View Full Version : good water/fuel mix for a home built system

03-03-2004, 03:40 PM
i have been looking at water injection for a while and i have finally got a car worthwile doing it too. a 1.8L '91 nissin pulsar, fuel injected. i was looking at the commercial units but i dont have a lazy $700 lying around. i am in my final year of an engineering degree so i was thinking of designing my own system that will use the air-flow metre to regulate the water flow.

So far i have collected this much information
? Proportional to engine load / air flow
? Only inject water at higher engine loads
? Do not use ethanol
? Mounted before the throttle valve but after the air meter

what i was wondering about, how much water do you inject into the engine? is there a way of accurately testing the optimum volume on a shoe string budjet? Any other important information that i may need to know would be useful, or comments on what i already have

i was also wondering wether the engine should be at operating temprature or close to it before the system turns on to prevent water damage to the engine and oil.


06-03-2004, 02:04 PM
Keep in mind that the "don't use ethanol" is Aquamist specific due to incompatiblity issues on the pump. There is nothing wrong with ethanol per se, you just have to use alcohol blends that are compatible with your system.

Here in the central plains states of the U.S. fuel grade ethanol is easily available, and is much less toxic or corrosive than methanol. I will be using it on my system for that reason.

You should also add to your list of axioms.

For optimum operation lean the AFR when injecting water for max power.


Charged Performance
06-03-2004, 06:07 PM
Ethanol can be used if the system components will tolerate it. The board was started by those with affiliations to Aquamist (but is meant for all) so that is why you will see a lot of no ethanol comments.

Water is being used for knock suppression only. You want to inject enough to suppress knock sensitive areas which are usually more correlated to torque/cylinder pressure than load. Load is more of a type of proxy for torque/cylinder pressure but not identical. You do not want to inject so much water as to displace fuel that would have been consumed in combustion. For instance you can displace fuel from a lambda of .6 to a lambda of .85 - but you would not want to reduce your fueling to higher than around .85~.90 which would be your optimum power AFR depending on your intake and cylinder mixing efficiency.

Alcohol in the mixture is good because you can reduce gasoline in the mixture further and the alcohol will burn in its place - making it a hybrid fueling. Once you do this though be careful using AFR and try to work and think in lambda since lambda is a consistent measure between different hydrocarbons.

The amount of water to inject will really depend on how much gasoline is currently being used for in cylinder cooling and detonation suppression. If a lot of gasoline is being used more water can be used while leaning out the gasonline mixture. If only a little bit of gasoline is being used to suppress knock - less water should be used. In any case injecting alcohol will permit gasoline to be leaned even further than water alone. As a rule 10% to 25% water to fuel can be used - but you will see applications all over the range.

Whatever system you use you want to bench test for how much flow your system actually generates no matter what the specs of the components are. Then take into account the reduced flow from boost across the jet face.

Injection should only be being used at higher torque generating conditions. You shouldn't be in those conditions before being warmed up in any event. Even without water injection generating high torque before being at operating temperature will damage your oil from the rich warm up fuel settings. Generating boost before getting to operating temperature is always a bad thing.

08-03-2004, 01:19 AM
Thanks for the info. i find that ethanol bit very interesting. about a year ago, some local fuel station started selling 10% ethanol in their fuel and the media cracked up a big stink about it saying it was dangerous. now every fuel station has signs saying 'no ethanol in fuel' i also read that 'an editors job is to sort the wheat from the chaff, and to make sure the chaff is printed' :? what do others think about this? is it safe for most vehicles to have 10% ethanol in their fuel? would it have the same effects as injecting it seperately?

This is my first fuel injected car, so i have just 2 small questions. what does "AFR" stand for and what is that lamda thing?

08-03-2004, 02:00 AM
The example you mentioned about the media is right on the money.
Here in Colorado U.S.A, we were one of the first states to shift to ethanol fuel blends back in the 1970's to control CO emissions during the winter. There was a brief period of minor problems caused by ethanol added to fuel.

There were a rash of cars that had plugged fuel filters --- translation, ethanol cleans harmful deposits out of your fuel system and if it is full of varnish and other deposits it will clog your fuel filter. Once that intial phase was past there were no problems. In addition, the ethanol keeps your fuel injectors cleaner than they would be with out it, it drastically reduces exhaust gas temps, it drastically reduces Carbon Monoxide emissions. Ethanol fuel blends will actually produce more power than straight gasoline blends, as the ECU adjusts for the new mixture.

There is some debate about oil compatibility and ethanol fuels, but practical experience over 30+ years shows its only a theoretical problem and especially with modern synthetic lubricants, it is a non-issue.

There is a slight reduction in gas milage. The ethanol has a lower specific energy content than the gasoline it replaces, however it actually makes some of that back due to the increase power you produce you spend less time in the throttle, to get the acceleration you want. The loss in milage is a minor problem if the consumer realizes he is not being "cheated" by running "bad gas".

There was also a brief period in the 1970's where there were fuel system incompatibility issues. Some fuel hoses and carburator float materials did not play well together with ethanol. This has long since been resolved my the auto manufactures. If you have a 1970's vintage VW and have never changed the stock fuel lines you have a problem, they will start leaking after a while. If you have a modern car it is a non-issue.

There are some folks that contend that ethanol fuel actually costs more energy to produce than it delivers this has been proved to be absolutely untrue. Those numbers are based on very old data that does not match up with the current art of manufacture.

The country of Brazil has made extensive use of ethanol fuel blends for decades. They run up to 22% ethanol in modern cars with no problems, the ECU can adjust for it. They also run some cars on 100% ethanol. Any cars marketed as FFV (flexible fuel vehicles) are designed by the factory to run on ANY blend of ethanol and gasoline.

Your going through the same bogus hype generated by folks that are trying to protect vested interests as we did. Some of these folks are just ignorant of the facts. This led to the large scale introduction of MTBE which has since been discoverd to be a major source of pollution to ground water, and is now rapidly being outlawed.

Ethanol is much less toxic than the gasoline that it replaces.

This site will give you some basic info you might find interesting.

Charged Performance
08-03-2004, 03:01 AM
From a nonchemist. A search on AFR lambda and stoich may yield more.

AFR is short for air-fuel ratio. It is the ratio of the mass of air in combustion and the mass of fue in combustion. Lambda is a measure of stoichiometry.

Chemical reactions have reactants and bybroducts - when the reactants fully convert to only the end by products the relationship of the mass of the reactants are at stoichiometry or stoich.

Hydrocarbons and oxygen in the chemical reaction of combustion will result in only water and CO2 as its byproducts. When air is involved there are other byproducts. Ignoring the other byproducts when gasoline and the oxygen in the air are mixed in the proper mass relationship they will be stoich or completely combusted with no other byproducts. This occurs with gasoline at around 14.7 air to 1 gasoline (by mass). This point of stoichiometry is lambda of 1.0 - for any fuel at stoich lambda is 1.0 no matter what the mass ratio of the air to fuel is. Lower than 1.0 is richer (less air than stoich to fuel), higher than 1.0 is leaner (more air than stoich to fuel). Some also use 1/lambda which just serves to confuse me personally.

Chemically inclined people please clarify any mistatements above.

Shade tree mechanics like myself rarely knew all this but without releasing it were using it when tuning even before fuel injection was on mass produced cars. It was always there in the background affecting us we just didn't have a way to monitor it and worse no way to manage it on such refined basis as we do now with EFI.