I am neither a mechanical engineer nor an expert in radiant slab heating. However, I did have it installed in my own home in northern Maine and was involved in the design of the system and the final adjustments.
In northern Maine the average winter temperature is 10°F with many days -20°F or colder. Each winter the ground freezes to a depth of 5 to 7 feet and the lakes freeze to a depth of a yard or more. Annual snowfalls over 15 feet are not uncommon.
What follows is my understanding of radiant heating, or, when placed in a concrete slab, radiant slab heating.
Heat energy is transferred in one of three ways; conduction, convection and radiation.
- Conduction occurs when a surface at one temperature comes in contact with a surface at a different temperature. Heat energy will flow from the hotter to the cooler surface.
Test – place your hand on a surface that is a different temperature than your hand. Conduction will be obvious. Your hand will either be warmed by a warmer surface or cooled by a cooler surface.
- Convection occurs when air is stratified by temperature with warmer, less dense air at a lower barometric pressure than cooler air. The normal situation is that barometric pressure decreases as the elevation increases, i.e. the pressure is higher at the floor than at the ceiling. Therefore, warm, less dense, air will typically rise above the cooler air.
Test: Start a fire in a wood stove in a cool room. The large difference in temperature between the stove and the surrounding air will cause the air near the stove to become hot through conduction. Convection will cause this hot air to rise [and the cooler air near the ceiling to fall] making the air temperature in the surrounding room near the ceiling warmer than the air temperature just above the floor.
Note that it is hot air that rises, not “heat”. Further explanation below.
- This brings us to radiation, the method by which a radiant concrete slab warms a room. Radiation involves a direct transfer of heat energy between two objects each having mass without regard to the temperature of the air or the distance between the objects. While the air does have a tiny mass it is small enough to be negligible in the equation and the air plays little role, or none at all. The surface temperature of a radiant concrete slab does not get hot. If it did it would not be comfortable to walk on. It is limited by the requirement of “creature comfort” to being barely warm.
Test 1: When the sun suddenly appears from behind the clouds on a cold winter’s day you instantly feel the warmth of the sun on your face. This is radiant heat. It involves a direct exchange of heat energy between the mass of the sun and the mass of your face which has nothing to do with the temperature of the air between the two or the fact that the sun is 93 million miles away.
Test 2: Another common example is the radiant heaters commonly found mounted to bathroom ceilings. When you step out of the bath, turn on the heater and stand under it you feel instant warmth. It is not necessary for the heater to warm the room or the air around you for you to be warmed.
Test 3 [The “bald man” test]: The bald man sitting in the recliner on a cold winter day with the back of his head facing a large window complains that his head is cold. But, if you check the air temperature between his head and the window it will be the same temperature as the rest of the warm room. His head is having a direct radiant exchange with the cold glass, just the opposite it would have with the warm sun in Test 1.
It is not necessary for the floor to be the surface that contains the radiant source. Radiant heating in non-directional. The source could just as easily be the walls or the ceiling or any other object in the room. In fact, it is commonly installed in ceilings. However, for general heating the floor has several advantages:
- It is equal distance from most objects in the room [as would be the ceiling].
- Concrete provides an excellent radiant surface.
- Horizontally placed concrete is relatively inexpensive and easy to install.
In fact, concrete is not required at all for a radiant system to work. However, it is a common and efficient material for radiant heating.
A concrete radiant floor system is made from the following components:
- A heat source. This can be whatever type of boiler/fuel source you choose. This source heats hot water.
- Hot water. This water can be either heated on-demand or heated and stored in a hot water storage tank. While we don’t need the water to be very hot to run in the system, hot water in a tank stores a lot more energy and boilers are not usually made to work efficiently if they were to produce only “warm” water. A secondary loop in this tank can also heat hot water for domestic use.
- A control system. More sophisticated systems include a thermometer to monitor the outside air temperature and adjust the heating water distribution temperature to accommodate the demand load. The system I had in Maine has this feature.
- A cold water source to temper the hot water created by the boiler to the appropriate, relatively low, heating distribution temperature.
- A manifold to distribute warm water through a series of loops [made of tubing] installed in the concrete floor.
First, it is necessary to insulate the floors, walls and ceilings to prevent absorbed heat energy from escaping to the surrounding environment.
As warm water begins to flow through the tubing embedded in the concrete floor it warms the concrete mass. As soon as the top surface of the concrete becomes warmer than the other surfaces in the room a radiant exchange begins between the slab and the other surfaces. These surfaces include the walls, ceilings and every object in the room. As mentioned above, while it also warms the air, the air has negligible mass and can be ignored in the equation.
