Running cars on water

Can you run your car on water?

There have been many stories of genius inventors making a car that runs on water, only to be silenced by the evil oil companies, never to be seen again.

Can you run your car on water? Well, not like this, obviousely.

I was intrigued by the sheer volume of technical claims, despite decades working at the cutting edge of automotive technology and being immersed in scientific theory I am acutely aware there is always room for doubt, and for new ideas to surprise and change the way we think. Ideas like these are rare but happen often enough to make keeping an open mind and essential part of the make up of the modern engineer.

The traditional view is that water is basically hydrogen that has already been burnt, so it has no usable energy content left. Because of this fact there is a tendency to think that all these people who believe in water power are all nutters who probably have been abducted by aliens and experimented on for a laugh and are a tiny minority. But nothing prepared me for the depth and complexity of the technical explanations and the shear volume of conspiracy theory’s involving governments, car makers and oil companies all in cahoots to keep us buying expensive oil.

Well, let me assure you that car makers would quite happily stab the oil companies in the back if they could sell a few more cars, and one that ran on water would sell rather well, don’t you think?

So, how do you run your car on water?

Well, you will need a big plastic jar (screw top with a good seal), a small plastic jar, some tubing, a few strips of metal, some small bolts, a fuse and some wire.

But before I go into details, I want to talk about electrolysis. I found a number of amusing web sites that propose the use of hydrogen made from passing current through water. So far that’s not a bad idea, many car companies are investing quite a lot in converting petrol engines to run on hydrogen.

Where these sites go off the rails is when they generate the gas on board the car, using electricity from the alternator, which is of course powered by the engine.

Here are some basic figures for you, a good car engine will convert 33% of the energy released from the fuel into usable power at full throttle, it goes down to about 10% at light loads. An old carb engine might be as poor as 20% at full tilt. Oddly enough big engines can be more thermally efficient than small ones, its all to do with heat loss and the ratio of volume vs surface area, the most efficient piston engines in the world are the cathedral engines in super tankers, such as the 25 thousand litre Wartsila-Sulzer RTA96-C turbocharged two-stroke 14 cylinder, which gets up to 50% efficiency. You may recal I did a blog post about it last month, its very impressive but at 2300 tons you would struggle to get it in a car.

So, going back to our in-car hydrogen plant, with the precious little power left at the crank we drive the alternator, which usually has quite a good efficiency, converting about 90% of the power fed into it into electrical energy.

Then there is electrolysis, at the molecular level, the energy you put in to separating water into hydrogen and oxygen is the same energy you get back when you set fire to it. Interestingly, some of that energy can come from heat from the environment, ie the heat from the engine.

So, when you add up all those efficiencies up, to produce 1bhp worth of hydrogen, the engine has to burn nearly 4bhp worth to make the process work.

Guess why it doesn’t work!

Mind you, if you generate the electricity away from the car, say from a wind turbine in your back garden, store the gas in huge explosive bags attached to the roof of your car then you are on to a winner. Until it imitates the Hindenburg.

So once again I have shown that you can’t run your car on water. So now I will finally get round to showing you how to do it.

One of the reasons that the efficiency of car engines is so low is that not all the petrol gets burnt, and some gets partially burnt. The actual combustion process is very complicated, with molecules decomposing into sub species before reforming into exhaust gasses. The flame front travel across the cylinder is also semi-chaotic and some molecules get passed by entirely. Some start burning then hit the cold cylinder walls and stop burning, some start burning at the end of the process when the piston is to far down to convert it into usable energy. Petrol mixtures burn at a rate of about 40 to 50 centimeters per second depending on loads of factors like pressure, turbulence and temperature. That’s one of the reasons that big engines run slower as it takes longer for the flame to travel across the bigger combustion chamber, for instance that super tanker engine I mentioned produces 5 million ftlb of torque at just over 100 rpm.

If we could speed up the burning process then more of the petrol’s energy can be converted into useful work on the piston. Also, if we could put in something that would mix more readily with the air and bridge the gaps in the fuel mixture we could avoid those dead spots.

Ooh, hydrogen does that, it mixes very readily and burns faster. So all we need is a very small amount of hydrogen and we can improve the efficiency of the petrol.

The difference in power at part load by introducing a sniff of Hydrogen.

