Incomplete reaction advice

[3DB]

In making recent batches of bio I have tried to do the reaction and let the product sit as long as possible before washing. I've found this to be beneficial in that more byproduct drops out with time and I have to wash less and seem to get less 'tarry substance' in the bottom of my cubies and filter sedimenter.

So I had what I thought was a reacted batch (that has been sitting for more than 2 months) but went to transfer it to the wash tank the other day it had the consistency and colour of unreacted WVO.

I just assumed that I'd forgotten that I hadn't already processed it and set about making some methoxide. I then realised I only had about 18L of methanol handy, but plenty of KOH. I am getting quite low on fuel, so I thought - I'll take a gamble here and try to use 18L of methanol instead of the normal 36L with the (fairly standard) 2 kg of KOH. I have about 2L of correctly mixed methoxide that I saved previously and thought I could just do a secondary reaction with it in the event that I didn't get 100% conversion.

So I went ahead and processed the batch with 50% less methanol than normal and it is clear that it hasn't worked. The oil still looks largely unprocessed. Fairly murky looking, similar viscosity to WVO, rather than biodiesel and no obvious heavier glycerol layer in the bottom of the tank. I thought I'd probably get a partial conversion, but hadn't gambled on what seems like a fairly homogeneous product.

What is my smartest move from here?

[WesleyB]

I would assume you refer to using 20% methanol as a normal amount to get a fairly complete reaction. I did the numbers on a stiochiometric amount of methanol to get 100% reaction in the past, it was about 12.5%. I'm not clear on the exact quantities present or if there's a possibility of wet oil or wet methanol. Test the oil with a hot pan test if a few drops are left in the container. I found my notes on the stoichiometric amount of methanol to react 1 litre of lard, it is 119.9 milliliters of methanol.

[smithy]

Hello 3DB. I think you will find that there has been some reaction. Sometimes if the reaction is small this can happen where the (small) amount of glycerol formed is reluctant to drop.

What you can try is to re-heat and add some glycerol from a previous batch, I think you will find when it all settles the amount of glycerol will be greater than the amount you added. (it is a little like seeding an over saturated solution with crystals, as the sugar boilers do when they seed a pan of sugar)

You would then need to conduct a 3/27 or 10/90 test to indicate the chemicals to add to get complete conversion.

[tillyfromparadise]

Hi 3DB,
Can you give us some numbers-
1. Amount of WVO reacted
2. Oil Titration
3. Amount of methanol used per litre of WVO reacted
4. Amount of NaOH/ KOH used per litre WVO

I have found that once you start using less than the stoichiometric amount of methanol in a standard single stage base reaction, separation becomes questionable.

As smithy has pointed out, there will have been some reaction happen, that is what happens when you mix methoxide with Veg oil

[3DB]

Hi all,

Thanks for the responses so far.

Tilly, to answer your questions:

1. 180L WVO
2. Titration was 3.4. Adding the baseline 7.7 = 11.4 g/L
3. I normally use 20% methanol: WVO, so in this case I would use 36L. But I only had 18L available this time.
4. KOH used = 2,000 g

[3DB]

So I heated the batch up to 55C again and added 7L of methanol + a similar amount of left over glycerol, mixed for 3.5 hrs and now have a good visible separation.

Still didn't pass the 3/27 test, so I'll now mix up the appropriate methoxide and re-process.

[smithy]

Pleased you got separation 3DB. How will you calculate the correct chemicals for full conversion as you can no longer titrate it (the bio/oil will be alkaline) you will have to conduct a 3/27 or 10/90 and calculate from the dropout amount.

[3DB]

Hi Smithy,

I don't own a reasonable graduated flask, but perhaps I should invest to get this calculation correct.

In the past, I have just done a rough 3/27 using an old syringe and have only ever had minor drop-out. In these cases, I've just mixed one 'scoop' (approx 500g) of KOH in 1L of methanol and added that to the offending (approx.) 160L of biodiesel and mixed it for a few days with my fish tank bubbler that I use for bubble washing & drying.

That usually does the job.

In this case it may not, so maybe I need to be a bit more scientific about it.

[smithy]

Hi 3DB, you can make an accurate 10/90 tube by sealing a 2ml syringe through the cap on a small coke bottle. The syringe plunger is discarded and the end hole sealed up with glue. Undo the coke bottle top add the 10.0 mls of sample and 90 mls of methanol, screw on the top, shake and turn upside down. It's cheap and works well.

Going into hospital now for a knee replacement so won't be posting for a few days, Dave.

[smithw]

Why dont you buy one of these?

https://www.aliexpress.com/item/100m...311.0.0.B1gqgk