I've done some reading, and it seems a lot of people are having good experiences with WI kits. Question is, what is a decent WI kit that isn't too expensive. I assume they all are very similar... Or am I wrong??
Also, do you think I could use the stock I/C sprayer resovoire? Will I still have to run lines, or will the stock ones be sufficient... Oh, and I don't use it since I got a FMIC.
Thanks
Chris.

Chris,

give me a buzz during the day and I'll give you some pretty sweet ideas regarding WI in your car. I have it in mine, I just wish I had hooked it up earlier. I have run it on every turbo car I've owned. I actually have a real trick setup on my STi. 818 414 6600

http://www.turboice.net/ lots of info...

talk to ed at charged performance...he has been using water injection for years

Thanks Tom! We have a great new URL for the information site though after our server debacle - http://waterinjection.info/ (turboice.net will alias to it for a while)

Installing our AquaXede this weekend on an SR50 charged STi - next weekend it will be dyno tuned at turbotrix.

Ed.

I think down the road I am going to do an alcohol injection kit for my STi once I get a bigger turbo. The kit I am looking at is used by Typhoon/Syclone, & Grand National people. My buddy has a Typhoon and says that alot of the people on his boards love this system they use from SMC Enterprise. Granted it has a big control, but I think this is what I am going to use. Plus alcohol is really cheap (rubbing alcohol from drug store) and it will cool your engine alot cooler than water.

Here is the link to check it out:

http://www.smcenterprises.com/injector_kit.htm

Just thought I would show you another option.

Matt

Originally posted by PA04STI
Plus alcohol is really cheap (rubbing alcohol from drug store) and it will cool your engine alot cooler than water.

:huh:

http://www.waterinjection.info/images/wateri2.gif

There is no liquid that you can readily inject into an engine that absorbs more heat than water. And the cheapest and most effective alcohol is a 5 gallon drum of methanol.

The benefit of running alchy is that you can lean gasoline AFR to levels leaner than 12.5~13:1 and the alchy will provide a higher octane replacement for the pulled fuel that would have otherwise been burned. So you can pull gasoline fueling to 14:1 and replace fueling back to 12.5:1 (approximate optimum power AFR) using alcohol. In effect you have hybrid fueling. But if you are at 12.5:1 AFR and getting detonation - water is the best solution to surppress knock. Any additional liquid injected below 12.5:1 will not burn anyway as there is no oxygen left to burn it - so you might as well use the most effective coolant - water.

While alcohol has a high octane rating the octane rating of water is infinite - it will not self ignite under any amount of pressure or heat.

Very few people run pure water (think of it as water mixture injection) - it is usually a water/alchy mix. You use enough water to pull gasoline to 12.5:1 and the alcohol to replace any gasoline you pull further than that. The amount of alcohol you run can be limited by your system. For instance Aquamist can run 1:1 water:methanol (and will not tolerate ethanol) - while the SMC system is much more tolerant of various alcohols and higher levels. Also I think you will find that if you are going to run alcohol seriously - methanol is the most effective to use.

Ed.

Ed's hooking me up with one of his 2-D kits tomorrow. I will post in about a week when I get it tuned. The STI is set up really nice for WI with the rear IC sprayer tank. You could probably go 2-3 tanks of gas before you would have to refill it. Plus you get a dedicated WI tank that is alreay mounted and finished by Subaru. I personally will only run straight water as I don't want to worry about buying/mixing alcohol all the time. If you run out of fluid away from home...what do you do? You can find water anywhere, anytime.

There are a couple really nice things about implementing water injection on the STi.

You have the 4~5L tank already wired, hose routed and present.
You have a pump in the tank that will prime the primary pump no matter where you locate it.
The efficiency and potential flow rate of the Aquamist piston style pump will be increased significantly by the light inlet pressure of the sprayer pump. (Increasing its flow potential from 300cc to probably 450cc/min)

Running straight water would be fine except for our recent weather - having a few pints of Heet in the yellow bottles (about 98% methanol) wouldn't be a horrible idea. I will bring you a four pack. Or you could turn the system off when the temps are below freezing and switch to your nonWI map.

