• If you need help identifying a pepper, disease, or plant issue, please post in Identification.

scovilles Let's talk Scoville scale

In Scoville's method, an alcohol extract of the capsaicin oil from a measured amount of dried pepper is added incrementally to a solution of sugar in water until the "heat" is just detectable by a panel of (usually five) tasters; the degree of dilution gives its measure on the Scoville scale. Thus a sweet pepper or a bell pepper, containing no capsaicin at all, has a Scoville rating of zero, meaning no heat detectable. The hottest chilis, such as habaneros and nagas, have a rating of 200,000 or more, indicating that their extract must be diluted over 200,000 times before the capsaicin presence is undetectable. The greatest weakness of the Scoville Organoleptic Test is its imprecision, because it relies on human subjectivity. Tasters taste only one sample per session.


First, this is not a "bash the good pharmacist Scoville" topic - I <3 the dude and love the scale. It's also led to several hilarious interpretations (the best by Cracked I think http://www.cracked.com/article_19187_the-science-behind-stupidly-hot-peppers-5Bchart5D.html) ....but since it's subjective and relies on something as variable as human taste buds and tolerance for peppers/capsaicin, I always wonder about Scoville units when they get into those ridiculously high stratospheres.


While browsing pepper stuff, I stumbled onto this on the Livestrong site of all places: http://www.livestrong.com/article/447263-peppers-with-most-capsaicin/

I see things like "1,001,304" - how do they determine it down to the single digit unit?!? Seriously - 1,001,304? :think: And how do they tell the difference between 1,001,304 and 1,001,507?

The other part that concerns me about the scale is that while it sounds pretty scientific and they call out one flaw, they seem to potentially ignore another.
Tasters taste only one sample per session

My good friend from Thailand used to take me to all his favorite Thai food restaurants. And man - I know the Thai chili isn't up there with the big boys, but he'd order in Thai, and order "Thai Spicy" . I'd order 3, sometimes 4 stars but i'd try his and those dishes made my eyeballs sweat. Just ridiculous heat. And he'd order a side of Thai chilis, minced - he'd grab a couple big handfuls and sprinkle it alllll over his food.

I asked him how he could do that and he said after a lifetime of eating food that hot he didn't taste it any more. He needed that much to get heat. We in the western world are reared on PB&J and milk, not 5 star massaman curry and peanut sauce.

So more than the interval for the tester between tastings, wouldn't you also have to check their background or lineage? It seems eating copious amounts of capsaicin as a child - in your mother's milk, and her mother's before her may have a cumulative effect? Granted, my method is way less scientific since I'm basing it off one guy at a restaurant, but you can't tell me Thai people don't eat up loads of bigtime fire on a regular basis. I know when you see uploaded videos of people tasting the "new hottest pepper" and rolling on the floor vomiting unholy fire from their shoes it's always, always goofy white dudes. I've never seen one of those with a Thai in it. heh

Anyone have thoughts on this? Is there really no way to chemically test the strength of a hot pepper? Would love any information (or wild speculation, or humor) anyone has on this. :dance:
 
wikipedia:
High-performance liquid chromatography

Spice heat is usually measured by a method that uses high performance liquid chromatography (HPLC). This identifies and measures the concentration of heat-producing chemicals. The measurements are used in a mathematical formula that weights them according to their relative capacity to produce a sensation of heat. This method yields results, not in Scoville units, but in American Spice Trade Association (ASTA) pungency units. A measurement of one part capsaicin per million corresponds to about 15 Scoville units, and the published method says that ASTA pungency units can be multiplied by 15 and reported as Scoville units. This conversion is approximate, and spice experts Donna R. Tainter and Anthony T. Grenis say that there is consensus that it gives results about 20&ndash;40% lower than the actual Scoville method would have given. Results vary widely, up to 50%, between laboratories.[sup][3][/sup]
 
Wow - very interesting, thank you.

Hmm....If 1= 15 then how do they get the 4 at the end? Still puzzled by that.
=/

why does the hot sauce industry continue to use scoville if this is more accurate?
 
LD, you might find this page an interesting read, too: http://www.chez-williams.com/Hot%20Sauce/chemistry_and_scoville_units.htm

I've been trying to find more details on the Scoville Organoleptic Test as well, and without much luck. For example, how much is the "measured amount of dried pepper"? What is the concentration of the sugar water solution? Subsequent dilutions are straightforward but if you don't know how much pepper or sugar to use to start with you aren't going to get valid results. I guess if you had a sample of pure capsaicin crystals you could devise your own calibration scheme. Even that would fail to account for the mixture of capsaicin and other capsaicinoids in a given sample, but maybe that wouldn't matter; heat is heat.

