Rocket stove efficiency/certification

Has the efficiency of the rocket stove been tested and if so what is its delivered heat efficiency (heat delivered to the home/potential heat in the wood x 100)?  Has a standardised design been submitted for certification anywhere (Canada, USA, Europe)?

The only "tests" have been amateur comparisons.

To transform this into heat-delivery numbers is kinda like the story problem from hell.  (Imagine your teacher going senile while trying to describe the speed of trains about to collide on a freeway to Hawaii....) 

But I'll give it a shot. 

(Erica taking over, because Ernie is tired of people wanting expensive answers for free.)

If you are an industry professional and can tell us how this testing data is obtained, and whether it can be duplicated outside of a lab, we would be very interested in getting preliminary data on a site-built stove.

The main measured data so far is temperature, and quantity of wood burned.

The original mass heaters have measured clean burn (1000F-1500F) in the combustion chamber, with over-fire getting hot enough to melt or warp thin steel tabs (1800 F).  Exhaust temperatures on the original systems typically measure in the 90's to low 100's  F.  (Shorter systems, or a hardwood / oil-enriched burn, more typically exhaust around 200-250 F).  The rest of the heat generated by the small amounts of wood in each load is stored in the bench / thermal mass, and thus delivered to the home over time (the next 8-48 hours). 

One load of wood is the amount that can be tightly fit into a 30 square inch, 1-foot-tall feed tube for a 6" system, or a 50 square inch, 15" feed tube for an 8" diameter system.  Assume about 25% air space, or for a final load, a 5" diameter log stuffed into a 5.5x5.5 square.  A typical burn session lasts 2-4 hours and consumes about 1-2 loads per hour.  And yes, the log does continue to burn cleanly with bright flame on its own coals; adequate temperature is maintained by the insulated firebox.)

57 lbs of wood produced 7 oz. of ashes when we tested our week's burn quantities.  (Ordinary moisture content for wood stored out of the rain in the maritime Pacific NW; very little carbon or unburned fuel, no visible smoke except during the first 3 minutes. The heat lost is in the exhaust; there is not a lot of chemical energy wasted in smoke, but I'm not sure how to factor in steam and evaporation.  Maybe temperature readings are enough.

As an experiment, we had Ernie's dad burn the same weight of wood in his woodstove in eastern Washington; he was left with 12 oz. of ash.  His fuel may also have been drier given the climate differences.  (He's certainly a competent fire man.)  Took us about a week to burn that much wood, and he was able to burn it overnight without overheating his house.  Similar-sized, equally idiosyncratic houses that are basically remodeled sheds; his area is colder in winter.

Other anecdotal comparisons: Ianto and Linda switched from a small woodstove to a rocket mass heater while living in the same small house in southern Oregon.  They had used 4-5 cords of wood per winter with the woodstove; they used about 1/2 cord with a rocket mass heater. 

Certification: By whom?  EPA?  ASTM?  Local building code officials?
Mass heaters are exempt from EPA, and local building officials don't like to permit wood-burning devices without the EPA (UL Labs) sticker.

We have submitted a proposal to the City of Portland's Alternative Technology Advisory Committee, to determine a permit process for rocket mass heaters.  Standardized designs are not really applicable since every stove is a site-built custom design, but we gave them 2-3 sample designs in late 2009.  We're still waiting for the next round of questions, which will probably include a request for testing data. 

At that point, we will be putting out a request for funding.  The main beneficiaries of such testing would be owners who want to build a licensed / permitted stove, and neighbors of those owners who might benefit from air quality & surplus fuel.  (Long-term benefits for sustainability are not dependent on the testing or certification, but might be improved due to increased use.) The designers and researchers are not likely to directly benefit, for reasons explained below.

Testing data will likely cost between $8,000- $30,000.
(Build a working unit at the OMNI testing labs, about 1 week's work for a qualified mason so say about $3000 if the builder does it at cost.  (Ideally you'd test both new, standard masonry, and the earthen and recycled materials favored in the original design.  Our hybrid stove took about 4 weekends to a rough finish, and cost about $1200 in parts (no paid labor).)

The lab's basic woodstove testing for EPA specs) would cost $5000 if you could do it on a mass heater, which you can't.
(Wrong burn rates, wrong heat delivery timeline, and it's excluded by weight from the EPA definition of a woodstove and exempt from their testing requirements). 

A custom test would need to be devised... which is why we would like the authorities to get specific about what tests are necessary, as we are really not interested in going through this twice.  We will need to fund-raise for the first set of tests as it is.

One reason that masonry heaters have been exempt from DEQ is that they are site-built, single-design units, too heavy to move, and typically cleaner and more efficient than the best woodstoves on the market. 

A woodstove manufacturer tests one in 5,000 to one in 100,000 stoves of a given model depending on performance; a masonry heater builder would need to test every single heater with a significant difference in dimensions, i.e. all of them. 
Masonry heater cores (for a different design) have been submitted for standard woodstove tests.  Inspectors are much happier to permit something with a UL sticker, regardless of the un-stickered device being legal & exempt from EPA.  You pay $6,000 or so for the core with a sticker.  It's basically an insulated firebox, like a wood-burning oven with exhaust chimney.

In Europe they are more comfortable with masonry heaters in the first place, so rocket heaters have been easier to introduce.  One of the master masons who works with rocket heaters is Fleming http://www.fornyetenergi.dk .

If the original designer wanted to patent and sell his work, or run a business building his proprietary design, he might be able to get financing to do the tests.  Several years of start-up costs could be parceled out with the 'approved' design to a limited number of owners who could afford such costs. 
  The design was instead produced by a collaborative process, has been published instead of patented, and the senior designer (Ianto Evans) is ideologically committed to off-the-grid sustainable living.  He educates people to provide for their own needs, by converting waste and natural materials into minimal-footprint lifestyle resources.  He is personally distrusts government agencies and corporate funding models (too much time on the ground in the third world, and seeing the results when traditional building methods are forbidden in favor of expensive industrial products).

The group of researchers helping him, and taking the idea more mainstream in spite of its ideological roots, are private individuals.  Most have day jobs or moonlight in freelance work that still allows us to run several workshops each year.

We would like to see these heaters become accessible to mainstream people, but it's a far-from-mainstream financial model that would be needed to take us there.

