitrogen without oxygen, because:

1) It requires separation of the oxygen, which is energy consuming.

2) It yeilds less output, 23% by mass to be specific.

3) More oxygen in the cylinder improves combustion.

Interesting concept, although rather complex. Maybe relevant for stationary equipment, and/or marine transport. That said, never underestimate the potential of inventing something smart with the prospect of selling it to millions of customers, rather than hundreds (in the case of marine engines) or thousands (in the case of trucks).

Note also on the IsoEngine concept that the water/fuel mass ratio is 1.0, indicating that the vehicle would require the same amount of water as fuel. This is not unrealistic, despite the fact that very pure water is required.

Note also that the shown concept for LIN production will clog from formation of ice and dry ice (solid CO2) in a matter of minutes, which is why air separation units have dedicated water and CO2 removal equipment prior to separation of nitrogen and oxygen.


Definitely is simpler and has more promise than a nuclear powered auto.

Account Deleted

Something fishy here, the liquid nitrogen is introduced to the compression cylinder. There are several issues here; first with the cold nitrogen being introduced, additional compression work need to be done to raise the cylinder temperature above the ignition point. Second,the liquid nitrogen will create cold spots which is not good for combustion.

the exhaust temperature is about 772 Celcius but i cant believe the gas coming out from the recuperator to be as high as 699 Celcius. I havent seen any heat exchanger being this good.

It also doesnt make sense for the turbocharger to return as much as 5.9 bar even though the exhaust gas coming out of the recuperator is about 348 Celcius.

There will be lots of pressure loss from the recuperator so I dont expect the 699 Celcius gas to be high in pressure when it enters the combustion cylinder. Bear in mind, the combustion cylinder is always in high pressure thus to introduce any gas, it has to be first compressed above the cylinder pressure and this requires a lot of compression work which is not free.


With low temperature at ignition, you also get low cylinder pressure (with reference to my previous comment). Ignition is not necessarily a problem since you can ignite both gasoline and diesel fuel at room temperature and ambient pressure by using positive ignition (e.g. spark plug or glow plug). If there is a problem it might lie in poor combustion at these conditions. However "cold spots" in the gases should not be a problem due to fast mixing.

I think you would have to be more specific in your criticism about the problems of turbocharging rather than relying on kind of a gut feeling. With the considerable experience Ricardo has in this area, I would rely on that they are able to calculate this quite easily. Everybody can make small errors and maybe set to ambitious conditions but I do not expect any major errors. Similar comments could be added about the recuperator.

Roger Pham

The LN2 is only injected into the cool compression cylinder, in order to reduce heat build up as the air is compressed. Recall adiabatic compression process requires work input to end up as heat. Preventing heat build up would reduce compression work input significantly. This cooled compressed air will receive recuperated heat from the exhaust gas and get heated up to 699 degrees C before entering the hot power cylinder, thus, only hot air will be in the power cylinder, and no cool spot whatsoever in the hot power cylinder.

You've raised an important consideration. There is no doubt that the cooled exhaust gas will lose a lot of pressure before reaching the supercharger's turbine wheel. However, Isoengine requires less air mass for a given work output than in a comparable diesel engine, such that you can size the centrifugal compressor much smaller than the radial turbine to compensate for the lower temperature hence lower pressure of the exhaust gas in the turbine inlet.

Remember that in a typical turbocharged diesel engine, more exhaust pressure energy is available than is being consumed by the turbocharger for compressing air intake, and this exhaust pressure energy will end up being wasted out of the exhaust. Unless there is a turbocompounder turbine included to harness this extra exhaust pressure energy and to return this power to the engine output shaft, or used to power an electrical generator.

Roger Pham

Thanks, Thomas, for the clarification. Agree with you regarding the complexity inherent in having an onboard LN2 generator. The LO2 should be separated out, since LO2 is high corrosive to many materials, IMHO. Hence, mobile application may find this engine concept a bit pricey and impractical.

For stationary application, there is already their demonstrated method of water injection. The water can further be cooled down to near the freezing point to reduce the volume of water injected, and to reduce compression work further due to the lower temperatures involved. The distilled water can also be recycled again and again in a closed system, when refrigeration is used to bring down the water temperature, hence minimizing evaparative loss.

