Mixing Wood And Water
The rural area of upstate New York where I live has many homes, and even some commercial buildings, heated by wood-burning appliances. Owners who want to improve the control of wood-produced heat over that offered by typical wood stoves often look to hydronics. Some are interested in using a wood-fired boiler as their sole heat source. Others want to use it as the primary source, backed up by a conventional oil- or gas-fired boiler. Both are possible, and to be successful, both must exploit the flexibility of hydronics to rein in the operating characteristics of a wood-fueled heat source.
Burn Hot/Burn Fast: One simple fact of wood-burning is that high combustion temperature equals high efficiency. When wood is sufficiently heated it gives off pyrolytic gases. Given high enough temperatures and the presence of oxygen, these gases combust to liberate a considerable amount of heat.
However, if the temperature in the combustion chamber is too low, or the combustion zone is starved for oxygen, a significant portion of these gases pass unburnt up the flue. As they cool on their way up the chimney some of these gases condense into creosote — a sticky, tar-like substance that eventually hardens in layers against the chimney walls.
Besides creating a mess, creosote left unattended is a disaster waiting to happen. Because it’s formed from unburned pyrolytic gases creosote has considerable fuel value still locked within its molecules. If reheated to a high enough temperature by other combustion products, or directly by flames, that dormant energy can quickly reappear.
Although I’ve never personally experienced the chimney fire that results, I’ve heard several people describe them as equivalent to a roaring jet engine, powerful enough to shake the entire house. Not exactly the kind of alarm clock most people want to wake up to. Every year the local news carries stories of houses lost to chimney fires, often on some of the coldest days of winter.
One of the best ways to minimize creosote formation and boost efficiency is to burn firewood as fast and hot as possible. Some “higher-tech” wood-burning units use a forced draft system feeding a ceramic combustion chamber to attain combustion zone temperatures over 2,000 degrees F, and efficiencies over 80 percent. About the only residue is a small amount of white powder. Nothing comparable to the ash left in a typical air-tight wood stove, much less the creosote it deposits in the chimney. This technology is available in pressurized boilers, but like the commercial says, “Bring your Visa card.”
Air-tight wood burners with catalytic combustors have expected operating efficiencies of 50 percent. The catalytic combustor lowers the ignition temperature of escaping gases forcing them to ignite and liberate more heat. Since more of the gases are burned less are available to form creosote. Take away the catalytic combustor and you can expect efficiencies in the 30-40 percent range, (assuming you’re burning “seasoned” firewood with a moisture content not more than 20 percent).
Overflow Parking: The minimum heat output of many currently available wood-fired boilers is in the range of 100,000 or more Btu/hr. When paired with a reasonably well-insulated house having a design heating load of say 40,000-70,000 Btu/hr., these boilers quickly run out of load before they run out of fire.
This is where hydronic systems present an opportunity, albeit at a cost. An insulated buffer tank can provide the necessary overflow parking for those extra Btus. Its size is determined by several factors that relate the total heat produced when a load of wood is burned, the time it takes to burn the load, and the space heating load during this time. Formula 1 lets you estimate the required volume of such a buffer tank:
Vt = required volume of buffer tank (gallons) vwood = volume of wood in loaded combustion chamber (ft.3) e = efficiency of the combustion process (decimal percent) d = density of wood being burned (lb./ft.3) Lbuilding = building load while wood charge is being burned (Btu/hr.) tburn = time to burn the loaded volume of wood (hr.) DT = useful temperature range of buffer tank (deg.F) 7,000 = Btu/lb. wood fuel value (assuming wood with 20 percent moisture content)
The following table lists the density of different wood for comparison and use in Formula 1. It’s pretty obvious that hardwoods rule as fuel.
For example, let’s say you load a charge of 3 cubic feet of oak into the chamber, and burn it at 40 percent efficiency over a period of 3 hours. During that time the building’s heat load averages 40,000 Btu/hr. Fortunately the building has a radiant slab heating system that can deliver the required output at supply water temperatures as low as 100 degrees F. The intended maximum operating temperature of the buffer tank is 200 degrees F. The necessary volume of the buffer tank would be:
The greater the fuel charge capacity, and the shorter the burn cycle, the larger the buffer tank has to be. The buffer tank in this example also has an exceptionally wide working temperature range made possible by a low temperature hydronic distribution system. If hydronic baseboard units had been used instead, the effective temperature range of the tank would be considerably reduced.
It’s also interesting to look at how long the heat stored in the buffer tank could provide the building’s load after the charge of wood has been consumed. If we assume the heating load in this example stayed the same, the “coast” time would be:
If the owners of the above (wood-only) system want to wake up to a warm house on a mid-winter morning, they better set the alarm clock for a midnight feeding! A larger buffer tank in combination with a greater fuel burn capacity would also reduce the chance of the occupants waking to the sight of their own breath. So would a high mass floor heating system.
The cost and logistics of installing a multihundred-gallon, well-insulated, pressure-rated storage tank capable of handling 200 degrees F water temperatures are both substantial. One possibility might be a reconfigured propane storage tank. Locally known as “lawn sausages,” these tanks are ASME-rated pressure vessels, typically available in 500- and 1,000-gallon sizes.
After the standard propane trim is removed they have a number of NPT-threaded connections available. With good planning such a tank might be incorporated into the heated space of a building assuming there’s plenty of room available, and (preferably) a means of removing the tank if ever necessary. Once piped the tank would have to be super-insulated to keep those Btus under wraps until needed. Skimp on the insulation and the tank becomes a powerful and uncontrollable heat emitter.
