While you make a valid point about the temperature/volume relationship it probably isn’t as dire as some believe it to be, particularly if you are refuling from underground tanks. Temperatures over a metre below the ground vary less than 3 degrees C annually and are virtually stable below 3 metres. Most bowsers are more than a metre underground so that the biggest effect on temperature will be the time gap between your filling up and the last visit of the tanker. Remeber that petrol is a relatively light distillate and that it evaporates (the tankers and undergrond tanks are ventilated) and that assists is keeping it cool. Here in relatively warm Oz most fuel deliveries are at night, mainly for logistics and safety reasons but it helps Oilco profits not to have it evaporate before it can be sold.

Relative expansion between the tank and the fuel is virtually irrelevant. Most cars coming out of Japan and Europe have impact resistant plastic tanks which are very temperature stable and even the old steel tanks were usually bolted quite rigidly to the vehicle indicating that tank expansion probably wasn’t an issue. The fuel overflows simply because of its own intrinsic expansion qualities. Back in the distant days of youth when endurance racing motorcycles we chilled the fuel with dry ice, there was nothing worse than having expanding freezing fuel seeping into your leathers and then evaporating from around your nether regions.

What you are bringing to light is really another source of experimental error, but given all the other unavoidable error sources that must be endured in this type of backyard experimentation I think it more important to rely on the largest possible sample you can generate. Again in a single vehicle experiment you run into further difficulties as seasons and engine tune change along with vehicle lifetime issues. No-one said being a backyarder was simple, you’ve got to be in it for the fun and the journey, any usable results are a great bonus.

Our experimentation back in the 70’s was all done without much electronics. Our location was semi underground, ie below ground level on two sides a sloping ground level on the third and normal access on the 4th. Although we took ambient temps at the begining of each session I think we were probably more concerned with the temperature of the water in the dyno tanks, that could easily be raised several degrees C in fairly short run times and needed to be taken into account to calculate the correct power numbers. One of the reasons we did a lot of short running (after getting everything up to temp first) was that our biggest metered flask was only one litre (doesn’t last long in a truck engine) and a lot of out tests were done using 250ml pipettes as our “fuel tank”. Timing was by both hand held stop watches and by a modified glider barograph, a modified mechanical tacho (with genuine bird feather quill inserted into end of needle) removed soot from acetylene blackened alfoil wrapped around barograph roll (I was very good at getting coffee when foil replacment was imminent). Despite sounding primitive it was accurate and actually worked quite well. It was certainly good enough for just below state of the art research. I can’t recall ever measuring fuel temp nor whether it was considered a variable.

Having spent a lot of time over the past couple of years trying to get a Megasquirt to run I know that at least my version of Megasquirt does not take fuel temperature into consideration. It only monitors throttle setting, manifold and fuel pressures in calculating injector cycling tmes. I think that in a carburettored engine evaporative effects might outweigh any effect that fuel temp might have and I’m just plain ignorant on what happens with injection systems other than the Megasquirt.

Craig,

I think to be realistic you have to say that your current results indicate a 10% variance between observations and as such no valid conclusions can be drawn. I think that you will probably find around the sort of variance you found on the first run every time you fill up, it’s completely normal, particularly with all the variables you have in milage accuracy, fuel pump accuracy. etc. Have you measured your cars odometer against those 5K tests that the RTA is so fond of putting on the highways these days? It’s just another quality control facet yu can cover and be able to correct for when doing your sums.

It might take a while longer than you’d like but I suggest that you institute a regime where on one fill you disconnect the hydrolyser and reconnect it on the next fill. If your wife is the driver don’t tell her which is what (or if you like don’t tell her the truth…you might just get away with it in the name of science) That way you will at least get some adjustment for seasonal, tuning and driver predisposition status. (assuming that the car doesn’t need tune-up each tankfull)

The Thylacine

I guess what it comes down to is, as suggested in previous comments is to use the average over a number of ‘fill-ups’ to come up with any really useful figures, which was not possible here becuase of the onset of our mixture problem but I do now wonder if the figure we achieved can be taken as any indication of a fuel saving?

When any new car is released the manufacturer has to supply fuel consumption figures and I believe there is a standard in place that dictates how these figures are calculated. I will investigate that further, I think it will be interesting to see how they deal with any variables.

I realise now that I have attempted to ‘over-simplify’ the procedure and given the small (9.2%) decrease in fuel consumption I wonder if we just got lucky!

