Does this sound right?

[Davedewhurst]

Hello good people!
I've been quiet for a while (I was urban warrior)
Been consistently making bio for almost 10 years now and looking to refine my method and increase my efficiency..
I've been reading some threads on enhanced glyc washing as I think I might have fairly high titrating oil (going to titrate as soon as my bits n bobs arrive)

For now I just wanna run my process past you folks and see if it sounds "normal"
I mean, some of you seem to be using such small amounts of chemicals whereas I seem to use a lot...

So I have a huge reactor now which is awesome as I only have to produce once every 6 months...

I settle an IBC of WVO and drain off 20L of white crud each process.
I start with 650L WVO already in the processor which has been settled for a further 6 months
I drain off 20L more of white crud from the processor before the process begins

I wash with my 130L of glyc from the last batch in 2 stages of 65L each (the methanol seems to evaporate off in the 6 months it sits there as the can isn't sealed)
I get about the same amount of glyc back as I put in but it's nice thick gloopy stuff and no longer runny suggesting I've extracted much of the biodiesel from it.

Then I start stage 1 reaction - I've been aiming for 80% conversion going on 15% meth and 10g KOH per litre.
So around 80L methanol goes in with 5.2kg KOH
(I mix for 1 hour and settle for 1 hour at 40 to 50degC)

90/10 test showed about 67% converted yesterday suggesting some high titrating oil.
It's always been pretty much like this...

I then mixed up the rest at 15% and 10g
So 32L methanol and 2.15KG KOH
Then got a fully converted 90/10 test
Bubble the lot overnight losing seemingly about 50L of methanol which seems a lot but possibly the volume also decreased as it cooled?

I ended up with about 585L of finished biodiesel so about a 90% yield from the WVO.

Mix 5% petrol and settle for 3 weeks

Does my 90% yield sound about typical in terms of efficiency?

If I back-extrapolate my chemical quantities it seems that my KOH is about 11.3g/L so if a base is usually 8g then can I assume it might titrate at around 3.3?
Obviously I'll be titrating the next batch to make sure I do it as accurately as possible but could you tell me if I'm thinking along the right lines with the enhancement of the glyc please
I'm wondering if I enhance my glyc wash with say 3.3g/L (actual amount to be worked out from the titration) split between the 2 stages of my glyc wash then it might bring my oil to neutral and give me a more efficient reaction with less FFAs and less water produced and therefore less soaps and less glycerol and a better yield?

So 3.3g x 650L =  2.145KG / 2 = 1.073KG KOH per glyc wash mixed with say 5L methanol each time (or maybe 10L each time since my glycerol looses it's meth content to evaporation?)

This should (depending on actual titration value) neutralize my oil right? Bringing the base level of KOH needed to around 8g?

Is 20% glycerol a typical amount? Can this be reduced to say 15% or even 10%?

Am I thinking along the right lines?

Thank you!!

Dave, Cornwall

[dgs]

A lot to think about there Dave.

Your calcs for the enhanced wash seem about right. Years ago I Realised that processing oil with no titration gives a really efficient process.

Your yield should be more like the 97% range.

67% conversion after S1 does seem quite low. My average used oil (for my heating) with the same chemicals (15% and 10gms/L ) would give about 0.8 mls dropout.

When you are removing whites after settling it makes me wonder how much of these are still suspended in the oil. These really can make your reaction inefficient.

my oil for vehicle use is glyc washed about 5 times in an outside ibc (I get glyc from other people plus my own) and doesn't titrate, it shows a conversion of around 15% I use about 11% methanol and 6.5 gms/litre for this and it produces about 15% glycerol.

Use 8 gms/litre for S2 as after S1 your oil won't contain any ffa's, so base is fine.

Hope this helps.

[countrypaul]

It is worth titrating you oil both before and after the glyc wash. The glycerine used for the glyc wash should contain a fairly high amount of caustic, however if it has been lying around for 6 months unsealed it may well have reduced levels of caustic meaning less able to neutralise the FFA.

Are you using both S1 and S2 glyc for the wash, or are you wasing twice once with S1 the  second time with S2?

If you can reduce the amout of KOH used you will reduce the amount of water introduced and hopefully the amount of soap produced.

Dave's (DGS) 5 glyc washes shows the results, with a titration of zero (or no titration), allowing a much lower level of KOH to be used in each stage.

As you have suggested, an enhanced glyc was where you have added KOH to the glyc prior to washing the oil can result in eliminating the FFA. This works best, I suggest, when the first glyc washes can be relied on to remove nearly all the water.

[Davedewhurst]

Thanks chaps - I'm asking around for glycerol - see if I can bring that titration right down..
I shall indeed titrate before the glyc wash and after


I'm using both S1 and S2 glyc mixed together, half at a time.

