Can you do math with Scoville units?
June 11, 2018 8:45 AM   Subscribe

Suppose I want a 7,500 Scoville Unit hot sauce. And suppose I'm looking at two kinds of hot peppers — one rated at 5,000 SU, and one rated at 10,000 SU. If the Scoville scale was mathematically well-behaved (and ignoring other ingredients), middle school math would say I should mix them 50/50 and I'd get what I want...

Does that actually work? Is there some other calculation that would work instead? Or are Scoville Units just so subjective or weird or unreliable that you can't do math with them at all?

More broadly, are there any ways besides "make it and see what happens" that hot sauce makers use to predict the heat of a recipe?
posted by nebulawindphone to Food & Drink (12 answers total) 4 users marked this as a favorite
 
Best answer: Since the 1980s, spice heat has been measured by high-performance liquid chromatography (HPLC) which measures the concentration of heat-producing chemicals, such as capsaicin

Wikipedia then describes how the results of chromatography are converted to the well-known Scoville scale.

The point is, though the old organoleptic method had some subjectivity (though it was designed to minimize that), the modern unit is derived from objective physical techniques.

So, TLDR: You can treat the Scoville as a concentration unit that will, to very good approximation, respect standard arithmetic for dilutions and mixtures.

(and yes, food scientists and industrial chefs can use the known info of ingredients, including concentrations of capsaicin and other well-studied "heat" compounds to predict how hot the sauce will come out)
posted by SaltySalticid at 8:53 AM on June 11, 2018 [7 favorites]


> TLDR: You can treat the Scoville as a concentration unit that will, to very good approximation, respect standard arithmetic for dilutions and mixtures.
Though it might still be a good idea to taste a bit before use. There may be some differences in varieties that look same-ish, or even in different harvests. I've had kind-of spicy peppers leaning to "tastes like bell pepper" as well as "oof that's pretty hot".
posted by farlukar at 9:01 AM on June 11, 2018 [3 favorites]


Ah, yes, indivual peppers are a bit different than a mass produced bottle of hot sauce; I was speaking about there latter. I doubt anyone does chromatography on individual batches of peppers, I assume they are labeled in terms of an average for the variety.
posted by SaltySalticid at 9:08 AM on June 11, 2018


Just FYI, there are plenty of mathematically well-behaved scales that aren't additive like that, such as the decibel scale for sound pressure level (and for other forms of power). Not all scales are linear.
posted by likedoomsday at 9:20 AM on June 11, 2018 [16 favorites]


I would think that your Scoville rating would be whatever the hottest pepper in the batch was. I don't see how putting in a less hot pepper would degrade the heat of the hottest.
posted by trbrts at 9:21 AM on June 11, 2018 [1 favorite]


a google search lead me to this 538 discussion including this line:

The Scoville heat unit (SHU) rating is then assigned based on the quantity of dilution, with the ratings working on a linear scale: a 350,000 SHU habanero is 100 times hotter than a 3,500 SHU jalapeño.

Your right in principle, but the variations in individual plants/peppers concentrations of capsaicin would introduce a bunch of error in your method. still, if you want 7500 mixing 50/50 quantities of 5k and 10k peppers should work. The one other thing to consider is the relative water content of the peppers because as I understand it the Scoville scale measures capsaicin concentrations relative to the water content of the peppers.
posted by Exceptional_Hubris at 9:26 AM on June 11, 2018 [1 favorite]


Best answer: Yes. As others have said, it appears that Scoville units are linear, so if you do a 50/50 mixture of a 10000 SU pepper and a 5000 SU pepper, you'll get a 7500 SU composite pepper.

The wrinkle, of course, is that there's lots of places where the human senses are not linear; they are logarithmic - this is why the decibel scale is logarithmic for sound and the magnitude system is logarithmic for light. Simply put, your body often doesn't interpret a doubling of some stimulus (sound, light) as a doubling. I don't know (and am interested) if taste is one of those things where your body is linear. My guess from what I know of sound and light is that it's not, but I don't know of any evidence either way.
posted by Betelgeuse at 9:27 AM on June 11, 2018 [5 favorites]


If it's linear, there's no need to mix. You could just concentrate or dilute whatever pepper you have to whatever heat you desire.
posted by paper chromatographologist at 9:31 AM on June 11, 2018


The reason not to dilute with water is that will change the consistency. A hot sauce made with 10 units of 5k peppers will be very different from a hot sauce made with 5 units of 10k peppers, though they should have comparable heat.
posted by SaltySalticid at 9:51 AM on June 11, 2018


Best answer: The wikipedia article indicates that there are standard "pungency units" which are equivalent to parts per million of capsaicin in dried peppers, and Scoville units are just 15 times this (presumably an arbitrary constant to get them on the same scale). So yes, you can do standard dilution math with them.
posted by madcaptenor at 10:11 AM on June 11, 2018 [1 favorite]


Best answer: BTW: If a recipe does come out hotter than anticipated, you can tone it down with sweet bell pepper during the ferment. This adds some nice fruit or grassy notes depending on the color of pepper. Apologies if this seems too basic, but it can also be tasted throughout the process. As you go, you learn what the mixture should taste like at the different stages of the process.
posted by jenquat at 1:03 PM on June 11, 2018 [2 favorites]


Worked on a farm, can confirm that heat in peppers varies significantly depending on growing conditions - sunny/hot weather seemed to yield hotter peppers, but our jalapenos were sometimes so mild the Mexican restaurant dropped us, so no robust data.

If it's not impratical, you could make hot sauce with each and then mix to desired heat? Tasting the peppers and calibrating with known-Scoville mass-produced sauces might work, too. Or cooking down the final sauce?
posted by momus_window at 2:48 PM on June 11, 2018 [1 favorite]


« Older How do I handle dog separation anxiety from a room...   |   Brake job priced different? Newer »
This thread is closed to new comments.