If the difference had been 20 or 30 percent then maybe the test would have been more conclusive.

Craig

Temperature has a large impact on the volume of gas you receive and to a lesser extent, the actual physical volume of your fuel tank.

Also filling style has a big impact, and if you are inconsistent, or ambient conditions work against you, your mpg calculations can be misleading. Having a very accurate volume usage is key to credible mpg stats, so outside of lab conditions, we have to do the best we can.

I understand you can expect about 1% more volume (1% less energy) for every 10C in fuel temperature. (supports the fillup at night premise.) There’s about a 30:1 expansion rate difference between gas:steel. Perhaps this is why a really full gas tank on a hot day will overflow, the steel shell can’t expand fast enough to keep up.

I’m beginning to note the temperature at time of fillup as a fallback position to my scientific method. I suspect temperature effects, but also feel the temperature of fuel recently pumped from underground fuel tanks probably doesn’t vary as wildly in temperature (and thus expand) as everyone assumes. It will expand once out of the ground and in the car’s tank, and possibly overflow and be wasted.

My primary concern is the actual fuel tank expands on a hot day, and thus has a bigger volume (taking more to fill), and this could be throwing off the denominator of the mpg calculation, especially if a prior filling was done under much cooler conditions.

Any suggestions for a reliable fill-up technique? I’m somewhat reluctant to just let it go on the “first click” where the pump automatically shuts off, as I feel like all automatic shutoff mechanisms or the general position of the nozzle in the fill tube may introduce some new error source.
Regards, USA Jim

Imagine this for a moment, someone that is ‘mechanically challenged’ gets a hold of one of these guides and follows the instructions to the letter and starts driving his/her car around. Performance is great(engine is running lean), fuel consumption is down(some of this can be attributed to the system but once again the engine is running lean)and he/she is a happy camper.
The trouble with this scenario is that from the factory the ECU is set to run the engine slightly richer than the stoichiometric A/F ratio of 14.7/1 to keep combustion chamber temperatures at a safe level and obviously these methods could have the engine running at any level - nobody knows because they don’t cover testing Air/Fuel ratio!! Running an engine leaned out for any period of time is going to do damage. Not good!

I have seen a number of people argue that the introduction of the HHO gas ensures that combustion temps are kept in check and that in fact the engine is not running any leaner than before but I’m not convinced.
Obviously the lubricating element of the fuel is reduced when leaning an engine out and I can’t see the amount of HHO gas that is used providing the same level of lubrication which brings me to the question of whether or not an additional upper cylinder lubricant needs to be used (as when converting to LPG)
If you were using this system on a carby engine I would say go for it, you have nothing to loose but using it on a efi engine, in my opinion, opens up way too many questions that so far have not been answered.
Don’t get me wrong, I have not given up on testing this out I just need to put some serious thought and time into how to best solve these mixture/lean out problems.

On another note - something that really gives me the sh**s is the way these ’systems’ are being marketed by people that obviously have no mechanical background and couldn’t care less(or have no idea) of the implications of running an engine lean and there is certainly no mention of the time and knowledge that it would take to find a viable solution for this issue. If you believe the sales pitch the system is simple and easy to install and requires minimal tuning to get the most out of it. Sure, on a carby or diesel engine this might be the case but as we found out there is more issues that need to be tackled.
Ok, I feel better now!

Thanks for that link Eddie, certainly looks interesting! I’m keen to find out everything I can about this subject and will have a read once the ferals are in bed.

Take Care,
Craig

Your explanation is good. Assuring to read the manual gives you pointers on the addressing the issue - are they easy or complex? To you have confidence in their suggested solutions? Your conservative approach (measuring amps & volts and oxy sensor output) to ensure you have a good air fuel ratio is very good which prompted to ask if you had confidence in their ’solutions’!

Also came across this website for a patented HHO generator. Nice unit but unfortunatley very expensive ($2500…) unless your a freight company/truck driver or owner of engines running long hours (burning a lot of fuel). http://www.H2fuel-save.net.au some interesting links on previous universities studies on this topic & existing user feedback.

Cheers,
Eddie

Craig

Regards,
Craig

Interesting comments. Which only puts questions in my head.
How is your computer having problems?
What do you mean by catalytic converter is fumey?

I agree adding a volt meter and amp meter will aid diagnostics.

Does the product manual offer much advice for above issues?

cheers