Heat loss through floor perimeter, or area?

[kryanl]

We have a large (400' x 750') warehouse, for which has now been decided, will require a heating system. I've been asked to estimate the cost. I've completed the walls and roof heat loss portion of my excercise, and it is a rather tight building, so infiltration is pretty low, but where I'm having my most trouble is figuring the heat loss through the floor. Going back through old notes, and searching advice on the internet, I've learned there are two schools of though for calculating heat loss through a floor:

1) calculate U for heat moving directly down into the ground, multiply by the floor surface area,OR
2) calculate U for heat moving through the perimeter and multiply by the perimeter.

The floor of our warehouse, is an 8" thick slab of concrete, with an uninsulated foundation (aside from the water stops, etc.)

Now, for the first case, I can figure that out easily, but come up with an outrageous number. The second case, I admit, I am not sure how to come up with a U-value for a perimeter (would it be considered at a point of the wall thickness as it sits on top of the slab?)

Can anyboldy help clear me up? When do you use case 1, and when do you use case 2?


Thank you so much,
Kayla

 

[stanlsimon]

There are two different heat losses, from the space into the ground and from the space into outside. The ground temperature will probably be warmer than outside. Figure a linear heat loss around the perimeter. If you figure 1 btuh / (ft Degree F) you should be covered. The heating setpoint of the warehouse will determine how much heat is lost through the floor. Generally you figure a ground temperature of 50 or 55 degree F. Don't know where you are. How warm does the warehouse need to be? If you are just trying to keep from freezing you could have a warming effect from the floor.

How about ventilation, do you need any? It can get humid and stinky without ventilation. Ventilation heating loads are signifigant.

Many warehouses suffer from open door syndrome. With loading docks and overhead doors there is lots of major infiltration, this can make conduction losses seem small.

So to answer your question, use both but slab loss may be minimal.

 

[GMcD]

Using the Carrier Load estimating handbook (which also parallels ASHRAE) the right answer is both (1) and (2) apply for a slab on grade, with no insulation under it or at the edges.

The heat loss via conduction through the slab to the ground is slab area x Rslab x (Troom - Tground). Ground temperature varies depending on location, for example in the coastal pacific NW, we'd use 55F. Rslab is the thermal resistance of the slab thickness of concrete.

The slab edge loss according to the Handbook is slab perimeter in ft. x perimeter factor x (Tindoor-Toutdoor). For slab on grade, the perimeter factor is 0.6 (Pages 1-81 and 1-82 for those of you who have the Carrier Handbook).

Consider a layer of rigid insulation under the slab and around the edge to reduce the heat losses significantly - do the numbers and I'll bet it's an instant payback for the cost of the insulation, if you are in a colder climate zone.

 

[DickRussell]

In that last reply, the statement:

"The heat loss via conduction through the slab to the ground is slab area x Rslab x (Troom - Tground)."

should use (1/Rslab) rather than Rslab. The "R" is 1/U, where the "U" is the heat transfer coefficient, in this case based on the thermal conductivity of the floor and any other material (such as rigid insulation) between the air above the slab and the ground below. The formula is the simple Q=A*U*dT.

For the air infiltration heat load, you would have to make some guess as to the volume of outside air that comes in over time. If the building is normally closed up and fairly tight, you could use perhaps 1/2 to 1 air change per hour, or more according to what "tight" means in your situation. Given a volume of air, at a density of about 0.076 lb/cu.ft. and heat capacity of about 0.25 BTU/lb-degF, then the heat required to warm inflowing air to inside temperature is Q=volume(cu.ft.) * 0.076 *(temp difference). As an earlier poster indicated, open truck doors may give you such an air infiltration load that heat loss through the slab may be rather minor.

 

[tombmech]

This all assumes warm air heating. There is another approach.

Gas-fired radiant tube heating can be a better choice in large aircraft hangars and warehouses. In effect, the radiant heating is irradiating the slab and surrounding surfaces. They become the heat source rather than a point of heat loss. If you have large infiltration losses, this can give you a tremendous leveraged advantage.

In one case, I had an aircraft hangar that had been heated historically with large gas-fired, rotating Wing heaters. The hangar's roof steel clearance was 90 feet A.F.F., and the doors were also that large. Needless to say, calculating for the infiltration loss of a 90 foot door opening was something akin to an infinite heat sink.

The radiant heating successfully replaced the existing Wing heaters and dropped the gas usage by a couple of magnitudes.

Try investigating "Reflecto-Ray" or "Re-Verber-Ray" radiant tube heating.

 

[GMcD]

Thanks Dick- good call- I've been doing heat losses for so long it was an oversight and I assumed that it was understood that "R" meant thermal resistance and that "everyone" knew that the "U" value was the number to use. Mea Culpa.