Rocket Stove

Maybe you’ve stored 100 pounds of wheat kernels, 75 pounds of beans, barrels of water and you’re all set for whatever municipal unreliability may bring. One question:

How will you cook the food you’ve stored?

In Food Storage and Refried Beans, back in April 2009,  I determined that to cook 100 pounds of beans would take about 4.7 gallons of white gas (Coleman fuel). Assuming the rest of a typical day takes a little less gas, you would still have to store about 10 gallons of gas to service that much food. While this is doable, it is certainly a hazardous amount of fuel. My personal solution is my own take on the Rocket Stove.


We have a wood burning fireplace, but it is not set up for cooking. However, we keep wood around for the cold winter nights. Obviously, mankind has cooked on wood basically forever. So, obviously it is possible. I was introduced to the rocket stove developed by Aprovecho, which is intended to reduce fuel consumption by more efficiently transferring the wood’s energy into the cookpot, reduce air smoke and soot inhalation by combusting more efficiently, and improve burn safety. They have ten design principals, summarized:

  1. Insulate around the firebox. I do this by using insulating kiln brick (about 0.65 g/cm3).
  2. Place an insulating short chimney directly above the fire. Mine is about 9 inches high, made by two races of insulating brick.
  3. Heat and burn (just) the tips of the sticks. The shortness of the firebox accomplishes.
  4. Heat is regulated by the amount of fuel. I have no damper, which their research shows is not effective.
  5. Maintain a good fast draft through the fire. This is accomplished both with the chimney and a recovered steel grate that creates an air channel under the burning fuel.
  6. Too little draft makes excess smoke. See principle 7.
  7. The opening, size of spaces, and chimney should all be about the same size. Specifically they recommend a 12 cm square opening (4.75 in), which is about 3.5 inches, and is probably slightly too small.
  8. Use a grate under the fire. See principle 5.
  9. Insulate the heat flow path. My entire structure is made of insulating kiln brick.
  10. Maximize heat transfer to the pot with properly sized gaps. I have not yet begun this phase of development.

For an initial design I used twelve insulating kiln bricks. Four make a floor—insulated enough, I think, to be used on a wooden stand. Two were cut into plugs to make the sides complete, and the remainder were stacked to make a square chimney and burn chamber. The burn chamber as seen through the chimney is filled with embers.


I scrounged and bent a wire rack to make a grate that retains a channel for air flow under the combustibles.


On top I used four small stones to make a burner. This is very much the wrong design for quickly heating water, but it worked for a test run.


Actually, it worked for three test runs. Yesterday we boiled water for tea.


This morning the kids helped make oatmeal on it. This evening I caramelized onions on it while grilling burgers.

My impression is that this device is the bee’s knees! The smoke was minimal (not as minimal as I would have liked, but largely avoidable). The amount of wood burned was about 4 linear feet of thumb-diameter sticks. The stove is stable, even with my crummy burner. I think it would make a nice patio fireplace for autumn evenings. Small, but controlling the smoke makes it much more pleasant to be around.

I do look forward to improving it. It is a pain to move, since it is about 45 pounds and doesn’t hold itself together. It is too low to cook on comfortably. Its heat transfer are near the pot is not well sized. This all requires work. I would also like to measure the efficiency of the stove. Always more fun to have!

Hot-Smoked Salmon


This afternoon I made hot-smoked salmon in my father’s MasterBilt smoker. The Web did not provide me with good resources on smoking salmon in a MasterBilt. Or rather, it provided many different and vague recipes. My result was excellent but not perfect. This article details exactly what I did and exactly what I plan to do the next time.

My target dinner time was 6 pm. I started with a 1.5 lb skinless salmon fillet about 1 inch thick at the thick part, still partly frozen, and 4 ounces of dry hickory chips. At 11 am I made a brine with 1 quart (4 cups) of tap water, 3/8 cup of iodized table salt, and 3/8 cup of packed light brown sugar.