I haven't talked at length with Junior yet - but based on the review of his last tune it does not sound like he is pulling so much gasoline that it isn't all still being burned by the oxygen in combustion. I would classify it as a good moderate safe/power tune that takes advantage of the knock suppression of water injection but isn't pushing any limits - perfect for this implementation. So certainly methanol will not be required.

Oh and for the first week to month of good weather you will probably use a bit more than a tank of water per tank of gas from the thrill of getting on it all the time. Once the novelty has worn off a bit normal driving will likely see two tanks of water per tank of gas.

Ed.

Absolutely no doubt about it, my car runs far nicer and makes more power with pure methanol injected rather than 50/50 water/methanol. EGTs are 30C lower with pure methanol, it makes about 5% more power, is far more detonation tolerant and much smoother.

I think I'm just going to junk the water injection kit and add methanol to the fuel tank in a 10% mix. That way I don't have to worry about kit like DDS2 to see when the flow stops, separate reservoirs, accurate delivery etc.

I gave the 50/50 mix a try because everyone bangs on about how good it is, pure methanol was far better.

I tried a range of AFRs and boost levels, and different ignition timing.

50/50 mix was a non-starter in comparison to the neat stuff :)

With 50/50 mix once it got to 12:1 on boost it did what felt like a lean stumble. Not very healthy at all. Screenwash mixture containing methanol was similarly inferior to neat methanol.

Pure methanol mixture was very forgiving.

I know it is not fashionable to say, but the water injection setup I've setup seems a bit pointless now.

Just my experience, but it was quite obvious how much nicer the car was with pure methanol, it also benefited from larger nozzles until it was delivering about 10% of the gasoline.

I may have missed some detail on setup, but I doubt it.

I can get 353 WHP at 19 PSI on a 20G with methanol and 98 RON which is my evidence for the neat stuff.

Anyone beat that with 50/50?

Always hard to diagnose over the internet, but what was your timing like with the 50/50? Also where was this stumble in relation to your engagement point. (i.e. engaged at 3,500 rpm and stumbled at 3,600?)

Was your WB lambda sensor calibrated for the water/methanol mix? Because everyone I have contacted has pretty much stated that water will dilute the partial pressure readings of the sensor on the rich side it will read leaner than you truly are. So you were likely still 11.5:1 or probably richer. Which is too rich for water to be mixed in without hurting the combustion process - just too many hydrocarbons. With rich settings and water poor results are common. I bet you if you take that fueling to 13:1 and then try the 50/50 your results will be completely different and if not better than pure methanol than certainly no different. The reason that pure methanol is working better is more than likely attributable to methanol interferring less with rich fueling than water does.

With water mixture injection AFR targets almost go out the window. Properly implemented you won't hit a knock threshhold before either losing power from pulling fuel to leaner than optimum power or advancing beyond MBT. I generally tune to optimum power and then see what the AFR is and say oh that is nice - because it isn't accurate unless calibrated or measured with a four gas analyzer. In one instance at 13.1:1 I found on a four gas analyzer I was actually at 12.55:1 using pure water and my CO was almost immeasurable. That is the real trick when you have almost converted all carbons to carbon dioxide but not quite you are pretty much at optimum power AFR - this was also at MBT with no signs of knock. Afterwards I add a touch of fuel pulled a degree of timing and was safe and happy.

This paper also shows how when wanting the best knock suppression a water/methanol mixture works better than others - http://naca.larc.nasa.gov/reports/1945/naca-report-812/

It illustrates when you reduce fuel consumption past a certain point you need to replace the gasoline that would have used the oxygen with a hydrocarbon - in this case methanol. But up until the point where just all of the oxygen has been used up water will not interfer with a non excessively rich fuel tune.

Just from my experiences.

Ed.