Remember too, that a large volume of water in a closed system undergoing a wide temperature change will require an expansion tank substantially larger than in other systems of similar heating capacity. In summary, a large buffer tank is plausible in certain projects, but it’s definitely not for the average house in the “burbs.”
Fossil Fuel Fallback: For all but hard-core Paul Bunyan types, wood-fired boilers are usually backed-up with a separate oil- or gas-fired boiler. Another possibility is a wood/oil “combination boiler” with a separate combustion chamber for each fuel. Personally I prefer two separate boilers from a servicing standpoint. In either case check your local codes. Most will require separate flues for each fuel.
If you’re designing a hydronic system that incorporates both a wood-fired boiler and conventional boiler give thought to the following guidelines.
First and foremost, do not create a system that moves heat produced by one boiler through the other. To do so just makes the cooler boiler a heat dissipater. Series piping is not an option. Even with parallel or primary/secondary piping, it’s imperative that system controls have the intelligence to take a non-operational (or lukewarm) boiler “off-line.”
Case in point, I once spoke to a fellow who wanted to route water heated by an oil-fired boiler through underground piping to an outdoor wood-burning unit then back through the building’s distribution system. I believe he cited freeze protection of the outdoor unit as the reason — talk about global warming!
A differential temperature control, such as those used on solar energy systems could make the decision as to when the wood-fired boiler goes on- and off-line. When there’s a call for heat the differential control would compare the water temperature in the wood-fired boiler to the water temperature supplied to the distribution system.
If the wood-fired boiler water is, say, 10 degrees F or more above the distribution system temperature, the boiler goes on-line. When this differential drops to, say, 5 degrees F the wood-fired boiler is taken off line, and the conventional boiler is fired — perhaps after a time delay of a few minutes just in case the wood-fired unit has a final hiccup of heat. A two-stage thermostat in the heated space could also be incorporated to determine the “urgency” of firing the conventional boiler.
Arguments can be made both ways regarding use of the conventional boiler to boost the temperature of a large buffer tank intended to store excess heat from a wood-fired boiler. Assuming the conventional boiler is not grossly oversized, there’s no benefit in converting the fuel’s easy-to-store chemical energy into heat, the storage of which is much more difficult. On the other hand, if there were many small independently controlled loads on the system, or if the boiler was considerably oversized, the buffer tank could be used to lengthen the boiler run-cycle and boost efficiency. This, however, would only make sense during periods when the wood-fired unit is off-line, and would remain off-line until the heat added to the tank by the conventional boiler is used up.
Since the latter case requires specialized circumstances, and more complex controls, I feel it’s best to simply bypass the buffer tank with conventional heat.
When Is A Boiler Not A Boiler: Some wood-fired “boilers” aren’t really boilers at all. They’re non-pressure-rated, wood-fired furnaces that happen to heat water. Think of them as stainless steel kettles supported over a fire inside an insulated box.
With careful design these unpressurized units can be part of an “open loop” hydronic system. Remember, however, that open loop systems bring with them a number of restrictions, one of which is not being able to use steel or cast-iron components due to the higher dissolved oxygen content of the water.
If you plan to use an unpressurized wood-fired “boiler,” interface it to a closed-loop distribution system using a plate-type heat exchanger. The lower temperature side of the heat exchanger becomes the pressurized heat source for the balance of the system. Since unpressurized wood burners often sit out in the back yard the heat exchanger also provides for an antifreeze-charged subsystem. The cost of the heat exchanger is usually more than offset by not having to use “open loop” circulators, flow-checks, etc. through the remainder of the system.
Figure 1 is a conceptual piping schematic that shows many of the above concepts in place. Several other piping/control variations are possible. For example, the plate heat exchanger could be eliminated if a pressurized wood-fired boiler is used. A diverter valve could also be used to control the on-line/off-line status of the buffer tank (although it’s likely to cost more than the small secondary circulator shown).
Abundant And “Free”: Years ago someone told me his firewood was available “free” from his uncle’s ample wood lot. I asked him if he bothered closing the door to his house during the heating season. His confused look prompted a better explanation. I reasoned with him that if his heating fuel was indeed free, and unlimited in supply, it made no sense to worry about excessive heat loss. Just keep the chain saw gassed up, and the ash pail from overflowing.
As he started to reconsider the word “free,” I suggested keeping track of every minute spent in attendance of the eternal flame, pay himself minimum wage for this time, and then re-compute his fuel cost.
Burning wood for home heating is to choose a way of life rather than just a fuel. I know because I’ve had a wood-burning stove in my home since 1980, and used to harvest a meager (by local standards) three face chords of firewood each year from a generous neighbor’s wood lot. I learned that you don’t need a Nordic Trak when you’re the sole custodian of a wood-fired heating system.
Nowadays I feel less compelled to load the chain saw in the pick-up and head for the woods. It’s just too easy to call one of the many local firewood delivery services and get it dropped in my driveway. The going price for seasoned hardwood is about $35 per face cord.
I put together the fuel cost comparison worksheet in Figure 2 to calculate the unit cost of heat derived from several common fuels including wood. Each formula computes the cost of delivered heat on a consistent dollar per million Btu basis. Plug in your own numbers (or assumptions) for cost and efficiency and judge the results for yourself.
Remember that the efficiency of wood burners varies widely depending on design and operation. For anything but a high-tech, forced-draft unit I suggest an average efficiency around 40 percent. The notes at the bottom of the worksheet give typical values of some efficiency factors.
Wood-fired heating systems definitely have their place. In my part of the country they’re literally part of the local heritage. It’s important, however, to compare both installation and operating costs before making a decision. And regardless of which fuel you choose to produce the heat, make sure it’s a hydronic system that delivers it!