Modern fuel on trial

The author of this article, Paul Ireland at speed (about 70 mph) in his TC with his son as passenger. They were returning from the Le Mans Classic trip.

Over the past three years I have written articles on the problems I have experienced using modern unleaded fuel in my TC. Talking to other classic car owners it appears as though these problems are widespread and not always attributed to fuel. Last year I shared the results of my rolling road tests with the Federation of British Historic Vehicle Clubs (FBHVC) and another MG car club, suggesting they were sufficiently conclusive to warrant independent, scientific tests; only to be told by the car club that it was unable to help and by the FBHVC Fuel Expert “there are no problems running classic cars on modern fuel”. To put it bluntly, my request was ignored.

Fortunately, club members have shown a great deal of interest and I would especially like to thank David Heath, John James, Chris Morgans, Colin Whitmore, BP Australia Fuel Research and Dr Robert Woolley from the Engineering Department of the University of Sheffield, amongst many, for their advice and help that ultimately led to the tests described below. Despite the difficulties of making scientifically justifiable measurements while bumping along busy, public roads in a classic MG, both David Heath and I have managed to gather what I believe to be definitive evidence. I leave it to you, the reader, to judge who is correct. I present the case:

Classic cars v/s Modern Fuel.


The charge against modern fuel is that it burns more slowly in classic car engines than the fuels these engines were originally designed to use. As a result, an engine set at standard tune operates less efficiently, possibly with still burning gasses leaving the cylinder. This causes elevated exhaust gas and cylinder head temperatures, which, in turn, lead to higher under-bonnet temperatures, overheating and fuel vaporisation problems. In some engines, these elevated temperatures, appear to cause damage to the cylinder head and valves.

Witnesses for the Prosecution:

I would like to call on the evidence gathered by:

1. (TA) David Heath’s TA, with a standard MPJG engine, a compression ratio of 6.5 to 1 using richer AV carburettor needles.

2. (TC1) My TC, with a standard XPAG engine, a compression ratio of 7.5 to 1, resulting from the head being skimmed, with standard carburettor needles.

3. (TC2) Mr A’s TC, whose XPAG engine has been bored out to 1350cc with a cleaned up head, a compression ratio 8 to 1 and a higher ratio back axle.

4. Colin Whitmore’s 3992cc Armstrong Siddeley Star Sapphire, having a compression ratio of 7.5 to 1, probably higher due to the head being skimmed twice.

The evidence:

The tests David and I performed are easily repeated and may be used by anybody to validate these findings or indeed to tune their own car; however, heed the warning at the end of this article before you do.

All you need is a stop watch, a digital multi-meter with a temperature measurement facility (e.g. 600.038 Digital Multimeter – cost about £10 – search for the code on the Internet), an insulating pad (e.g. asbestos) and a straight, level, quiet, de-restricted stretch of road about 1 -2 miles long. A timing light with the means for measuring engine advance (e.g. Gunson Supastrobe) is useful if you want to compare your results with ours, but not necessary.

Fasten the thermocouple onto the rear of exhaust pipe as close to the manifold as possible using circlips and insulated from the air stream using the insulating pad (see photo below), install a passenger with a notepad and the stopwatch and you are ready to go.

The results, presented as evidence, are from a set of runs measuring exhaust temperature and timing full throttle acceleration. The first run was performed using the manufacturer recommended ignition advance setting and advanced each by around 3 – 4 degrees using the vernier on the distributor for each subsequent run. This gave acceleration time and temperature measurements for a range of different ignition advance settings.

After advancing the engine, the tick-over advance was measured using the timing light. Every attempt was made to ensure the data was as accurate as possible by starting the run at the same place on the road, taking the average of more than one reading or by using two stop watches to time the acceleration and repeating the each run, in opposite directions.

Exhibit A: The full throttle top gear acceleration time from 2000rpm to 3000 / 3500 rpm (approximately 30 to 50 mph). (Note: On the TA, a wooden block was fitted underneath the accelerator to reduce engine load and possible damage). I fitted a modern digital rev counter to my TC (TC1) for these measurements. In the other cars, the original rev counters were used.

Exhibit B: Exhaust temperature. (see opposite column) After a constant speed run of around ¾ mile on a flat section of road, the temperature reading taken once it had stabilised towards the end of the run.