Account Deleted

HI Peter,

Low cylinder temperature is a problem during compression and ignition as it slows down the flame speed. Agreed that spark plug may help but with the use of spark plug, the risk of knock is there and it will limit the power output.

Have you seen the liquid nitrogen injection, it is not premixed but it is direct injected into the cylinder. Cool spots will still be there if it is not premixed and this will make the flame velocity gets worse.

I tend to disagree with your claim that Ricardo has considerable experience in this area. I happen to work for an automaker and based on discussions with lots of colleagues from other automakers, I havent heard anything positive about them. Others also told me that they are excellent in presentations but poor in project deliverables. For turbocharger, FEV is where people go to, not Ricardo

Account Deleted

hi roger,

when it comes to compression, we need only to compress the charge to get it to the required auto ignition temperature. Nothing more or less. Adding cooling substance does not make sense especially if it is an inert gas like the nitrogen.

Are you saying that it is good to add recuperated gas after the compression cylinder? in the first place the recuperated gas needs to be compressed before it will move into the already high pressure gas passage.this requires a lot of additional compression work.

Regardless of whether it is isoengine or not, the principle or turbine and turbocharging remains the same. Once the pressure and temperature are reduced, there is not much else that you can extract. Somehow, the thermodynamics law are being overruled here, i bet that the actual engine will give a very different outputs.


I think this is all BS.

Do I have specifics?

Way to complex.

Don't need any more than that; this is just a press release/proposal of what they will model and evaluate.

Account Deleted


There is a limit on how much I can absorb BS, in this case, the BS is ridiculous. Reminds me a lot of the 2S/4S BS from them. Not sure what happen to it but I am glad the 2S/4S BS has already stopped too

Roger Pham

Well, umm, without going back into further technical details, let's all agree that this CoolR stuff is a bit too complicated to be practical, and leave it at that...


Flame speed is not an issue with compression ignition engines. It is diffusion controlled combustion and to a large extent also controlled by “large-scale” air-fuel mixing. What I am referring to is a spark-assisted (or glow-plug assisted) diesel engine. Knock is not an issue for these types of engines either. I do not know what you base your assessments on. If you work for an automotive manufacturer you should have more basic knowledge. Must I assume that any small detail is unknown to you?

I asked you to be more specific about turbocharging for one particular reason; I wanted to assess your knowledge in this field. But you did not provide any more information. How would the laws about thermodynamics be overruled? Frankly, I think Ricardo knows a lot more that you do about turbochargers. Ricardo has been around for a very long time. Sir Harry was already retired when FEV started their business. For sure, FEV might be better in many areas than Ricardo and vice versa in other areas but I have not seen any indication that this should be the case in the turbocharger field. Have you? The point is that all the three big consultant companies (including AVL) generally know what they are talking about. To receive commissions from the automotive industry, they have to be in the front line. If they are not, the will get no work. As you can see, they are still around.

The main problem here is that the proposed engine is so complicated that you would have to do the same work that Ricardo already did to fully assess it. No one of us who has commented on this site has done that work.

Account Deleted

Hi Peter,

you are entitled to whatever you want to think about me. it doesnt really matter to me..

So which type of combustion are you referring? spark ignition or compression ignition. if you're referring to spark ignition, there will be a point of ignition where it will propagate to the outer edge of the cylinder. Contrary to what you said, knock will be an issue if the at certain point.

i thought i was being specific enough by highlighting the numbers that do not make sense. if you're so confident about it just provide clarification on how on earth 1 bar ambient pressure can be raised to 5.9 bar using exhaust gas which has just came out of a recuperator? care to explain?

To me, harry ricardo was a good engine researcher. Having a founder who was good will not guarantee the next generations of employees to be good too.

Roger Pham

With a temperature of 699 degrees C and at 170 bar, the intake charge of the power cylinders can combust spontaneously with finely injected diesel fuel without requiring spark ignition. With high-pressure injectors and the right cetane rating, there should be no problem. The Isoengine is designed to run at ~600 rpm, so it has more time for combustion in comparison to a big-rig diesel running at 1300-1500rpm, even when fuel injection is done after top dead center.

