methanol recovery from biodiesel glycerine - solar powered

[Ivan]

Benefits:

• No fuel costs
• No sparks or flames = safer
• No moving parts or controls=more reliable
• carbon-neutral!

Glycerine is considered a waste product of the biodiesel reaction. It's a large potential energy source, but recovering it is tricky. One thing that can be recovered from the glycerine is the excess methanol which ends up there - estimates in the region of 40% of the overall volume (in my experience, it's probably a bit less than this). In any case, that's a fair bit of methanol (several litres per batch) that could be recovered.

I've basically done this by filling a solar tube (remember the 'Navitron Solar Kettle from last year?) with glycerine, cork in the top and leaving it in the sunshine. It heats up very rapidly (see pic below from 13th Feb). The cork has a hole in the centre, through which pipe can be inserted, running to the recovery unit (basically a water-cooled coil and/or reservoir. If you want to get fancy (I haven't, as I've just been doing some experimentation), you can make it a reflux still - get some 15mm tube, put a small thickness of insulation on it (we're not talking gas burners heating this thing!), and some steel wool inside - this should increase the purity of the recovered methanol. There's plenty of examples of 'normal' reflux stills on the web that you can pinch ideas from.

The solar still is slower than a conventionally heated still at this time of year, but give it a month or two and you should be able to distill most of the methanol off within a couple of hours. I'd say, it takes a full afternoon and perhaps a little more to get the bulk of the methanol out, in February. I'm really trying to purify the glycerine rather than recover the methanol, so I need to run extended periods at 160C - not easy in my location, as at ground level, the sun is still partly shaded by mature trees.

Depending on which tubes you chose, you can use 0.5litres - 2.2litres of glycerine per tube. If you want more, simply connect several tubes in series.

This method is great for experimentation and semi-serious use. If you want to do it in greater volume, use the Navitron SFA solar panel - this can process 80-300litres of glycerine per batch, depending on which size you go for. It might take a couple of days to recover the methanol, but you haven't got any naked flames or electric sparks (=much safer than most DIY stills), and it's not costing anything to run, so just walk away from it and come back later. If you are intending to purify glycerine, then perhaps the Navitron SFH panel might be more suitable, as it has heatpipes and doesn't allow any glycerine into the actual tubes themselves. I haven't got many pictures, so if this is unclear, let me know and I'll clarify.

[Ivan]

I'm also working on a design for a solar-powered biodiesel reactor. At the moment, my processor consumes the following energy (for approx 100litre batches):

Drying oil prior to reaction (if required)                  ~6kWh electricity
Heating Oil and maintaining throughout reaction      =6.4kWh electricity
Washing                                                          = 6kWh (hot water - usually solar-heated)
Drying after water wash                                     = 10-15kWh electricity

Worst case scenario, with fossil-fuel/electricity, this is over 30kWh of energy - which could cost £4 per batch (adding 4p to the cost of each litre produced

It's perfectly feasible to adapt a Navitron SFH solar panel to make a biodiesel reactor, which should reduce electrical requirements to pumping costs only, although heating time might be somewhat slower than a normal reactor. On the plus side, no chance of ignition of methanol from dodgy heater elements, sparks on heater controls etc.

[jason]

Ivan nice idea and any thing to bring the cost down

[Twenty4Seven]

Ivan

Thanks for elaborating - looks very interesting. I've got about 120 litres of glyc to play with ATM so I guess I need a solar panel. When the methanol is removed, the viscosity increases - does this pose any problems?

Cheers

Nick

[Ivan]

Once you've removed the methanol, assuming you are using NaOH as the catalyst, you will find that the glycerine sets solid. But this only happens when the temperature drops. I haven't actually checked the melting temperature, but I guess it's around 20C or so. I have left the glycerine in the tubes overnight, and it freezes solid, but half an hour or less in strong sunshine, and it's pretty runny again.

I've just checked, and we don't have SFH panels in stock at the moment, but the SFA panels should work just as well, the only problem is that you wont be able to drain every last litre of glycerine from the system, but that shouldn't cause any problems. In fact, the larger heating area of the SFA panel may be an advantage - and it's cheaper! We should have SFA panels from 104litres to 330litres in stock

[PEMTEK]

One way to maybe reduce your power consumption is to use ion exchange resin to wash the diesel.

A few of my friends are now using the columns and literally just leave the diesel to separate and then pump it through the ion exchange column and then through some fine filters and its ready for use.