In any case, there is no way a subjective test such as the SOT can yield valid results with seven significant figures. Such a number could be had by averaging the reported dilutions of the five tasters, but the average is only significant to the same level as the individual values. HPLC is much better but still does not have that kind of precision. A digital readout might show those numbers, but that doesn't mean they are all significant. These numbers may come from a lab tech just writing down the raw number from the display without giving due consideration to the actual capabilities of the instrument. And yes, that would assume an internal ASTA-to-Scoville conversion with some other unlikely number of significant digits.

It sure seems like an area ripe for more investigation.

To answer your last question, if only in part, tradition carries a lot of weight.
 
To answer your last question, if only in part, tradition carries a lot of weight.

First, stellar link, great post and fascinating subject - far moreso than the tongue-in-cheek topic I started (but then what did I expect at a hot pepper forum? Heh)

As for the quoted part I honestly don't have a problem with the Scoville Organoleptic Test, or that folks rate their sauces with it - Any marketing approach is good.
:)

I found how the US military rates LTL hot pepper weapons...
http://en.m.wikipedia.org/wiki/Pepper_spray

They use the HPLC method for what it's worth.

I think they grow ghost peppers for that primarily?




Ps - to be honest to my taste anything over "wow, that's really, really freakin' hot!!" (chug chug chug) is too hot. And I think that's probably somewhere well south of 1,000,000 Scoville units for me.

I've tried the nuclear ones at the SF Winter foods show (next weekend!!) - drop on a toothpick, 10 min burn. I honestly couldn't tell the difference between "ohmygodthatssohotIcrappedmyself!!!" and "ohgodogodogodkillmenowtomakethehotgoaway!!!!"

It must be something subtle? ;)
 
Its been debated whether pure cap is 15 or 16 million SHU. Some tests use 15, but I believe these days its usually 16
HPLC results are converted into scoville and these results can be quite different depending whether 15 or 16 million is used for calibration.
 
Yeah, I've seen the 16 million number referenced far more often than the 15 million, but I haven't seen how it is determined. Is it by administering the SOT with pure cap as the test material? Seems unlikely that if that were the case you'd wind up with either exactly 15 or 16. Maybe the tests average out to something between 15 and 16 and then for convenience the number is rounded off. If that's the way it's done. I'd love to get my hands on the relevant papers.

Does anyone here have, or have a reference to, the detailed procedure for administering the SOT (other than the original 1912 publication... I'm working on getting that one) or one of it's modified successors? It seems like something the average chilehead could perform at home with a minimum of extra hardware (scales, beakers, graduated cylinders, droppers). It certainly wouldn't be official, but it could give folks a ballpark idea of where their peppers fall on the Scoville scale and seems like something fun to do. I've queried the CPI about it and got a polite, but curt, response that they don't have that information; they use the HPLC method. I'd have thought they'd have the information for historical purposes, if nothing else. I'm usually pretty good at find any information I want, but my google fu is weak here. I smell a conspiracy... :scared:
 
I found how the US military rates LTL hot pepper weapons...

I think they grow ghost peppers for that primarily?

I know I've read the Indian government does this; I don't know about the US. If so, that could be some sweet action to get in on. Where do I bid?

Ps - to be honest to my taste anything over "wow, that's really, really freakin' hot!!" (chug chug chug) is too hot. And I think that's probably somewhere well south of 1,000,000 Scoville units for me.

I've tried the nuclear ones at the SF Winter foods show (next weekend!!) - drop on a toothpick, 10 min burn. I honestly couldn't tell the difference between "ohmygodthatssohotIcrappedmyself!!!" and "ohgodogodogodkillmenowtomakethehotgoaway!!!!"

I'm pretty much a wuss when it comes to eating this stuff. Trying to build up my tolerance. I made some powder recently out of toasted ghost pepper placenta. All I get is heat from too little to even see on a toothpick, but it sure smells good.


Cool, thanks. I'll look at that in more detail.
 