The fish soaked for 90 minutes in the brine. I removed the fish from the brine, and dried both sides with paper towels. I placed the fish on a cooking-spray coated double-layer of aluminum foil. The fillet air-dried for 30 minutes, until 1 pm. I put the fish in the smoker, and then turned the smoker on to 210 F. The smoker started producing smoke with residual chips after about 20 minutes. At 1:30 I put in about half the smoked chips, and at 2:00 I put in the remaining chips.

I left the fish completely alone until 4 pm—at which time the fish had been cooking for 3 hours. I brushed the fish with maple syrup, about two teaspoons. I tested the temperature and at the thickest part it was already up to about 170—well over the target temperature of 165 F. Dinner was still two hours away, so I dropped the temperature to 140 F and left the fish in the cooker until about 5:45 pm.


The result was beautiful to look at, and smelled enchanting. The thick parts of the fish, oddly, were overcooked and dry—a texture like papier mache. The thin parts were moist and rich. My guess is that the brine did not penetrate and protect the inner parts of the fish. Of course it was obviously overcooked too.

You can see the dark smoky coating on the fish. The inside stays the pale color of poached salmon. The flavor was excellent. My 5 year old and 8 year old both ate it and wanted more. They seemed to like the smoke and the slightly sweet coating. My wife and I enjoyed it too, so the meal was a success. Next time it will be better.


Next Time

  • Start with fully thawed salmon.
  • Make only two cups of brine, and use a freezer bag to brine the fish.
  • Start brining the night before.
  • Use 4 oz hickory chips (same as this time)
  • Give the fish about an hour to dry after pulling from the brine
  • Put fish in cold cooker at 2 pm (4 hours before serving)
  • Cook with MasterBilt set at 210 degrees F.
  • Put first load of wood chips in once temperature reaches near the set point, about 30 minutes.
  • Put the second load of wood chips in after the smoke stops, about 3:30 pm.
  • Go running (optional)
  • At 5 pm, glaze with about 1 tablespoon of maple syrup
  • At 6 pm, remove and serve immediately

Barbecued Ribs and Smoke Metrology

I made smoked ribs yesterday and tested out a new measurement system based on an Arduino, data logger shield, and a Sharp GP2Y1010 dust measurement unit. I didn’t have a long enough cable to get the smoke measurement unit inside the smoker, so I put it in my own contained pod above the smoker. By “pod” I mean the red re-purposed steel can on top.


The smoke extension pod has a port for sensor access and an adjustable damper. You can see the ribbon cable coming from inside the can to the data logger.


The sensor access port is on the bottom front of the picture below.


The smoke sensor, shown below, has some very small deposits around the sensor hole—the big hole in the center. In addition, I used the RIMU to log the temperature inside the smoker.


The temperature, oddly, was much too high. It is supposed to be 225F, but I appear to have set it at 265F. Perhaps there is a problem with the RIMU, but more likely there is a problem with the built-in sensor. The error in temperature control did not hurt the ribs, they were excellent. Further investigation is needed.temp_in_plot


The smoke voltage measured with the Sharp GP2Y1010 appears to distinguish between noise (between 0 and 1 volt) and total saturation. It is, nevertheless, a pretty good indicator of when smoke was applied. I loaded the smoker with chips at 10 am, and there was no smoke noticeable until 10:30 am. The full load of chips produced smoke for about 1/2 hour and then stopped almost completely. I put in a second, smaller, dose of chips at 11:30, which produced about 20 minutes of smoke from 11:45 until shortly after 12:00.

The period of time from about 14:30 until 15:30 is unexplained. I would guess that the ribs were slowly dripping, and each drip would burn and generate a short dose of smoke. It started at a time when I had moved the smoker onto the porch to avoid threatening rainstorms, and perhaps an angle changed. The smoker was shut off briefly, which explains the temperature drop around 14:20. The pattern of increased smoke does appear to match the temperature cycle.

The amount of smoke, or at least its duration, seems to relate well to the amount of chips. Smoke seems to start about 20 minutes after chips are applied, and lasts for up to 30 minutes.