Let me give you a semi-quantitative analysis of the turbocharging system. First of all, you are rightly concerned that at 5.9 bar of boost, a typical automotive turbocharger will not be up to the task. But remember that this engine is in the tens of thousands of horsepower class, or tens of Megawatt of power, not the typical automotive engine that you are accustomed to. This high pressure requires probably one axial stage followed by one centrifugal stage to avoid surge and other issues. 2 stages of axial turbine would be required to harness enough power for the compressor.

Now then, 5.9 bar of compressed air from one compression cylinder will be heated to 699 degrees C, or 972 Kelvin, before entering 3 power cylinders. In the power cylinders, the intake charge will combust to above 3000 degrees K to ensure 3:1 increase in volume before adiabatic expansion while taking account for in-cylinder heat loss also.
At the end of the expansion strokes, the 3 power cylinders will discharge exhaust gas at ~5.9 bar but will provide 3 times the volume of exhaust gas in comparison to the volume of intake air entering the single compression cylinder. This means that the exhaust gas contains 3 times the energy required by the compressor. Assuming ~compressor having 70% efficiency and turbine having 80% efficiency, .56 x 3= you will still have 1.68 times the energy required for the turbine compressor. With exhaust gas heat recuperation, the temperature of the exhaust gas at 772 C (1045 Kelvin) will be reduced to 348 C, or 621 K. 621 / 1045 = ~60%. This means that the exhaust gas volume after recuperation will drop down to 60% the volume of the gas before recuperation while still at the same 5.9 bar pressure. .6 x 1.68= 1.008. Woaa, you will just barely have enough exhaust energy post recuperation to power the turbine compressor.

Your concern is quite relevant, since the IsoEngine will require high-grade turbine equipments with high efficiency level to do the task. Automotive turbocharger simple ain't up to snuff!! Almost every single Joule of exhaust heat energy is utilized to maximize energy efficiency.


I clearly stated that I was referring to a spark-assisted diesel engine. I anticipate that you are not familiar with that concept. Since there are some practical problems with spark plugs in diesel engines, I would prefer glow plugs instead. Furthermore, you might only need this assist at certain operating conditions (cold start, low load, etc…) and glow plugs are more practical under such conditions. I was referring to spark assist only to convince you that it is possible to ignite fuel under very “severe” conditions, i.e. with low in-cylinder temperatures. Remember, you an even use gasoline in a diesel engine with ignition assist. You also referred to “cold spots”. If you look at the PV diagram you will see that admission of nitrogen is under the whole compression stroke, not only at top dead center. So, you will have both high impulse from the nitrogen injection and long time for mixing. Besides some need for optimization, I see no “show-stopper” in this area either.

Thank you for a very good analysis of the turbocharging system (that I did not have time to do myself this time…)! In contrast to azmio, I think Ricardo can master similar calculations. The fact that the company has continued after sir Harry died is proof of that.

Account Deleted

Hi Roger,

Have you noticed that the natural gas is introduced at 190 bar. So while you're saying that the recuperated gas enters at 170 bar and 699 C, i assume that the natural gas is introduced when the piston is at the bottom or otherwise the pressure will be too high for the fuel to get in. In this case, considering that the recuperated gas is introduced at 25.3 kg/s, it will be 97:1 AFR, will it ever combust?

You mentioned about the turbocharger, do you expect the exhaust gas exiting the recuperator will have its pressure increased from 5.1 bar to 5.9 bar? How will it ever be possible?

This whole thing consumes a lot of power for injection and gas compression, do you think that with just 0.26 kg/s of natural gas which is equivalent to 10.9 Mj per second, will it return excess of 60% efficieny? -60% efficiency is more like it.

Account Deleted

Hi Peter,

Care to explain on how will it be possible to ignite a charge with your glow plug or spark plug when the air+nitrogen comes in at 25.3 kg/s and the fuel is only at 0.26 kg/s. We are talking about 97:1 AFR here and assuming that the natural gas is injected at 190 bar, it can only be injected when the piston is at the bottom of the cylinder before the cylinder pressure build up.

