Flex fuel methanol, is it possible?

[Quattroking]

Hi all, I’m working on using blended petrol / methanol as a way of getting a higher octane fuel here in the uk, for use in a road car (but i must say obviously not on a public road, don't come looking for me Mr tax man!). For lots of other countries that have access to E85 were stuck with 99 ron (from shell or Tesco) and obviously we don't have access to bulk duty free ethanol!

I thought this would be the best place to put the info i find, theres not much on the web about flex fuel with methanol.

My 20v has all oem combustion chamber parts (for now) so comp ratio of around 9/1 (cranking comp 13.5~ random old gauge) with the reinz mls gasket and oem pistons / head / ABY cams. With the IAT around ambient temp 15C~ and a lambda of around .75~.73, 1.5 bar of boost and above (gtx3584rs boost threshold is 4600rpm), i keep getting different cylinder random single knock events even though timing is easily 3-4 degrees off MBT. EGT stays below 750C and a cylinder to cylinder spread of 30C. It's a bit of a concern as oem pistons, and it's like tuning it on a razor edge. Its all run on a MAXX race ecu, and right now its time to upgrade and probably over spec the fuel system (460kpa fuel pressure it starts to drop from 1-1 badly with a 300 kpa base). I'm using genuine NGK bkr8eix and about to try 9s if theres no slow speed fouling.

If anyone is gonna suggest to just have a duel fuel rails with 2 different fuel systems, there is no space to fit a second set and keep the rad in the front of the coupe, especially with the BST intake. Plus the meth is a PITA to keep it pure as a fuel, in my opinion not for a road car.

For me this is all a learning exercise to apply to a "fully built" engine and try to avoid the complexity of water/meth injection for knock control in a spark engine. I already have good experience with water/meth, my TDI drinks gallons of the stuff. If I have trouble controlling IATs with lack of intercooler size I will resort to using water/meth injection but only with a thought to IATs, not flooding the engine to deter knock. Depending on my piston choice i'll probably end up with a comp ratio of 8~8.3, and hopefully plans on using all that turbo will offer out to when the piston speeds get silly!

thoughts ill try to cover: might take me a bit to experiment and post it all
how it ‘should’ actually help
Fuel system requirements and compromises
How nasty is it on parts
Will it even stay mixed
Tuning blends of M5 to M40, iv got no experience

I’m just a guy in a garage trying to make some power reliably, far from a professional! If anyone has any info or corrections or just don’t do it, it doesn’t work, please jump in!

[Quattroking]

how it ‘should’ actually help

so what is the 99 ron petrol mostly made of?

low boiling point naphtha 80% ish
(benzene/toluene/hexanes/xylene, predominantly carbon chain numbers C4 to C12)
mtbe 0-15% (I bet they use this as much cheaper)
etmb 0-15%
ethanol 0-5% (V-power 2.7%)

so it’s basically the top fractions of crude oil and up to 15% oxygenated octane boosting component and a bunch of additives to keep the fuel stable and useable. Its these additives that should help keep the meth blended and in check when it starts to absorb small amounts water from the air. The UK already allows up to 3% meth in petrol anyway.

Blending to around 40% meth can bring the total octane rating up to potentially 130ron and bring some extra oxygen along with it. Also it pulls a lot of heat out of the intake charge and combustion chamber when it goes through phase change before its burnt, but I dint think it will have as big effect as running pure meth. Theres a bonus to running a blend, theres almost no risk of pre ignition that pure meth can have troubles with (would have to be wildly off with the tune, knocked it to death or have a glowing plug to do that on petrol). Hopefully i'll still be able to run a stoic mixture at cruse without issues. China are already using M15 in special fuel friendly cars.

There are some drawbacks though, phase separation being the biggest. Its where the meth absorbs so much water it cant mix with the naphtha and mtbe and separates into 2 layers. Like oil on water. This is obviously bad, very bad, car won’t run drain all the fuel bad. The only way to avoid this is use all the blended fuel before it happens or lots additives to keep it blended but this will throw the tune off. So the more meth mixed in the more water absorbed, but the more meth the higher the octane. Im planning on mixing up a 50/50 mix and leaving it open to the air in the garage, giving it a shake every day and seeing how long it takes to separate. Can’t do gravity measurements as temp and evaporation will change the results anyway.

Corrosion of parts is another problem, meth has a nasty habit of destroying most of the things it comes into contact with. Especially alli oxides and rubber parts. I'm hoping in lower concentrations the additives in the 99ron fuel will help with this as its already designed to have some alcohols blended in. Iv made many 50/50 water/meth tanks out of 5000 series ali and not had any problems yet, hopefully that continues. I'm going to take that same 50/50 mix and leave some alli and buna-n rubber in to see if it damages it.

Range is another problem as meth per the same volume has roughly half the energy content of petrol. This also means that twice the fuel volume needs to be injected. Cost comes into this too, buying in bulk 205l drums delivered I can get the cost down to the same as petrol but it only goes half as far.

Meth’s vapour pressure is also double that of petrol so the tank will Breathe more.