Witness 4: In tuning his Armstrong, Colin has found he needed to advance it around 2 degrees above the recommended setting. “If I didn’t advance the ignition, the car would overheat” he said, “I use 97 octane fuels to stop it pinking”.

Summing up: When an engine is running at its optimal efficiency it is delivering the maximum power from the inducted fuel load. Using a consistent throttle setting for the acceleration tests ensured the inducted fuel was the same for each run; hence the acceleration time is a direct measure of the power generated by the engine over the rev range (Exhibit A). I direct, you the jury, to consider that the fastest acceleration time is achieved when the engine has delivered its maximum power. You will note the differences between the worst and best cases are around 1.5 seconds, much larger than any measurement errors.

Cruising at a fixed speed on a flat road ensured the volume of inducted fuel was constant, therefore any variations in exhaust temperature are only due to differences in the way the engine is burning the fuel. The jury should consider the thermocouple is on the outside of the exhaust pipe and does not provide an accurate measurement of exhaust gas temperature, however, as it is insulated the temperature reading will be a good approximation of the exhaust pipe temperature. This in turn will reflect the exhaust gas temperature. When an engine is running less efficiently, less of the heat energy from the fuel is being converted into motive power and the additional unused heat will raise the temperature of the exhaust gasses. The temperature of the exhaust gives a good estimate of how efficiently the engine is running. I direct the jury to consider that lower exhaust gas temperatures are an indication that the engine is running more efficiently.

An engine runs at optimal efficiency when the maximum cylinder pressure on the “bang” stroke occurs 15 degrees after Top Dead Centre (TDC). Those who have read my “Suck, Squeeze, Bang, Blow” article (published in the April 2010 Octagon Magazine) will know the ignition is fired in advance of TDC to provide sufficient time for the fuel to burn. The jury should be aware that a slow burning fuel will require a higher advanced ignition setting to allow it to run at optimal efficiency. Additionally they should consider, when the ignition is retarded, there may be insufficient time for all the fuel to fully burn before the exhaust valve opens. In this case still burning fuel will leave the engine, further raising exhaust gas temperature.

Mr Whitmore’s evidence was arrived at independently from the research of Dr Ireland and Mr Heath. Initially he did not consider the need to advance his ignition timing to be related to fuel issues and once stated “I do not have problems with modern fuel”. While Mr Whitmore has not presented any data to justify his findings and his figure of 2 degrees advance is “from memory”, the jury should be aware this statement is from an independent source and relates to a more modern (1960s) vehicle.

I ask the jury to note the correlation between exhibits A and B. As the ignition was advanced both the acceleration times improved and exhaust temperature fell. These are independent measurements of engine efficiency and the fact that they agree adds significant credence to the presented evidence.

I now ask the jury to retire and consider its decision. Should you find modern fuel guilty as charged, please contact the Secretary of the FBHVC (Tel: 01865 400845 E-mail: [email protected] ) stating you believe there are problems running classic cars on modern fuel which should be investigated. If you find the defendant not guilty, I suggest you skip the warning below and read the other, more interesting articles in the magazine.

Conclusion and Warning:

Originally, I assumed the XPAG’s lower compression ratio was the reason modern fuel burned slowly. For all the three cars tested, the optimal ignition advance occurs at a similar degree of advance even though they have significantly different compression ratios. As there is very little difference, other than compression ratio, between the XPAG and the engines in “modern” classics, it is probable that the heat related problems seen in these later cars is also related to the burn rate of modern fuel.

These results are frightening. You can see from Exhibit B, running your engine with the standard ignition advance gives an exhaust temperature of 390˚C which decreases to 325˚C as the ignition is advanced, i.e. it is running 20% hotter. If this increase is reflected in the exhaust gas temperature, the 1200˚C gasses leaving the cylinder are probably nearer to 1440˚C. It is difficult to imagine how such a large temperature increase is only due to the decreased engine efficiency, adding further weight to the suggestion that fuel is still burning when the exhaust valve opens. It is no wonder some cars suffer from exhaust valves disintegrating and dropping into the cylinders. I am sure you can imagine the damage this causes.

