I've been worried that, although my koi pond goes below the exert-advice limit of 50 degrees for feeding, it may not get cold enough to keep their metabolism near zero for the five months it will be until it warms again. So far, the pond has been averaging about 44 degrees, nowhere near freezing, and one of the three has died. (He was by far the least aggressive eater.)
So I bought a 500w aquarium heater, hoping that a pair of those might be enough to keep the water 10 degrees warmer than it would be otherwise, allowing me to feed them again. I did some math based on heat loss tables for water surfaces, and it seemed reasonable. But I hadn't yet done the math on the speed of reaching equilibrium, so I've been checking the temp every 2 hours. (It's a 4500 gallon pond, half covered by the screened-in patio, and the surface area is 144 sf.)
If I've done the math right, it will take almost exactly 24 hours for the heater (drawing 438 watts right now) to change the temp 1 degree F from equilibrium...and longer for the following degrees as it approaches equilibrium Guess I don't have to worry about thermal shock.
For once I did a science thing in british units, since I was measuring temp in F and had the volume in gallons. 1 gallon is 8 pounds, so it's 8 BTU to move one gallon one degree F, and all I needed was BTU per watt hour. Felt so weird and archaic doing it that way. 3.4 BTU per watt hour in case you're wondering.