PS - there are few implementations around you can compare your outputs with NavyBlueSubaru's who uses washer fluid (not something I recommend myself - I use distilled water and pure methanol) and a recent tune by Jr. Not sure if they are close to your setup. I don't have a direct comparison in front of me right now.

john banks,

I use the Aquamist 2D and have so on all my turbo cars. I was under the impression that the Aquamist components would not work well with more than a 50/50. I don't know that you are using, but I love methanol injection and would certainly run a bit more methanol if the components could handle it. I would still run a little distilled water, but not nearly 50/50. What is your take on this?

Thanks

Sincere question - I really don't know the answer.

Did a quick inadequate search - will methanol even remain in emulsion with gasoline? I know winter gasoline comes with ethanol in it along with the midwestern states that think it is a fuel. I mean if it were that simple why would anyone inject methanol and no just dump it in like it was toulene?

Ed.

Methanol will mix in all proportions with gasoline as long as its dry. At high concentrations >10% you run the risk of seperation of the alcohol/water/gasoline in the tank. Generally >10% mix of methanol also has corrosion issues. methanol is much more corrosive than ethanol. It does not like a couple metals that are commonly used in fuel systems in some cars, and it is electically conductive and promotes electrolytic corrosion. Methanol also has toxicity issues that are much worse than ethanol.

As posted above, research by NACA showed that max power was achieved with 50/50 mix and true air fuel ratios in the 13.x :1 range.

The down side of mixing methanol in the fuel is you burn it all the time, even when you don't need the extra octane.

I will however agree that the subaru engine appears to like alcohol mixes with the gas. I run a mixture in the summer time, and straight gasoline in the winter. (local gasoline here is blended with ethanol for emissions reasons due to the altitude, during the winter months.)

The unlimited air race folks use the 50/50 mix for their ADI systems (Anti Detonation Injection) but at much higher percentage of fuel than you tried John.

They run 0.8 - 1.0 lb of ADI fluid for each pound of fuel injected when no intercooler is used, and 0.4 - 0.5 lb ADI fluid per lb of fuel. This is about 4x the flow you mentioned trying.

The Buick GN folks used a seat of the pants tuning approach where they kept adding water/alcohol until the engine felt soggy, then they started pulling fuel, adding boost and timing. They kept up with this round robbin process until they stopped gaining performance.

You may not have had enough ADI flow for your fuel as you leaned it out.

Larry

Given that I replaced half the methanol being injected with water, then the 50/50 mix was leaner than the neat mix as methanol is a fuel.

I took the wideband as lean as 12.8:1 and there is no way I am going leaner, it detted and just would not take timing as well as getting very hot and did not make power. It was also not at all smooth but this was not related to the trigger point.

I am using about 10% of my injector flow for the 50/50 injection.

I have taken loads of fuel out of the fuel map (about 15%) from a wideband reading without additives of about 11.2:1.

It is all well and good quoting reports from aircraft engines with different compression and geometry, but water makes my engine horrible - makes no more power than well tuned pump fuel and runs hot.

Don't know what the Aquamist kit tolerates in terms of methanol concentration, but 10% methanol in the tank is fine without issues talking to people that have done it on Imprezas - AndyF got an 11.66 without nitrous using it and a 20G.

Is anyone running an Impreza engine at 13:1 on 50/50 at decent power - ie 350 WHP+ without blowing up?

The study quoted is using:

Different fuel
Supercharger not turbocharger
2 valve per cylinder not 4
7.0CR
2500 RPM
Excessive charge temperature

It does not appear to compare higher concentration of methanol than 70%, and does not demonstrate that I can see that the MEP is higher on 50/50 than 100% methanol.

If this is the basis for the 50/50 assertion I think it is flawed.

I can't find the thread I was looking for in search. I think NavyBlueSubaru may be the example you are looking for if he comes by this thread.

The paper was not listed as THE answer or even the basis for it. It was an example of a tangently related objective resource away from just my own experience.