Hopefully I will only need M15 or M20 too keep knock and EGT in check when tuning close to MBT.

any thoughts welcome!

[Quattroking]

Fuel system plans

Injectors, big ones. Too big to use with regular 99. The best ones that I can find that easily fit are the 1700cc Bosch CNG injectors. There way too big to reliably get stoic at hot idle, so that means an injector swap for MOT time. The internet also says they change flow rate with fuel temperature and that mtbe swells the valve seat and they randomly change flow rate. I’m gonna find out for myself what they get up too.
My new balanced set tested at 1810cc and I’m setting the base fuel pressure to 2 bar to help the sizing on straight petrol. There basically garden hoses anyway so lack of atomisation isn’t as much of a problem as you’d think. I did test it, pressure at idle was 140Kpa.

1810cc/m x5 = 9050cc/m

I need to put a supply system together that can support 9L/m or 540L/h at 6bar to be able to hang the injectors open and not have the pressure drop. But also flow a small amount at no engine load to not over work the fuel.
First thing is can a 60mm pump that flows 340L/H @ 3bar (like a 044 with no check valve), with properly sized cables and relayed in the ABY coupe oem basket with oem internal fuel lines supply 540L/h with no load to a surge tank under the car (AN-8 lines with 30-micon filter external from the tank). If anyone knows that would be great and probably save me making a big mess. Might be not a big deal as no idea if the oem plastics and hose can withstand up to M40.

That will feed a small surge tank that I’m fitting where the 1st silencer in the old exhaust system went, next to the battery box. This will have 2 60mm pumps, one 340L/m and one 400L/m. I’m using old school 60mm pumps because the basket in the tank already has an adapter and it means I can swap pumps around if they fail on me on a road trip, plus have you seen the price of fancy pwm pumps…
The pumps y into the flex fuel sensor (in -10 adapter) then the 6 micron filter and up to the rail. Its going to be ali lines under the car with the return bent in an s shape to hopefully provide enough cooling. All lines and fittings in an-10, I did some ‘quik flow maths’ and the id of the -8 hose fittings i use brings the gross flow rate at 6 bar to just below 9L/m. plus I only have the stuff to work with -10 hardline anyways.
The ecu will switch the 400L/h pump on when needed, and check valves inline will keep the fuel going where it’s supposed too, including one to stop back suction from the main tank to stop it sucking air out the return.

Flex fuel testing, the magic of this whole thing. The E85 sensor used by GM made by continental can measure temperature and the alcohol/water blend in petrol. Quick lesson on how I understand it works, it sends an AC signal across the fuel to find its capacitance, it then outputs a square wave signal. The frequency is the % of detected alcohol/water and the duty is the fuel temp. The calibration of this sensor is key as the different alcohols have different densities which causes the sensor to read different values between ethanol and methanol. Even the amount of absorbed water will change the reading. So I’m going to mix up a bunch of different blends, from M5 to M40 and see the freq output to scale the sensor for as dry as I can meth at room temp. At the same time I’m going to measure the density at different temps/%’s to build a fuel density table to help the fuel model deal with such big injectors. Won’t go above 40C as need to do in sealed container and don’t want to have an explosion hazard. If anyone already has this info please share it! Would save me a bunch of work.

With all this hopefully sorted, ill be able to tune 4D tables in the maxxecu, and find the minimum blend needed to keep knock minimised at a sensible ish power level on oem pistons.

hopefully the fuel testing wont take me too long

[Quattroking]

Will it even stay mixed:

M50
yes, yes it does. stayed very stable over a 3 month period in a sealed container. (kept in my kitchen so 20c average and wild swings in humidity with cooking)

In an open container the fuel had evaporated to 2/3 of its original volume and was still phase stable. This was in a cold dry garage, averagely 10C and relatively low humidity.

The one thing i will note is when the fuel is spilt in small splashes as it evaporates you can clearly see the separation and the last to evaporate is mostly water/meth.

So with handling this stuff quite a bit now, i think it will be safe to use in the uk climate. I would not keep it in the car for more than a month, without flushing with fresh 99. Also this would be with a valved breather setup not just and open to atmosphere tank.

[Quattroking]

The GM continental flex fuel ethanol sensor:

The setup i used to test was the sensor wired into the maxx ecu in the car and an external plastic tank with a 044 and -10 lines to the sensor. The sensor was mounted in a nuke -10 bypass style holder that i'll use later in the feed side of the fuel system. The only downside to this setup was no pressure, only the resistance of the sensor and hoses.

I started with 5L of winter V-power and progressively added meth% (accurately using graduated cylinders), letting the pump run to blend the fuel for a couple of minutes between readings.

flex sensor meth data.png

I was hoping to custom scale the sensor output to find a linear line to make it easier to scale the fueling in the ecu. Unfortunately that's not possible as the density of meth is too different from ethanol.

Its also important to note that after 55% methanol (M55) the sensor exponentially gets less accurate. I would have a safety in place if the blend goes above M50.

