Titrating oils

Notice: Don't pay too much attention to the butchering of chemistry terms and procedures I'm (probably) about to do and possibly already did with the title :mrgreen:. I'm more of an alchemist than a chemist :think:.

OK, so, I've learned (so far only in theory, but soon to be tested on my supposedly 90.5% pure KOH) how to titrate/measure caustic purity from Deanna's excellent instructions. Long explanation provided in Kevin Dunn's presentation (which she also cites as a resource) here.

Which leaves the other, so far unknown, variable for "precision" soapmaking: the actual SAP values of my oils.

How the heck would one go about that? The first thing that came to mind was a similar procedure:
1. Add water
2. Dissolve known amount of KOH in water
3. Slowly add oils until pink turns to clear
4. Record oil weight and do math magic
5. Results

I was in the process of writing this to ask if this is the way to do it, but the obvious donned on me: the pink will indeed never turn clear, because alkaline soapy science. I stand before you, a wide-eyed chemistry noob, anxious to get schooled on the procedure, if anyone happens to know it.

I had a hunch the answer might be provided in Scientific Soapmaking, and it turns out it is, from it's Amazon page the first review reads:
"The first time I cracked this book open, all of the technical information was lost on me. Then I started selling my soap, and I realized that if I wanted a product that wasn't superfatted, which I didn't (that unsaponified oil goes down the drain, after all), I would need to titrate my own oils. Otherwise my soaps could turn out lye heavy even running it through a lye calculator, which uses average SAP values, and following the recipe exactly. Yes, I own just about every soap making book on the market today. No, none of them tell you how to titrate your oils. This book does. It also tells you how to measure properly, how to test your bars for various qualities, what to do to avoid the dreaded orange spots, and why it's a waste of time adding "superfatted" oils to your soap last. "

Alas, I don't have the book quite yet, and I'm wondering if the process is pricier and more elaborate than the caustic purity check.

The purity check I describe is a rough test -- it's good to maybe 2-3% accuracy. That's close enough for most soapers. If you want better accuracy for your alkali purity, however, a more rigorous method adapted to kitchen chemists is presented in Dunn's book Scientific Soapmaking.

His book also presents a method adapted for kitchen chemistry to measure the saponification value of fats. Dunn's method of measuring the sap value does not require expensive equipment, except for a decent scale that reads to 0.01 gram.

The methods in his book are more complicated than his rough alkali purity check that I have in my video, but more rigor is the only way to get the accuracy you appear to be wanting. I really recommend you get the book -- your questions will be answered in detail.

Will do that!

Proof is in the pudin'. Supplier says 90.9% KOH purity, I measured 65-85% with the CA test.

Yes, the margin is huge, I'm not gonna pretend my current crappy scale is even close to accurate, but it's somewhere in that range. Actual measurement was 73.01%, but as I said I don't trust the numbers to be very accurate. This is a brand new box of KOH, mind you (used once, open for about 2min in total, then sealed to it's original state).

This also explains why my previous batch of LS had sky high SF. Apparently their quality control is beyond crap, because the previous box of KOH I used from the same supplier never gave me a failed soap, so it must've been close to 90% which I used for calculations.

I for one am never using KOH without previously testing it again.

A 20% range in results is a wider variability than I would be comfortable with. Having done many titrations as a chem lab technician, I know this kind of test takes some patience and practice to do well. I think you've made a good start, but perhaps there's a bit of room for improvement?

The strength of this rough purity check is also its weakness -- you are using the solid alkali to titrate with. Normally you wouldn't do that in an analytical chem lab. You'd make a solution of the solid alkali, so you're doing the test with diluted alkali as well as diluted acid. That reduces the potential for error in a lab situation.

But using the solid alkali is an acceptable compromise between simplicity and "do-able-ness" versus high accuracy. You just have to remember to only add little bits of solid alkali and mix well after each addition. Your first test will give you a rough idea of about how much alkali is needed, so the remaining tests will go faster and easier.