I also never said that 50/50 was even the answer.

What I have said and will maintain is that when fuel is pulled to the point where the last amount of oxygen is just used in combustion water is a better knock suppressent than alcohol.

Previously I would have said that alcohol is only preferable for injection when there is still oxygen left to burn the additional hydrocarbons.

I guess I will add to it methanol is preferable when the gasoline AFR remains too rich for water.

I use 50/50 for freeze protection and to make absolutely sure all the oxygen is going to be used in combustion. I also don't know that I would be comfortable pulling enough gasoline to run a richer mixture of methanol optimally for street use. High levels of methanol injection are generally only done for drag racing - I would not tune in that manner for the street or track.

However each person has a different objective - and what they pursue should be compatible with their own independent conclusions and comfort levels. I would say that adding 10% methanol to gasoline is preferable to dumping excess fuel and if it suits the user great - but I would not let go that it would be more knock suppressing than a water mixture injection or that it would inherently result in a more powerful or safer result if both methods were optimally tuned.

Ed.

Just put 10% methanol in the tank and mapped for it results are dramatically improved.

Running about 12.2:1 AFR according to wideband is the sweet spot, go leaner and the EGTs rocket and it dets. Go richer and it loses power.

It is now making 364 WHP on Delta Dash which is equivalent to 376 WHP on our local dyno. Best on neat methanol injection was 341 WHP running the same boost. So in tank methanol mapped to the best of my abilities on my car shows double the gains of injected methanol.

Just uninstalled my water injection kit so a bit of weight saved :D

Oh well, it was worth trying I suppose.

BTW the methanol in the tank is taking 5 degrees more ignition timing than injecting water/methanol 50/50 allowed.

Knock suppression? I have to go on what I see, there is no way you could convince me to use 50/50 water/meths. The only reason for the WI kit was to allow water as you can't put that in your tank.

Theory is all well and good, roll on the Subaru results :D

Try actually getting to 13:1 at 20 PSI on a small turbo on an EJ257 and keeping the EGTs down. Isn't going to happen.

John:

No worries, I have run a very similar setup to what you are now, but think WI ADI has a lot more to give when folks get the right recipe. In aircraft engines it can yield 88% increases in power on lean mixtures when you get it right.

I always defer to actual experience, so if you get the preformance you want with that setup that is a good thing.

Just posting some info that may be food for thought for folks who want to explore the limits.

The Buick GN folks are running a lot more water/alcohol than most folks think is proper, and getting stunning times for a car that size.

The Unlimited air racers are running those .5 lb ADI sprays to 1 lb fuel on engines that are very similar to the subaru engine in some respects as they are 4 valve pent roof engines. They are running slightly lower compresion ratios of 6.0:1 to 6.9:1 or so.

The engines are much larger (translate more detonation prone -- with 6" bores). They are running up to 50 psi boost on 160 octane fuels so there are some differences as well.

By the way those NACA papers do not consider the alcohol as fuel when they compute the fuel air ratio. When they run up to 17:1 fuel to air, that is the computed ratio of gasoline to air.

Larry

You'd need a pump a lot more potent than Aquamist for those ratios though?

If only running about 10% of the fuel load (which is all I have tried) then it appears that 50/50 is nowhere near as good as methanol for detonation control, torque, power and EGTs.

And if you can't run 100% methanol through the Aquamist pump then what is the point of it?

I understand some of the high pressure bosch EFI pumps are alcohol safe and some are rated at output pressures up to 170 psi.

Maybe a trip to a breakers yard and a used pump is the way to go.

