Experiments in Cookware Efficiency
- kaleokahu Sep 7, 2010 04:45 PM
Over the weekend (yeah, it was raining here in Seattle) I ran a little real-world experiment to gauge the heating/cooling efficiency of saucepans of varying composition. I'm interested in sharing the results, and also in finding out if others have run their own similar--perhaps more scientific--experiments.
I had 4 same-diameter (5.75 inches), straight-walled saucepans, one each of tinned copper (3mm), enameled cast iron (3mm Belgique), stainless/copper clad (1 mm Revereware), and monolithic aluminum (1mm mongrel). I put 3C of tapwater in each, and let them all come up to room temp, then put them, in turn, uncovered, onto an electric radiant heat hob of the same diameter, preheated and maxed out on HI.
The times to boil were as follows:
Cast Iron 9:21
SS Clad 12:10
This order of finish was not surprising to me given the published conductivity numbers of the various metals, but the time differences were surprising--what those differences mean not only in terms of cooking time, but also in terms of energy used. Comparable time differences were also consistently noted at 100, 125, 150, 175 and 200 degrees F. Also surprising was the upper limit of temperature measured in the pans: the copper registered a max of 210F on my dairy thermometer and attained a roiling boil, whereas the others maxed out at lower temperatures (Al 202, Cast 207, and SS 206) and had less vigorous boils (the aluminum pan's was particularly feeble). While this embarrassingly proves the weakness of my hob, it also proves the relative efficiencies of the materials.
Following each boil test, I also timed cooling. The boiling pan was moved to a cool trivet and the time to cool in 25 degree steps was measured. The times to cool from boil to 125F were:
SS Clad 38:46
Cast Iron 40:51
The surprises here were that the water in aluminum cooled substantially faster than that in the copper pan, and the SS pan cooled it faster than did the cast pan. However, I did not have access to 3mm-thick aluminum or clad pans, so I think the much-thicker copper and cast pans tended to retain more heat than the thinner materials, and perhaps skewed the numbers.
Finally, I tested the pans to assess what hob settings would maintain a gentle simmer, both lidded and uncovered. The results were:
Copper 3 (2 lidded)
Cast Iron 4 (3 lidded)
Aluminum 4.5 (3.5 lidded)
SS Clad 5 (4 lidded)
If someone DOES have access to 4 same-sized pans of SAME bottom/wall thickness, and wants to run a similar test, I'd be interested in the results.
"The surprises here were that the water in aluminum cooled substantially faster than that in the copper pan, and the SS pan cooled it faster than did the cast pan."
First off, thank you very much for taking the time to post your results..
As to the above quote, I cannot say for certain, but I suspect the reason for the strange results may have been quite simply that the aluminum pan had less mass (and as such retained less heat), and likewise the cast iron had more mass than the ss clad. To compare the specific heat characteristics of the different materials, I think you would need samples of equal mass AND surface area, which may not be practical for a cookware test.
That said, your test is actually quite nice in that it compares the heating and cooling characteristics of different constructs of pan in a more realistic kitchen scenario. As such that data is easier to apply to actual cookware choices than raw figures about the conductivity and specific heat of various metals might be.
Yes, of course you're correct. But in the world we live in, we buy and use pans by portion SIZE. Isn't that the only comparison that makes practical sense, regardless of mass and specific heat and gravity?
I DO have one thick-walled saucepan in aluminum, and I MAY have a similar-sized one in 3mm copper, but they are larger than my 5.75" trial sizes, so we'll have to extrapolate a bit, use a larger hob, and fill to the same level as the initial test. What results do you predict for a 3mm-walled aluminum pan versus copper of the same size?
Ok, I'll criticize the experiment a bit.
You used a radiant element hob, which means that it wasn't just the conductivity of the materials that mattered, but their properties relating to how well they absorb infrared radiation, *and* the specific condition of the bottom surface along with its color.
Try comparing the same size, make and model of enameled cast-iron pots but where one is black and one white. You *will* see a difference. Was the aluminum pot anodized? Scratched and rough or polished and shiny? Try polishing the aluminum pot's bottom to where you can see yourself and try the experiment again. I'm willing to bet the aluminum winds up taking even longer than the 12 minutes for the SS pot.
Remember, most mirrors these days are made by plating aluminum onto glass... how effective do you think trying to heat a mirror with radiant heat would be?
Truly, your results of the cooling experiment probably more closely resemble the results you'd get from using a gas hob to boil water, meaning the aluminum would probably outperform the copper, with the enameled cast iron coming in last place.
So, my request... try the same experiment with the same pots, however this time clean and polish the bottoms.
The aluminum pan's bottom was clean but not mirror polished--pretty matte. As were all of the rest. The cast piece has a yellow enamel bottom. If anything, the copper pan was the most "mirrorlike".
Sigh... I've already spent TWO evenings trying to get results that are meaningful [see Post #2 below]. I guess now I'll try it again on a gas hob. But I do not understand the basis for your prediction that aluminum would outperform copper at boiling water. Can you explain the theory?
I'm not sure I'm willing to blacken my pan bottoms with high-temp black woodstove paint--unless you're willing to strip and polish 'em!