I think the Cracked scale (linked in the OP) is the most realistic for me. :lol: :dance: :lol: :dance: :lol: :dance: :lol:

I'm pretty much a wuss when it comes to eating this stuff. Trying to build up my tolerance. I made some powder recently out of toasted ghost pepper placenta. All I get is heat from too little to even see on a toothpick, but it sure smells good.

I think I'm somewhere north of "wuss" and south of "teensy bit on a toothpick has me begging for a swift death". I LOVE the habanero pepper. Not the R.S., just the straight hab. It has terrific citrus undertones and a wonderful heat quality.

An interesting anecdote - I recently made a fire-roasted Serrano/Jalapeno/dried Cayenne experimental sauce recipe that dozens of folks who tried it thought was "too hot". Yet I sent those same people my finalized Orange Label (a fire-roasted Habanero/Serrano/Jalapeno with much less dried Cayenne, which I am positive is a MUCH hotter sauce) - and they loved it. One fan described it like this, "The Cayenne is a 'rough' heat and really intense, while the Habanero is a 'warm' heat and much easier on the palate".

While unscientific, my point is that while you can possibly measure the chemical composition of a pepper, there would also seem to be an interpretation piece that I think is overlooked in all this talk. How the human palate interprets different types of heat is very interesting to me - that people with low heat tolerance would prefer a hotter pepper would seem to be evidence that "not all hot is created equal" - does that make sense? But then these are mild sauces comparatively - I admit that my sauces are never going to "out-hot" anyone. My goal is to "out-flavor" everyone, which seems like a tall task since there are so many great hot sauces out there. ;)

Also interesting (to me at least) is that I've experienced tremendous variety even within a single type of pepper and also within a specific cultivar of pepper.

I actually have an ag/hort background, and I know from experience with veggies & fruits that seasonality, geography, soil, nutrients, weather - all these environmental factors can and do affect the flavor of a fruit, whether it's tomatoes, cucumbers, grapes, oranges, etc. It's got to be the same with hot peppers - a single varietal habanero type grown in California is going to be stronger/weaker/have more citrus undertones, rougher/thinner/thicker/smoother flesh, variation in color intensity, and have more or less seeds/capsaicin oils than one grown in Nebraska or Wyoming, or Chicago or Texas, even if all are from the same seed.

All of those environmental factors would seem to make it very difficult to say, "I use X pepper so my sauce has Y # of Scoville units". Again, once you get up to the 1,000,000+ range, how can you really tell anyway unless you have "super tongue of steel". At the food show the giant posters of the Scoville scale with the arrow pointing to some astronomical number are good marketing, but seem questionable in accuracy/relevance.

Just more food for thought (no pun intended?) :dance:
 
Here's the thing with the scoville scale. It really can't be used as an exact measurement for any given species. Each plant can produce incredibly different heat level pods depending on an enormous amount of factors. One pod on a plant might be several thousand scoville points higher than another just because the leaves around it got thirty minutes more sun per day. This is why it is so hard to trust a scoville scale test as an accurate measure of heat for a STRAIN. It would be much more accurate if there was a standardized indoor garden where all plants of all species are given the exact same conditions and THOSE plants were tested for average heat. Then at least you could compare strains for heat in a more controlled test study.

That being said, high tech chili institutes like the one at New Mexico use incredibly accurate machines to detect every last nanogram of capsaicinoid molecules in a given sample. They can, in fact, test at higher accuracies than 1ppm. I couldn't give you an exact number, but if they managed to find a scoville rating accurate to the single digits, they must have the ability to detect finer measurements than 0.1ppm. Now whether or not they CAN give that accuracy is a completely different thing than whether or not they SHOULD. The sample of ground up pods that gave that exact number will never be replicated and so each measurement will be different. Add that to the fact that you can't expect to get within even 100,000 SHU of that number consistently and you have a bit of a problem.

Most importantly, the human pallet detects capsaicinoids on a logarithmic scale. This means that munching a pepper many times higher on the scoville scale might only be twice as hot. That's why jalapenos are hot at 5000, and a cayenne seems much hotter at 50,000, but not TEN TIMES hotter. The higher we get on the scoville scale the harder it is to detect small changes in the rating.

Think of it this way: When you have six pennies, and add three more, you notice a pretty big size difference between the two. But when you have fifty pennies and add three more, you can hardly notice the difference.

I'm not even going to get into the idea that there are different capsaicinoids that different people will probably react differently to, like LD had mentioned. Maybe another day.