The ones I'm interested in checking out are:
Airtex # ----- car ------ flow ----- shutoff pressure
E8203 98-87 Porsche 60 gph 130-160 psi
E8185 78-77 BMW 4 55 gph 160-180 psi
E8176 77-74 Audi & VW 40 gph 150-170 psi
E8169 94-77 Porsche 70 gph 160-180 psi

Larry

:D

[QUOTE]Originally posted by PA04STI
[B] Plus alcohol is really cheap (rubbing alcohol from drug store) and it will cool your engine alot cooler than water.
.................................................. .....Please do some more research, to save you're self some probs. Ethanol is a big no no, in an aluminum engine, Methanol is good as long as it's at 50% or less. The other guys are right no liquid is as good at absorbing heat as water. Alk. is added for a bit more power in a properly set-up system. I've been running the Aquamist 2d system for several months and it's better for me than any system I've run before.

364 WHP with controlled EGTs on a 20G, don't care what is SUPPOSED to work, 10% methanol in the tank does work and is safe long term that is well proven.

John,

Adding 10% Methanol to your fuel will increase your octane level and will resist engine knock.

For effective in-cylinder cooling, you need to either dump fuel or injecting water. Putting 10% methanol into you tank will not have any addition cooling effect useless you are running a over-rich fuel mixture.

346 WHP with controlled EGTs must require quite a bit of fuel without WI, what a/f ratio are you running?

Originally posted by peter head
John,

Adding 10% Methanol to your fuel will increase your octane level and will resist engine knock.

For effective in-cylinder cooling, you need to either dump fuel or injecting water. Putting 10% methanol into you tank will not have any addition cooling effect useless you are running a over-rich fuel mixture.

346 WHP with controlled EGTs must require quite a bit of fuel without WI, what a/f ratio are you running? You must have somehow missed his very recent post in this thread where he claimed 364hp, not just 346hp.Originally posted by john banks
Just put 10% methanol in the tank and mapped for it results are dramatically improved.

Running about 12.2:1 AFR according to wideband is the sweet spot, go leaner and the EGTs rocket and it dets. Go richer and it loses power.

It is now making 364 WHP on Delta Dash which is equivalent to 376 WHP on our local dyno. Best on neat methanol injection was 341 WHP running the same boost. So in tank methanol mapped to the best of my abilities on my car shows double the gains of injected methanol.

Just uninstalled my water injection kit so a bit of weight saved :D

Jon, I think he transposed the last two numbers. ;)

John, thanks for the additional input and more for me to consider as I am always looking for new information. Your findings may be of interest to those who visit our information website. Always looking for those with something to contribute to provide to those coming to look for information.

http://waterinjection.info/phpBB2/

I have thusfar found that my results have always exceeded my expectations from using straight water to 1:3 water/methanol ratios on various applications and straight water to 1:1 on the WRX. In the coming months I will add more from results when used on the US STi.

Ed.

Sorry, I'm new to the whole water injection craze. When you talk about adding 10% Methanol to the tank are you talking the fuel tank, or wi tank? Sorry for my ignorance:confused:

Fuel tank, yes it is dramatically lowering EGTs. I am running about 12.2:1 AFR, you don't control EGTs with a small turbo at 20 PSI at that AFR on gasoline! You melt.

John,

Are you saying that you can reduce EGT by adding 10% methanol in your fuel tank - running the same turbocharger?

Yes. Dramatically so.

Originally posted by john banks
Just put 10% methanol in the tank and mapped for it results are dramatically improved.

Running about 12.2:1 AFR according to wideband is the sweet spot, go leaner and the EGTs rocket and it dets. Go richer and it loses power.John, despite what your meter might say, I doubt you are "running about 12.2:1 AFR" because your meter probably presumes that your fuel's stoichiometric AFR is 14.7:1 which, in your case, is not true. I'll guess that a blend of 10% methanol and 90% gasoline would have a stoichiometric AFR somewhere in the vicinity of 13.8:1. This is because straight methanol is about half oxygen (by weight) and has a stoichiometric AFR of only 6.45:1.

Your AFR meter probably has an option to display Lambda values instead of AFR values. If so, you should train yourself to think and work in terms of Lambda because, regardless of the fuel or blend, stoichiomteric is always 1.0, max power is always around 0.9 (±.05) and max fuel economy is always around 1.15 (±.05).