Since the air in the room has such a tiny mass relative to the other masses that make up the room, the air temperature almost instantly matches the temperature of the mass of the room. Since all of the individual surfaces and objects in the room are approximately the same temperature, the air temperature near the floor is approximately the same temperature as the air near the ceiling. I use the term “approximately” because not every object in the room has the same mass density. Denser objects absorb radiant heat energy more slowly than less dense objects. However, relative to the air, all of the surfaces and objects in the room will be nearly the same temperature.
The floor will continue to radiate heat energy to all of the surrounding surfaces until all of the mass in the room is approximately the same temperature as the slab. This will bring the room temperature up to the target temperature. The circulating pump will then stop and water will no longer flow through the tubing in the floor [although we will discuss later why it is best for the pumps to stop only briefly at best].
As heat energy leaks from the mass of the room to the environment beyond the room [insulation isn’t perfect] the mass temperature of the room, and therefore the temperature of the air in the room will begin to cool. When the air temperature reaches a pre-determined low the circulating pumps will once again pump warm water through the slab and the process repeats.
In my home there was a large space with two levels open to each other. I monitored two thermometers for nearly two years – one placed near the ground floor and the other placed just under the upper level ceiling. There was about 18 feet vertically between them. I never witnessed a temperature differential between the two thermometers of more than 2° F. “Heat” doesn’t rise – it is hot air that rises, and there was no hot air in the room.
This is also why radiant slab heat is perfect for large open buildings such as a firehouse that has large doors to accommodate vehicles. When all of the doors are opened suddenly on a cold winter day the warm air escapes quickly. However when the doors are closed the air is warmed again almost instantly as the mass of the room has not had time to cool.
The usual construction is a cast-in-place concrete slab containing tubing through which warm water will flow. The top surface of this slab is the heating source. The slab also contains steel reinforcing to control concrete shrinkage and cracking. There is insulation under the slab to ensure that the only radiant surface is the top of the slab.
The radiant concrete slab [or screed] can be constructed on top of any structural concrete, steel or timber deck at any floor level. It is often constructed directly on the ground with a vapor barrier under the insulation to prevent the migration of ground moisture into the system.
The most important variables to be considered are the temperature of the outdoor environment and the desired indoor temperature. The temperature of the environment is usually expressed in degree-days and reflects both the average annual temperature as well as the temperature variation throughout the year.
The system is designed based on the reasonably coldest day of the year. “Reasonable” is a judgement call. It would not be reasonable to use the coldest day in 50 or 100 years as you would to design against floodwaters. If you were resigned to wear a sweater for a day or two on rare occasions you would have a much more economical system than if you design for the worst all-time case. And the more days you agree to wear a sweater the more economical it will be.
The indoor design temperature needs to be chosen with the knowledge that during the typically coldest days of the year a properly adjusted system will not allow the indoor temperature to be increased above the design temperature, and, during those few hours when the actual outdoor temperature is below the design outdoor temperature, the indoor temperature will be slightly below the indoor design temperature. These situations almost always occur at night when you are snug under the covers.
You can have as many zones as you wish and each zone can be a different temperature. Each zone will be served by one or more loops distributed from the manifold. The slabs in different zones should be separated by insulation. Because zones are almost always separated by walls, it is easy to hide these insulated seams.
Once the target outside and inside temperatures have been determined, the length of tubing for each loop can be calculated. Warm water leaves the manifold at a particular temperature and returns at a lower temperature, having distributed a portion of its heat energy to the slab. The quest is for the temperature differential between the water leaving the manifold and the water returning to be minimized. You can, however, make the loops too short so that little energy is distributed during the trip. It is the job of the system designer to determine the optimum loop length based on the design temperatures, building configuration and energy losses caused by the physical parameters of the system, especially the tubing.
Once this loop length is determined then all of the loops in a particular zone should be the same length so that the slab is evenly heated. You add as many loops as required to cover the size of each zone.
The pattern for the tubing can be varied. The tubes should be closer together where the demand for heat is highest, for instance near large areas of glazing. I have seen sophisticated calculations made to determine the proper spacing of the tubes but I have rarely seen this done in actual practice. I have seen many installations and the tubes have usually been evenly spaced at 12 inches on center. I know that this is not always the optimum situation.
The thickness of the slab and location of the tubes vertically within the slab are also important. The surface of slab has to be warmed to the appropriate temperature before the indoor design temperature can be achieved.
In a properly adjusted system the thickness of the slab is less important than the distance between the tubes and the top of the slab. This distance wants to be just enough for the surface of the slab to heat evenly. Too close to the surface and you evenly heat only strips of the surface. Too deep and it takes too long for the heat energy to be released, although this is only a problem if the slab is allowed to cool too much between demand cycles for heat. This is discussed in more depth in a following section.