Well, if its so good why hasn’t it been done before? Well, it turns out that this method has been used for many years and some reasonable research has gone into it, have a trawl through the SAE web site and you will see that many respected institutions and big companies have published papers on the subject. Some use methane which is broken into hydrogen and carbon dioxide by the use of rather hot steam. Of course you then have a hydrogen car that produces co2 which is sort of bad really.

So here is the theory; you get a bucket of water strapped to the car, stick two bits of metal in it (electrodes) connected to the battery (via a switch and fuse). The lid on the bucket has a hose to transport the explosive hydrogen and oxygen gas to another bottle where any water is removed (don’t want to hydraulic the engine). Finally the hose is stuffed somewhere in the intake to allow the gas in.

Great, but how much gas do we need? Well, to make 1g of hydrogen you need to apply 285Kj of energy to the water, luckily the water tends to use energy from the environment during the process, so potentially about 48 Kj of heat will come from the engine bay, leaving our electrics to provide about 237Kj per gram.

Now, power in Watts is Joules per second. So 1.4Kw (1 HP) is 1.4Kj per second. Which works out equivalent to burning about 0.006 grams of hydrogen per second, or in volume terms that’s about 0.066 litres per second, a steady stream of small bubbles.

So for a cars electric system to put in 14Kw of power at 14v the current will be 100 amps, which is a lot.

But remember we are not talking about using the hydrogen to power the engine, but to help get more useful energy from burning the petrol. So the big question is how much do we actually need? The web sites suggest one litre of water will last up to 900 miles. That works out at about 1.1g per mile, and of that 1.1g of water there is only 0.12g of hydrogen per mile. At average speeds this works out at about 0.0013g/s, about one fifth of a bhp, less than 1% of the overall fuelling and will draw about 10 amps in the electrolysis bucket.

To get a sensible answer to all this, I nailed some scrap metal into an old washing powder tub and tried out some combinations.

First attempt, stainless steel wire in plastic formers.

The first version was one recommended on an American web site, claiming up to 40% gains in efficiency. It consists of a one litre plastic container with two bolts in as electrodes. But even with a little salt added to my copy it only managed to draw 0.1 amps and no detectable gas flowed out of the outlet hose. In short, it was useless and had no effect on the engine.

Clearly we need more power, one of my favourite sayings. So next we have a system with drastically increased conductor length by using wire, wound round a cross shaped former, still in the one litre pot. This has the effect of drawing 10 amps, about where most of the internet products are, and a steady stream of bubbles.

Second attempt after a few thousand miles. Third version ran cleaner.

This very small amount of gas will have the most dramatic effect on the engine at lower loads and idle because that is where it will be the biggest proportion of the total mixture, so I fitted the system to the intake and watched the injector pulse widths and lambda compensations so see what effect it had. Well, the web sites suggest a 30% fuel saving overall, so at idle it must be huge, but no, there was absolutely no difference on average.

I even drove it round for a few weeks, and at first I though I could just detect an improvement in power and the fuel bill seemed to be dropping. But then it went up again; it turned out that it was just me driving more carefully as I paid more attention to what I was doing after fitting the kit. This is a very common phenomenon.

First instalation, about to test current and voltage on my long suffering Land Rover, subject of a great many experiments...

So how much gas do I need then? Well, some very useful research has been done by NASA and various universities. Basically to get a 30% increase in efficiency (power out vs fuel in) we need to run 90% hydrogen/10% petrol!

At 50 % hydrogen the efficiency improvement is only 15%, but what does that mean in a normal car? Well, if you are cruising down the motorway you might be using about 20kw of engine power, if you only run 50% HHO then that’s 10kw of electricity running through your jam jar, at 14v that’s over 710 amps out of your alternator! But then that power comes from the engine so it would now have to produce over 30kw, which would mean a 30% increase in petrol use in order to gain a 15% efficiency saving….. Hmmmmm..

Now, bear in mind that the average driver can improve their fuel economy by up to 30% just by learning better driving techniques, and that on an average commute fuel economy can vary by 20% easily depending on what mood the driver is in. So subjective assessment of 10% economy gain is meaningless, it has to be checked on a proper test facility.

There are of course other ways of generating hydrogen on board a car, using chemical reactions, and this would take the alternator problem out of the loop, but generally the chemicals are rather nasty/expensive and leave a chemical waste problem.

One of the favourites being explored by the car industry is called ‘reformate’ where the petrol is partly separated into CO2 and Hydrogen using a catalysts and exhaust heat. At the moment the fuel savings don’t justify the expense of the extra equipment, but I am sure that will change in time.