John,

I am not familiar with you car's management system, do you see a dramatic ignition advance when Methanol is added via some diagnostic equipment?

I can't quite understand this since there are no other hardware change.

Originally posted by Jon [in CT]
train yourself to think and work in terms of Lambda because, regardless of the fuel or blend, stoichiomteric is always 1.0, max power is always around 0.9 and max fuel economy is always around 1.15.

Good point - I need to make this transition as well. I wasn't sure if the lambda sensor read stoich or not for all fuels - good to have you confirm that.

I don't think it solves the partial pressure dilutions of the additional steam from the water side of my injection. (Which you first pointed out to me and I have since spoken to many lambda experst on.) There is no error at stoich but the further from stoich you actually are the larger the error is back towards stoich from the introduction of the present water in the induction. I think this can only be accounted for with calibration with a four gas analyzer.

BTW, John, do you worry about the methanol's solubility in gasoline or about phase separation? If not, maybe you should read:
http://www.hawaii.gov/dbedt/ert/afrw/afrw-05.pdf

John,

When you were injecting more water, couldn't it be possible it was too much water causing the mixture to ignite badly? As you know, the ignition system on the Subaru isn't the most powerfull.......

Just a thought.

Mark.

It could be Mark, but with a smaller nozzle with the methanol based screenwash it was not very effective at doing anything.

Peter I am using the factory ECU but it does not work like the USDM ECUs you may be more familiar with. I know what ignition I am running as it is not running any knock correction of relevance, it just runs the main ignition map plus the compensation map all the time, Delta Dash reports figures that agree with this. Rather than repeatedly reflashing I try to keep the shape of the ignition map nice and then use Delta Dash's live retard feature to retard up to five degrees if I am getting knock.

Thanks for the input on the mixtures, I was hoping someone would help on that.

What about a bit of acetone to help blending?

My AFR meter only reports AFR for gasoline.

Sorry I am on brain block, does the lower stoich mix mean that I am really running richer than I think by about 1 AFR point?

Temperatures have been down to freezing over the last week and with two tanks so far I've found no driveability problems.

AndyF has tried ethanol and had similar results, so if I can find a good source I might try some.

So if I want to run 0.85 lambda for a bit of cooling beyond maximum power then I need to see 0.85 * 13.8 = "11.73" on my WBO2?

Confused!

Originally posted by john banks
My AFR meter only reports AFR for gasoline.

Sorry I am on brain block, does the lower stoich mix mean that I am really running richer than I think by about 1 AFR point?

Temperatures have been down to freezing over the last week and with two tanks so far I've found no driveability problems.

AndyF has tried ethanol and had similar results, so if I can find a good source I might try some. First, let me say that it's amusing that Brits are experimenting with Gasohol/Gasahol (registered trademarks :)). My estimate of 13.8:1 as the stoichiometric AFR of a 10% methanol blend was basicly on the back of a coctail napkin and derived from simple volume proportions and not as a result of balancing a molecular relationship. This approach usually leads to incorrect results. For instance, if one looks at Table 3 in http://www.westbioenergy.org/reports/55019/55019_final.htm, one finds that the stoichiometric AFR for a 10% ethanol blend is 14:1. But if one increases the ethanol to 15%, then the stoichiometric AFR of that blend drops precipitously to 9.7:1.

Jon:

I think your alcohol percentage in the above quote, is incorrect.

one finds that the stoichiometric AFR for a 10% ethanol blend is 14:1. But if one increases the ethanol to 15%, then the stoichiometric AFR of that blend drops precipitously to 9.7:1.

As I read that chart, the 9.7:1 stoich A/F applies to E85 which is a mix of 85% ethanol and 15% gasoline.

That follows in line with E100 (pure ethanol's stoich A/F of 9:1. If you plot the stoich A/f ratios of the 4 values commonly available, ie pure gasoline 14.7:1, E10 @ 14:1, E85 @ 9.7:1 and E100 @ 9:1 you get essentially a straight line. I think you could simply interpolate from that the proper stoich A/F for other mixes. The same could be done for Methanol as well.

