This article will discuss the preliminary engineering tasks that underlie any hvac project: calculating the loads and determining what type of system to design. It is based on a residential project where a hybrid (in-floor radiant/baseboard) hydronic heating system was installed to heat an addition. However, the principles apply to any project. The analysis is by a practicing engineer and presents practical engineering considerations intended for fellow practicing engineers.

Although it is called in-floor radiant heat, Figure 1 shows that at typical room and panel (floor surface) temperatures, about one-third of the heat output is by convection.

Various magazine articles, manufacturer promotional literature, and other sources have advocated modifying the heat loss calculation procedure for spaces heated by radiant heat. There is only one real way to calculate heat loss: Q (heat loss, Btu/hr) = U x A x Delta T. The competing methods are all simplifications or variations on that basic formula. The following paragraphs describe some of the modifications proposed for radiant heating systems and explain why not to use them.

1. "68 degrees feels like 72 degrees." Some radiant heat proponents advocate reducing design capacity based on the claim that radiant heat achieves the same comfort as convective (baseboard or warm air) at a lower room temperature. This claim comes from the theory that radiant heat increases comfort by raising the temperature of the surrounding surfaces. Applying this philosophy can get a designer into trouble:

• Reducing design capacity on the basis that "68 degrees feels like 72 degrees" raises a design criteria dilemma: Does the design promise a specified indoor temperature at a specified outdoor temperature, or does it promise subjective and unmeasurable "comfort"?

This debate is primarily about business practices, not engineering. Designers and installers who promise "comfort" run a risk that occupants will say they are cold at 68 degrees F/20 degrees C or that 68 degrees F/20 degrees C does not feel like 72 degrees F/22 degrees C to them. An owner who claims 68 degrees F/20 degrees C does not feel like 72 degrees F/22 degrees C raises a legitimate breach of contract claim against a designer or installer who promised comfort. On the other hand, if the designer/installer promises a system that maintains an agreed indoor temperature under agreed conditions of temperature and wind outdoors, objective measurements can determine whether the system fulfills that contract obligation.

• ASHRAE comfort research confirms that people lose body heat by radiation to surrounding surfaces. Raising surrounding surface temperature reduces body heat loss by radiation. We've all experienced this phenomenon by going outside on a cold but sunny winter day. Standing in the direct rays of the sun feels warm, even if the ambient temperature is cold.

One problem with applying that reasoning to space heating systems is that mean radiant temperature does not account for local conditions or cold spots from cold surfaces. This condition is called radiant asymmetry. Those cold spots can be thought of as analogous to drafts from an air distribution system. Overall conditions are okay, but the draft or radiant asymmetry creates local discomfort. ASHRAE comfort research recognizes that local conditions affect comfort. Someone walking by or sitting in front of a large window is likely to feel the cold, especially at night, even with radiant heat and even if the person's mean radiant temperature is satisfactory. After all, 60 degrees F/16 degrees C and 80 degrees F/27 degrees C average out to a satisfactory 70 degrees F/21 degrees C, but neither 60 degrees F/16 degrees C nor 80 degrees F/27 degrees C is satisfactory. Depending on the type of room and the amount of glass, 68 degrees F/20 degrees C might not feel like 72 degrees F/22 degrees C.

• There is no convincing reason to believe in-floor radiant heat raises mean radiant temperature more than a system like continuous baseboard circulation with water temperature reset. Continuous circulation with water temperature reset is common in commercial buildings. It warms wall and glass surfaces, raising mean radiant temperature. It has also been identified, at least since 1980, as a viable concept for residences. See "Technical Topics Number 7A," available from the Hydronics Institute division of GAMA. Baseboard with continuous circulation seems to achieve the same functional result as radiant heat (warmer wall and glass surfaces). It therefore should provide comfort equal to radiant heat, except for floor surface temperature, which is important only if the floor is over an unheated space. No one advocates reducing capacity or lowering room temperature setpoint on baseboard systems with continuous circulation and water temperature reset. What makes it appropriate for in-floor radiant systems?

• Although it is called in-floor radiant heat, Figure 1 shows that at typical room and panel (floor surface) temperatures, about one-third of the heat output is by convection. The argument that in-floor heating systems are different from baseboard or warm air systems falls apart as the convective portion of heat output increases. This analysis might be different for radiant ceiling panels or high intensity infrared heating systems that operate at higher surface temperatures.

• There appears to be no published, generally accepted engineering basis for reducing capacity or expecting lower room temperatures with radiant heat. On the other hand, one published report concludes that occupants of radiant heated homes do not set their thermostats lower than homeowners with other types of heating systems. "Research Highlights" Fact Sheet Number 01-106 from the Canada Mortgage and Housing Corp. reports on a survey in Nova Scotia that compared thermostat settings in fifty houses with radiant in-floor heating and 25 houses with other heating systems (forced air, hot water baseboard, etc.). Thermostat settings for the 50 houses with radiant in-floor heating systems averaged 20.4 degrees C. Main floor thermostat settings for the 25 control houses with other heating systems averaged 19.8 degrees C.

The risks associated with promising "comfort" as opposed to a measurable room temperature may outweigh any benefit from reducing capacity based on the choice of a radiant heating system.

