Over on Yahoo Backpacking Light forum I was asked about alcohol stove performance by a guy planning to use an alcohol stove while hiking with his wife. He asked me how one an alcohol stoves would really perform when trying to boil very cold water. Last night I did the testing and thought I would share it here in case anyone is interested:

-----------------------------------------
As promised, more stove testing.

Challenge: Test the Brasslite Duo ability to boil for a pair of hikers.
To make the challenge a little harder, and maybe more realistic, the
water and alcohol were both chilled to a temperature between 32*-40* F.

Brasslite Duo
Price: $55
Weight (manufacturer): 1.2oz
Weight (tested): Duo = 2.1 (with measuring bottle, simmer ring, and
windscreen per my instructions)
Fuel type: Alcohol
Capacity: Duo = 1.75 ounces (52 ml)

Tests
Standards:
1. 1 Liter of water at refrigerator temperature – 33-40 degrees.
2. Pot used is a Wal-Mart 1.25 Quart aluminum mug with aluminum
foil lid.
3. Stove used the windscreen per instructions from my web site.
4. Alcohol was tested in 6 ml increments, starting at 24 ml
alcohol. Priming fuel came from the measured amount.
5. Each amount was double checked using a scale. The weight of one
fluid ounce of alcohol is .82 avoirdupois ounces.
6. Each Test was done only once - shortage of chilled water for
the test at this time.
7. Starting time was when the stove was lit.
8. Stoves were allowed to completely cool between tests.
9. Barometric pressure here was 30.15, the boiling point was
determined to be 212.43 degrees.
10. Altitude is 90 meters above sea level.
11. For temperature readings, a Thermocouple was used hooked to a
digital multimeter.
12. Air temperature was 65 degrees with the stove fan running to
simulate a light wind.
13. Scale used was a Royal EX3.
14. To further simulate possible field conditions, the alcohol fuel
was also chilled in the refrigerator to 33-40 degrees.
------------------------------
24ml Test:
150* - 12:10
175* - N/A
200* - N/A
Max Temperature - 160.0*
Time to Max - 13:40
Burn out - 13:45
Time over 175* - N/A
------------------------------
30ml Test (this test was actually a little colder - ice was in the
water):
150* - 14:00
175* - 17:20
200* - 20:55
Max Temperature - 200.0*
Time to Max - 20:55
Burn out - 21:10
Time over 175* - 3:50
------------------------------
36ml Test:
150* - 13:15
175* - 16:30
200* - 19:30
Max Temperature - 212.5*
Time to Max - 20:20
Burn out - 23:45
Time over 175* - 7:15
------------------------------
Conclusions:

1. The Brasslite worked about as I had figured it would for a couple.
36ml fuel to boil 1 liter water is consistent with earlier tests.

2. The best way to speed boil times is to warm the fuel up before use.
Although not tested to the same exact standard - I found that fuel
warmed to room temperature achieved much better results.

3. Time over 175 is important because that is the temperature most of
the "real cooking" is happening. Most coffee makers only heat water to
about 180 degrees, so at this point most hikers would be pleased and not
need to further heat water or waste fuel.
------------------------------
Weight over time:

Base Weight = 2.1 ounces + 1.0 ounces for 20 ounce soda bottle fuel
container.

9 days hiking between re-supply with a planned meal schedule of two hot
meals a day. Daily fuel usage 72ml of alcohol which has a carried weight
of 1.8 ounces.

Start Weight: 19.3
End of Day 1: 17.5
End of Day 2: 15.7
End of Day 3: 13.9
End of Day 4: 12.1
End of Day 5: 10.3
End of Day 6: 8.5
End of Day 7: 6.7
End of Day 8: 4.9
End of Day 9: 3.1 ounces (empty)
Average weight carried: 11.2 ounces.

Does this answer all the questions or do I need more tests and info?
-----------------------------------------


In this case I was going for fuel efficiency. I could test again later going to 52ml - and I bet it would heat faster. I just wanted to start with the base known for me - 18ml can boil a pint, so could 36ml boil a quart. In my experience with open stove designs it does not work- so hats off again to the Brasslite.

There wasn't much ice BTW. I actually miscalculated when I put the water into the fridge the other day, so I ran out before I did the 30 ml test. In attempt to rush I put water in the freezer for 1.5 hours. When I pulled it out there were very small chunks of frozen ice. So what happened in the 24 and 36ml test was water temp of approximately 40*F, but in the 30ml test the water temp was about 32*F.

Only one small question - why not record the exact starting temp. of the water based on the T-couple reading? Or are you using a T-couple that does not go down that low?

One way to get a fairly good starting standard temp is to put the water to be used in a ice water bath - wait a few minutes for it to stabilize the temp and you have water at 32 deg F. Other option is to just have a bucket of ice water - strain off the ice as you pour it into your test container and you have 32 deg. water.

For the Thermocouple to work, it must have an ice bath for the coupling to be immersed into. My thermocouple works fine at any voltage my multimeter works at, temperatures below 32* F. come out as negative numbers. I have a conversion chart that came with the thermocouple for determining the temperature based on voltage.

The water for the 24ml and 36ml test was placed into a 100 ounce platapus container and stored for appoximatly 24 hours. Water temperature upon removal was determined to be approximately 38.5* F. When I did the 30ml test, the water temperature was determined to be 32*F. There is a margin of error of +/-2* F if I remember correctly.