"As you have suggested, an enhanced glyc was where you have added KOH to the glyc prior to washing the oil can result in eliminating the FFA. This works best, I suggest, when the first glyc washes can be relied on to remove nearly all the water."

Ahh okay so I may not get great results if I only have my own glycerol to wash with?
Unless I wash with half of my glyc as normal and then enhance the second half perhaps?

Good point about the whites - I might try to leave them out from now on if there's not too much of it.... would it really stay suspended after 6 months plus?

I'LL try 8g/L on S2 then and see if it's enough - that should make things more efficient hopefully.

Cheers!!

[Davedewhurst]

Hello fellow grease monkeys!
Just make a nice 650l batch of beautiful golden nectar and wanted to share how it's going.

So my oil titrates at 5.5
I'm enhancing my glyc wash using the 120l ish of s1 & s2 glyc from last batch mixed together and split over 2 glyc washes)
I enhanced with 2.5g/l Koh into 8l methanol.
I actually forgot to split the amount into 2 and when I realized after it had gone into the processor, I just added the rest of the glyc so this time I actually did a 1 stage glyc wash.

This brought the titration down to 2.25 at first but when I titrated the next morning it came out at 3.5
I tested twice to make sure.
Can a reaction reverse like this? Why would titration be 2.25 and then go to 3.5 by morning?

Anyway I then reacted using 15% methanol and 8g/l base + 3.5g/l
Aimed for 80% and got exactly 80%!

Definitely getting a better more efficient reaction I'm just confused about the titration increasing over night.
Also it'll be interesting to see how my 1 stage glyc wash compares to a 2 stager next time.
It'll certainly fit better in the tank (it almost overflowed!) and maybe I'll get an even more efficient reaction?

We aim to bring titration down to 3 or below correct?
Could I double up on my enhancement next time then and aim for a titration of 1.5?

Anyway thought it would be nice to post on here.
Appreciate this forum sooooo much!

Dave. Cornwall.

[countrypaul]

One problem with using a high amount of NaOH in stage 1 is that dissolving the NaOh produces a lot of water, for each 10g of NaOH you dissolve you produce 4.5g of water. Water present during the processing catalyses the production of soap which obviously can reduce the yield significantly. Having oil tirating at 5.5 and adding NaOH at 10 means you are effectively trying to process with 4.5 g/L of NaOH but with the 4.5g/L of water also present ! At the end of your S1 processing you will have less that 4.5g/L left as some will be used to form soap.
A lot to consider, I'll try explain some of the consequences of the glyc wash.

When you do a glyc was with the Glycerol from the above you will have much less than 5.5g/L of NaOH present in the Glyc, probably well under 4 so very little (read no) chance of neautralising all the FFA. If you split the glyc into 2 then you probably have the equivalent of under 2 in the first glyc wash leaving over 3.5 in the oil for the s2. Chances are you will still have oil titrating at 2 or more after the second glyc wash.  Doing both stages together would have the same results so your figure of 2.25 sounds pretty much what I would expect. In you were to do an enhanced glyc wash I think you would need to be adding the equivalent of at least 3 g/L NaOH to minimise the amount of FFA left. Well washed and clean oil titrates at 0.

Regarding the titration increasing overnight are you sure both the samples taken were from a homogenous mix? When you got a reading of 3.5 the next morning how had things been left? Was the processor effectively closed to the outside, had the glyc been removed, did you take a fresh sample of the oil, did you mix the oil before taking the sample of was it just from the bottom/top of how the processor was left? If you clean all the titrating equipment before use, did you blank the titration reagents before measuring the FFA level, I assume you used fresh caustic solution and did not just leave it in the burette overnight?

I would not expect the oil to increase in acidity that much without some other interaction. Given that 650L is a large volume compared to the area exposed to the air oxidation would be unlikely to explain that change (even at higher than ambient temperature). 

[dgs]

Hi Dave, don't expect your titration results to be so accurate. The test is approximate at best due to the colour of wvo and fading end points. In lab conditions the norm would be to have a standard to refrence to (a titration flask which is buffered so the end point doesn't fade, so you know what you are titrating to)

The way we titrate is a guess as regards the end point colour and with a fading end point colour the test is very approximate at best.

[Davedewhurst]

Thanks Paul! But I'm not using sodium I'm using potassium.
I did mix the oil thoroughly next morning after drianing so it should have been pretty much the same result. However I did leave out my jug of caustic solution uncovered overnight so perhaps that skewed the result?
It was a fresh sample, the machine was effectively sealed.
Clean equipment yes but what is blanking the reagents?

