Storage is easy, delivery is the problem
Water stores 1 BTU per pound per degree F, and a gallon weighs 8.34 pounds, so every gallon holds 8.34 BTU for each degree it rises or falls. That is the best figure of any material anybody is likely to put in a greenhouse — concrete is around 0.2 BTU per pound per degree and stone is similar, so a cubic foot of water outstores a cubic foot of masonry by a factor of roughly two and a half.
Twenty 55-gallon drums hold 1,100 gallons. Through a 15 degree swing that is 1,100 times 8.34 times 15, or 137,610 BTU, which is a real quantity of heat and comparable to a couple of gallons of propane. Fill them only to 95 percent to leave expansion room and it drops to 1,045 gallons and 130,729 BTU, which is the honest version anywhere the water might freeze.
Then look at how it gets out. Heat leaves a container across its surface at a rate of roughly the surface coefficient times the area times the temperature difference between the water and the air. A 55-gallon drum standing upright has about 20 square feet of surface air can reach — sides plus the top, since the bottom is against the floor. Twenty of them is 404 square feet. At a coefficient of 1.2 and a 10 degree difference between water and air, that is 4,848 BTU an hour.
| Same 1,100 gallons, three shapes | Stored over 15°F | Surface | Release at 10°F difference | Against a 20,000 BTU/hr loss |
|---|---|---|---|---|
| 20 x 55-gal drums | 137,610 BTU | 404 sq ft | 4,848 BTU/hr | 24% |
| 4 x 275-gal totes | 137,610 BTU | 278 sq ft | 3,336 BTU/hr | 17% |
| 220 x 5-gal buckets | 137,610 BTU | 990 sq ft | 11,880 BTU/hr | 59% |
All three rows hold the same heat. They deliver it at rates that differ by a factor of three and a half, purely because of surface area per gallon. This is the single most useful thing to know before buying containers, and it is why the shape of the store matters more than its size.
The swing is smaller than people expect
Fifteen degrees of overnight swing is on the optimistic side for barrels that are not being deliberately charged. To take 1,045 gallons up 15 degrees requires putting 130,730 BTU into them during the day, and that only happens if the sun is actually landing on the containers. Barrels along a shaded north wall, or tucked under a bench, charge mostly from warm air, which is a slow process across the same limited surface that limits the discharge.
The honest way to find your swing is a thermometer taped to a drum and read at dusk and at dawn for a week. Most people who do this find something between four and ten degrees, which is a third to two thirds of what they assumed, and the stored heat scales directly with it.
What the mass is actually good for
Not as a heater substitute. As a peak shaver, it is genuinely useful and the numbers support it. A house that would have hit 105 in the afternoon and 33 at dawn instead hits 99 and 37, and the second of those numbers is the one that decides whether a crop of transplants survives an early April night. Frost protection in the shoulder seasons is where mass earns its floor space.
It also flattens the rate of change, which matters for anything sensitive to a fast drop. And it costs nothing to run, does not fail in a power cut, and does not produce carbon monoxide, which are three advantages no heater has.
For the load figure that goes in the form, run the house through the greenhouse heat loss calculator at the night conditions you care about. If you are trying to decide between mass and a second glazing layer, the comparison is not close in a cold climate — halving the loss helps every hour of every night, and the mass helps for a few hours after a sunny day. If you are trying to decide between mass and nothing, in the shoulder season, mass wins easily.
Questions people ask
How many barrels do I need per square foot of greenhouse?
There is no such ratio that survives contact with a real calculation, which is why this page asks for your heat loss instead. The number depends on the loss rate of your envelope, the length of your night, the temperature difference you are trying to hold, and above all on how much surface area the containers present. Two greenhouses with the same floor area and different glazing can differ threefold in what they need. Put your own heat loss in and read the coverage percentage; that is the only version of this answer worth having.
Are black barrels better than white ones?
For charging, yes, where direct sun actually reaches them, since a dark surface absorbs more of the light landing on it. For discharging at night the colour makes very little difference to convection and only a modest one to radiation. The larger effect by far is position: a barrel in direct winter sun along a north wall charges enormously better than the same barrel in shade, regardless of colour. It is also worth remembering that a barrel in the sun is a barrel casting a shadow on something, and that shadow has a cost.
Is water better than a concrete floor or a stone wall?
Per pound, water stores about five times what concrete does, and per cubic foot about two and a half times. Water also moves heat around inside itself by convection, so the whole volume participates, whereas heat has to conduct slowly through the depth of a masonry mass and only the outer few inches take part on a daily cycle. Where masonry wins is that it is already there, takes no floor space of its own, and does not leak. A concrete floor in a greenhouse is doing useful thermal work whether or not anybody counted it.
Will thermal mass keep my greenhouse above freezing on the coldest night?
On the numbers here, on a house losing 20,000 BTU an hour, twenty drums cover about 57 percent of a twelve hour night on stored energy and can deliver only about a quarter of the required rate at a 10 degree water-to-air difference. So no, not on a genuinely cold night. On a light frost night in April, where the loss is a fraction of that and the drums went into the night warm after a sunny day, it can easily be the difference between 34 and 30 degrees. Those are two different questions and they get very different answers.
Should I circulate air across the barrels with a fan?
It raises the surface heat transfer coefficient, which is exactly the constraint the third section of this calculator identifies, so in principle yes and the effect can be substantial — moving air across a surface can multiply the coefficient several times over compared with still air. Rerun the calculation with a higher coefficient to see what your own setup gains. The trade is that the fan runs on electricity all night and a horizontal airflow fan is doing other useful things anyway, so it rarely comes down to the barrels alone. It also does nothing for the stored quantity, only the rate.