Open Source Biodiesel Tutorial

Here we show the complete documentation of a biodiesel production run. The video shows the process from oil settling to fueling the car – with titration and addition of reactants in between. We used 5 parts of waste vegetable oil to one part methanol, with potassium hydroxide as the base. We produced about 42 gallons in this run.

Open Source Biodiesel Tutorial from Marcin Jakubowski on Vimeo.

Used vegetable oil containst partially decomposed oil molecules, which must be neutralized. To determine the quality of the oil, we do a titration. This determines the amount of additional potassium hydroxide that needs to be added for the biodiesel reaction to go to completion.
The video shows how we determined the amount of potassium hydroxide via titration, and the details of this process are shown at the Collaborative Biodiesel Tutorial. Our result indicated that we should use 1.60 kg of potassium hydroxide for our 42 gallon batch of oil, with 8.4 gallons of methanol. We obtained about 42 gallons of biodiesel, and about 8 gallons of dark glycerol byproduct. We ran the reaction at about 120 degrees F for 4 hours, and let it settle over night. We had no soapy layer, just 2 clear fractions of biodiesel and glycerol.

Our reactor is based loosely around the open source Appleseed processor, to whom we are indebted for their clear documentation of the required plumbing. Of course we made some modifications, but the basic concept is sound – and we recommend replication. The two noteworthy points are that a steel drum works fine if you can braze on the required fittings. The only disadvantage is the relatively flat bottom, which adds an extra ~15 minute requirement for separating the biodiesel-byproduct fractions. But, if you can’t afford a standard $180 conical bottom tank, then a $5 metal drum will do. You will have to insulate it if you are running the process in winter temperatures. The second point is that it we used a 1/4 hp electric motor with a roller pump, as shown in a former video. The type of pump shown with the Appleseed processor lasted us no more than 2 hours. We found that a separate motor with a roller pump is a robust solution – up to the garden house coupler between the motor and pump.

Here are some more details of our experience. This stuff is great – if you are running small batches for yourself. It takes about 2 hours of hands-on time with a system like ours per 45 gallon batch. That’s about $50/hour value generated, given the 70 cent per gallon production costs. The main challenge is securing and handling used vegetable oil. It’s easy for small production like ours, but larger volumes require significant transportation and handling effort – turning into a laborious job. Plus, the supply of waste vegetable oil is not stable.

The real answer for the long term is to go beyond biodiesel production, which relies on unstable industrial detritus feedstock, and go to pyrolysis oil production. That, to us, is the real biodiesel. We can produce 500 gallons of pyrolysis oil per acre per year of sustainably-harvested grass biomass. Our goal is to show that a replicable economy can be based on such a biofuel, as part of a resilient community infrastructure.

So, if you have a metal drum, roller pump, electric motor, heater element, and some plumbing – you’re ready to set up your own biodiesel reactor.


  1. Inga


    Marcin, when you pick me up at the airport in Kansas City, MO July 27th I’ll have my first taste of a car run on homemade Open Source Biodiesel! Let the real life adventure begin!

  2. Abe

    Marcin, When you say “We can produce 500 gallons of pyrolysis oil per acre per year of sustainably-harvested grass biomass”, have you actually done this, or is it theoretical? Do you have an resources for actually making pyrolysis oil? Everything I have read is that it is really not a home-scale project.

    Instead of grass, or in addition to, you might think about fast growing coppice crops. Trees are more sustainable than grass crops, are are less prone to droughts and diseases. There are lots of fast growing trees like mulberry, willows, tree of heaven, eucalyptus, etc that make excellent coppice crops. Plant the trees, then harvest every 2-4 years. Stagger your plantings for continuous harvest. No need for planting every year. Once the get established they will be very drought and disease resistant.

    Just something else to think about…

  3. Marcin

    Just as a quick update, the present direction of Factor e Farm is pelletized biomass burned in modern steam engines. This is 2x as efficient as gas or diesel from a systems perspective, even if steam engines are 2x as inefficient thermodynamically.

  4. Sofia Baker

    i always use Biodiesel on my car to help the environment. Biodiesel is cleaner and is reneweable.,’-

  5. Marcin

    Abe, the proven figures for Missouri are 4000 dry pounds of biomass per acre. With additional measures for retaining rain runoff, I think we could double that. Proven yields of pyrolysis oil are up to 75% by weight, from what I recall. This gives us the 500 gallon per acre yield, if a gallon weighs about 10 pounds. This is not fiction or magic, but it does require that the production ergonomics are sound for this to be an effective route.

    At present, the above point is moot until we test out the potential of pelletized biomass for modern steam engines. Yes, coppicing may be a great boost here as well.

  6. […] Source:¬† […]