Better Bread from the Bread Machine


We have a Cuisinart convection bread machine (CBK-200) that was a given to us for Christmas one year. It is an attractive and well-featured machine, but I haven’t had much loaf success with it. We have had bread machines in the past. The Cuisinart tries to produce a loaf which is horizontal, like a loaf from a conventional pan. Overall Cuisinart’s design was not a good compromise. The wide loaf pan accumulates clumps of unmixed flour on the edges. These remain in the baked loaf, and are inedible and unsightly.


The loaves also come out, usually, concave on the top. They rise high then collapse to produce an ugly loaf with a flavorless crust. I set out to improve on this using 3-factor, 2-state screening design to study the influence of sugar, water, and yeast. Foolishly, I started with the Cook’s Illustrated recipe from the mid 90’s, rather than starting with the Cuisinart book’s recipe.

I hypothesized that our high altitude (1 mile above sea level) was the cause. My results with the Cook’s Illustrated recipe were like the abhorrent results from the Cuisinart book (see results ). I used the results of the screening experiment for guidance by extrapolating the fitted equation along the curve of steepest ascent. Unfortunately, I had two different extrapolation curves–one for height of the loaf at its center and another for the height of the loaf at its edges. One curve said to add more yeast, the other said to add less yeast. So, I read more of the advice available online. Probably I should have spent more time reading at the start. For the sunken loaf problem all advice was consistent: add less sugar, less water, and less yeast, and maybe add more salt.

I used my curve which predicted less yeast to extrapolate the amount of water, and yeast, and sugar. The result of that extrapolation was almost the same amount of sugar and the same amount of yeast recommended in the Cuisinart book. The Duh! moment. However, my prediction wanted less water, much less. The Cuisinart recipe wanted 1.42 flour:water by weight, but my extrapolated recipe wanted only 1.96 flour:water by weight.

I am not done with the experiment yet. Because I autonomously concluded that I should use the same amount of salt and yeast as the Cuisinart recipe, I believe that those factors do not need further exploration. The flour:water ratio, on the other hand, is worth exploring.

Flour and Water

I am horrified by the lack of consistency in the specification of flour weight and volume. One thread indicates that the weight of a fixed amount of flour will change by about 2% due to humidity variation. I’m more than satisfied with 2% accuracy. Measurement by volume, on the other hand, may vary by over 10%.

Because of this variation you would think that the assumed weight of a cup of all purpose flour would be a standard. But no. Consider these sources:

4.25 oz King Arthur
4.375 oz USDA
4.625 oz Gold Medal
4.7 oz
4.75 oz
5 oz Artisan Bread in 5 Minutes
5.5 oz Cook’s Illustrated Rustic Italian Bread (bread flour ~5% denser)

When a recipe says “one cup of flour”, it is talking about somewhere between 4 and 5.5 oz, a variation of +/-15%. The people worrying about the 2% variation due to humidity are clearly insane, since unless the recipe specifies, no recipe reader has much better than 15% accuracy just understanding the author. I use 4.25 oz/C in my recipes.

Since nobody seems to publish actual data on the net, I will. I used dip-level-pour measurement for 17 trials. The sample mean is 4.76 oz, and the sample standard deviation is 0.17 oz.

Num oz Num oz
1 5.00 9 4.50
2 4.80 10 4.65
3 4.70 11 4.65
4 4.80 12 4.75
5 4.90 13 5.05
6 4.65 14 4.80
7 4.50 15 4.80
8 4.65 16 5.05
17 4.75

2 The Recipes

Add the ingredients in the order listed. Set the machine for 1.5 lb loaf, with a light crust. Note: my recipe developed at 1 mile above sea level.

Mine CI Cuisinart
6.5 fl oz 10 fl oz 9 fl oz water, fl. oz.
1 Tbsp 1 Tbsp 2.33 Tbsp oil, tablespoons
1 tsp 1 tsp lemon juice
2 tsp 3 tsp 2.25 tsp granulated sugar, teaspoons
1 tsp 1 tsp 1.125 tsp salt, teaspoons
12.75 oz (3 C) 12.75 (3 C) 3 C flour, C
1/4 C 1/4 C 3 Tbsp nonfat dry milk
1.25 tsp 2 1/4 tsp 1.5 tsp instant yeast

I have tried recipes with 6.5, 7, and 7.5 fl oz of water. Both 6.5 and 7 produced nice loaves, but 7.5 again produced a saggy loaf. This provides a fairly contained range to consider for a response surface method. Though perhaps the next experiment will include some whole wheat.