Try to answer my question first and if you cant find any answer to it, you have to agree that Ricardo as it is now is full of incompetent people

Oh by the way, how big do you think the recuperator is when gas pressure entering it is about 170 bar? what sort of heat exchanger fins will have to be used to raise the gas temperature from 217 to 699 when the exhaust gas is only 772 C? will the fins have large surface enough area and how can it withstand the 170 bar?

Account Deleted

Hi Peter,

Understanding that the nitrogen will be introduced at in liquid state, it will immediately boils once entering the cylinder. From this point onward, how much compression work does it need to compress it to 170 bar?

oh by the way, how easy it is to transport hot gas at 170 bar and 699 Celcius from one place to another?

Roger Pham


The NG is introduced the power cylinder immediately after dead center, whereby it will mix with the air and combust. There is a 20-bar pressure differential for the highly compressed NG to enter the combustion chamber. Air mass intake is only 6.25 kg/s, not 23 kg/s as you mentioned. Spray water is introduced at 19kg/s, when combined with intake air mass of 6.25kg/s, will give a combined mass of 25.3kg/s of water and air mixture exiting the compression cylinder. 6.25/.26= 24:1 AFR. I would guestimate stoichiometric AFR for NG:air at ~18:1 because of higher hydrogen content of the NG, so the 24:1 AFR should give quite robust combustion at at least 2500 degrees Kelvin or higher. No LN2 was used in the diagram in the article, only spray water was used to achieve isothermal compression.

I was wrong about the 3000 Kelvin final combustion temperature in the power cylinder. Given a initial pressure of 170 bar expanding to 5.9 bar of exhaust pressure at 1045 Kelvin, I calculate expansion ratio to be 11.8 in the power cylinder, and combustion temp. before adiabatic expansion to be ~2500 Kelvin. Quite within reasonable range for an AFR of 24:1.

WRT the actual exhaust gas pressure at 5.1 bar, I used a hypothetical situation of exhaust gas exiting the power cylinders at pressure 5.9 bar to be equal to the 5.9 bar pressure of air intake into the compression cylinder to simplify the calculation of power balance between the energy available in the exhaust gas volume vs. the energy required for compression of intake air.

Ricardo somehow must have used even more efficient turbomachinery than what I have in mind, in order to enable them to extract enough exhaust gas energy out of only 5.1 bar of exhaust gas pressure for compressing the intake air to 5.9 bar. Still within the realm of possibility.


The whole idea of introducing liquid nitrogen in this way is to decrease the compression work compared to, for example, using injection of gaseous nitrogen (or any other inert gas) in the inlet port. You see a problem with compression work. How much compression work do you think a normal diesel engine need to achieve 170 bar at TDC before fuel injection? Future state-of-the art diesel engines have to go up to 250 bar and beyond. Natural gas and biogas is compressed to ~250 bar at the refueling stations. Does this require a lot of work? Of course! Is this done is one stage by a piston compressor? No, of course not! It is done in several stages, with intercooling in between the stages. Of course they want to reduce energy use and one way to do this is to approach isothermal compression. What is wrong by using a technology to reduce compression work in a piston engine?

If you have an IDI (pre-chamber) diesel engine, you transport ~50% of the cylinder charge at similar pressure levels, e.g.170 bar, but at higher temperature from the pre-chamber to the main combustion chamber. Is that difficult? No! The hole required for this transfer has a diameter of a few mm. It is done near TDC at high pressure but at low piston speed, so pressure loss is low. Precision casting of the pre-chamber and cooling is of course used, as in any cylinder head of a combustion engine. What about transporting high temperature gas at high pressure in the chemical industry? Nothing novel here either and nothing that would become a show-stopper…

Frankly, I do not understand what you are after… You seem to have decided that your “gut feelings” tell you that this engine will not work. However, you do not present any information to support these feelings.

Just when I had finished my comments above, I saw the comments from Roger. I have not had the time to look at those yet. Maybe I will do that later…

The comments to this entry are closed.