Minimum amount of methanol

[dgs]

I completed a test this morning just to see what little methanol I could get away with. It was an answer to Rossey who asked the question on VOD.

I glyc washed 300 litres of average oil. This oil would be mostly rapeseed but as it comes from different places I do not know the exact composition.
I glyc washed it in my M67 set up. The glycerol I used had been used twice before for pre-washing, in fact it was set aside ready to be returned.
All I wanted to do was to neutralise any FFA's and dry the oil.
A titration after the pre-wash confirmed no FFA's were present and a 10/90 showed no conversion.

200 litres were transferred to my 220 litre conical. No additional heat was used and the 3 reactions I did were as follows;

1st reaction 17 litres of methanol+1000gms KOH 2.2mls dropout.

2nd reaction 5 litres of methanol+300gms KOH 0.4 mls dropout.

3rd reaction 1 litre of methanol+ 70gms KOH Zero dropout.

Total methanol used 23 litres (11.5%)

Total KOH used 1370gms (6.85gms/litre)

I realise the above result would not have been as low in chemical usage if the oil had not been glyc pre-treated. For the purposes of the test it was important that I started with dry, neutral oil with no conversion.

If I had used ex process glycerol for the pre-wash the chemical usage would have been even lower.

The fact that I also use a TAM 120 fitted with one of Uberveg's 1/2" eductors is also a bonus to the process.

[Dickjotec]

Interesting, what temp do you take it to in the M67 and what was the process temp? I assume you must be using a larger than standard tank on the M67 with a pump attached to be able to glyc wash?

[dgs]

I test the temp by 'feel' on the M67, probably between 50 and 55degs. I have a cone bottom drum with the M67 in and a 1000 litre conical a few feet away with a TAM 120 in the middle. I can pump from either into either vessel or together.

The process temperature  (in the 220 litre plastic conical ) starts at ambient, but the 120 (another one) increases the temp by anything up to 15degs per hour.

I now do all my processes at ambient, but after the last reaction I leave it to sit with the last reaction glycerol present until ready to de-meth. I then heat and mix for 30 mins (simply to reduce or get rid of any mono's present) then use this heat to also de-meth.

The rule of thumb that the reaction takes double the time for every 10 degs less is not correct. At 30degs it probably takes up to 50% longer than at 50degs but certainly no longer.

After the 1st reaction methoxide is added Which I do by sucking it in with the pump in about 10 seconds (I suppose you could call this a venturi, it is a 1" diameter pipe) The contents of the processor turn opaque (methanol and oil not being miscible) At temp (55 degs) this effect lasts about 3 mins before clearing. At 15degs it lasts about 4 mins. When I have done a 10/90 immediatly after the mix is clear it shows around 85% conversion If I added chemicals to convert in one reaction.

[Drum]

Great data as usual Dgs. I don't expect you to rush out and buy 200 litrs of fresh oil but I would really like to see the minimum methanol usage with fresh oil. To set an absolute minimum so to speak. Your mixing must be excellent. I could not get close to 11.5% and still get zero dropout, even with three stages. Or four as some night describe your experiment. Are you sure the stoichiometric value isn't actually more than 11.5%?

[dgs]

Hi Drum,
The minimum amount is probably greater than 11.5%. Which leaves us with more questions. Even though the glycerol I used to neutralise and dry had been used twice before, I realise it would still contain some methanol, but how much is anyones guess, it was really thick stuff. What I didn't want to do was start the test with wet oil or a titration of ? (maybe 2.0)

I still have the feeling that I slightly overdosed with the methanol to what I could possibly achieve.
For instance on last reactions where we are adding 1 litre or less + KOH, I have added just KOH into the processor before on a few batches (something like up to 50gms) and it does work, but for some reason I didn't do it with this test.

I have titrated the new soy that I get and it is 0.3 (KOH)
I will have to save up some of this and do a process on only new stuff (good old Bookers)

[dodge50]

This thread is perfect timing for me. Glycerol washed, then used the last of my methanol to mix 20 litres with 1500g KOH for first stage [in 200 L oil plus 10 L Glycerine that didn't return after putting 30L in] and was worried I didn't have enough meth. Be a few days till today's methanol order arrives to complete process [if required.....usually is] so thanks for the reassurance.

[dgs]

Good Stuff Dodge50,
Let us know how you get on.

[dodge50]

Not good. Only 30% conversion. Pump mixed for 2 1/2 hrs at 57C. It did contain about 50 or 60 litres of fat which had to be hot-air gunned through the sieve however [before processing]. Used the last of my Trinity reclaimed meth. Ordered new lot from Sidwil which is pure as wasn't too impressed with the reclaimed stuff as it had a brown tinge which made it hard to see if the KOH had dissolved. In this case though it had.

[dgs]

Thats a shame, I wonder how much the off colour methanol had to do with the low conversion or if the fats titrated quite high.

[dodge50]

Finally found the problem with my conversion rate. It wasn't water or bad chemicals....just Time. I've been hampered for years by the belief that the reaction can go into reverse but was assured this year by a uni chemistry lecturer that it can't. Manuals from 3 different processors plus the guidelines for the ASTM flask all suggest a process time of 1 hour to 2 hours. The last 3 batches [200L each] have been processed for 5 hours in one go and produced 97-98% conversion and 190-195L Bio. Methanol's been reduced from 40L to 30L with no ill effect and I threw in 1800g of KOH [without titrating although usually it titrated between 1500 and 2000g].
Glycerol wash, process for 5 hours [57 deg] in Chinese stainless processor, water spray wash [moves to Green Genie processor for this as it automatically dumps while spraying], heat/spray dry and pass through resin tower on way to IBC tank. Might well be able to reduce the methanol further as Dgs does but I'm more than happy with the 25% reduction for the most expensive ingredient.
Old Merc C250TD coped with whatever conversion rate I had anyway for over 80,000 miles [still going strong] and I was happy enough too in blissful ignorance before I decided to buy an ASTM flask last year. Chuffed to bits now though.

[dgs]

Thars good dodge50. Pleased you have sorted it out.

As regards the reverse reaction thing. Over the years I've seen posts on various forums where people have reported this. I must say in 7 years or so it has never happened to me.

[dodge50]

It was actually a poster on another forum who I met years ago who told me so it seems to be a common misconception. It's been balanced though by your experiments on reducing the methanol so thanks for that.

[neisel]

I did a batch the other day which used similar amounts of chems.

150 litres of mixed, pretty decent liquid oil from various sources, double glyc washed (1st time with glyc used once before, 2nd wash with 1st stage glyc from previous batch - about 18-20 litres each time), normal 2 stage deliberately underdosing on 1st stage (16 litres, 700grms), all at 60-65 degrees.

All told I used 18 litres of meth & 770 grms of KOH which works out at 12% & 5.13grms.

I swear by the double glyc wash now.

[dgs]

Well done neisel, like it.

I've been thinking about your 5.13gms all day, you must have achieved @ 20% conversion with your 2x glyc washes to use so little KOH. Don't suppose you did a 10/90 before your 1st stage.