Your aluminum pan is 1mm thick. The copper is 3mm. Copper conducts twice as well as aluminum, but it's 3x as thick. So if aluminum's conductivity is 200, 200 divided by 1mm is 200. If copper's conductivity is 400, then 400 divided by 3 is 133. Thus the thinner aluminum pan will allow heat to more easily reach the water than the thicker copper pan.
Now, I suggest the aluminum pan will win on the gas hob because of the cooldown results. Assuming equal surface area of the water (since evaporation will be the biggest contributor to cooling) the only factor that would differentiate the pans is heat loss through convection (heating up the air) which is directly influenced by conductivity.
Here's Post #2, in which I attempt to remove any effect of the initial aluminum pan being very thin (1mm). As I only had one thick aluminum pan (3mm), and one copper that was the same size (7 inches diameter), there are no comparable figures for cast and clad. I used the same hob, the same heat, same trivet, and the same procedure, except that I filled the pots to the level as in the previous test (about 3/4 full), 7 Cups instead of the 3 that the smaller pans held.
Here are the times to boil along with the max temperature reached:
Copper 13:55 (206)
Aluminum 28:08 (201)
Bear in mind that for consistency, the hob was the same small one used in the original test. A larger hob would have yielded faster times, but probably the same winner. The times to cool from a boil back to 125F were:
The hob settings to maintain a gentle simmer (uncovered/lidded) were:
I am concluding that, so far, copper is the clear winner as far as fastest heating and cooling, as well as energy use on my radiant stovetop. However, the cast pan actually outperformed the aluminum in heating to boil and (at 207 F) attained a more vigorous boil. The SS clad pan was dismal at heating and simmering, but did edge out the cast pan at cooling (probably by virtue of its thinness).
As time allows, I'll try the original experiment again using a gas hob, to test ThreeGigs hypothesis.
It occurs to me that a 220V Watt meter would be extremely useful to actually measure energy used in these tests. Anyone have one or willing to carry the ball and do your own experiment?
OK, here's Post #3, in which I report results on using all 4 original pans (5.75 dia.) on a gas hob instead of my radiant hob as originally reported. I also repeated the tests using the two 7" pans to show a direct comparison between 3mm copper vs. 3mm aluminum. The gas hob in this case was my triple-ring crab boiler, so you will see in the heating times that there were MANY more BTUs dumped into this test than with electricity in the first test.
Times to boil and max temp reached on gas were:
Copper 5.75 3mm 4:50 212
Aluminum 5.75 1mm 5:25 207
SS Clad 5.75 1mm 4:46 212
Cast Iron 5.75 3mm 4:47 210
Copper 7" 3mm 8:10 211
Aluminum 7" 3mm 11:53 209
These were the times to max temp--all 4 smaller pans made it to 200F within 10 seconds of each other, but the thin aluminum pan took almost a full minute to come up the last 7F!
Cooling times were again measured (because the max temps were higher with gas) and from max back down to 125F the times were:
Copper 5.75 3mm 27:40
Aluminum 5.75 1mm 17:28
SS Clad 5.75 1mm 27:39
Cast Iron 5.75 3mm 30:10
Copper 7" 3mm 32:18
Aluminum 7" 3mm 39:59
Because my crab boiler does not have a numerical dial on the valve, I could not replicate the simmer test. So instead, I tested the copper and aluminum 7" pans' time-to-boil, but with lids in place to cut surface heat loss. Those times, recorded when the lids first spat out steam were:
Copper 7" 3mm 7:30
Aluminum 7" 3mm 9:32
From all these tests, I have concluded:
1. For equal thicknesses, copper clearly outperforms aluminum by a substantial margin in both heating and cooling.
2. Thinner pans--of all materials--will boil quickly and all about the same speed, provided you dump enough heat into them.
3. Gas hobs tend to extend their heat (but NOT the flame in this case) past the pan bottoms, up the sides of the pans, and even directly above the water's surface (as witnessed by the fact that I had to use welding gloves to hold the thermometer), whereas the radiant heat was applied only to the bottoms. On gas, copper pan actually could be seen boiling from the SIDES, as opposed to just the bottom.
4. Cast iron DOES have a definite "heat holding" effect, as shown by the longer cooling times. Notably, the cast iron pan attained the second-highest (to copper) max temperatures, and sustained an active, if not vigorous boil, more active than either SS or aluminum).
5. Aluminum lived up to its 2nd seed ranking for conductivity, but with a twist: from 200 heating to max, and from max back to 200 cooling, aluminum was far less responsive to the heat than at lower temperatures. Moreover, the aluminum pans registered the LOWEST max temperatures of all the pans. In that 200-to-200 range, the aluminum seemed sluggish in conductivity.
6. The simmer and covered boil tests proved convincingly that copper uses far less energy. This held true even with the thin 1mm aluminum pan. Query whether the simmer result would have been the same on gas, where the heated combustion gasses continue up and over the pan walls and lips. And until I get a watt meter, quantifying the energy savings of copper will have to wait.
I hope the 'hounds find this info useful in making their consumer choices.