* * *

If you are in the industry and know how to obtain the data you want, we have a heater you are welcome to examine.  We can give you most measurements if we know what to measure.  Maybe you could look up the testing procedure, or heat values for wood, and let us know what the missing information is?

We would be thrilled to have some grad students or off-duty environmental lab techs bring their toys over for dinner, and get some decent-looking preliminary test data.

Yours,
-Erica Wisner (and Ernie)

www.ErnieAndErica.info

see also www.rocketstoves.com

Thanks for the lengthy reply, the teacher reference was rather appropriate as I teach high school science and math though my trains would have collided in Saskatoon in -30 C weather.  Delivered efficiency is a simple calculation  (1 - (mass exhaust x temperature)/(kg wood x energy content/kg)) x100 to give a percent.  Mass of exhaust gasses can be estimated by exhaust velocity x area of the exhaust opening to get the volume and using temperature based density characteristics of ideal gases.  Alternately put a big plastic bag over the exhaust, collect and cool gases to ambient over a measured time and weigh then multiply by the ratio of the burn time to your collection time.  Basically you are measuring how much heat escapes vs how much stays in the house.  This becomes a "delivered heat" efficiency as opposed to a combustion efficiency which is simply how much of the fuel becomes heat.  The latter isn't very useful especially when a LOT of heat goes up the chimbley.  I'll continue reading and see if there is anything else to address in your response.   I was interested because of the difficulty in getting insurance if you build your own stove.

So you still want me to do the math, eh?
I will if you will. 

(Using a thermometer, a postal scales, and a stopwatch, observing a local woodstove. Measure the temperature at the exhaust (or creosote or lack thereof), and the amount of wood burned in an hour, and eyeball the exhaust rate in feet/second.  Then do the same math.  No looking at the mfr specs first; the point is to see whether armchair estimates have anything to do with lab ones.)

My answer: 75%, +/- 85%.  (That's a range of about 10%-100%).  Work is explained below.

My steps in solving this problem:
1) Recognize that what you want is a theoretical, high-school-science answer. 

A much more relevant answer is the one we already gave. 
   In a particular building, a rocket stove may use about 1/8 the wood, produce almost no smoke, for the same level of indoor comfort in the same building, as a small woodstove circa 1975. 
   You could take the wood you use in a year, divide by 5 for a safety margin (your woodstove stove might be more efficient than the test example), and calculate the rocket stove size needed.  Or build a one-room heater and see if you need another one later.  This would be the practical answer to "how big does my stove need to be."
   For insurance purposes, your problem is not the heat efficiency, but the building regulations.  Basically, if you can get a permit from the local code officials, the insurers feel safe. 
  We're in conversation with the City of Portland to develop a code-friendly writeup comparable to masonry heater or woodstove specs, so an inspector could permit the stove.  Will cost $$.  Masonry heaters are exempt from DEQ, and can be built/designed by a licensed mason or architect who can certify its safety. It might be worth a conversation with your insurer about whether they're willing to add a rider given that it's a non-statistically measurable risk (few existing examples, and I don't know of any reported fatalities outside of maybe Hansel and Gretel). 

Back to the recreational math:

We are comparing data with very low reliability (the weight of a bag of gas for Pete's sake).  There's a 10% range of error in the fuel value for wood alone; air moisture content adds a weight variability of 60%.  What error percentage would make the results meaningful? 

I'd add another 50-100% error range for actual usage factors, like the operator's standards for 'dry' wood, the frequency of operation and cleaning, heat loss through walls, the tendency to forget the fire or damp it down, etc.

So I don't trust the industry data either, and I prefer practical answers. 

But I do enjoy math.
So I'm willing to chew this cud through, as long as you recognize that the results are ... (what normally results from recreational mental digestion of unreliable data).

Solving the Story Problem:

1) Guess the weight of exhaust gas based on observations, using memory of observing two 6" systems. 
(My assumptions are likely to be just as reliable as any bag-weighing experiment I can do with equipment on hand.)

Air weight vs gas coming out:
Our foggy flue gas at 95-100F (310K) is about equally dense with air at 70F(295K).  (Very hard to run the stove then, though we can do it if the stove was pre-warmed in cool air the night before). 
Flue gas at 110F(315K) is less dense than air at 55F(286K), even with condensed water; rolls upward at about walking speeds.
Weight of air at 70F(20C)(293K) = 1.2 oz / cu. ft; (saturated with water vapor e.g. in vacuum above a pool is 1.9g/cu ft; Wikipedia says "Humidity ranges from 0 gram per cubic metre in dry air to 30 grams per cubic metre (0.03 ounce per cubic foot) when the vapor is saturated at 30 °C;" nobody will tell me what fog weighs, but dry ice vapor CO2 is 1.5 times denser than air. 
One patent does claim that water droplets in a cloud are a small weight compared to the (saturated) vapor... Water compared to air: 780 times denser.  ... I glipmse on a Google search that a gram per cubic meter is about normal for "wild" fogs ... that would be another 0.001 oz/cu ft, but what is the volume of vapor which squeezes down into that 0.001 oz?   How much steam does an ounce of water make? ...
http://www.elmhurst.edu/~chm/vchembook/123Adensitygas.html

1a) I ran only the first comparison; got 1.4 oz of flue gas per cubic foot at 95 F.  This is almost twice the weight I found online, which makes sense (maybe) due to fog dew and CO2?  The weight compared to ambient air would be a negative number at high temps, but heavier at STP. 

1b) Rate of flow observed in a typical 6" system is about 1 foot/second for moderate burn, up to 3 feet/second for "rocketing" (fast, rushing burn).  Cross section is 30 in^2; cubic inches per second = 360 in^3; 0.2 cubic feet/second = 750 cu ft/hour. 

1c) Weight of gas coming out: 1.4 oz/cf * 750cf /16 oz/lb:
I'm getting about 66 lbs of exhaust per hour. I guess a few lbs of that is combustion products, and the rest is air in, air out.   

