I beat the Prius!
A more professional title of this page may be, “SVO Hybrid vs Electric Hybrid,” to highlight the fact that to compare hybrids you are comparing cars that can switch between petroleum based fuel (e.g. gasoline, diesel) and something else (e.g. electricity, straight vegetable oil). As stated in my page on the energy balance of fuels, running on straight vegetable oil is the only way to ensure our human efforts are taking advantage of the energy inherent in our resources (also SVO is carbon neutral).
A GMC Hummer owner once told me her car was more environmentally friendly than a Toyota Prius because the biggest environmental impact is in the manufacturing of a vehicle. The Prius’s batteries are expensive and when their warranty time is up, the technology will be obsolete (relative to newer generation, cheaper electric hybrids of the day) so it will be crushed and recycled first. From the other camp I’ve heard the same argument made about nuclear power plants, but I digress.
I choose to drive a luxurious 1982 Mercedes Benz diesel with a kit to make it a straight vegetable oil (SVO) hybrid. Like the Prius and it’s brothers, it is a parallel hybrid, it runs exclusively on costly fuel part of the time and an alternative the other part. In a future page I will identify the different ways to convert an engine to run on SVO. The following analysis pertains to my current system (no looped return). My commute is 25 miles one way, and my car gets 23 miles/gallon.
The way a SVO hybrid works is based on 2 tanks – the original one full of diesel/biodiesel and an aftermarket job that filled with regular old vegetable oil. The car is started on diesel and after the engine temp reaches operating temp (80 deg C ~ 10 minutes), I flip one of two switches that changes the source tank supplying the engine to fry/donut grease. Diesel engines have a primary fuel pump that introduces a pressure in the fuel lines ahead of the injection pump. This primary pump moves much more fuel than is burned and all that extra fuel is flushed through a series of lines and returned to the tank. The second switch controls where this extra fuel is returned. So when going on a journey, how much of a delay should there be between flipping the source switch and the return switch? What is the best delay when switching back?
After doing the following analysis, I filled up with 14.1 gallons after 712 miles and drove to work and back 35 times. Thus I got 50.5 miles per petroleum gallon and beat the best electric hybrid on the road. It more than doubled it’s unsustainable-fuel efficiency – a newer model that starts with over 30 mpg would get 70 miles per petroleum gallon.
So what is the contamination tolerance? Mixing SVO and Diesel No. 2 in 85 deg F at various concentrations, I found a relationship between VegDensity to Viscosity, the horizontal red line marks the viscosity of diesel at 32 deg F – the same a 50% veggie at 85 deg F.
the data is fit pretty well with a simple parabola: viscosity = c * ( % SVO )^2. This is plotted in blue below:
Then I got my car hot and running on veggie, pulled the return line off the engine and began dumping it into 4 oz babyfood jars. The fuel at the start was less viscous than the pure WVO I used when generating the blue curve above as there is diesel in my WVO tank so I adjusted the parabola to fit my fuels as the magenta line shows.
Just by looking at the jar, you can estimate the length of the purge cycle (when the color stops changing). I measured the viscosity with an Ubbelohde viscometer (Capillary No. II which is accurate for fluids with viscosities in the range 4-60 cSt) to get the following curve:
Note the last few data points show there is no more SVO being purged – the viscosity of diesel at 85 deg F is about 4 centiStokes. Now from the parabolic relation I found between viscosity to percent WVO, the above data can be converted to percent of vegetable oil:
From this data, you can find at any time after the switch was flipped how much of the total veggie oil in the engine has been purged which is plotted in the magenta line above. For example, about 85% of the oil has passed through after 20 seconds (assuming the car is merely idleing).
I measured the flow rate through the return line to be about 1/2 oz per second on my car (again at idle). Thus about 10 oz of veggie is in the engine when the purge cycle begins. Furthermore, after 20 seconds about 2 oz. of vegetable oil is still in the engine.
I did this same measurements for Jeff Coombe, my partner in veggie fuel shenanigans, who runs a 1999 Dodge Cummins. The charts follow all the same trends, but has twice the flow rate and twice the veggie oil in the engine when the purge begins. The Fraction hits 50% at the same time, so the optimal purge times are the same.
Analysis:
Combining the knowledge of how much oil is in the engine upon purge initialization with the data of the percent of the total that has passed gives an estimate how much SVO and No. 2 have passed as a function of time:
Points can be pulled off this chart in a spreadsheet model. Assuming it takes 1/3 of a gallon of diesel to heat up the system every time I warm up the engine, I made estimates of how many purges to expect from a full 20 gallon diesel tank and how high the veggie density gets at the end of that tank. The model is based on how much of each fuel is transferred when switching to SVO at the beginning of the journey and back again at the end which I read right off the Fuel Transferred plot above.
The left plot is Number of Purges I can expect for different purge times (both on the initialization purge of diesel and on the final purge of vegetable oil). The right plot is the resulting contamination of vegetable oil in the diesel tank. As you would expect, the model indicates longer veggie purges result in lower contamination of the diesel tank when it is time to fill up. Same goes for shorter initialization purges.
Conclusions: The viscosity of diesel at 32 deg F is the same a 50% veggie at 85 deg F. So in the Summer I can tolerate more than 20 second diesel purge and less than 10 second veggie purge. But that is assuming 100% diesel in the primary tank which is for condensation reasons is not to be drained that far regularly. The rule-of-thumb I take away from this is to maximize separation. In the Winter, the system has a much lower tolerance as can be extrapolated from the following plot:
Relationship of temperature & viscosity in Diesel Fuel and Vegetable Oil
As I do not purge at idle I’ve had to adjust the times found in the analysis based on the sounds of the engine. My sense is that the 50% mark in the Frac Veggie vs Time plot happens at about 15 seconds (which corresponds to the conventional wisdom). It makes sense to use this “sound metric” because of how quickly the fuel type changes (as seen in how fast the colors change in the jars in the pic). For this whole tank I purged diesel for about 12 seconds and veggie for about 18. I plan to continue to beat electric hybrids throughout the winter. Will be listening at the end of this tank in the dead of winter for any discernible changes on cold starts and try to measure viscosity.