One problem with using a high amount of NaOH in stage 1 is that dissolving the NaOh produces a lot of water, for each 10g of NaOH you dissolve you produce 4.5g of water. Water present during the processing catalyses the production of soap which obviously can reduce the yield significantly. Having oil tirating at 5.5 and adding NaOH at 10 means you are effectively trying to process with 4.5 g/L of NaOH but with the 4.5g/L of water also present ! At the end of your S1 processing you will have less that 4.5g/L left as some will be used to form soap.
A lot to consider, I'll try explain some of the consequences of the glyc wash.

When you do a glyc was with the Glycerol from the above you will have much less than 5.5g/L of NaOH present in the Glyc, probably well under 4 so very little (read no) chance of neautralising all the FFA. If you split the glyc into 2 then you probably have the equivalent of under 2 in the first glyc wash leaving over 3.5 in the oil for the s2. Chances are you will still have oil titrating at 2 or more after the second glyc wash.  Doing both stages together would have the same results so your figure of 2.25 sounds pretty much what I would expect. In you were to do an enhanced glyc wash I think you would need to be adding the equivalent of at least 3 g/L NaOH to minimise the amount of FFA left. Well washed and clean oil titrates at 0.

Regarding the titration increasing overnight are you sure both the samples taken were from a homogenous mix? When you got a reading of 3.5 the next morning how had things been left? Was the processor effectively closed to the outside, had the glyc been removed, did you take a fresh sample of the oil, did you mix the oil before taking the sample of was it just from the bottom/top of how the processor was left? If you clean all the titrating equipment before use, did you blank the titration reagents before measuring the FFA level, I assume you used fresh caustic solution and did not just leave it in the burette overnight?

I would not expect the oil to increase in acidity that much without some other interaction. Given that 650L is a large volume compared to the area exposed to the air oxidation would be unlikely to explain that change (even at higher than ambient temperature).

[Davedewhurst]

That does make sense but I'm pretty sure it's as accurate as needed coz I did the test twice and used the same sort of "fade" point and I'm just convinced the result was different by 1.25
I did leave the caustic solution out overnight uncovered though maybe that was my problem?
In which case, 2.25 from day 1 would be the more accurate result however my reaction responded as expected using chemical amounts dictated by the 3.5 result. hmmmmm

Hi Dave, don't expect your titration results to be so accurate. The test is approximate at best due to the colour of wvo and fading end points. In lab conditions the norm would be to have a standard to refrence to (a titration flask which is buffered so the end point doesn't fade, so you know what you are titrating to)

The way we titrate is a guess as regards the end point colour and with a fading end point colour the test is very approximate at best.

[countrypaul]

Thanks Paul! But I'm not using sodium I'm using potassium.
I did mix the oil thoroughly next morning after drianing so it should have been pretty much the same result. However I did leave out my jug of caustic solution uncovered overnight so perhaps that skewed the result?
It was a fresh sample, the machine was effectively sealed.
Clean equipment yes but what is blanking the reagents?

One problem with using a high amount of NaOH in stage 1 is that dissolving the NaOh produces a lot of water, for each 10g of NaOH you dissolve you produce 4.5g of water. Water present during the processing catalyses the production of soap which obviously can reduce the yield significantly. Having oil tirating at 5.5 and adding NaOH at 10 means you are effectively trying to process with 4.5 g/L of NaOH but with the 4.5g/L of water also present ! At the end of your S1 processing you will have less that 4.5g/L left as some will be used to form soap.
A lot to consider, I'll try explain some of the consequences of the glyc wash.

When you do a glyc was with the Glycerol from the above you will have much less than 5.5g/L of NaOH present in the Glyc, probably well under 4 so very little (read no) chance of neautralising all the FFA. If you split the glyc into 2 then you probably have the equivalent of under 2 in the first glyc wash leaving over 3.5 in the oil for the s2. Chances are you will still have oil titrating at 2 or more after the second glyc wash.  Doing both stages together would have the same results so your figure of 2.25 sounds pretty much what I would expect. In you were to do an enhanced glyc wash I think you would need to be adding the equivalent of at least 3 g/L NaOH to minimise the amount of FFA left. Well washed and clean oil titrates at 0.

Regarding the titration increasing overnight are you sure both the samples taken were from a homogenous mix? When you got a reading of 3.5 the next morning how had things been left? Was the processor effectively closed to the outside, had the glyc been removed, did you take a fresh sample of the oil, did you mix the oil before taking the sample of was it just from the bottom/top of how the processor was left? If you clean all the titrating equipment before use, did you blank the titration reagents before measuring the FFA level, I assume you used fresh caustic solution and did not just leave it in the burette overnight?

I would not expect the oil to increase in acidity that much without some other interaction. Given that 650L is a large volume compared to the area exposed to the air oxidation would be unlikely to explain that change (even at higher than ambient temperature).

Hi Dave,

If you are using KOH the same rules apply only it is for every 56g of KOH you get 18g or water when dissolving in MeOH.