2) Weigh some wood.
Our stove burns about 4.5 lbs of wood/hour in relaxed mode (one load of typical mix: cherry, fir, and kindling), this would go faster in rocketing mode so we might use 7-8 lbs per hour.  Weight in kg is: 2 to 3.5 kg of wood/hour, error range 2.5 kg +/- 40%
(For practical consideration:
A 4 hour burn (20-25 lbs of wood) heats our home for 24 hours, to 65-73 degrees in the various rooms and corners, with outside temperatures in the 40-55 range at this time of year.  A 1-2 hour burn keeps the bench alcove comfortable for warm seating, and leaves the back bedroom at 55-60 degrees.)

3)  Look up fuel values for cherry, doug fir, and alder.  (There are no fuel values for punky sticks with the bark on, so we'll assume the oily paper from our fish and chips brings those up to average.  We burn what we got, as dry as we can get it, and adjust.) 
http://www.consumerenergycenter.org/home/heating_cooling/firewood.html
Cherry and alder both turn out to be 8,200 BTU per lb, despite their different densities.  Doug fir is 9,100 BTU/lb.  Average fuel value for our 1-hour, moderate load: about 40,000 BTU/hour. 

What was the question again?

mekennedy1313 wrote:
...  Delivered efficiency is a simple calculation 
(1 - (mass exhaust x temperature)/(kg wood x energy content/kg)) x100 to give a percent. 
... Basically you are measuring how much heat escapes vs how much stays in the house.  ....
I was interested because of the difficulty in getting insurance if you build your own stove.

I gotta say that my mass exhaust number feels unreliable - I got nothing to compare it to.  If you don't care about the integrity of my exhaust number, skip these italics:

4) Question the integrity of my work in #1), and re-do it a different way:

Going with the chemistry, the wood is say
20% water (unbound & resin-bound),
1% mineral ash, and
80% carbohydrates and hydrocarbons (resin, cellulose, etc). 

By dry weight, paper pulp woods are about 45% cellulose by weight, 27% lignin, 8-9% pentosan, and 3% other (plus that 1% ash).  (Sorry, Internet source uncited)
Cellulose like sugar is basically CH2O;
lignin varies, but it's generally richer than cellulose as a fuel; (http://www.jbc.org/content/114/2/557.full.pdf) reports it as C40, H42-48, O 15-16. (C2,H2-2.4,O)
and some resin or oil would boost this further.

By a rough combination of these molecular weights, I make wood by weight:
10-20% water (unbound)
33% carbon (bound)
34% water (bound)
2% hydrogen (bound)
1% ash
10-20% other (oils? resins? variability in other components?)

Using the 33% carbon, and 2% available hydrogen, I make out that wood requires just over its own weight in oxygen to burn completely. (104% of the wood's weight in oxygen to burn 100% of the wood.) 

Air is 20-21% oxygen, 78-79% nitrogen; 1% CO2; and less than 1% of other stuff.
 
The resulting flue gasses (of a pure wood reaction with no excess air) would be:
75-90 parts water by weight, plus atmospheric water, plus "other" - this might be either fog or vapor, denser or lighter than air, and releasing additional heat as it condenses...
130 parts CO2 by weight (includes about 10 parts from air)
520 parts nitrogen (inert) by weight
10-20 parts other (if inert).
(If the "other" is rich fuel, it could combine with another 80-150 parts oxygen and bring along another 300 - 700 parts nitrogen and a pinch of CO2.)

Total predicted exhaust weight/hour of a perfect wood/air reactor (perfect mixing, no excess of any reagent, just keep it on the heat until it's all incinerated) would be 7.5-15 kg exhaust per kg of wood.

Total predicted exhaust for a real-world fuel burner, where excess air must be provided to ensure that fuel is completely burned: say 2x to 4x that, with reasonably good mixing at high temperatures; so 20 to 60 kg exhaust/kilo of wood.

So any clean stove must pull at least 6x its fuel weight in air to burn clean, and 20-60x is more reasonable to get complete combustion.  For 2 kilos/hr of wood, 40-120 kilos/hr of exhaust.

My weight guess in #1) (66 lbs exhaust/hr for burning 4-5 lbs of wood) represents about a 15x factor, on the low side but possible.

5) So what the heck, let's run the formula. 
66 lbs is 30 kilos. (convenient.)

(1 - (mass exhaust x temperature)/(kg wood x energy content/kg)) x100 to give a percent. 

(1- (30 kilos x 310K)/(40,000 - 46,000 BTU)) x 100

Ok, can we convert kilo.K's to BTU?
9300 kilos*K;

A BTU is 1055 Joules...
a joule is 1 kg*m2/s2 ... 
Is a m2/s2 really equivalent to a degree Kelvin?
That would open up a whole new line of scientific inquiry about the universe...

Specific heat capacity is measured in joules/kg*K; that's the closest related unit that I can find online. 

So please check the units on your equation. 
It may just be "engineer math," in which you leave out anything inconvenient for estimating purposes.  (As a scientist, when the units don't agree, it makes me nervous.)

But putting on my marketing / engineer hat, and running the equation as given:
(1- (30 kilos x 310K)/(40,000 - 46,000 BTU)) x 100
(1- (9300 kilo*K/43000BTU) x 100
(1-(0.216) x 100
(0.78) x 100
78 %

Efficiency of 78%, plus or minus....

(1-(100% variability in exhaust weight * 3% variability in temp)/(20% variability between fuel weight and fuel value/weight)) 

Aggregate error of +/- 85% ? 
(I no savvy statistics, is this a meaningful way to combine error estimates?)

So the real value could be anywhere between
10% efficiency (or -7% ... )
and
130% efficiency (.... 82%? 91%? Or something....).

(Taking 100 lbs for exhaust weight gives efficiency of 70%.  Taking a theoretical value of the exhaust mass - 2x extra air for complete burn - is pretty close to the first value listed above: 79%.) 
  Trained smoke observers (firefighters, DEQ) confirm that there is no detectable smoke in the exhaust, and the ash weights pretty closely correspond to the theoretical ash weights for the wood, about 1%.  So clean combustion is definitely occurring.)

For comparison, a good auto engine has about 38% efficiency, a fuel cell about 78% efficiency, using a somewhat different definition of efficiency (conversion of fuel energy into useful motion).

Pellet stoves claim 75-90% efficiency, discounting the work (and expense) of the fuel manufacture.