Smoking Good Ribs—Temperature Management

I have smoked ribs about four times in the last couple years. I’ve been fortunate enough to have my father’s MasterBuilt smoker. It is a box about the size of a dorm fridge, and can readily smoke two racks, each cut in half. It has electronic temperature management and is well insulated. I’m fond of the smoker, and totally committed to the ribs from it.

At the bottom of the smoker is a pan, like a porkpie hat upside down. MasterBuilt’s instructions say to put water in the pan. Other sources suggest putting something solid in the pan instead. The argument is that boiling water is around 212°F (205°F  at my altitude), and that the smoker’s temperature will not rise above that. A very reasonable argument. Consequently, I have used bricks wrapped in aluminum foil in the bottom of the pan.

What does the thermal mass do, whether bricks or water? One idea is that it causes the temperature to be more even, especially for it to bounce back faster after the door has been opened. For this to work well the thermal mass should a) be quite conductive, and b) have a high specific heat. Conductivity makes the heat inside the mass available to the air in the smoker. High specific heat means that a modest amount of mass can hold lots of heat.


Conductivity (W/m K)

Specific Heat (kJ/kg K)

Brick 0.8 0.9
Water 0.58 (but can convect) 4.2

It is hard to beat water for specific heat, and convection within the water pool should make the effective conductivity much higher than the intrinsic conductivity.

Which thermal mass provides faster temperature bounce back? It turns out they are functionally identical. The following figure shows four trials, two with water and two with brick moderator. During each trial I held the door open for one minute. There were differences in breeze and in ambient temperature (not analyzed). However, the chart shows that in all cases the temperature returned to its pre-opened temperature in the same amount of time, around 8 minutes. Rebound time is  not much affected by the choice of moderator.


Which thermal mass provides more even temperature? This is harder to measure because I opened the door for the rebound test. Nevertheless, by removing the open-door data I can see the thermostat’s temperature management wave, and estimate the variability.

The next figure overlaps these two sets. In both sets the temperature swing is quite high, either 13°F or 17.6°F as measured by twice the sample standard deviation. Water provides more even temperature, but one that still varies quite widely.


The chart reveals the tie-breaker fact: the average water temperature is almost 10°F closer to the thermostat’s set point. The set point is 225°F, as recommended for ribs. The thermocouple was positioned at the top 1/4 of the cooking chamber, in the center front/back and left/right. This is where the top slab of ribs goes.

Presumably the bottom half of the cooking chamber is cooler; rotating the food top-to-bottom may help. The full data record is plotted in the final chart. The warm-up and cool-down bracket the data set. The record started with brick moderator, until about 19:05, when I replaced the brick with boiling water. The deep dips in temperature come from opening the door.


The next experiment is to instrument the box at several heights in the column, but that is low priority. Spring is coming and pruning must be attended.

Freezer Swan Song

The last few weeks our ice cream sandwiches have been awful. I’m throwing away ice cream sandwiches, and I’m only marginally pickier than my dog. They’ve been gummy, almost squishy.

We have been suspicious of the refrigerator since we smelled a hot electric odor—like burning dust. I monitored the fridge with the RIMU, and it looked more or less reasonable. I monitored the freezer too, for one night. It is looking less good.


In addition to the short-lived odor, there is another reasons to be suspicious. The fridge gets warm on the outside, notably warmer than it used to.

Normal fluctuation in freezer temperature should be about 5 degrees, measured with a mass-damping technique…that I did not employ. I estimated the air temperature variation, not the food temperature variation. The air temperature variation is larger. The air temperature swings through a larger range than I would expect.

The average temperature, though, looks altogether too high: 16°F. The normal target temperature is 0°F. I’m assuming some of the spikes are because we were accessing the freezer, but they may all be defrost cycles.


We’ll continue fiddling with the temperature control, but likely the freezer is singing its swan song. That’s one note, about 0.25 millihertz.

