"Degree Days" rule for BBQ?

[Syzygies]

Recall the notion of a degree day. For example, growing degree days add up the surplus heat over a base temperature, often 10 C. Three days averaging 15 C, 20 C, 25 C add up to 5+10+15 = 30 degree days.

What is the best rule of thumb for degree hours in barbecue?

For example, in the thread Dinosaur Beef Ribs, Hector cooked 10 hours @ 235 F. Dave cooked 8 hours @ 285 F. Different pieces of meat, but it was rather evident that Dave overcooked in comparison to Hector. Could one have predicted this by an easy rule of thumb?

A base temperature of 35 F makes these two cooks the same: 10 * (235 - 35) = 8 * (285 - 35) = 2000 degree hours, either way. But 35 F is obviously too low, and we know Dave overcooked compared to Hector.

One can sous vide pretty much indefinitely at 135 F. We have 10 * (235 - 135) = 1000 degree hours, while 8 * (285-135) = 1200 degree hours. That difference is closer to what we observe, studying the two cooks in the thread.

The rate of ideal heat transfer is proportional to the temperature difference, but real world roasting is far more complicated. One follows the water as much as the heat, to understand what is happening. For example, one can model the dwell in a pork butt cook by watching what happens to a wet towel. Nevertheless, a rule of thumb like I propose could be useful for anticipating the effect of small changes in protocol. If I know how long I like to cook a pork butt at 225 F, how do I adjust my cooking times for 240 F? That sort of question could be easily handled by a rule of thumb like I propose, at least for a first guess.

[cschaaf]

This thread is WAY over my head... but I am interested in seeing the replies.

[_Ed_]

so the art would be to establish the correct baseline temperature, no? And that could only really be done as a practical, with three identical pieces of meat cooked at, say, 225, 250, 275, and pulled at optimum doneness (which is not necessarily a constant, but rather subjective).  This gives you your 'time to doneness' and allows you to establish your baseline temp, which in turn allows you to extrapolate timings to other temps.

[wilburpan]

Although there’s something to the idea of finding equivalents of temperature x time with varying conditions of temperature and time, there are going to be limitations to that. To take an extreme example, you could cook a moderately thick 1/4 burger in a KK grill at 200ºF for as long as it takes to cook it through. That’s not going to be equivalent to cooking the same burger for a short amount of time over a 700ºF fire, no matter what the time period is. Over the hot fire, the outside is going to be cooked to a greater degree than the interior.

My feeling is that making those sorts of adjustments happen on the fly, and come with experience. If I want to cook my pork butt at 225ºF, and I see that Smaug settled in at 250ºF that day for whatever reason, I’ll know to check on it sooner than expected, while dialing down the vents a bit. 

This is one of the things that makes cooking different from baking. If I’m sautéing some garlic on my stove, I can use the same pan, the same amount of oil, and the same setting on my stove, and some days the garlic just cooks faster than I expect. I’ll adjust for that as it happens by pulling the pan off the flame and turning down the burner.

Ed also alludes to one other key factor when he mentions the three identical pieces of meat. There aren’t three identical pieces of meat. The difference between one rack of ribs and the next, or one brisket and the next, probably overrides any consistency you’re looking for with time and temperature.

[_Ed_]

yes, agree entirely. that's why you'd have to set the notional experimental temps within 'accepted' bbq ranges for it to be meaningful. And, as we always say, we're cooking to doneness, not to time - and the variances in the meat itself will cause significant noise in the data with such a limited sample size. 

[mk1]

Interesting concept but it will be dependent on the device used as well due to differences in airflow....

Edited May 31, 2016 by mk1

[Syzygies]

Interesting replies. Let's just stipulate that nothing beats experience, and there is a boatload of gotchas and nuances here that any rule misses.

For an interesting parallel, human stock traders have always valued judgment. Yet there was a time when no one had any idea how to value stock options, then there was a rule. People still bent the rule using experience and judgment, but the rule was a good starting point. The traders that started with the rule now live in bigger houses than us. The traders that ignored the rule sooner or later lost their shirts.