If you try washing the diesel with water immediately after this process the water stays clear. I believe as long as you dont contaminate it with glycerine then it lasts very well and reduces time considerably. The resin columns can be initially expensive but a friend of mine i believe has found a relatively cheap source of them. I believe they are used in allsorts of processes and one being the mass purification of water.

http://www.biodieselwarehouse.co.uk/index.php?cat=Low_Price_Biodiesel_Ion_exchange_Resin


Phil

[Ivan]

The resin is great to use, from what I've heard (although I haven't tried any myself) - essential for commercial production. However, it is quite pricey, and I'm sure that reflects a lot of embedded energy. The resin can be regenerated to a certain extent by passing methanol through it (when I was in the lab, you'd need 10 column volumes to regenerate an ion exchange resin). I'll try to get round to testing resin at some point, but yet to find any at a realistic price, unless I buy several years supply at once!

BTW, the cylinders in the link look like dosing cylinders from pool suppliers. I wonder whether a cylinder could be cheaply produced from some soil pipe and a couple of caps

[Twenty4Seven]

I use the GL process - no water washing, no resins, magnesol or filtration required - just gravity to drop out soap after removing the methanol from the raw bio...

http://www.graham-laming.com/bd/ecosystem/state_diagram_new.htm

Nick

[jason]

Hi Nick
thanks for the link 
how do you find the process just seems to be a lot of gloop in the bottom of the settling tank i suppose this could still be washed in the same way with hot water or is their no need?
Jason

[Ivan]

I've heard people say if you leave the biodiesel long enough, you find the soap at the bottom, but having left biodiesel for five days, then done a wash I can confirm there's still plenty of soap in the biodiesel and I've also found that any soap that settles out does not drain easily, due to it's semi-solid nature.

Graham's process takes advantage of the 5% prewash process. I think this is very effective, but it does put water into the glycerine, which makes it more difficult to recover the methanol, and virtually impossible to purify the glycerine

[Twenty4Seven]

Hi Jason

No, you can't wash the gloop as it is mostly soap. Graham puts the gunk in a pillowcase and hangs it over the settling tank to drain the last of the bio. The stuff that remains in the pillowcase can be used as liquid soap.

Hi Ivan

If you remove the methanol from the raw bio, as per the GL process, the soap will settle out in a few days (2 or 3). It is somewhat temperature dependant, taking a bit longer in the cold weather.

The 5% prewash is not essential though it does make the glycerine drop much quicker and means that less soap remains in the bio. Graham is currently investigating methanol removal from the whole batch, immediately after reacting and before removing the glycerine. To prevent a reverse reaction, the NaOH is neutralised with sulphuric acid. Oh and no 5% prewash is necessary so the methanol purity is not impaired. It does require additional energy input though so I still favour the solar still approach.

Cheers

Nick

[Ivan]

So it's the removal of methanol from the biodiesel which is causing the soap to crash out in the GL process?

Have you tried washing your biodiesel after the methanol recovery/soap precipitation to check that all the soap has been removed? if so, it sounds as if it could save a lot of waste water!

[Twenty4Seven]

So it's the removal of methanol from the biodiesel which is causing the soap to crash out in the GL process?

Yes, exactly so.

Have you tried washing your biodiesel after the methanol recovery/soap precipitation to check that all the soap has been removed? if so, it sounds as if it could save a lot of waste water!

Indeedy..... I'd never go back to water washing now. The soap clumps together once all the methanol has gone. You can re-circulate through progressively finer filters if you're in a hurry to finish a batch in one day but I just let gravity (and a little time) do the work.

Here's a pic from Graham's site of the result of a shakem-up test.....

    Nick

[JamesTrike]

Ivan, List,
               I also have been using the GL process for around two years now.. and I can confirm that if the methanol is removed then the soap just simply drops out. I have the addition of a bubbler unit in my settle tank to remove the very last traces of methanol as It is very expensive (takes a lot of heat) to evaporate it. To "demeth" the gylc by solar is an interesting proposition, especially if its a start and forget process.

James

[Ivan]

Well drying is also an energy-intensive process, so de-meth sounds like a viable alternative.

An SFH solar panel would probably produce more than enough heat on an average summer day. Of course, oil has lower heat capacity, so will heat up faster than water. Also, the SFH panel is well insulated - 75mm polyurethane (which is equivalent to around 110mm of fibreglass, I think).