By the way thanks for pointing out that paper, it has a lot of info in it I've been trying to chase down for a number of years.

Larry

Gentlemen: if I want 0.85 lambda on my mix then I just need to read of 0.85 * 14.7 on my wideband for whatever mix I choose correct? So my indicated 12.2 is 0.83 lambda.

http://www.techedge.com.au/vehicle/wbo2/wblambda.htm

Should I just go off EGTs to determine how lean I want to be?

What lambda would you want to run with 10% methanol?

No John,

Lambda value is independant on the fuel used, AFR is. On petrol lambda 1 = 14.7 AFR, on a 10% methanol mixture it is 14.0 AFR.

You can easily calculate: real AFR = (reading wideband / AFR stoich from fuel mixture) * AFR stoich petrol

Eg. 10% methanol mixture and 13.0 AFR reading on the wideband: (13.0 / 14.0) * 14.7 = 13.65

Mark.

If you look at the Techedge calibration though it would show "14.7" running stoich on pure methanol. Which means that on any fuel if it shows 14.7 that is lambda 1. I was dreaming about this last night. Have another think about it and let me know if you still think as you posted. If you are sure tell me so.

I am not sure if I am correct to say that Lambda sensor is just an oxygen sensor, its output is directly related to the excess oxygen left after combustion.

If I am right, then it doesn't really matter what fuel you are burning?

That is what I am thinking too. So whilst I run 0.78 lambda on 98 RON for cooling and safety, I might want to run 0.83 lambda on the mix, which is my "12.2".

Jon in CT, AndyF pointed out to me that the Subaru fuel tank is quite flat shaped, and the fuel pump is constantly churning the fuel and mixing it nicely. Does this allieviate your concerns at all? Separation in static tanks could be more of a concern? What do you think?

Originally posted by john banks
That is what I am thinking too. So whilst I run 0.78 lambda on 98 RON for cooling and safety, I might want to run 0.83 lambda on the mix, which is my "12.2".

Jon in CT, AndyF pointed out to me that the Subaru fuel tank is quite flat shaped, and the fuel pump is constantly churning the fuel and mixing it nicely. Does this allieviate your concerns at all? Separation in static tanks could be more of a concern? What do you think? Yes, I imagine that AndyF is correct and there is much turbulence in the fuel tank containing the fuel pump, much of it due to the fact that fuel from the other side of the tank is continually pumped over via fuel return line pressure.

To re-scale an AFR meter to Lambda, just divide by 14.7. Here's the reason I raised this issue. Because methanol is half oxygen, you don't need to mix in as much air with it to achieve stoichiometry. Alternatively, for a given amount of air, you need to increase the amount of fuel supplied to the charge and, hence, injector pulse width. The stock ECU, relying on the front AF sensor, will, within certain limits, make this adjustment automatically. This adjustment will be reported as Long Term Fuel Trim (or AF Learning #1 by DeltaDash).

Thanks, Hotrod, for correcting my mistake on the percentage of ethanol in E85 fuel.

One last note on "excess" fuel. The main reason for doing this is to increase the cooling effect which the fuel's evaporation has on the intake charge, which increases its density and makes it less prone to knock. Straight gasoline can cool the charge by 40 degrees Farenheit when it evaporates. The cooling effect of straight methanol is triple that of gasoline. So, with a 10% blend, you won't need as much excess fuel to achieve a particular level of cooling when compared with straight gasoline.

A wideband device calculates lambda value from the O2 content in the exhaust gasses. This shown lambda value is independant of the fuel used. If you run the engine on a stoich 100% methanol mixture (which means an AFR of 6.4:1) your wideband device will show you a lambda of 1.0, if your wideband device only shows AFR and it is supposed to work with petrol only it will show AFR 14.7 which should be 6.4.