Let's confine interest to the practical range of low & slow for large hunks of meat, say 210 F to 285 F, cooking in a KK. You had a plan, but you're starting an hour late. What's your best guess how to adjust your plan? Bend this guess using all the experience and judgment in the world, but start with a baseline guess. What's your baseline guess?

My bias here is that cooks who proudly refuse to measure are foolish. This is most so in baking, but also in salting meat. One needs as much experience as Aaron Franklin to salt as accurately by eye as anyone can salt with a scale. Bend the note using judgment, but start with a number.

I could start a similar thread on a running forum. I kept race records for many years, including my weight, and my race times were best explained by foot pounds per hour. Of course there were nuances, who was I dating? How was I sleeping? What running shoes? Was I training enough? But the rule worked. The philosophy of calculus predicts this; zoom in on anything that curves, and it looks flat.

Asserting a pair of protocols yield the same result is asserting a value for the base temperature in determining degree hours.

To give a formula, asserting m hours @ x degrees = n hours @ y degrees is asserting that (m * x - n * y) / (m - n) is the base temperature.

For example, asserting that 10 hours @ 220 F = 8 hours @ 240 F is asserting that (2200 - 1920) / (10 - 8 )  =  140 F is the base temperature. Using this base temperature, 10 * (220 - 140) = 8 * (240 - 140) = 800 degree hours either way.

So you were planning to cook 10 hours @ 220 F but you find your KK stuck at 240 F. To hell with my math, use your judgment. What is your initial guess how long the cook will take, before you begin prodding and using your judgment, to actually decide when it's done?

Edited May 31, 2016 by Syzygies

[CeramicChef]

@Syzygies - I like your ideas.  I've gotta think about this for a few.  I used to do a lot of heat transfer and thermo back in the day and I've gotta knock the rust off. And I like your analogy of the options traders.  The guys who discovered that time is a wasting asset correlated with risk and return won a Noble Prize for their discovery.  I think their equation took the form of a second order PDE as I remember.  This equation was very familiar to anyone who ever took a course in heat transfer at the graduate level.  I also tend to agree with your philosophy of calculus analogy.  Economists have survived for decades drawing straight lines in a curved world.  Are you saying that to a first approximation, linear interpolation is a valid methodology for adjusting cook times?

A couple of questions come to mind.  Let's talk butts and briskets.  Both are quite heterogeneous.  How do you take into account that the cook time for similar 9# butts or 15# packers is so different?  How do you account for differences in geometry, i.e. surface area available for heat transfer?  I've had a 9# butt be done in 12 hours and I've had 9# butts take 14 hours. The same is true with briskets.  And over the past 20 years of low-n-slow cooking on kamados at 225F, I've seen different cuts of meat take time that was unreal.    (I've never kept any recent cook data; I'm the guy who cooks to an Internal Temperature and then probes for tenderness.)  That's why I like using electronic thermometers.  I follow the temp of a cook and look to adjust accordingly.

I'm wondering if meat just isn't too heterogeneous for your undertaking.  Fat content varies tremendously.  So does water content.  The protein content will also vary tremendously.  All these factors taken together mean that cooks happen idiosyncratically given that heat transfer through a nonstandard medium (meat) will differ.  I know my kamado cooking mentor, a Professor of Meat Science, would love to chat with you about this.  Alas, he is now deceased and cooking great BBQ in the sky.  

I will agree with you that to a first approximation, your methodology may not be too bad.  I think one has to consider the geometry (since weight really means nothing).  A butt cut into quarters will cook much quicker than the same butt left whole.  

I've really got to sit and think about this whole notion.  You're off to a good start.  Might there be a publication in this someplace for you?  If so, GOOD!

 

[Syzygies]

6 minutes ago, CeramicChef said:

I've seen different cuts of meat take time that was unreal.

Yes.

I've seen calculus students, when asked to calculate a volume, give a negative answer. They weren't being idiots, they were just forgetting to ask the question, "Is my answer reasonable?"

We all have to guess plans, 10 hours @ 225 F. And the best plans go awry, in all the ways you note. But we still need initial guesses. Yes, I'm saying that to a first approximation, linear interpolation is a valid methodology for adjusting cook times.