But as we read here, http://www.consumerenergycenter.org/home/heating_cooling/fireplaces.html

Robert McCrillis of the EPA says, "...In the field, it's the installation and how the stove is operated that has the largest effect on how it performs..."

Examples of how operators abuse a lab-designed stove are given - the commonly practiced errors add up to an efficiency profile kinda like using your refrigerator to heat your house.

Rocket stoves built in the earth-friendly original materials are designed specifically to account for these human factors, and environmental ones rarely considered:
- very little transportation of materials (local earth and rubble),
- very low energy cost in manufacture of materials & components (re-used scrap metal and masonry components)
- no need to overheat and "damp down" to maintain comfortable temperatures; this practice alone is responsible for most woodstove pollution and wastes enormous amounts of fuel compared to a clean burn.
- burn at maximal efficiency for 4 hours, yet experience the resulting warmth gradually released over 24-36 hours.
- heating the people directly, so that the entire house does not need to be kept as hot, reducing heat loss through walls & roof.
- can be fueled on 100% local biofuels; we use mostly yard debris from a less than 2 acre property, with a little shop scrap and free local utility trimmings.
- low-cost building and fuel leaves money in the budget for insulation, etc.

You may quote me on "75% ... plus or minus 85%"; but please don't let people mistake that for accuracy.

Sooner or later, I'm sure someone will cough up to get one of these things lab tested. 
If you got $6000 to burn, we'll build you one here at OMNI-labs and tell you what they say.

I'm interested to see their math, and how they measure actual flow rates without impeding draft.  These things are touchy on draft (part of their efficiency comes from letting the exhaust wander through until it's almost sluggish), so putting a bag over its chimney might just stop it drafting altogether.

And now I realize that I should have been running my stove while I type.

Good night and good luck.

-Erica Wisner
http://www.ErnieAndErica.info

Impressive, the only thing I can see that was missed, perhaps because I didn't state it clearly originally, is the specific heat of the gasses involved, air would be a reasonable approximation at .72 kJ/kg K in a constant volume throughput (ie a steady combustion state) which would resolve the units issue but change the energy loss to .216 x .72 = 0.15552 thus improving delivered efficiency to about 84.5%.  This too is quite impressive especially considering that the efficiencies reported for other stoves including pellet stoves, are combustion efficiencies, ie energy in leftovers (smoke, creosote, charcoal)/potential energy in wood and thus not a good measure of whether the stove is a good heater.  Considering that condensing natural gas heaters, currently the most efficient common technology using a consistent fuel thereby promoting best efficiency, are 90-95% efficient in heat delivered (80-90% if one includes the energy for controls and fans) this puts the relatively low tech rocket stove on par with the best the petroleum industry has to offer.  As you so elegantly point out though this doesn't account for energy used in fuel processing and transportation which is much higher for natural gas, oil etc so the rocket stove comes out ahead.  Also, the way you combined statistical error isn't the way to do it, you've vastly overestimated it, I'd put your actual error about 15% worst case scenario and probably less than 10% as an average.  This is amazing.  I didn't have fuel numbers for a rocket stove so couldn't do a calculation myself but the rate of fuel use you report is incredibly low.  You also speak of a high and low burn of the stove, may I ask how you do that??  The few designs I've seen, and the video's, don't seem to have a control for the burn other than perhaps putting in fewer sticks of fuel, but would they simply not burn faster then?  A cover for the air intake perhaps.  Just goes to show seemingly old Idea's aren't necessarily worse than new fangled ones.  Interesting thought exercise.

I have heard that round pipe is not the most efficient thing for transferring heat to mass. Something skinny and flat with the same cross area being better. Also easier to crush. After reading some of the stuff in the masonry heaters stuff, I was wondering if a sideways ladder-shaped network with bottom feed and exit would help. hot air is able to rise and sit against the top... cold air falls and is pushed out exhaust... kinda like a "bell" heater. Gases from stove enter T with middle pointed up to an L next two Ts make an H on its side... add more Hs for length of bench then use L on T to end. Even a dead end upside down T in the line will get more heat where you want it.

So far as I can tell, the rocket heater is the "cheap" version of the masonry heater. The burning chamber is good... the heat to mass part is a work in progress. efficient is measured as heat transfered to room over energy available in the fuel. 70% sounds high.... but maybe if the bench is long enough. Still need 200C or so to make the chimney draw. Just an idea.

mekennedy1313 wrote:
Thanks for the lengthy reply, the teacher reference was rather appropriate as I teach high school science and math though my trains would have collided in Saskatoon in -30 C weather. 

My God!  What is it with you people and colliding trains!?  Are your railroad tracking people so whacked out on Molson (or would it be Moosehead?) that they don't realize putting two trains on the same track moving in opposite directions is a bad idea?  Are your train engineers so enamored of the beauty of the Canadian wilderness that they don't notice all the warning bells and whistles screaming, "Hey stupid!  Pay attention!"?  Why don't you also have your students calculate the estimated number of dead and crippled from this accident, eh?  Maybe try to figure out the estimated amount each surviving family member will get from the legal settlement, after the lawyers take their cut.  Sheesh...

(Here's to planning on flying next time I travel from Winnipeg to Calgary... Or maybe I should just stay home...)

Ernie and Erica, thanks for the all the math!  My concern has to be with the amount of air the typical system draws in order to maintain combustion.  Your figure of 750 cu ft/hour (12.5 cu ft/minute) caused me to raise my eyebrows.  If all that extra air is being provided naturally, that is definitely a leaky house!

According to research done at Lawrence Berkeley Laboratory (LBL) in the mid-80's, each occupant of a building needs 15 cubic feet per minute of natural air exchange (15CFMn) or 0.35 ACHn (Air Changes per Hour natural) to meet the Minimum Ventilation Requirements.  The mean all the air in the conditioned air space of a structure needs to be changed at a rate of 15cfm for each person living in the structure.  The air change can be accomplished either through natural ventilation ("leaky" house, open windows, etc.) or through a whole-house ventilation system.