It´s all dependant on the setup of the software of your wideband device. I can enter the stoich AFR of the mixture I am using and my wideband will show the exact AFR. If this is not the case with your device, you have to use my calculation as above.

Mark.

The techedge calibration is shown, it reports AFR for gasoline. So when it shows 14.7:1 it is lambda 1 regardless of blending. So I am correct to say "12.2" is lambda 0.83 agree?

It's nice to see a post where there is an actual discussion instead of a pissing match. Back on topic. Please continue so, I can learn.

Originally posted by john banks
So when it shows 14.7:1 it is lambda 1 regardless of blending. So I am correct to say "12.2" is lambda 0.83 agree? I agree.

Jon wrote: Straight gasoline can cool the charge by 40 degrees Farenheit when it evaporates. The cooling effect of straight methanol is triple that of gasoline. So, with a 10% blend, you won't need as much excess fuel to achieve a particular level of cooling when compared with straight gasoline

Jon,
I agree straight alcohol has three times the latent heat of gasoline (by weight), can you give us some indication if this can be equated to John's claimed of 30C lower EGT with only 10% alcohol ?

Don't forget the higher octane also allows more timing advance which may contribute to the lower EGTs.

Yes, John, I did ask about the ignition advance earlier but didn't get any reply from you.

The mystery is now solved.

I looked at Jon's link and the corrosion due to alcohol is not pretty.

Most Gas in California is oxgenated (with alcohol) to boost the octane, it seemed that manufacturer has slowly use alcohol friendly components. Pity that the only half of alcohol molecule produces heat for power, the rest just exits the exhaust like nitrogen in the air.

How do do mix the alcohol to the right proportion?

I don't know about oxygenated fuel as it sounds like you guys like it, but the oxygen I'm getting seems to be good as it is free air. So I get charge cooling, octane enhancement and free oxygen :)

I mix it from a 5 liter can which I store in the trunk, I nearly always end up putting 50 liters total in as I usually have about 10 liters left in the tank when I fill up.

The techedge calibration is shown, it reports AFR for gasoline. So when it shows 14.7:1 it is lambda 1 regardless of blending. So I am correct to say "12.2" is lambda 0.83 agree?

Reread my posts John!

If your wideband shows AFR for gasoline, it doesn´t show you the right value when using different fuel. If your wideband shows you lambda value, than it is independent on the fuel used.

Mark.

You agree with what I posted then?

NO!

Perhaps it's the difference in language which makes it difficult for me to explain.....

If your wideband shows AFR you have to correct the reading. If your wideband shows Lambda it doesn't need to be corrected.

Mark.

My wideband is calibrated to give 2.5 volts output at lambda 1. However, the display reports this as 14.7.

When I read my wideband I am not reading "12.2" and thinking I am running AFR 12.2, I look at it and think "I am running lambda 0.83".

As Jon suggested I am now thinking in lambda.

Do you now agree this is OK?

Update:

Played some more and found that lambda 0.80 on the 10% tank methanol mix gives 2 degrees more timing compared with lambda of 0.82, and keeps the EGTs to 870 instead of 895 C just before the (overworked) turbo.

Boost at 6000 RPM is at its sweetest from 18-19 PSI and it consistently does 351-352 WHP on Delta Dash road dyno.

The 364 was a best blip with higher EGT and the timing was a bit too far advanced for what I would consider reliability.

The typical with injected methanol was 341 WHP on Delta Dash.

If I want to go further I need a bigger turbo :)

i know some guy that has developed a fairly inexpensive water injection system for wrx's or sti's. supposidly it is supposed to show tremendous temp decrease under the hood. something like 300deg. i have seen about 3 rex's with it installed and have heard good reviews from all of them. i saw one of the wrx's run at the track with it. he had other mods like boost controller, tb exhaust, front mount intercooler but i think it was still stock turbo and he ran 12.07. i was pretty impressed and i have been looking into getting a WI system for myself.