Advancing your ignition timing gives modern fuel more time to burn, increasing engine power and reducing running temperatures. For the cars David and I tested, the ideal is 15 degrees advanced at tick-over. For my TC this is 13 degrees advanced static on an engine where the static advance should be set to 0 degrees. This may not sound a great deal until you look at the advance curve (Exhibit C) and realise 13 degrees static corresponds to 38 degrees at 4000 rpm. A standard TA would be 6 deg at tick-over with a maximum of 38 deg at 3000. Tell anybody who knows about engines what advance you are using and they will say “what!!” and when you repeat the figure suck air between their teeth, shake their heads and say “you know, that is very high”.

Modern cars run at as little as 5 degrees advance. People who have read the Suck, Squeeze, Bang, Blow article will be aware of some of the problems running such high levels of advance can cause.

Classic car owners are between the devil and the deep blue sea. They have a choice, destroy your engine through heat damage to the head and valves by running at standard advance or run it at a large advance and risk a hole in your piston or damage to the bearings through pinking. Remember the overheating and petrol vaporisation problems we all have seen are just symptoms of the slow burning properties of modern petrol. Don’t think your car is immune. Unfortunately, there is no compromise. Exhibit D shows the exhaust temperature for my TC across the rev range running at 7 and 15 degrees tick-over advance. As you can see it is still running significantly hotter with only 7 degrees advance.

Given this choice what option have David and I adopted? We are both running our engines with an advance of 15 degrees at tick-over and in my case using 97 octane fuels to offset the effects of kerosene lowering the octane rating. On my drive to Le Mans 2010 Classic, TC showed no signs of heat related problems despite the very hot weather and 35C ambient temperatures. On the downside, it does not run as smoothly on some brands of fuel.

A strong warning for anybody else thinking about following this example. Running your engine so far advanced can be dangerous. Certainly you should ensure the advance springs in your distributor are not tired otherwise the advance curve will be too steep (Exhibit C that shows the differences in my advance curve between new and old springs). David has fitted new TC springs in his distributor so his TA advance curve will be the same as a TC’s. Also be aware that it is very difficult to detect pinking in low compression engines; you should heed Chris Morgan’s warning. While running with high levels of advance mitigates the overheating problems, it carries with it its own risks.

I recommend you take the time to measure your advance curve. This is very simple using the Gunson timing light and a volunteer to hold the engine at a fixed number of revs. You set the dial on the timing light to align the mark on the front pulley with the TDC marker and read off the advance. I usually take measurements every 500rpm (e.g. 1000, 1500, etc.). This should give a curve something like Exhibit C.

What I personally find appalling is that in 1989, more than 20 years ago, Chris Morgans published an article in the Octagon Bulletin describing the running problems he was experiencing which could be directly attributed to the slow burn rate of unleaded fuel. He also described the serious engine damage these problems can cause. Why has this warning been ignored for so long? And why have both the car club, mentioned above, and the FHBVC chosen to ignore the problem? My impression was that both these bodies were created to help classic car owners so why are they not taking this problem more seriously?

Heat related problems affect a large number of classic cars to a greater or lesser extent. If you have read this section and you have found Modern fuel guilty as charged, I urge you to write to the secretary of FHBVC (E-mail: [email protected] ) asking they take this matter seriously. If sufficient people contact them, we may be able to elicit some action. My hope is that through an understanding of what causes modern fuel to burn slowly, it may be possible to find acceptable solutions for the benefit all classic car owners.

This article and associated graphs are the copyright of Paul Ireland and may not be reproduced without his explicit permission.

So over to the jury: what are your thoughts on Paul’s case? Add your comments below!

6 thoughts on “Modern fuel on trial

  1. Donald M. Lawson says:

    Hi John, great article by Paul! I am the tech-ed for the NEMGTR and have been putting together some research material for an article on modern vs vintage fuels and their impact on T-Type engines. Anyone out there know what was available in the 30s and 40s and what the octane was. I heard of “pool gas” and was told it was around 60 octane. Once again, thanks for the wonderful site and the information it will provide to all people who love these cars!