Natural Air Changes per Hour (ACHn) are estimated by using a blower door to either pressurize or depressurize a house by 50 Pascals (ACH50) and measuring the resulting airflow.  This airflow is noted as CFM50, or Cubic Feet per Minute at 50 Pascals.  It can then be converted from ACH50 to ACHn using a factor "n".  This factor depends on the number of stories in your house, what climate zone (different from the USDA zones) you live in, if your house is shielded from wind, etc.  If you live in most of the Pacific Northwest, you're in zone 3 (the southwestern quarter of Oregon is in zone 4 -- n will be higher.)  Assuming that your house is in zone 3, one-story normal (neither well-shielded nor exposed) and about 8000 cubic feet (1000 square feet of conditioned air space with 8ft tall ceilings), n=21.5.

So, assuming that there are 4 occupants in your house (two parents and a pair of rugrats), the Minimum Ventilation Requirements at CFM50 = (15CFMn)(4)(21.5) = 1,240 CFM50.  Converting that number to the natural rate of air flow per minute means taking out the "n", or (15CFMn)(4) = 60 Cubic Feet per Minute.

So for your house to have the MINIMUM amount of breathable air, you have to have a system there that is capable of bringing in 60 CFM of fresh, outside air AND REMOVING THAT SAME AMOUNT.  That's without running the RMH.  If you're going to be running that, your system has to be bringing in an ADDITIONAL 12.5 CFM fresh air.  Otherwise, it's going to be sucking a combination of fresh air (reducing the amount available for the humans) and carbon-dioxide-rich "used air" (which will reduce the efficiency of the rocket stove.)

I hope when people are building an RMH, they're taking the ventilation requirements into account and providing some outside source of air for the combustion.  It might not be a concern with some of the older houses, but if they've built their houses with a lot of insulation, it rapidly becomes an issue.

While this is all nice, its not tested for certification. I have a report by OMNI Environmental Services, Inc.
10074 SW Arctic Drive
Beaverton, Oregon 97005

from 1992 for a masonry heater. This document gets them past council approvals for installation in homes. Similar would need to be done for an RMH

The report is very detailed, but the exec sum lists:

Particulate (PM) and carbon monoxide (CO) emissions using scientific sampling equipment.

Description of the unit in this case a combination overfire and underfire combustion air unit.

PM emissions averages 2.96 g/kg, 1.26 average daily g/hr and 2.96 normalized average daily g/hr. These PM values are in between those obtained from already certified pellet stoves and EPA certified Phase II woodstoves in the field.

CO emissions averages 82.7 g/kg, 25.2 average daily g/hr, and 72.7 normalized g/hr. The values are comparable to Phase II EPA certified non-catalytic woodstoves.

The average net delivered efficiency was 61.8%, which is in between EPA certified Phase II woodstoves and conventional stoves. Average heat output was 4915 BTU/hr and daily burn rate averages 0.43 dry kg/hr.

...

What follows are a lot of pages defining the test methodology and equipment (standard fare in scientific reports). Schematics of the unit design that was tested.

Emission Results
inc. comparisons to woodstoves, pellet stoves, and tested heater

Efficiency Results
inc. comparisons to catalytic, non-catalytic stoves, and tested heater.
References

List of Figures

List of Tables

and stack temperatures and oxygen levels in the stack at all levels.

This is what is required. Giving numbers on a forum won't get it past most councils I know of.

If you get some extra time, will you calculate;
How much wood would a wood chuck cut
if a wood chuck could cut wood?

Len wrote:
I have heard that round pipe is not the most efficient thing for transferring heat to mass. Something skinny and flat with the same cross area being better. Also easier to crush.

Crushability depends on how you build it. I think you'll want to build a metal table, a plancha (griddle), see wood-fired-cacao-dryer-in-nicaragua.pdf

He build the table top out of 1/4" steel plate but I don't think you need to go that thick if you add internal supports  which don't have to impact cross sectional area, but if they do they're negligible
6" diameter pipe >>> 1" x 28.274" , 2.356 feet wide
8" diameter pipe  >>> 1" x 50.265 ", 4.188 feet wide

Lay a few inches of mud on the plancha and you're cooking mud
The only problem with this approach is it involves steel bending/cutting/welding which could cost you a few hundred bucks

After returning from Nicaragua in summer of 2002, Dr. Winiarski had his students build the same dryer for $300 US from recycled components (aprovecho probably has a metal shop).

careful of the assumption that structure in flow has little result. the exhaust stream is a chaotic system one little deviation can cause a whole lot of results. Test test test every time make sure the system dont do something you dont expect.

OK, all this math  makes my head spin.  Couldn't one calculate the amount of energy that it takes to warm a given amount of mass a given amount, measure the amount of mass, take temps in enough different places to estimate an average, calculate the amount of energy in a given amount of fuel, and arrive at a reasonably accurate calculation?

Erica Wisner wrote:
careful of the assumption that structure in flow has little result. the exhaust stream is a chaotic system one little deviation can cause a whole lot of results. Test test test every time make sure the system dont do something you dont expect.

I agree with you in principle, and like all mass heaters tweaks will be necessary, esp with new adaptations, but i'm confident it won't make a difference because

1) all this happens after combustion has completed (heat storage)
2) there are NO TURNS!!! (round chimney pipe has at minimum four 90 degree turns)

these supports would be placed inline with the flow, streamlined not peripendicular, maybe 2mm thick 1inch wide (size of a  U.S. quarter) , the only possible effect is they might create 1inch x 2mm hot spots, maybe 5-15 celsius hotter than the rest of the steel bench

2mm deviation (which isn't even necessary, just make wider bench) in beyond negligible, 90 degree turns (round pipe design) affect the draft 10 times more
, and regular rocket mass heater has at minimum four of these, four !!
your proposed permit drawing has at least 4

what I forsee as a small  issue with this approach is hotter/faster exit temperatures  than with regular mass heater,
because you'd be achieving faster heat transfer to the bench
and you would be heating less cob , only the very top of the bench where the buts go

so say instead of a 100-200°F exit temp you might get with round pipe mass heater, i expect something around  300°F

so to recover more of this extra heat and not wast it you would
* burn less sticks at a time (could be a hassle)
* down size combustion chamber diameter to 3inch (not desireable)
* add more bench/battery length (the food dryer is ten feet long),
* add an expansion chamber to slow down the draft, like talked about in "Designing Improved Wood Burning Heating Stoves"
* extend the lip of the steel plate to go to the floor on front (heated by conduction) which might not help enough (i guess, safe bet, at least 50 farenheit )
* or here is a radical idea, extend this chimney to the floor or wall if you've got enough bench area


and oh yeah, all this talk of inline obstructions is ONLY if you go for thinner steel plate  ....... imagine if the steel bench wasn't flat but was an arch, with external ribs (where the cob goes),
oh boy , my head is spinning now

Muzhik wrote:
My God!  What is it with you people and colliding trains!?