Leaving an open jug of caustic overnight will allow it to absorb CO2 and for NaHCO3 / KHCO3 both of which will increase the results of your titration as the solution is weaker. It is not a simple relation as the FFA in the oil will act as a weak acid and may/may not react with the bicarbonate at the same rate as the caustic The result can be that the caustic is used for the reaction during the titration depending on how long the titration takes and the fade poit of the indicator can be different.  There is no way to tell easily so it may or may not have had an effect on your titration.

Blanking the reagent is normally done to eliminate any effect of the solvent being acidic. If you use IPA (or even acetone) as solvent then you add the indicator to the solvent before adding the oil. You then titrate the solvent to the point where the indicator change is the same as you will be using when you titrate the oil. This can be difficult as there may be a very rapid change when no oil is present. After you have titrated the solvent you add the oil and titrate as normal. If the solvent was very acidic for any reason you can find that it titrates significantly higher than clean fresh solvent which should have a very low (even almost zero) value.

KOH is normally much more forgiving than NaOH, but if you want to make your process more efficient such as using less reagents, producing higher yields and having less soap to get rid of then the arious details can make a difference.

[Davedewhurst]

Ahhh ha well I shall blank my IPA next time then!
And I shall seal my reference solution too
thanks for the tips!
So I will assume that my 2.25 titration was the more accurate one after my enhanced glyc wash however I used the chemical quantities for a titration of 3.5 and it worked out exactly right.
So now I'm wondering if perhaps something else is going on to make the process less than perfectly efficient.
Could my KoH be too old? It's been sat in it's unopened bag in shed outside for over a year

[Davedewhurst]

Ooooh - question:
why do we do the glyc wash in 2 stages?
Does this produce a better result or is it simply that it fits in the processor better than all in one go?

[countrypaul]

Ahhh ha well I shall blank my IPA next time then!
And I shall seal my reference solution too
thanks for the tips!
So I will assume that my 2.25 titration was the more accurate one after my enhanced glyc wash however I used the chemical quantities for a titration of 3.5 and it worked out exactly right.
So now I'm wondering if perhaps something else is going on to make the process less than perfectly efficient.
Could my KoH be too old? It's been sat in it's unopened bag in shed outside for over a year

If your KOH is properly sealed then it should not degrade. It is probably not worth testing. If you use a caustic solution made from the same KOH for titration as well as processing things should be consistent. If the KOH has gone off slightly (aborbed CO2) then you titration figure may be slightly high, but that will compensate for the fact you need extra for processing.

I mentioned earlier that KOH is more forgiving than NaOH, what this refers to is that if you use excess NaOH you can quickly end up with a jelly (as mentioned in the other thread you recently read) but that is unlikely to happen with KOH. The problem is that you may be using more KOH than required but there are no obvious signs when you do so. If you have a consistent oil that will be used for multiple batches you can try using less KOH on S2 until you find the optimum amount.

There are various reasons for multiple stages in glycerol washing the oil - it depends on quantities and quality. If you oil is quite wet with water then most of that water can be removed using older glyc that has virtually no caustic left in it. If you oil is quite acidic then you need the caustic to neutralise that acid however the caustic in the presence of water will lead to more formation of soap as a competing reaction to the neutralisation of the acid reaction. That suggests that washing with almost caustic free glyc on wet high ffa oil might be a worthwhile first step. When removing ffa from the oil using glyc you need enough free caustic pesent to neutralise all the acid so if you have a low caustic level then either an enhanced wash (with added caustic) or washing with multiple loads of glyc will be required.

Washing with glycerol from a previous batch will also realease bio held in that glycerol. If the glycerol is quite fresh or has been stored sealed the it will have not only free caustic present but also methanol present. If this glycerol is used for a later or last wash not only will and bio present be released but he presence of the caustic and methanol will also result in some conversion of oil to bio. If you have a significant amount of bio in the oil after glyc washing then that bio will also then allow the methnol to dissolve (bio and methanol are totally miscible at all concentrations) so you can save on methanol.

If you read some of Dave's (DGS) posts on glycerol washing I believe he got upto 20% conversion just from glycerol washing - though I believe he uses around 5 washes for each batch of oil.

Incidently, if you had clean neutral oil and converted 100% to bio, then you could get 104% yield based on measured volumes due to the lower density of bio.

If you use ASM (anhydrous sodium methoxide / methylate) instead of caustic there is no water produced in the conversion of oil to bio as a result the glyc is water free and soap free with very high yields. Unfortunately, ASM is quite expensive and now very difficult to obtain.

[Davedewhurst]

Awesome! Thanks for the intel
I think I might glyc wash next batch firstly using already used once (for glyc wash) glyc and then go for an enhanced glyc wash using previous batchs glyc.

Thanks again