I hope this helps
 
I'm not even going to get into the idea that there are different capsaicinoids that different people will probably react differently to, like LD had mentioned. Maybe another day.

I hope this helps

Interesting post - agree completely with the environmental variables. I dunno if the quoted part is true or not but different people have different tolerances for capsaicin, so it seems likely. Whether its genetic or conditioning I dunno.

I know my Thai friend could eat a gasoline sandwich & ask you for a Strike Anywhere match to wash it down. That's more in the "building up a tolerance" category though.
;)


 
HPLC is the best way to test SHU but it is not perfect, and not every lab has the expertise to provide accurate results. There always seems to be errors in HPLC tests, especially with smaller labs not used to testing food in this manner. As said already the best way to test pepprs is to grow all the superhots in the same unstressed conditions and environment and test only random pods from random plants, then have them re-tested at another lab for varification. Results likely won't be as high as the records since those plants are likely stressed nearly to death, but you'll still learn which varieties are hottest.
Even the chile pepper institute(NMSU) usues the analytic food lab to test or re-test their peppers since it is the most accurate lab known in the US for this process.
Testing fresh peppers in not everone's thing, but try using them in cooking or sauces. I like to make exact same recipes for sauces and sausages etc. with the exception of the chiles. Great way to compare heat (and flavor)without blowing up
 
High performance liquid chromatography (HPLC), the technique used by NMSU CPI and other analytical facilities measures the concentration of capsaicinoids, typically in units of micrograms/milliliter of solution. This can be a very precise technique depending on the quality of the instrumentation, calibration schedule, and the competency of the technician, but as POTAWIE notes it is prone to error. Even in an error-free case, though, it is somewhat misleading to quote a single measurement to 7 significant digits. Running another sample of the same starting material will almost always yield a slightly different value. The most honest way to quantify the results in a single number is to quote the average of several measurements along with some measure of the variability such as the standard deviation or variance or some other generally accepted quantity. Since it's usually prohibitively expensive to run every sample several times, it's more common to periodically calibrate the instrument with a known quantity of a standard material. Then any quantity quoted with n significant digits is understood to be accurate to +/- 5 in the n+1st digit. This site has a good, brief intro to these ideas.

But I digress from my point. In one sense, it really doesn't matter how accurate or precise the HPLC measurement is with respect to quoting a Scoville rating, because in this case the Scoville rating is derived from the HPLC measurement via a multiplicative factor. The end result is only significant to the number of digits contained in the factor with the fewest significant digits. One number I see mentioned frequently for that factor is "15", so any SHU rating derived from HPLC can only validly contain 2 significant digits. Unless the factor is 15.0 or 15.00, etc., or is treated to be an exact number. Given that the Scoville scale itself is a subjective scale, I don't see how any conversion factor could be considered "exact", though I'd be interested to hear an argument that it is exact.

But none of that really matters for our purposes. It's fun to talk about which pepper is hotter and a few extra digits make the game more enticing.

I found this interesting link while writing all this: http://www.cqbsupply.com/history.html
 
The Scoville scale is inaccurate at best. Even the HPLC method is highly innaccurate. Myself and other companies I know have sent some of our hotter products for testing. We sent 2 bottles of the EXACT same sauce, and the results were stupidly different (yeah, it was a complete waste of money). I even sent in the ZERO Sludge, a serious concentrate of Habaneros (10 pounds reduced to .87 ounces), the HPLC result was 35,000SHU, the equivalent of a Cayenne pepper. I laughed so hard I almost crapped myself.

We get asked these questions at virtually all of the shows we do, "How many Scovilles are in this". I tell them, "A whole bunch". If something is hot, it's hot, if something is friggin' hot, it's friggin' hot. I do like the people that tell me they can tell the difference between 500,000SHU and 501,000SHU. They are usually they ones with the "L" tattoo on their forehead.

I have never had any trust in the Scoville scale, and with the personal experience with HPLC testing, it holds about as much water as a sieve.
 
Okay okay I'll bite one more time now that I have some down time.

There are plenty of different capsaicinoids. They are basically all similar organic compounds that will react to the same pain receptors, but all of which will react in a slightly different way. Just to name a few, you have your major 'Capsaicin' and then of course Dihydrocapsaicin, Homocapsaicin, Homodihydrocapsaicin, Nonivamide, Nordihydrocapsaicin... it keeps going.