Not to get too off-topic, but I understand the US railway system was built quickly to cover large distances under conditions where steel was relatively expensive, and so it relies on sidelines to allow traffic to pass in opposing directions, to an unusual degree. Most routes here involve at least a few stretches of track designed to pass trains headed in both directions.

PeterD -
You may have picked up from my tone that I don't put a lot of stock in the efficiencies we described here.  We are swapping numbers here, not to get a permit, but in response to a high school math geek curiosity about what results we might get if we could afford the tests you (and others) have proposed.

We are now accepting donations to a fund specifically dedicated to those testing fees.  Current total is in the dozens of dollars.  When it gets up to the thousands, we'll call OMNI-Labs again and see what set of tests they can do, and what the costs will be.  ($5000 is the usual cost for a woodstove, and I don't know if masonry heater testing is more or less expensive, or if this system with its unique draft and slow self-regulating feed will be able to undergo their normal testing regime.)

c605311 -
Please note that the RMH  draft mechanics are substantially different from both woodstoves and masonry heaters.  RMH's can be built with no vertical chimney aside from the combustion unit's 'heat riser' itself.  Exhaust can be 'pushed' up a vertical chimney if desired, even at low temperatures, or out a horizontal opening, by the thermosiphon combustion and downdraft cell. 
It might be wiser to ensure additional draft with hot (200+) exhaust up a vertical chimney, but the current systems are designed to maximize heat retention in the mass with exhaust temperatures generally in the 70s to low 100s F.

That is why Ernie (writing from my login) recommended test, test, test.
We've seen several examples where mixed sizes or shapes of ducting in the heat-exchange area reduced the draft to unworkable sluggishness.  You can compensate by ensuring a hotter exhaust for secondary draft, but you will essentially be creating a different type of stove at that point, and the rules of thumb for rocket mass heater proportions may not remain accurate.

  I have a much simpler method that you might want to try when determining rocket stove efficiency. Bagging and weighing exhaust gases seems like something that your average homesteader would get terribly wrong. It seems that the important matter is not the exact burn efficiency, exhaust temperature or exhaust component measurements but rather you need to answer the question of how much heat is delivered to the home and thermal storage for a given weight of wood consumed. If you have a cob bench and are able to calculate how many cubic feet of cob there are, from that you should be able to calculate a total weight of your structure. I'm sure there's a table somewhere that will tell you the heat capacity of cob, that is the number of BTUs per pound, per degree of heat rise for a typical mixture. If the entire structure is covered with blankets and possibly straw bales or some other good insulator you could run a fire using say 5 pounds of bark free firewood of known species and moisture content. Once this wood is consumed close all baffles and prevent airflow through your system. You might want to hold off on taking temperature measurements until the inner heat has time to reach an equilibrium with the outer surface of the cob. A system with a cobbed over heat riser would be best for this test since you want to store the heat and not let it radiate to the room. By taking many temperature measurements of the bench in different places you will be able to come up with a fairly accurate average temperature rise. So suppose it turns out your bench weighs 2 tons. It's now a simple matter of figuring out how many BTUs it takes to raise the temperature by whatever rise you get. BTU tables are available for most species of wood.       

So in summary what does it weigh? What was its temperature before the fire? What is the temperature after the fire? How much wood did it take to attain this temperature gain? This will provide a relatively accurate measure of your systems efficiency. The biggest of variable would be any screwup in calculating bench weight. But after we have all of these measurements that is the weight, temperature gain, weight of wood consumed and heat capacity of cob it comes down to grade 5 math skills. No fancy formulas just multiplication and division.  Feel free to call me if you need clarification on any of this. I live in Victoria British Columbia and would gladly conduct this test for free if there is someone local who has such a system and is willing to participate.

  I'm very wary of high-tech answers to simple questions. This is the simplest, non-laboratory method I can think of. Thank you: Dale 250-588-3366

dale hodgins wrote:
  I have a much simpler method that you might want to try when determining rocket stove efficiency. Bagging and weighing exhaust gases seems like something that your average homesteader would get terribly wrong. It seems that the important matter is not the exact burn efficiency, exhaust temperature or exhaust component measurements but rather you need to answer the question of how much heat is delivered to the home and thermal storage for a given weight of wood consumed. If you have a cob bench and are able to calculate how many cubic feet of cob there are, from that you should be able to calculate a total weight of your structure. I'm sure there's a table somewhere that will tell you the heat capacity of cob, that is the number of BTUs per pound, per degree of heat rise for a typical mixture. If the entire structure is covered with blankets and possibly straw bales or some other good insulator you could run a fire using say 5 pounds of bark free firewood of known species and moisture content. Once this wood is consumed close all baffles and prevent airflow through your system. You might want to hold off on taking temperature measurements until the inner heat has time to reach an equilibrium with the outer surface of the cob. A system with a cobbed over heat riser would be best for this test since you want to store the heat and not let it radiate to the room. By taking many temperature measurements of the bench in different places you will be able to come up with a fairly accurate average temperature rise. So suppose it turns out your bench weighs 2 tons. It's now a simple matter of figuring out how many BTUs it takes to raise the temperature by whatever rise you get. BTU tables are available for most species of wood.       

So in summary what does it weigh? What was its temperature before the fire? What is the temperature after the fire? How much wood did it take to attain this temperature gain? This will provide a relatively accurate measure of your systems efficiency. The biggest of variable would be any screwup in calculating bench weight. But after we have all of these measurements that is the weight, temperature gain, weight of wood consumed and heat capacity of cob it comes down to grade 5 math skills. No fancy formulas just multiplication and division.  Feel free to call me if you need clarification on any of this. I live in Victoria British Columbia and would gladly conduct this test for free if there is someone local who has such a system and is willing to participate.