2. Add 0.5 to R-values. One radiant heating system manufacturer advocates increasing R-values by 0.5 hr-ft2- degrees F/Btu when radiant heat is used. The theory is that radiant heat induces smaller air currents in the room, so the inside air film resistance increases. The theory makes sense, but little or no scientific evidence has been presented to substantiate the theory or to show that 0.5 is the correct adjustment. As a practical matter, the adjustment makes almost no difference for well-insulated opaque surfaces. Consider a wall with 6" fiberglass batt insulation. If R-total is 20, the U-value is 0.05. If R-total is increased to 20.5, the U-value becomes 0.049. The difference in calculated heat loss is 2% (1-.049/.050). Adding 0.5 to the R-value makes a bigger difference on glass or poorly insulated walls. Unless the room has a lot of windows, the overall impact on the size of the heating system is quite small. Therefore, until the basis for the adjustment is substantiated, published widely and generally accepted in the engineering community, business sense dictates sticking with conventional heat loss calculation techniques.

3. Cut I = B = R heat loss by 20%. Some commentators, especially those who target trades people, advocate cutting the I = B = R heat loss by 20% when using radiant heat. The recommendation may have some merit, but not because of radiant vs. convective heat.

The I = B = R heat loss calculation method is essentially the same as the ASHRAE method. The only difference is that the tables in the I = B = R Heat Loss Calculation Guide H-21 have infiltration heat loss built in. The infiltration heat loss is based on an outside air infiltration rate of one air change per hour (1 ach).

For a typical house of modern construction in the northeast quadrant of the United States, the infiltration rate at heating system design conditions is about 1/3 to 1/2 ach. The infiltration heat loss typically winds up about the same size as the transmission heat loss. For example, if the transmission heat loss is 23,000 Btu/hr, the infiltration heat loss will probably also be somewhere around 23,000 Btu/hr, making the heating load 46,000 Btu/hr.

If the calculated infiltration heat loss of 23,000 Btu/hr with modern construction techniques represents 1/3 to 1/2 ach of infiltration, the 1 ach infiltration heat loss built into the I = B = R tables is two to three times the actual infiltration heat loss. Therefore, the 1 ach infiltration heat loss built into the I = B = R tables represents 46,000 Btu/hr to 69,000 Btu/hr of infiltration heat loss for our example. The total heating load (transmission plus infiltration) determined from the I = B = R tables will be between 69,000 Btu/hr and 92,000 Btu/hr. Cutting the I = B = R heat loss by 20% results in a design heating load of 55,200 Btu/hr to 73,600 Btu/hr for a building with an actual heating load of 46,000 Btu/hr. Obviously, these numbers and the ratio of infiltration to transmission heat loss will vary with each project. The point is that with modern, energy-conscious residential construction techniques, cutting 20% off the heat loss derived from the I = B = R tables is unlikely to cause a problem, not because of radiant heat but because today's infiltration rates are lower than when I = B = R developed the tables.

System Selection

Once loads have been calculated, the next step is to select a system type. This is when the designer decides whether to use hot water baseboard, in-floor radiant, a warm air furnace, a heat pump, etc.

Proponents of radiant heat claim many benefits over other types of heating systems. Some of those claims are truly attributable to radiant heat. Others result from comparing well-designed radiant systems to poorly designed other systems. The following considerations compare perimeter baseboard to in-floor radiant heat:

1. There is no question that in-floor heat keeps the floor surface warm. This advantage is particularly important when the floor is over an unheated space. The cold floor problem can exist even if the unheated space is enclosed and stays well above the outdoor temperature. In commercial buildings, floors over unheated space are often kept warm by insulating and heating the soffit below. That technique could be used in residential construction but is not often seen.

2. In-floor radiant heat makes wall space available for furniture placement. This advantage is less important at windows because there is usually free space in front of windows for access to open them. Bookcases would not back up to a window. A couch or a chair might back up to a window. That type of furniture is often not a big problem for baseboard heat because it usually has openings near the floor that allow air to reach the baseboard inlet.

The flip side of this coin is that baseboard heat might be an advantage in high heat loss areas like under a window. Baseboard heat typically delivers more heat per linear foot than radiant heat.

3. Comfort considerations based on floor surface temperature limit the heating capacity available from in-floor radiant heat. Figure 9 in Chapter 6 of the 1996 ASHRAE Systems and Equipment Handbook shows a maximum available heat flux of about 25 Btuh/ft2 (79 W/m2) from a carpeted floor without exceeding the recommended 85 degrees F (29 degrees C) maximum surface temperature. If the heating load exceeds that density, either a different heating system or some form of supplementary heat will be required. A "great room" with a cathedral ceiling and a lot of glass could easily exceed the limit, as could a small addition with three sides, a roof, and a floor over an unheated basement or crawl space.

In-floor radiant heat is considerably more expensive than perimeter baseboard. For the systems and equipment used on this project, hydronic baseboard heat cost \$20/mbh (materials only, purchased at a home center). In-floor radiant heat using a sandwich over subfloor design cost \$86/mbh for materials only (tubing, manifolds, etc.) without labor, special materials like under floor insulation, or the extra subfloor. Therefore, the total installed cost for in-floor radiant heat will likely be four to six times as much as a hot water baseboard system.

The observations and conclusions presented here result from the author's experience, engineering judgment and evaluation of data from this case study. They may need further testing before they can be generalized or extended to other situations.