Each one will affect you slightly differently. One might creep up slowly and then burn you for a long time, while another might slam you all at once. There are ones that cause a more stingy burn and some that are a warmer, rounder burn. Some nail your throat and some nail your tongue and mouth. Each strain of pepper will produce different quantities of each of these compounds. Some have very trademark burns, like the Fatalii, mainly because of its high Dihydrocapsaicin content I believe (correct me if I'm wrong). This is why no two peppers will burn exactly the same.

The problems come when you have to rank each of these compounds on the scoville scale. Pure capsaicin is rated around 15-16M SHU, while Nordihydrocapsaicin is said to have a heat of somewhere around 9M SHU. That, of course, is assuming that we all react the same to both compounds. We don't. So to you NDHC might clock in around 7M SHU but to me it might seem like 11M if we compare it to a standardized 16M pure cap. So each pod will be different for each person even if you have equal tolerances.

Just one more reason to loath the scoville scale.

Maybe one day a chili pepper institute will have the balls to grow a few hundred strains out in a standardized lab and test for ppm of each of these compounds. Then we can use that as good lab data to judge heat. There should be no standardized, scientific scale used to declare what pepper will cause the strongest burning reaction, because quite frankly, it's impossible to standardize.
 
Each one will affect you slightly differently. One might creep up slowly and then burn you for a long time, while another might slam you all at once. There are ones that cause a more stingy burn and some that are a warmer, rounder burn. Some nail your throat and some nail your tongue and mouth. Each strain of pepper will produce different quantities of each of these compounds. Some have very trademark burns, like the Fatalii, mainly because of its high Dihydrocapsaicin content I believe (correct me if I'm wrong). This is why no two peppers will burn exactly the same.

I don't know the science behind it, but I absolutely agree with the bolded bit. I've experimented with dozens of fresh and dried peppers and there are vastly different aspects to the heat. I refer to them as layers in casual conversation and people seem to get what I mean.

For example the Chili Tepin or Bird's Eye - dried Tepin has this great quality of an explosive heat thats really strong, but very short-lived. Habanero has the slight sting, and Cayenne has something of a "slow burn" - get all those qualities together and you end up with a "presentation" (when it first hits your mouth) of Tepin, followed by the burn of the Hab while you're eating it, followed by the lower, slower Cayenne burn. All seem like independent kinds of heat. Your explanation matches what my "gut" has been telling me for years - there are a lot of different types of burn, and not all heat is created equal.

great post - very informative (again) - thanks!
 
I agree entirely with the points being made regarding the various types of capsaicinoids and their varying impacts on different human palates. Anytime you try to quantify a subjective experience you are skating on thin ice with hot blades, but I believe the Scoville scale is still useful, as long as everyone understands its limitations. HPLC, on the other hand should be a much more precise and objective test of the capsaicinoid content of a pepper or pepper powder. (There might be some interfering components in a sauce that could complicate sample prep, but those complications can be overcome with a little effort.) If done correctly, the test report should include the concentration (mass of capsaicinoid per unit volume (or mass) of sample) for each detected capsaicinoid. Things get a lot more iffy trying to relate those objective numbers to the subjective Scoville scale.

Even the HPLC method is highly innaccurate. Myself and other companies I know have sent some of our hotter products for testing. We sent 2 bottles of the EXACT same sauce, and the results were stupidly different (yeah, it was a complete waste of money). I even sent in the ZERO Sludge, a serious concentrate of Habaneros (10 pounds reduced to .87 ounces), the HPLC result was 35,000SHU, the equivalent of a Cayenne pepper.

DEFCON, it's unfortunate you (and others) have had such a poor experience with HPLC testing. I firmly believe the problem lies with the testers, not the test itself. I suppose it's inappropriate for me to ask who did your testing for you (here anyway, maybe in the vendor forum?), but may I ask if they were university labs or commercial labs? Also, what was the form of the report you received? Did it give a capsaicinoid breakdown as I mentioned above along with the formula used to convert to Scoville? Did the lab include a copy of a recent calibration run? Finally, may I ask how much the tests cost?

My questions are motivated by more than idle curiosity. I'm looking into getting my own HPLC system to support my own research (peppers, small grains, oil seed, etc.) If I do, I will almost certainly make testing services available commercially as well. (I'll run a poll on interest here on THP if/when I get closer to actually acquiring a system.)
 
Back
Top