   I'm very wary of high-tech answers to simple questions. This is the simplest, non-laboratory method I can think of. Thank you: Dale 250-588-3366

That does sound like a much more useful measurement of heating effectiveness.  Heating 'efficiency' comparisons are almost meaningless as the various methods each neglect different factors.  By measuring approximate BTU's stored per load of wood, you'd get a general idea how to compare it to an existing stove.  And the measurements themselves would be much more useful.  You'd actually document the warm-up, cool-down cycle, which would give new builder/operators an idea what they can look forward to. We could also compare the indoor and outdoor temperatures, and wood usage, during a typical winter heating cycle when you don't let the mass cool completely.

The figure we are using for mass is 95 lbs per cubic foot-  same as adobe.  This can vary with local rock/soil composition and building methods, but it's probably a good average.  I don't know that heat capacity has been tabulated, but maybe in the UK.  I think I've seen R-value somewhere, which might be convertible. 
To get the mass volume, we measure the total volume, then calculate the volume of the ducts, voids, and insulation, and subtract those.  Next time heating season rolls around, we can see about getting some temperature readings from one of the stoves we've built.

Yours,
Erica

  Hi Erika, I've done some digging on the Internet and located all the information necessary to calculate heat absorption by the thermal mass.  When wood is burned in perfect laboratory conditions you get almost 7000 BTUs per pound. These laboratory conditions include absolutely dry wood and a pure oxygen environment. Of course the air running through our stoves is only about 21% oxygen.    But this is a good starting point for the purpose of this discussion. Unfortunately my computer does not have the symbols for multiplication and division so I spell out the words instead. You mentioned earlier that heat capacity may be somehow related to R value. This is not the case. For instance granite and water are both very poor insulators with negligible R values but they have vastly different heat capacities. We don't want to contaminate the herd with faulty information

      30 seconds on the net lead me to the heat capacities of every solid you can imagine. Clay has a heat capacity of point .22 BTU per pound of heat rise. Dry sand has a heat capacity of .19 BTU per pound of heat rise. A typical cob mix will have three parts sand to one part clay.              22+22+22+19=79 round to .8    .8 divided by 4=.2  so the heat capacity of cob is .2 BTU per pound of heat rise. Therefore one BTU would raise 5 pounds of cob by 1°F.    For purpose of explanation I'm going to use a cob bench which weighs 5 tons or 10,000 pounds. At 95 pounds per cubic foot this would be 105 cubic feet of cob.  We're going to burn 10 pounds of dry firewood. 7000 BTUs times 10 is 70,000 BTU's. 10,000 pounds times .2= 2000 so it takes 2000 BTUs to raise the temperature of our bench by 1°F.    We have total available energy of 70,000 BTUs divided by 2000 equals 35.

    Therefore in optimal conditions the temperature of our bench could be raised by 35°F if 100% of available energy were absorbed.

    I've used a 10,000 pound bench and 10 pounds of firewood in the example because these values are so easily multiplyable or divisible. Depending on the cob mix used and combustion variables such as relative humidity, mineral content of the wood, altitude and moisture content the math is probably accurate to within 10% or so. Not perfect but certainly not the sort of malarkey the Internet is famous for

    Hi Erika, after completing this exercise I decided it might be wise to read the older posts and I discovered that someone named Tinknal was also paying attention in grade 8 science class. He expressed the basic concept but didn't do the math. When I got to Ronie's crack about woodchucks I had to nod in agreement

    But that's not what I'm writing about this time. I live in Victoria British Columbia and I own a 24 passenger tour bus which is set up with all of the equipment necessary for 24 people to go camping. This resource might be quite useful for anyone building an alternative structure using the workshop model. I could haul everyone to the site and house them for several days during the process. I have all cooking facilities, tents, mattresses and entertainment facilities to ensure that participants remain happy campers. Normally a bus like this goes out for $800 per day but since I am keenly interested in promoting green building I would do it for $300 per day which works out to $12.50 per person. This would include all fuel for both the bus and the cooking facilities and all other supplies. The bus effectively turns into a movie theater at night with a 46 inch TV and a good sound system. Videos regarding the workshop could be shown in the evening and later everyone could watch a movie or it could devolve into a jam night. I could see this working well for someone who doesn't have adequate facilities for a large workshop crowd. If you know anyone in the Victoria area who hosts such events please forward this e-mail to them. Once I get license to drive into the US I plan to take tour groups to visit various eco-villages and other points of interest to those involved in green building. It just occurred to me that this might be considered advertising which is probably not allowed on this site. But with the price I'm talking about it's hardly a business venture. I only signed up on here yesterday so hopefully they won't kick me off.        On another completely off-topic note I do all of my writing with my Dragon speech program which converts my voice to written text. If you do a lot of writing you could save your fingers and concentrate more on what you want to say rather than focusing on the typing process. This program and all things computer are relatively new to me. So I've gone from being a four fingered typist to being faster than the court stenographer. This fact and my relatively slow workweek have contributed to the lengths of my postings . Hopefully the information will be put to good use. I enjoy solving problems like this so if you find that my system works for you, let me know. Good night and good luck.

The Dragon speech software is one Ernie would like to find a current version of, that's compatible with our not-quite-current computer arsenal.

I don't mind knowing about a resource like that, we are getting some interest from the US area of the Okanogan valley, so having a workshop-in-a-can facility is nice to know about.  Might be a better place to post it on another forum, too.

Cob may mix at 3 sand : 1 clay, but the resulting volume is significantly less than 4.  I'll have to keep track of how many buckets it takes to hold a (finished) "4-bucket batch," I think it's about 3.  It would be worth testing the actual heat capacity, as the mixture is denser than either of its parts.  Varies with moisture, too.
If you can find figures for adobe, it's a pretty similar range of densities.

     Hi Erika, the calculations concern only the weight of the material. Density variations and volume reductions when you mix don't matter at all. And since clay and sand have almost identical heat capacities it's not terribly important to track the components of various mixes. As long as your stated weight of 95 pounds per cubic foot for finished dry cob is accurate then my example will be accurate. The heat capacity of adobe is nearly identical to cob on a per pound basis. Calculations are for dry material. During the heating season I can't see that the cob would hold much moisture since it's constantly being cooked out of it. The heat capacity of water is approximately 5 times that of your other materials based on weight not volume. The huge heat capacity of water warrants the use of water tanks embedded into or adjacent to rocket stoves whenever space is too tight to allow a large enough bench. The tank would require an open vent to the exterior as a safety measure in case the system is overheated and boils. The tank tank could have an insulating curtain placed on it whenever stored heat is unwanted. The maximum temperature that such a tank could reach is 212°F at sea level so there would be no danger of scorching a covering blanket or curtain.