Instead i have a table that scales the stoic AFR to correct that instead.
sensor scaling to AFR.png

I'm going to run this again with the summer fuel (in the summer) to check the work.

Next is building a density table, it's a bit more involved than i thought. Getting all the blends at -20C to 50c is a challenge.

[Ross]

I would go for the tertiary 'GEM' blend, it'll be within a few % of ethanol sensor readings. I ran approx 1000l of Tesco99/Ethanol/Methanol blends through my TTRS over a period of a few years on various maps. I've got all the flex fuel kit but never got round to utilising the flex map. The TLDR is you can pretty much replace Ethanol with Methanol at a 3/2 ratio on a given Ethanol map.

Search up some of the James Turner (Bath Uni) papers on the topic, all very insightful, has done extensive work with Methanol/Ethanol blends in collaboration with Lotus. I had a bunch somewhere but this is the only one bookmarked unfortunately: https://journals.sagepub.com/doi/pdf...54407014529424

The most convenient for me was 20% Methanol / 30% Ethanol mixed with 95RON - ran clean timing logs on an E60 map with that and easy to then just take 20l in a jerry can and add to 20l at the pump! I did some evaporation and phase separation tests and it holds up very well too at these ratios. Also, the bioethanol fuels that are hydrous (96% ethanol) perform just as well although the water may throw it off by a %. I ran mostly anhydrous fuel-rated Ethanol (99.9%).

It's also really useful to throw a gallon in the tank with 95RON to bring it up to 99RON if needed, methanol/ethanol or any blend.

I keep the alcohols in dry, dark storage in certified containers. Actually have about 120l of anhydrous Ethanol and a bunch of containers in my garage I need rid of if anyone interested.

[Quattroking]

I would go for the tertiary 'GEM' blend, it'll be within a few % of ethanol sensor readings. I ran approx 1000l of Tesco99/Ethanol/Methanol blends through my TTRS over a period of a few years on various maps. I've got all the flex fuel kit but never got round to utilising the flex map. The TLDR is you can pretty much replace Ethanol with Methanol at a 3/2 ratio on a given Ethanol map.

Search up some of the James Turner (Bath Uni) papers on the topic, all very insightful, has done extensive work with Methanol/Ethanol blends in collaboration with Lotus. I had a bunch somewhere but this is the only one bookmarked unfortunately: https://journals.sagepub.com/doi/pdf...54407014529424

The most convenient for me was 20% Methanol / 30% Ethanol mixed with 95RON - ran clean timing logs on an E60 map with that and easy to then just take 20l in a jerry can and add to 20l at the pump! I did some evaporation and phase separation tests and it holds up very well too at these ratios. Also, the bioethanol fuels that are hydrous (96% ethanol) perform just as well although the water may throw it off by a %. I ran mostly anhydrous fuel-rated Ethanol (99.9%).

It's also really useful to throw a gallon in the tank with 95RON to bring it up to 99RON if needed, methanol/ethanol or any blend.

I keep the alcohols in dry, dark storage in certified containers. Actually have about 120l of anhydrous Ethanol and a bunch of containers in my garage I need rid of if anyone interested.

Defiantly a good option! Lots of really good info out there about GEM.
Its getting the pure ethanol in bulk that's the problem, if it was easy I'd just run E90 and forgo the headache.

I get Virgin Methanol 99.98% delivered at well under a quid a liter, admittedly in bulk.

Hopefully I'll manage to get to the point where i can just add 20l at the pump and it'll still drive and start really well.
(Like the misses can just jump in and drive it on a weekday, even after a weekend of gapping teslas on the M40)
It's all in getting the ecu calibration right, and it's tricky with big injectors, wild swings in alcohols content.

[Ross]

The price of Methanol was always attractive! Once covid hit the Ethanol prices went a bit nuts and never recovered. Re: the tax man, I kept a small spreadsheet showing I mixed my blends on a few separate days using all the ethanol toward my allowed limit of personal biofuel, just in case. The law just requires a log to show production, not consumption, and I think it's 5000ltr/yr duty free. It used to be biodiesel but they changed the wording to include any biofuel blend or additive. If i recall i produced an additive for 50/50 use in tank

The Ethanol i have is from here; https://www.richter-chemie.de/bioeth...anol_217_1015/

Back then it was about £1.30/l shipped to UK for 600ltr but looks like you'd have to buy an IBC and ship to get it under £2 now. The stuff for bioethanol fires is what I started with, 96% but the water content didn't cause any issues for me (Although my garage fridge freezer was useful to prove that in sub-zero it did phase separate about 5-7C earlier in a GEM blend).

For what it's worth - cold weather starting I found the Methanol blend helped, would usually go on the first crank and if not always the second. I guess that extra 10% petrol in the blend helped.

My motivation was more about running the E85 map with reasonable convenience, cost and a daily that could run 10.0 sec quarter and 200mph VMAX in street trim (me and Sonic managed 10.1s and 198mph at VMAX , close but no cigar lol). It provided some fun over Covid, there was no way I was letting that stash get turned into hand sanitiser!