Water has a number of other benefits in a traditional engine, I am sure you know about water injection which turns a bit more of the heat energy from combustion into pressure energy. It also reduces cylinder temperatures and so reduces knock, allowing a bit more advance. But also a small amount of water, up to 5%, if well mixed in the fuel before its injected can increase power by up to 7%. Unfortunately you cannot just chuck a cup of water in the fuel tank, because it wont mix, you need some reasonably clever and accurate mixing machinery.

In short, there is a lot you can do with water. However, if you google ‘water car’ then the myriad of websites that appear generally talk complete twaddle and make excessive claims for there own brand of hocus. Whilst I am talking about the web sites, there were a few claims that I feel are potentially harmful.

First is the matter of your cars warrantee, unlikely anyone thinking of doing this will have any, but the point is that despite the various web site claims, fitting absolutely anything to your engine will invalidate the warrantee. In the car industry we spend a huge amount of time and effort checking the car works under all sorts of conditions, it takes years for each model to complete all the tests before being ready for production. Modern engines are so finely tuned that any disturbance will cock things up, and that includes those plug in chip tunes by the way. So plumbing in a bottle of water to your intake will definitely lose any warrantee.

Next, some sites sell you electronic devices to bias engine sensors to make it run lean, best fuel economy generally happens when you run 10% lean of stoich, which is fine at part load on a non cat car, but obviously not at full load because things overheat and fail, and if you have a cat it will stop working eventually. And as these are sold without checking the engine on a dyno the potential for detonation or catalyst damage is rather significant.

Then there is the matter of additives in the water, remember it all goes somewhere and some chemicals will put quite nasty things into the atmosphere. Some sites advocate volatile fuel additives too, again the reason the car industry doesn’t do this is that it poisons the catalyst and puts very nasty stuff into the air. Fuels with super chemicals that don’t pollute as much are readily available, such as Shell Optimax or BP Ultimate which have over 200 chemicals in to get the best performance.

Safety seems to get a fairly low priority too. Remember that as we separate the water we get a perfectly explosive mix of hydrogen and oxygen, albeit in very small amounts. If the pot you generate the gas in is not ventilated then any slight spark could leave you pot-less and potentially destroy things under the bonnet. Luckily most of these systems produce so little gas that it is unlikely to be a problem!

So in summary, yes you can run your car on water, but not using a jam jar and some wire.


Don’t just take my word for it; here are some of the research papers I used in my research:

SAE 841399 1984

Water/fuel mixtures

SAE 740187 1974

Lean burn with Hydrogen supplementation

General Motors Corp.

SAE 810921 1981

Lean burn with Hydrogen supplementation

University of Michigan

SAE 2004-01-1270

Hydrogen reformate (H + CO2)

Robert Bosch GmbH

SAE 760469 1976

Hydrogen reformate in aircraft engines

Jet Propulsion lab

What are the losses?

The heat released from fuel is mostly wasted, about a third goes down the exhaust pipe (turbos can recoup a small fraction of this) and another third goes into the oil and coolant. A smaller amount is wasted as noise and vibration.

At part load the throttle causes an obstruction that wastes power too. A better solution at part load is to replace the unwanted air with an inert gas, this allows the throttle to open up and reduce losses. That is why Exhaust Gas Recirculation (EGR) can improve economy by 5%.

What is reformate?

This is where the petrol is broken down into hydrogen and carbon dioxide. The CO2 is inert and at part loads is used as ‘padding’ allowing higher throttle openings, this reduces the throttle losses and also improves efficiency. This double benefit is why the car industry is concentrating on this way of making Hydrogen. But with the petrol engine’s days numbered, the technology may never mature.

What is HHO, Oxy-hydrogen or ‘Browns Gas’?

This is what you get from electrolysing water, it is simply oxygen and hydrogen in a gas. There is a lot of myth about it, but here are the facts: It burns at about 2800 C, compared to about 2000 C for petrol in air, or about 3000 C for oxyacetylene welding kit. In fact it was one of the first gasses for welding, but in slightly richer proportions. When HHO is burnt in air, as in our engine example, the flame temperature drops to similar values to petrol.

When it burns it expands, and so can be used to power piston engines, then it forms water vapour and starts to condense, rapidly contracting into just water.

The gas has been the subject of many hoaxes, frauds and misguided optimistic claims.

It’s not magic, just nature. Although personally I thing nature is pretty magic.

Leave a Reply

Your email address will not be published. Required fields are marked *