   The Dragon speech program that I'm using cost me $100 including the headphones and microphone. There are lesser programs available free on the Internet. I'm very pleased with this one since it becomes more accurate over time as it learns my speech patterns and word usage. I am using a Toshiba satellite laptop which has a Pentium dual core. My knowledge of what this means is almost nil but my computer guy told me that this is the very minimum strength system which can support Dragon speech. When I'm using it I can't navigate the Internet properly and if I tried to do security updates or other things that make the computer think a lot the whole system freezes. Everything gets warm and the fans work at full capacity when I'm running the system. He suggested that I get something twice as powerful as what I currently have.  Used computers are a dime a dozen these days.

dale hodgins wrote:
     Hi Erika, the calculations concern only the weight of the material. Density variations and volume reductions when you mix don't matter at all. And since clay and sand have almost identical heat capacities it's not terribly important to track the components of various mixes.

My thought is that density does matter. A low density patch not only stores less heat, but also acts as more of an insulator. That is, if close to the flue is low density, and farther away is more density, the heat will not get to the high density materiel as easily and less heat will be stored. On the other hand, if the material close to the flue is higher density than that close to the skin, more heat would be stored, but it would take longer to get to the skin and the temperature over time would be different. This doesn't actually affect the calculations that much. As ...

As long as your stated weight of 95 pounds per cubic foot for finished dry cob is accurate then my example will be accurate.

This makes sense. As far as I know, most RMH are built the attention to trying to make the mix pretty even though out. The only time measurements may be skewed would be in the case of a portable RMH where loose mass is being used.

It will be interesting to see the results. Temperature over time with a good selection of probes and an idea of total surface area will probably give a pretty good idea of efficiency even with differences in materials. I wish I had one to measure as I am pretty close (Comox valley)... no ferries to catch. However, my family is keeping me busy with other things...

    There is an online resource called the engineering toolbox. If you search heat capacities you'll see that almost all common solids that you are likely to incorporate into a cob mix will average out to about .2 which is 1/5 the heat capacity of water. There are many uncommon items such as rare salts, waxes and metals which stray far from this average but they are unlikely to be incorporated in cob mixes. Whether it be limestone, clay, granite, sand, brick or almost any other earthen material this average works pretty well. Since heat capacities are only concerned with weight it makes no difference whether they are highly compacted or fluffed up into aerated concrete, the heat capacity per pound remains constant. I'm not advising anyone to create voids within a cob mix, just trying to convey that heat capacity of the cob is strictly a function of its weight and not it's density.

  To anyone who is reading this. If you live on southern Vancouver Island or in Vancouver and if you have a rocket stove or are building one I would gladly conduct this test. An efficiency test like this would help greatly when trying to sell the idea to building officials. If you live somewhere too far from here I'm sure we could converse by phone or through this forum and figure out how to run a test on your system. Thank you: Dale 250-588-3366

Dale here. It's been 16 months since I joined the forum so that I could get in on this particular thread. Just wondering if anyone has done or would like to do one of these tests. I've re read everything and although I've learned quite a bit about rocket stoves since this first discussion, I wouldn't change anything to do with the testing method or the math. It's still the only reasonable method I've seen.
..........
Once I learned how to start a thread myself, I got carried away with calculations concerning cob vs water as a heat storage medium ---- an exerpt ---

Here are some very useful figures for anyone who is trying to choose whether or not to include a water tank within an RMH. Cob weighs 95 pounds per cubic foot. Water weighs 62 pounds per cubic foot. The heat capacity of cob is .2 which is 1/5 that of water which has a capacity of 1.00 So supposing we want to build a RMH which occupies 100 ft.³ of space.

First the cob - 100x95 equals 9500 pounds. 100 ft.³ of cob will weigh 9500 pounds. 9500x.2 equals 1900. So our cob bench has the same heat capacity as 1900 pounds of water.

Water weighs 62 pounds per cubic foot, therefore the tank containing 100 ft.³ of water weighs 6200 pounds

6200 divided by 1900 equals 3.26

A given volume of water can store 3.26 times as much heat as the same volume of cob.

Here's the link - https://permies.com/t/9933/energy/Water-cob-thermal-storage

Let me touch again on the certification issue. Peter D., you posted some very useful comments regarding certification of other stove types and referred to a document you had that provided a great deal of detail on methodology. Could you make that available to Ernie and Co. to serve as a model for certifying a RMH. {I would like to see the face on the reviewers when the read PM emissions < .1g/kg.... That would be entertaining!} Given the standard measures you mentioned, an RMH would blow other stove types out of the water. I believe there was already mention that there was a case where fire fighters measured the characteristics of an RMH exhaust, those same folks and their handy instruments might be brought to bear again to provide a baseline estimate for the lab to work from.

As for the ability of an RMH to generate heat, the measurement could be vastly simplified by changing out the Cob for a water jacket, well inslulated. [if only as a test model] Raising a given volume of water one degree is standard in both scientific and industry practice. Start off with water at a given temperature and volume, run the stove, measure the change in temperature and convert it to BTUs. Also, by measuring the surface area of the heat riser and measuing the skin temperature at several places we could get a fair representation of the radiated heat from it. That would leave only the heat lost out the exhaust. Again temperature of air in compared to temperature of air out and volume of air in will yield the values needed to calculate that. To give the measured performance some relevance we might agree on a "standard model RMH" that could be replicated by all manufacturers. [We do have an industry association, right?] We could compare the performance to the amount of potential in the wood burned for a relative performance figure. Compared to other heat appliances I feel certain the RMH would stand well above the rest.

We touch on rocket mass heater efficiency and what "75% efficient" means in this new video:

Peter makes a summary about the batch box rocket mass heater project with details about the efficiency testing.

To see the full build, check out the DVDs.