My 8 hours @ 285 F felt at odds with Hector's 10 hours @ 235 F, even though the pieces of meat were different. Degree hours makes this precise; one would need to posit the ridiculous base temperature of 35 F for those two protocols to be similar, whatever the shape of meat.

Let's back off and say that two protocols are plausibly similar if they correspond to a base temperature anywhere near 140 F. If the computed base temperature is nowhere near that range, then one is being an idiot. At least I was being an idiot. We're screening for plausibility here.

[CeramicChef]

@Syzygies - LOL!!  I was always in for a treat whenever I gave an exam, especially when I was teaching Thermo.  Answers were so far from reality I almost had a red rubber stamp made that said WHAT WERE YOU THINKING?  Alas, I didn't.  Too many tender feelings.

I'm really liking your ideas here.  And I think that they have some real merit.  In fact, you got me thinking (watch out!) and I think I'm going to start keeping notes again on my cooks.  I also think I'm going to start using your linear interpolation technique to see what we can see after a baseline is established.  This should be fun.  I was always a theoretician versus an experimentalist.  Now I'm living on the ragged edge and it's all your fault!

Stay tuned for further announcements ...

[mk1]

I agree about geometry as well, but as a rule of thumb I think this would work. I also believe it would be more accurate if you did a separate formula for each cut of meat. I know that baby backs take 5 hours at 250 but 6 hours at 225. I have done both many times. I think it would be possible to develop a nomogram for different cuts by gathering data collectively.

Edited June 1, 2016 by mk1

[Syzygies]

1 hour ago, mk1 said:

I know that baby backs take 5 hours at 250 but 6 hours at 225.

(5*250-6*225)/(5-6) = 100 F, and 5*(250-100) = 6*(225-100) = 750 degree hours either way. 100 F is a lower base temp than I would have expected, but plausible for this rule. This supports your point that it depends on the cut.

Using instead a base of 125 F, 5*(250-125) = 6.25*(225-125), suggesting that 5 hours at 250 = 6.25 hours at 225. So any reasonable base gives reasonable predictions.

Edited June 1, 2016 by Syzygies

[egmiii]

I find this thread fascinating and feel compelled to chime in, despite my mere 7 months of experience. While solving for the baseline degree temperature may make sense as a starting point, I believe the formula must become a bit more complicated to be useful. Forgive me for using very loose numbers below.

I break my cooks down into three phases, specifically tied to the meat temperature.

35F-150F - "Smoke Ring": During this phase, I do my best to maintain a very clean (light blue) smoke and low temperature. The top damper is choked way down to maximize moisture, smoke condensation, and development of the smoke ring (from NO and CO). I firmly believe this phase can't be rushed. Lets assign it a constant, A=4 hours.

150F-180F - "The Stall": Smoke condensation drops significantly and the meat temperature stalls somewhere in this range. This is the phase where you can generally run the grill hot to accelerate the moisture loss, or wrap it to avoid moisture loss. The weight, shape, and water/fat content are significant variables when unwrapped, and would amount to a unique BBQ DegreeHour for a given permutation. Lets assign this phase variable B, where B (in hours) = BBQ DegreeHours / (Grill Temp - Base Temp). When wrapped, a universal constant could be substituted for BBQ DegreeHour, versus the highly subjective number for unwrapped.

180F-210F - "Collagen Breakdown": Collagen breaks down into gelatin, giving BBQ its wonderful texture. Arguably, this is the most critical phase. All the wonderful smoke flavor and moisture in the world won't help if you can't chew your food. I believe an exponential relationship exists, similar to the bacteria kill curve. Anyone who cooks sous vide is generally familiar with the concept. Everything dies instantly > 160, but with enough time, it will also die in the low 130s. Gelatin will form eventually in the 180s, but a brisket will be ready in short order at 208. I believe this is why probe tender rules for doneness, not the final temperature. The journey is more important than the destination. Unfortunately I have never taste tested the difference in texture between a brisket held for 8 hours at 190, vs 30 minutes at 208. I have, however, eaten many corned beefs boiled for a couple hours, and can say the corned beef simmered at 190 for 10 hours was far superior. I suspect this is true for smoked meat as well. Lets call this final period C (in hours).

