One number describes the whole building
UA is the heat a building loses per hour for every degree of difference between inside and outside. Add up each path — area times U-value for anything you conduct through, perimeter length times a per-foot factor for the ground line, and 1.08 times the leakage airflow for the air that swaps itself out — and you have a single figure in BTU per hour per degree F. Multiply by the temperature difference and you have the load. Multiply by degree days and by 24 hours and you have the season.
The 1.08 on the air leakage line is worth knowing rather than copying. It is 60 minutes per hour times about 0.075 pounds per cubic foot of air times 0.24 BTU per pound per degree, which comes out at 1.08 BTU per hour per CFM per degree F. Every sensible-heat airflow calculation on this site uses it, and the same constant reappears when the same house is being cooled in July.
Worked on the numbers the form opens with: 1,500 sq ft of single film at U 1.20 gives 1,800. Two hundred square feet of insulated knee wall at U 0.35 gives 70. A 136 ft perimeter at 0.80 gives 108.8. A 7,540 cu ft interior at one air change an hour is 125.7 CFM, and 1.08 times that is 135.7. Total UA 2,114.5. At a 45 degree difference the house is losing 95,153 BTU an hour.
What a second layer actually buys
Going from U 1.20 to U 0.70 on the glazing takes that 1,800 down to 1,050, which is 41.7 percent off the glazing line. The total only goes from 2,114.5 to 1,364.5, which is 35.5 percent off the load. The missing six points are the perimeter and the air leakage, and they sit there unchanged no matter what film goes over them.
| Path | Single film | Inflated double | Change |
|---|---|---|---|
| Glazing | 1,800 | 1,050 | −41.7% |
| Insulated wall | 70 | 70 | none |
| Perimeter | 108.8 | 108.8 | none |
| Air leakage | 135.7 | 135.7 | none |
| Envelope UA | 2,114.5 | 1,364.5 | −35.5% |
| Load at 45°F difference | 95,153 BTU/hr | 61,403 BTU/hr | −33,750 |
Thirty-five percent is still an enormous number. Over 4,000 degree days at a base of 50, the single-film house needs 203.0 million BTU and the double-film house needs 131.0 million, a difference of exactly 72 million BTU. At 91,500 BTU per gallon and 80 percent efficiency that is 2,773.1 gallons against 1,789.5 — 983.6 gallons saved. At $2.50 a gallon the season saving is $2,459.01 against a $900 upgrade. It pays back inside the first winter with $1,559.01 left over, which is the rare case where the arithmetic is not close enough to argue about.
The same arithmetic explains why the air leakage line deserves the first afternoon. Halving the air changes from 1.0 to 0.5 takes 67.9 off the UA — a third of what the whole second film layer bought — for the price of weatherstrip, a vestibule at the door, and closing up the lap at the end walls.
Degree days are wrong for a greenhouse, and it matters which way
A degree day total counts every hour the outside temperature sits below the base, including the daylight hours. A house is dark inside and that is roughly fair. A greenhouse spends those daylight hours flooded with solar gain that is often several times its heat loss, and in many climates it is venting heat away at noon in January. Feeding raw degree days into a UA gives an upper bound on fuel, sometimes a generous one.
There is no honest general correction, because the answer depends on latitude, on how many of your winter days are clear, on the glazing transmittance, and on whether you have anything inside able to absorb heat and give it back after dark. The solar credit field exists so you can see the size of the effect rather than pretend it is zero: a 20 percent credit on the single-film case takes 2,773 gallons to 2,218. If you have delivery tickets from a past winter, running the degree day calculator backwards from those gives you a rate per degree day for your own house, which beats every assumption on this page.
Getting the inputs
The glazed area is the one people underestimate, because they measure the floor. A semicircular hoop of 20 ft span has 31.4 ft of arc over 20 ft of ground, so the skin is 57 percent larger than the footprint before the end walls are counted. The hoop house sizing calculator produces that area, the end wall area and the interior volume from the span and length, and those numbers drop straight into this form.
U-values come from whoever sold you the covering. Film manufacturers publish them, twin-wall panel makers publish them, and the two are not comparable to a wall R-value without care — the U-value of glazing is a whole-assembly figure that already includes both air films. The perimeter factor is the vaguest input here; if nobody detailed the base, setting it to zero and reading the answer as glazing-plus-leakage only is more honest than inventing a number.
Once the winter numbers are done, run the same house through the summer ventilation calculator. On the default house here the July solar gain is roughly double the January heat loss, and it is the summer number that usually decides whether a small greenhouse is usable at all.
Questions people ask
Why is my greenhouse heat loss so much higher per square foot than my house?
Because almost the entire envelope is glazing, the envelope is large relative to the floor, and there is no insulation anywhere in the load path. A code-built house wall might be U 0.05. Single greenhouse film is around 1.2, which is roughly twenty-four times as much heat per square foot per degree. On the default numbers here the house is losing about 99 BTU an hour for every square foot of floor at a 45 degree difference, which for a dwelling would be an extraordinary figure and for a single-film tunnel is unremarkable.
Should I include the floor as a heat loss surface?
Not as an area times a U-value, no. Heat leaving downwards into soil under a greenhouse behaves nothing like heat leaving through a wall — the ground under the middle of the house warms up over a season and then loses very little, while the ground at the edge is in contact with outside air and keeps losing. That is why the calculation uses a per-foot-of-perimeter factor instead of a floor area. If you have no insulation at the base and no figure from anyone, run it once at zero and once at a guess and treat the range as the honest answer.
How do I find my air changes per hour without a blower door?
You mostly cannot measure it, which is why it is a field with a wide hint rather than a constant. What you can do is bracket it. Run the calculator at 0.5 and again at 1.5 and look at how much the answer moves; on the default house that spread is about 136 BTU/hr/degree, or 6,100 BTU/hr at design conditions, which is the size of the uncertainty you are living with. Then spend an evening with a smoke pencil on a windy day and fix whatever it finds, because that is cheaper than resolving the number.
Does the calculator tell me what size heater to buy?
It tells you an output in BTU per hour that the steady-state physics asks for, plus whatever margin you set. It does not know your wind exposure, whether you run a night curtain, how the heat is distributed, what happens during a defrost cycle, or what the appliance actually delivers at your altitude. Take the number to somebody who installs greenhouse heat in your area and let them adjust it. An oversized unit short-cycles and stratifies; an undersized one loses the crop on the one night it mattered.
Is a second layer of film worth it if I only heat in spring?
The payback scales with degree days, so a short season stretches it. Rerun the season section with the degree days you actually accumulate between the dates you heat, rather than a full winter. On the default house a full 4,000 degree day season pays a $900 upgrade back inside one winter; a 1,200 degree day spring shoulder saves around $738 and takes something over a season. The other half of the answer is that the second layer also cuts the summer light by a few percent and adds a blower running continuously, so it is not free in either direction.