I gleaned from a radiant heating design manual that the optimum depth to the bottom of tubes spaced at 12 inches on center is about 2 inches below the surface of the slab. This works well where the reinforcing steel is placed on chairs in the middle of a 4” slab and the tubing is tied directly to the steel.
I should note here that in very frigid climates there are instances where radiant slab heat cannot meet the demand at specific locations within the building on the coldest days of the year, such as near a large wall of glass. These isolated areas may require a secondary heat source adjacent to the glazing. At the very least the tubes should be more closely spaced adjacent to large glazed windows. A knowledgeable designer should be able to determine if spacing the tubing closer together is enough or if supplemental heating will be required.
We have noted that the most efficient system is designed for the reasonably coldest day of the year, not the coldest day to ever be expected, and that the water leaving the manifold should have the lowest reasonable temperature drop upon its return to the manifold.
For this to occur the temperature and velocity of the distribution water needs to be adjusted so that it is supplying heat at the same rate as heat is being lost from the structure, i.e. supply = demand.
This means that in a properly designed and adjusted system the circulating pumps would run almost continuously and the slab has little opportunity to cool.
If the system monitors the outside air temperature and adjusts the water distribution temperature accordingly, this situation will occur for most of the heating season.
If outside air temperature is not part of the equation then this situation occurs only during the coldest days of the year and a considerable level of efficiency can be lost.
If there is a significant time delay between demands for warm water then the slab will cool and when the next charge of warm water flows there will be a lag before any heat energy is released to the surrounding mass of the room because the slab will have to be heated first. This results in an inefficient and more costly operation.
The final adjustments to the system should be made during the coldest part of the heating season. If the system does not monitor outside air temperature, then these adjustments must be made during a period matching, or at least nearly matching, the outside design temperature.
What follows is how I adjusted my system.
The control panel has a dial to adjust the water distribution temperature for each zone. When the system was installed the installer made a guess [on the high side to be safe] and let the system initially warm my home to the various indoor design temperatures. There are 3 zones with one zone set at one temperature and the other two set at a lower temperature.
I turned the system on the first week in December. It was -20°F. It took an entire week and a tank of oil to bring the mass of my 4500 sf home/office/shop/garage and unfinished first floor up to design temperatures. Of course the shop, garage and unfinished first floor were only heated to 50°F. Having a heated garage in a frigid climate is the best present I ever gave myself.
After the design temperatures were reached I lowered the water distribution temperature at the control panel in small increments over a couple of days until the indoor air temperatures slipped just below the design temperatures. This meant that even with the circulating pumps [one for each zone] running continuously there was not enough heat energy being distributed meet the demand.
I then turned the thermostats all the way up. That made no difference in the room temperature since the system was already running full out.
I then began raising the water distribution temperatures at the control panel in very small increments. I waited until the room temperatures levelled off before I made the next adjustment. I continued this process over several days until the room temperatures were just barely over the indoor design temperatures.
I then lowered the thermostats back to the indoor design temperatures. The supply was now just meeting demand, the circulating pumps were running almost continuously, the differential water temperature leaving and returning to the manifold was minimized and the slabs did not have a chance to cool off before the next call for warm water.
And because the temperature of the outgoing distribution water was adjusted automatically as the outdoor air temperature varied, it made no difference what the outdoor temperature was – the system continued to operate almost continuously and therefore, economically.
I owned that home for 5 years and never touched the thermostats again. There is no sense in turning them off or even adjusting them in the summer. They will call for heat when it is needed and won’t when it’s not. Besides, you won’t get them properly adjusted again to the most efficient settings until the coldest part of the next winter.
It is also my recommendation that the sensors that monitor the temperature and the actual temperature controls be two separate devices. You put the temperature monitors where they need to be to work efficiently BUT you should put the temperature controls down in the cellar behind the axe [an old New England expression] so that you are never tempted to touch them again.
In the right situation, radiant heat is wonderful. The room temperature is constant and there are no “cold” corners. There is no hot air in the room so ceiling fans are not needed when they are installed for the purpose of distributing hot air near the ceiling to the rest of the room. There isn’t any hot air near the ceiling. After the initial adjustment you can forget about adjusting the temperature – forever.
However, radiant heat is not for everyone or for every situation. For me it does not work at all in my bedroom where I like a window open at night even during the middle of the winter. In the morning I like to close the window and have the room warm in a few minutes. Alternative heating sources must be installed for such areas. As long as you keep the radiant heating system temperature monitors outside of these areas then the radiant system can supplement these rooms. But the most efficient systems will leave radiant heat out of these rooms altogether.
Radiant heat won’t work in any situation where frequent air temperature adjustments are required.
And it certainly will never work efficiently if you insist that the thermostat be turned down at night or when you are on vacation.