Sourdough Chowder Bowls

In San Francisco, on the wharf, you can buy a hollowed-out loaf of sourdough bread, filled with New England style clam chowder. You can buy it from street vendors, or sit down in a restaurant and eat it. Sourdough and San Francisco go together. I love the bowl, chewy, flavorful, damp from the chowder. I even like the chowder, which is pretty much like what you get from a can. Unfortunately, Albuquerque does not have a wharf on the Rio. What to do when I want a chowder bread bowl on a cold late autumn day.


I made them today, to moderate success. Two weeks ago I made them to miserable failure, having forgotten the salt. The recipe is simple, the challenge is in the baking. I bake in a convection oven, but the process is almost the same without convection. I put the boules in a pre-heated covered Pyrex baking dish at 425 for 15 minutes then uncover the dish and lower the temperature to 325 for between 20 and 35 more minutes. I spray the boules with water before putting them in the dish, to help the crust develop.


The crumb is dense, the crust is chewy, the taste sour but not overpowering. The chowder, from a can, hearty.

  • 3 cups of flour (4.4 oz each, by weight)
  • 1 tablespoon salt
  • 1 cup sourdough starter
  • 1/2 cup water, around 100 degrees F

Sushi Dinner

Tonight we made a sushi dinner, and for the first time included kid-specific items. I used sweetened egg, tamago, sheet cut and pressed with rice in cookie cutters. We made stars and sushi boys. They love the California rolls, shrimp pieces, and vegetable-tamago rolls as much, but it was still fun to make.

My rice recipe and the idea for the shaped pieces came to me from Barber and Takemura’s Sushi, Taste and Technique. It turned out a double recipe of rice was not enough, so you can see at the back pinwheels. Probably we should call them tortilla maki.


Food Storage and Refried Beans

A solid storage plan includes food, and also anything else you need to make the stored food into something edible. The difficulty is that none of the food storage guides or recipe books I have list how much fuel to store in order to make their recipes.

To rectify this, I have begun measuring fuel consumption when preparing recipes roughly as I would in no-gas and no-power emergency. Perhaps others will find this useful, and quantify their recipes.

The first step is to reduce the total fuel as much as possible, and the first target food is beans. Beans are a nice target because I have a lot of them—something like 15 lbs of kidney beans that are probably older than me. Furthermore, they offer excellent nutrition, and finally they cook for ages. The two techniques I know to reduce fuel use are pressure cooking and pre-soaking.

The test recipe today was for refried beans. The result was more delicious than any canned refries I have ever eaten, though the texture was much lumpier. The recipe is below. I cooked on my Coleman dual-fuel stove, using white gas. Cooking is performed in two stages, the first cooks the beans under pressure, and the second cooks the onions and “fries” the beans.

Total Fuel Use: 122 g (about 6 fl oz) of white gas

Pressure Cooking the Beans: 70 g (about 3.4 fl oz) of white gas

Refrying: 52 g (about 2.5 fl oz) of white gas

Pressure cooker seated on the stove in the back yard.

Pressure cooker loaded with the beans and other ingredients, prior to pressure cooking.

Pressure cooker at pressure; the heat is too high as shown by the copious steam jet.


Adapted from Vickie Smith’s recipe for Refried Beans

Step 1

1 lb dried pinto beans, soaked at least 4 hours

4 cups pork or beef broth, stock, or water

Add beans and broth to pressure cooker, plus enough water to cover beans by about 2 inches. Stir to mix, lock the lid, and bring to pressure (15 psi) on high heat. Reduce heat to lowest setting that will maintain pressure, and cook for 12 minutes—I cook for 13 at 1 mile altitude. Remove from heat and let pressure drop naturally. Drain beans and mash them with a potato masher until they are to your taste in lumpiness.

Step 2

1/4 cup bacon fat

2 onions, finely chopped

1 mild poblano, pasilla, or Anaheim chile, seeded and chopped

2 cloves garlic, crushed

1 1/2 teaspoon ground cumin

Heat the fat in the pressure cook, add the onions, chiles, and garlic, cumin and cook, stirring, until they are very soft. Add the mashed beans in two or three batches, stirring to mix.