(And a Few Things I Still Wonder About)
When I started making soap, I didn’t see any reason to dream up my own recipes. For one thing, I belonged to a couple of soapmaking mailing lists, and they were thick with emails from people who were trying to develop recipes and having a lot of trouble with it. Also, what with books, Web sites, and soapmaking friends, I had about a million recipes to try. I didn’t see much point in struggling to think up new ones.
Fast forward a few years, to my decision to write a soapmaking book. I’d been making soap, teaching other people to make soap, and working out some new methods to make things easier. I’d written about half of the first draft when it occurred to me that I was going to have to design my own recipes.
Looking over my books, I came upon something I’d read before but hadn’t paid much attention to: the concept of INS. According to Robert McDaniel, using a simple formula would produce “ideal soap.” Could it really be that easy? I decided to find out.
A lot of experiments later, I had about enough soap to wash an army. And I had an answer to my question. Let me tell you what I did, what I learned, and what I still wonder about.
First, I analyzed the recipes I’d been using. The INS values of my two favorite recipes turned out to be identical, both the “perfect” value of 160.
I studied McDaniel’s discussion of INS a little more. He didn’t claim to have invented the idea, but he didn’t say where he got it, either. I had a lot of questions about what INS might be and whether I could really make an ideal soap by using a formula. I tried to contact him, but wasn’t successful.
Researching further, I found a professional article from the 1930s that described INS in a way that sounded like it was a newly-developed concept. According to this article, INS values would predict the hardness of soap made from a mixture of oils. Obviously, it wasn’t going to be possible to contact the author of this article, either, so I decided to just experiment on my own.
I developed some recipes that approached an INS of 160. Then I made soap, soap, and more soap.
It all came out fine. In fact, it was comparable to my best recipes. But before I decided that INS was the solution to all problems, I decided to see how far I could push the limits.
First, I made a soap with an INS of 160 and no solid fats. It saponified nicely, and I looked forward to trying it when it came out of the molds. That turned out to be the problem: unmolding it. After about a week, I gave up and peeled off the molds in shreds, happy I hadn’t used my best ones. The soap didn’t hold a firm bar shape, and I would have written it off as a failure, except that I tried it first. It made my face feel better than any soap I’d ever used. I gave some chunks of it to friends, who immediately dubbed it “the weird soap”—and asked for more.
So I made it again, but with a minimum amount of water. That batch unmolded easily. Though still a bit soft, it’s a wonderful facial soap.
Then I put together a recipe with an INS of 160 and no liquid fats. As I’d hoped, it holds up well in the moist area next to my kitchen sink.
Everything I’d tried worked well. But the various soaps were very different in their lather, cleansing and emollience. I wondered what, exactly, Robert McDaniel had meant by “ideal.” My experiments so far had given me good soaps for different purposes, all with the supposedly perfect INS.
Next, I deliberately got as far away from the good INS range as I could. I made a soap of 100% canola oil (INS value of 56, as listed by McDaniel). The mixture did make soap, eventually, but it took a long time to trace, a long time to solidify in the molds, and a long time to saponify. When the pH of the soap was finally in the usable range, I washed with it. It was a mess—I had to take the goo off my hands with another soap. It certainly was far from ideal, by anyone’s standards.
I was still mulling about this as I made a batch of Castile. When I learned soapmaking, I was repeatedly warned against the traditional all-olive-oil formula. I suppose it would be difficult to make if you’re stirring by hand—even with a stick blender, it’s half an hour before I’m satisfied that it’s sufficiently mixed. In fact, the time to trace was similar to that of my dreadful canola soap. Of course, the end product was fine soap, not an unwieldy mess. That gave me an idea about the “ideal” INS.
The INS value of olive oil is about 109. It makes good soap, but with difficulty. The value for canola is 56, much farther from the ideal INS of 160. And canola is even more difficult to make soap from than olive oil—and the soap is no good.
Could it be that the “ideal” McDaniel discussed was more a chemical ideal than an esthetic one? That what he meant was that the ideal INS range will give you functional soap with very little trouble?
The more I thought about it, the better sense it seemed to make. There’s no way to control the esthetic qualities of soap by manipulating numbers. Clearly, a gardener’s scrub soap will need different qualities from one for sensitive skin, for example. Whatever the INS value of your fats, you still have to choose them according to their individual properties of lather, cleansing and emollience.
Also, you need to consider the proportions of solid and liquid fats when you decide how much water to use. INS values may have been developed to make soap hardness more predictable, and they are helpful in doing that, but the “weird soap” convinced me that they’re not infallible. Whatever the INS value of your mixture, you need to use your judgment about this.
I’m certain that measuring errors are not the explanation for any of my results, but I’m not a chemist as McDaniel is, and I don’t pretend my experiments are scientific.
To recap, here’s the way my test batches pointed:
• Design your mix of fats to provide the emollience, cleansing, hardness, and lather you want for the particular purpose. Judgment, experience, and a certain amount of trial and error are necessary here—you can’t do it “by the numbers.”
• If the mixture of fats has more than about 60% liquid oils, reduce the water in the recipe to the minimum prescribed by your lye calculator.
• INS values predicted ease of saponification more than anything else.
The whole exercise was extremely valuable to me. It made it possible for me to design easy, high quality soap formulations for my book, and to explain to beginners how to design their own.
I still wonder about a few things, though.
I wonder if the high-INS oil I used in the “weird soap” was really a fair test. The soap is composed of almond oil and fractionated coconut oil—the coconut is probably necessary to get a value as high as 160. But fractionated coconut oil is an artificially manipulated product. Possibly INS value actually is a good predictor of hardness, if naturally occurring oils are the only ones considered.
I wonder about cocoa butter. With its close-to-ideal INS value, it should make a good single-oil soap. But according to other sources, pure cocoa butter soap would be too hard and would have poor lather. I suppose I’ll try it someday, just to round things out.
Finally, I wonder if I interpreted McDaniel’s work correctly, or if I’ve added a spin of my own that he didn’t intend. My theories work for me, and that’s the main thing I needed.
Anne Watson is the author of the forthcoming book, “Smart Soapmaking.”