Putting it all together and you get A + B + C = Total cook time. A traditionalist may go 18 hours at 225 unwrapped. An optimizer may go 225F x 4 + 300F X 3 + 225F x 3 for a 10 hour cook (with a wrap in phase 2). And a late sleeper might go 250F x 3 + 325F x 3 + 350 x 1 for a 7 hour cook (with a wrap in phase 2).

Ideally I'd love to produce consistently good BBQ. Knowing all the variables involved makes that a difficult proposition. Understanding the phases above, their contribution to overall quality, and the techniques used to reduce the time spent in each has been a focus of mine. I've eaten more pork butts alone after midnight than before, so consistency has become a priority. Hopefully my new CyberQ will allow the exploration of these concepts in a consistent way. I look forward to your thoughts.

[CeramicChef]

@egmiii - wow, I've learned something with your post.  I just throw the meat on and pull when it proves tender.  And i understand a little better now the three stages. I've never run temps up and down during a cook.  But I do have a question: after 3 hours at 300°F, how you you get you temp back down to 225°F?  TheBeast simply doesn't bleed heat that quickly!  What I have done is 225 to the stall, wrap at 160, and then turbo cook at 300 to an Internal Temp of something like 200-205 where it probes like warm butter.

I think to a first approximation, @Syzygies method works and shows promise as Phase 1.  @egmiii, I think your techniques would be Phase 2 after the validity of Phase 1 is shown.  Geez, this sounds like a research program!

[CeramicChef]

Ya know, this thing has interesting possibilities.  Getting good datat would take a lot of it,e individually, but corporately, we could knock this out pretty quickly.  We could all agree to do pork butts on a specific weekend.  Record the weight of the butt that goes on the grate and the weight at the end of the cook.  First cook is at 225° and record the ambient.  Report the time to prone tender, say 203°.  Then 2 weeks hence, same thing but crank up cook temp to say, 275, and record the same data as before.  No Texas crutch at this stage.  Two weeks hence we again cook butts only at different temps between 225-275.   Same data recorded and reported. 

@wilburpan, you've got some access to SAS or SPSS or some big statistical package.  Could you and @Syzygies crank the data and see how his model presents?  You guys up for some modeling "fun"!?

[tinyfish]

Don't know if I'll be any help, my wife married me for my looks...lol

[ckreef]

I've been reading this thread and trying to keep up without going to sleep.

.

BBQ isn't science it's an art form. This is sucking the fun out of BBQ.

Just my opinion don't hate me for it.

.

Sent from my XT1585 using Tapatalk

[CeramicChef]

2 hours ago, tinyfish said:

Don't know if I'll be any help, my wife married me for my looks...lol

@tinyfish - poor woman!  Born blind and then she marries Tony …. There is simply  just no justice in this world!

[Cookie]

2 hours ago, ckreef said:

 

I've been reading this thread and trying to keep up without going to sleep.

 

.

BBQ isn't science it's an art form. This is sucking the fun out of BBQ.

 

Just my opinion don't hate me for it.

 

.

Sent from my XT1585 using Tapatalk

What he said...but if someone comes up with a standard base temp, I'll be in line to try it...

[wilburpan]

11 hours ago, CeramicChef said:

Ya know, this thing has interesting possibilities.  Getting good datat would take a lot of it,e individually, but corporately, we could knock this out pretty quickly.  We could all agree to do pork butts on a specific weekend.  Record the weight of the butt that goes on the grate and the weight at the end of the cook.  First cook is at 225° and record the ambient.  Report the time to prone tender, say 203°.  Then 2 weeks hence, same thing but crank up cook temp to say, 275, and record the same data as before.  No Texas crutch at this stage.  Two weeks hence we again cook butts only at different temps between 225-275.   Same data recorded and reported. 

@wilburpan, you've got some access to SAS or SPSS or some big statistical package.  Could you and @Syzygies crank the data and see how his model presents?  You guys up for some modeling "fun"!?

If I wanted to put that much work into smoking a pork butt, I’d be cooking on a Weber grill instead. 

***ducks***