You should be able to get much closer results even with this method. For example, I tested some NaOH yesterday after I read your post; I was getting results within 1%. But I have a nice scale that reads to 0.01 g. I'm sure that helps.

Sapo said:

...I measured 65-85% with the CA test. Yes, the margin is huge, I'm not gonna pretend my current crappy scale is even close to accurate....

Click to expand...

I applied the huge tolerance "manually/in theory/post-test", after having done 2 tests getting similar results, to account for the crappy scale I used. The actual results were relatively tight, calculated at 73% and 78% respectively. Sadly with that battered old kitchen scale, that is about as accurate and trustworthy as it gets .

Have more accurate and trustworthy gear on the way.

Obtained 0.01g scale and M3 (+-50mg) calibration weights. The results are now tight enough to have a decent idea of the KOH purity:

Test 1: 86.646%
Test 2: 86.466%
Test 3: 86.299%

Manufacturer's "this KOH is 90.9% pure" is a load of bull confirmed :mrgreen:. Also confirmed Deanna's statement that this technique doesn't provide absolute correctness, but is close enough for our purposes.

Sweet! Nicely done!

...back to the original point...

I was pondering about actual sap values of oils, and I've now read on how to do it from Dunn. BUT!

I never stopped to consider that I might not have to :think:. Usually its expressed on a certain range, for example 185 - 195 for apricot kernel oil. But if pull up the certificate of analysis for such an oil from a soap materials supplier, the COFA states a sap. value of 191.

I'm assuming their number isn't just an average/approximate? COFA attached.

It's still approximate because fats and oils are natural products, so their chemical characteristics are not fixed in stone. That said, the sap values from a particular supplier are likely to be more accurate for that supplier's products compared with the sap values you will find in a general table. That's why I use sap values from Soapers Choice for the fats I buy from them rather than the sap values from Soapcalc.

Another issue to consider is the kind of supplier. It is likely that fats from a large supplier like Soapers Choice will be more consistent because they buy in bulk to resell to the food industry as well as to soapers. Any variations in chemical properties will be averaged out when Soaper's Choice buys a tank car of sunflower oil. Variations in chem properties are likely to be greater if buying from a small or artisan producer. Or if you render your own lard or tallow.

I see. I thought that it would accurately represent the actual values and not averages, since the analysis is done for each batch, presumably in a way that ensures stable results?

For example if you would please take a look at the attached file (pdf) for an old (2014) certificate of analysis for their apricot kernel oil. It states averages are 185-195 and that this particular batch had 190.

I suppose there could still be fluctuations, if the batch was compiled from different farmers/sources and not thoroughly mixed, and the test sample taken from such a batch, resulting in the top of the tank having a sap value of 189 and the top 191 or something.

View attachment 11111.pdf

Are there batch numbers on the fats/oils you are using? If so, that's great! Get the batch analysis results and go to town.

So far I've been using grocery store oils and relying on averages. But I plan to switch completely to this supplier, which has batch/lot numbers and analysis for them, yes.

PS: any idea what kind of superfat you would be looking at if you simply took the lower end sap value every time? For example a 185-195 sap value oil, but instead of averaging out to 190 and using a 5% superfat for calculation, you put 0% superfat into the calculator and ran with 185 sap? Seems even more reliable in terms of preventing excess alkali than a fixed superfat percentage.

You'll need to pick a typical recipe and do the math, Sapo. It's not that difficult.

That sounds good, I recently got some argan oil samples (http://www.fabfreesamples.co.uk/free-argan-oil-sample/) that I would like to create a soap with, would I just simply add it to the soap mix?

jennacandle said:

That sounds good, I recently got some argan oil samples (http://www.fabfreesamples.co.uk/free-argan-oil-sample/) that I would like to create a soap with, would I just simply add it to the soap mix?

Click to expand...

What kind of soap do you make? CP or MP? Also, Argan is too expensive to put in soap in my opinion. I use it in body butter though.

If you do CP you would need to run it through a soap calculator as part of your recipe. For MP not sure.