    • Paul Ireland says:


      Thanks for the positive comment. As for Octane rating, firstly beware there are two measures RON and MOM, older fuel was usually quoted as MOM, modern fuel as RON. The difference is in the way they are measured. For a fuel its MOM rating is lower than the RON rating (e.g. 85 MOM is approximately equivalent to 95 RON).

      In practice octane rating is only a measure of the knock resistance of the fuel and providing it exceeds the minimum required by the engine, running with a higher octane rating fuel does not cause any problems, it is just a waste of money.

      My original trials using kerosene (paraffin or 28 second heating oil) at a ratio of 1 part kerosene to 10 parts petrol were based on the incorrect assumption that the higher octane rating of modern fuel was the cause of my heat related problems. In practice my rolling road tests using kerosene did show it improved the way my TC ran and as yet I have been unable to find an explanation.

      On the basis of my findings, I have negotiated a concession with HM Revenue & Customs to allow vehicles manufactured on or before 1956 to add untaxed kerosene to their petrol. People should write to:

      Mr John Loughney
      Excise, Stamps and Money Businesses
      HM Revenue & Customs
      3rd Floor West
      Ralli Quays
      3 Stanley Street

      Requesting a “General Licence to mix hydrocarbon oils under Regulation 43 of the Hydrocarbon Oil Regulations 1973 (SI 1973/1311)” giving your name, address, model and dates of production of the model of your vehicle.

  2. Peter Roberts says:

    Your article references only “unleaded fuel”, which is a rather large field. “Leading” petrol to my understanding, has more to do with lubrication than burn rate, so I assume your concerns stem from the burn rate behavior of the alcohol addition. You would be quite helpful in commenting on exactly what component(s) of modern fuel produce the retarded burn rates. Through my racing experience, I have learned that there are many formulators who would be quite happy to produce a fuel that would satisfy your requirement….at an unbelievably unrealistic price! On the other hand, many in our MG-TABC community are regularly buying and pouring in lead additives, in lieu of installing hardened valve seats.

    In a related matter I have noted that setting lean/rich with modern fuels has become quite a guessing game. The usual standard of plug condition no longer seems to apply. Black sooty is normal when lean/rich has been properly set! One engineer has opined that this is a condition resulting from the addition of alcohol. Do you have any thoughts on tjhis?

    • Paul Ireland says:

      Lead was originally added to petrol for the single purpose of increasing its octane rating. Although I cannot remember the details, it was found early on that a second compound was needed to prevent the lead metal from being deposited on the valves. One side affect of the combination the lead and this compound is to protect exhaust valve seats, the second is to create a white deposit in the exhaust. Remember the light brown exhaust pipes of properly tuned cars? It is the lack of this white deposit that makes it more difficult tune a car using plug colour, but it is still possible to produce biscuit brown plugs with a correctly tuned engine.

      As for burn rates. At a given pressure, temperature and mixture, the experimental burn rate of a wide range of hydrocarbons, including alcohols are very similar. The same applies to the addition of lead and although it partially catalyses the combustion process, it does not significantly influence the burn rate of the fuel.

      The combustion process in the cylinder of an engine is very complex and while the addition of ethanol to petrol does not necessarily explain my findings as yet I have been unable to find a plausible explanation for my results.

  3. Steve says:

    “We are both running our engines with an advance of 15 degrees at tick-over and in my case using 97 octane fuels to offset the effects of kerosene lowering the octane rating. ”

    Since Higher octane fuels burn slower, using 97 octane is counter productive. It has long been a known fact you get the best performance out of lower faster burning fuels. I suspect the real issue may be that modern fuels are higher in octane, then what was available when the T series cars were made. My understanding is fuels of the past were much lower octane and that is one reason the engines were so low on compression. This would explain the issue, more then the recent fuels formulations being different.

  4. Paul Ireland says:


    Thanks for your comment. Unfortunately, your statement “Higher octane fuels burn slower” is a common misunderstanding. Indeed, I initially shared this view which I why I started adding kerosene to my fuel to reduce the octane rating.

    In practice there is virtually no difference in flame front propagation speed through a hydro-carbon / air mixture over a wide range of different hydro-carbon compounds, this includes fuels with different octane ratings.

    The only difference between higher and lower octane fuels is that higher octane fuels are more knock resistant, i.e. less susceptible to auto ignition.

Comments are closed.