What size heater your tank actually needs

A 5-gallon betta tank in a room that drops to 68°F overnight needs 15.1 watts to hold 78°F. Nothing is sold at 15.1 watts, so the answer you buy is a 25W.

That is the answer. The rest of this page is where it comes from, because the number is only worth anything if you can check it - and because nobody else in this category publishes their working at all.

Step one: your tank does not hold what the box says

A tank sold as five gallons is five gallons of glass. Put an inch of substrate in the bottom, add hardscape, and stop filling an inch below the rim, and you are holding less.

Our assumption, stated: substrate, hardscape and the fill line take about 15% of rated volume. That is not a manufacturer figure and we are not going to pretend it is. It is a judgment, it is the same judgment everywhere on this site, and if your scape is heavier you should redo every sum below with your own number. A rated 5-gallon on our assumption holds 4.3 gallons of water.

Step two: the constant, and where we got it

You will read “three to five watts per gallon” everywhere. It is not wrong, but it is useless on its own, because it does not tell you which end of the range applies to your room. Three watts per gallon and five watts per gallon differ by two-thirds, and the difference is entirely about how cold your house gets.

So we derived a constant that includes the room. Fluval rates each of its P Series heaters to a tank capacity, and separately publishes that if the difference between ambient room temperature and the heater’s set temperature is more than 12°F / 8°C, additional heating capacity is likely necessary. That makes every rating a triple - watts, gallons, and a maximum lift - and three data points is enough to divide.

We use 0.35 watts per gallon per degree Fahrenheit - the largest of the three, not the average. Undersizing a heater is the failure that kills fish; oversizing one only makes it cycle more often. When two defensible numbers disagree, take the one whose failure mode is survivable.

Step three: the target

We size to 78°F. Aquarium Co-Op publishes that bettas prefer temperatures around 80°F and that a betta without a heater in a 67-69°F house is lethargic and stressed. The range in common use is 76 to 82. Seventy-eight sits inside both with a degree of slack either way, and a degree of slack is what stops your heater running flat out.

Step four: the sum

Watts = actual water gallons × 0.35 × (target − coldest room temperature)

Note coldest, not average. A heater that holds 78°F on a mild evening and gives up at 3am in January has not done its job, and the fish spends the winter at 73 getting slowly sicker. If you do not know how cold your room gets at night, that is the single most useful thing you could measure this week.

Worked, for the three tanks people actually own:

• 3 US gal rated, less 15% for substrate and fill line, is 2.6 gal of water. A 78F target minus a coldest-room 68F is a 10F lift. 2.6 x 0.35 x 10 = 9.1W required. Nothing is sold at 9.1W, so the answer is the 10W.
• 5 US gal rated, less 15% for substrate and fill line, is 4.3 gal of water. A 78F target minus a coldest-room 68F is a 10F lift. 4.3 x 0.35 x 10 = 15.1W required. Nothing is sold at 15.1W, so the answer is the 25W.
• 10 US gal rated, less 15% for substrate and fill line, is 8.5 gal of water. A 78F target minus a coldest-room 68F is a 10F lift. 8.5 x 0.35 x 10 = 29.7W required. Nothing is sold at 29.7W, so the answer is the 50W.

The table

Find your tank, find the coldest your room gets, buy the wattage in the last column. An asterisk means the lift you are asking for exceeds the 12°F that Fluval publishes as the point where one heater stops being enough - the recommendation still clears the arithmetic, but expect the heater to run close to continuously and consider warming the room instead.

If your answer lands between two sizes, round up. The cost of a heater that cycles more often is a few cents a year. The cost of one that cannot keep up is the fish.

What this sum does not account for

Be clear about the limits, because a method you cannot see the edges of is not a method.

• Tank shape. Heat is lost through surface area, not volume. A long shallow tank loses more than a cube of the same volume. Our constant is derived from Fluval’s ratings across three quite different heaters and averages that away rather than modeling it.
• Lids. An open-top tank loses a great deal more heat by evaporation than a lidded one. If you run open-top in a cold room, treat the asterisked rows as a warning rather than a footnote.
• Drafts and windowsills. A tank on an exterior windowsill is in a colder room than the thermostat says it is.
• Thermostat accuracy. None of this tells you whether a given heater actually holds the temperature it claims. That is the question that decides whether a heater is any good, it can only be answered by running a lot of them for a long time, and we have not done that. We say so rather than implying otherwise.

Buy a thermometer, whatever the heater says

A separate thermometer at the opposite end of the tank from the heater is five dollars and it is the only thing in this article that tells you the truth. A preset heater is a claim; a thermometer is a measurement. If they disagree, the thermometer wins.