click to return to the Solar Utilities Network home page Solar Utilities Network projects

Solar Space Heating

by Michael Potts

     For a civilization addicted to forced air heating, radiant heating offers some lovely qualities. It has been observed that in the heating theatre of the battle of the sexes delivering heat from below bequeathes a peaceful and one might even say aesthetic result. Babies and children love to play in rooms with radiant floors; Dr Doug remembers his youth in an early Frank Lloyd Wright home warmly.
     From an energy perspective, forced air heating and cooling are just about the most ridiculous solution imaginable, air being a terribly inefficient heat transfer medium. The most direct application of energy is always best. The puzzle of harvesting the sun's heat and using it to keep a home warm overnight drove physicists and chemists crazy in the '60s and especially during the petroleum crisis years in the late '70s. They tried pebble beds which mildewed, and eutectic salts which worked well in the lab but failed in the field, and trickle trays of black plastic which leaked... Rube Goldberg designs liberally sprinkled with sensors and active elements which seldom worked properly.
     As early as the '30s and '40s architects, led by Frank Lloyd Wright, realized that the Roman idea of a hypocaust was right on the money: the best way to deliver heat to a room is through the floor. By the time capturing solar energy as heat graduated from experimental to practicality in the early '80s, radiant or hydronic floors were making a similar transition. Thirty years or less after the first costly installations were placed in service, they were plagued by massive and spectacularly costly failures. Concrete is not good for the copper tubing originally used to move hot fluid through active floors. The problem is that the lime in cement reacts with the copper, corroding pinhole leaks; the worst-case failure case results in puddles on the floor, but generally failures just leak costly hot water beneath the slab. In Scandinavia, where nasty winter weather makes active hydronic floors especially desirable, chemists developed cross-linked polypropylene plastic tubing with long life (probably in excess of 50 years although they only guarantee 30) and good thermal qualities.
     By 1990 the parts were firmly in place. Solar thermal capture technology, hot water panels, had been very efficient and cost-effective for a decade. The tubing had also been tested for long enough to show that it was far superior to copper. Pumping hot water requires special materials and technology, but the strong crossover between the solar heated domestic hot water and hydronic flooring technologies had refined the plumbing. Since a slow flow is the best for thermal transfer, a tiny magnetic pump works best, and such pumps use a small amount of electricity. The only remaining puzzle: what kind of battery do you use to store heat?
     The solution to the puzzle of thermal storage turns out to be quite simple: include adequate thermal mass within the insulated envelope of the building. One good method is to pour a thick cement floor over insulation and hydronic tubing, and then pump the finished slab full of heat, which radiates out at a comfortable rate. Since the pressures and temperatures involved in a solar-powered system are moderate, the whole apparatus is simple and cheap. The capacity of the battery is directly related to the mass of the insulated slab; in Wisconsin, builders have installed 60-inch slabs with two and three levels of hydronic tubing so that excess solar energy harvested in September can still be radiating out of the floor in late February.
      Here's a schematic of how the system works:

The sun shines on this end This is our double-headed power source: heat is directly harvested with thermal panels, and electricity to run the pump is generated by a little photovoltaic module.
the energy is pumped from source to destination The hot water is pumped down through a well-insulated pipe to the 'thermal battery' -- the active slab below. Most of the heat transfers to the slab before the transfer fluid runs out into an insulated storage tank. Cooler replacement fluid is pumped back up through the wall to the panel in the sun.
heats is delivered where it's needed -- to your feet!

     This illustration shows a drainback system in which fluid is present in the solar panel only when there is enough sun to drive the pump. If the panel and its support pipes are dry, freezing temperatures are no threat. Pressure in the tubing is very low, and so even when leaks occur they are not very serious. Over time, the combined effects of hot water, heat, and pressure do not attack plumbing parts, and so the longevity of the system is likely to be longer.
     One noteworthy feature in this minimalist design is the absence of controls. In practice, controls are the parts which most commonly fail in any hot water system. Here's a nightmare scenario: the wires to a seldom-used thermocouple react with weather, environmental pollution, and the tubing whose temperature the thermocouple is meant to measure. One cold morning, this thermocouple fails to warn the system that its pipes are about to freeze so it can protect itself by pumping hot water through its exposed parts. The pipes freeze and burst, and the resulting damage costs three times as much as the system's total savings to date. The owners decide to move back to a cave.
     In practice, due to weather's irregularities, an uncontrolled system like this is unlikely. To be practical, there would have to be a perfect match of the weather, the thermal capture, and the thermal mass, which is unlikely. In the real world, there are two conditions which must be anticipated and controlled: too much heat, and too much cold. In a simple system like the one illustrated, too much heat is controlled by a simple bimetal thermostat which disconnects the power to the pump when the floor gets warm enough. Too little heat is a more likely eventuality, since good solar design is always a little smaller than the most extreme requirements. In this case, the tank shown could easily be a small batch hot water heater set to maintain the lowest acceptable slab temperature, typically 80 degrees Fahrenheit. Most of the time this back-up device will consume little or no energy ... but its presence surely will be appreciated on cold mornings!
     Laying out a hydronic floor, you put the heat right where you want it -- in front of the kitchen sink, beside the bed, in front of the toilet, beneath the windows which despite their being best-available glass still import cold -- and keep it away where it isn't needed -- the pantry, closets, under the couch. Larger households can be 'zoned' so that different areas are heated according to occupancy patterns, thereby reducing heating costs dramatically while improving comfort.
     There is also a caution. Heat stored in a slab isn't immediately available to the room above. Heat travels through a slab about an inch an hour. This is perfect for diurnal heat shifting: late afternoon sun's heat is injected at the bottom of a four inch slab and makes its way into the living space four hours later, just when needed. We are used to instant solutions, however, and so this latency can be irritating if we must wait four hours after turning up the thermostat to feel any warmth. This can be compensated for easily by using one more control: a setback thermostat programmed to anticipate the family's needs.

     Radiant flooring can be retrofit anywhere an existing floor can bear the weight, and enough headroom will remain when the floor level rises four inches. Special low-mass concretes are available, but since heat storage capacity relates directly to mass, these seem self-defeating. Pouring a slab inside an existing building is, it should be noted, an artform. But there is not a doubt possible that energy efficiency savings shifting from forced air to radiant floor will pay many times over for such a dramatic effort.
     Frequently in retrofitting an old building, any new floor will contend with two or more competing predecessors. In one restoration we encountered floors differing in elevation by an inch, in thickness by another inch. Ripping up and regularizing such disparate floors is a costly proposition, and produces a mountain of toxic waste. Our solution was to double up the floor joists, leave the linoleum, pine flooring, and (probably asbestos) tile in place, lay a grid of 6-6-10 reinforcement tie down the hepex tubing, and pour in four inches of forgetfulness. In our climate, this simple addition to the area's thermal mass decreased the range of unheated temperature extremes by 50%.

 click for the Solarnet home page    click to the previous article TopB.gif - 904 Bytes  click for Projects list click to the next article click to the next topic     click for Solarnet Opinions  click for the Solarnet Bookshelf click for relevant Links

Michael Potts, webster
updated 29 August 1999 : 16:28 Caspar (Pacific) time
this site generated with 100% recycled electrons!
send website feedback to the Solarnet webster

© 1999 Solar Utilities Network. All Rights Reserved.