Two numbers on a storage water heater get treated as one number, and they answer different questions. The water heater first hour rating says how many gallons the unit can hand over during its busiest hour, starting from a fully heated tank. The recovery rate says how fast it makes more once that tank has been emptied. One is a sprint with a running start. The other is the pace it can hold all morning.
Which one matters depends on the shape of your household. A family that drains the tank once and then leaves the house is buying first hour rating. A family whose second round of showers lands ninety minutes later is buying recovery, and a flattering FHR will hide a tank that cannot serve them. If you have not settled on a gallon figure yet, work through the sizing worksheet for storage tanks first, then come back and check the second column.
The short answer
- One concentrated rush, then an empty house: buy on first hour rating. A 50-gallon gas tank rated near 84 gallons covers three back-to-back showers and a dishwasher without anyone noticing.
- Showers spread over two or three hours: buy on recovery. FHR stops describing your house at the sixty-minute mark, and 20 gallons per hour against 40 is the entire argument.
- Electric only, no gas at the meter, four or more people: you are buying stored gallons to compensate for half the recovery rate. That usually means stepping from 50 up to 66 or 80 nominal.
- Hybrid heat pump tank: read the FHR for the mode you will actually leave it in. The published rating usually assumes the resistance elements are allowed to help, and heat-pump-only recovery is the slowest figure in the catalogue.
- Tankless: neither number is printed, because neither applies. The equivalent spec is gallons per minute at a stated temperature rise, and it never runs out.
What each number is actually measuring
First hour rating comes out of the federal test procedure. The tank starts fully heated, the rig draws water at a fixed rate until the outlet temperature has fallen 25 °F below where it began, the burner or the elements fire, and the sequence repeats until the hour is up. Add every gallon that left the tank in those sixty minutes and you have the FHR. It is printed on the yellow FTC EnergyGuide label stuck to the side of the box, near the top, in gallons. It is also listed model by model in the AHRI directory.
Recovery rate is not on that label at all. It lives on the manufacturer specification sheet and on the data plate riveted to the tank, usually written as “recovery at 90 °F rise, GPH”, sometimes at 100 °F rise instead. Check which rise the sheet used before comparing two brands, because the same heater looks ten percent worse at the higher figure. Recovery is also the only one of the two you can reconstruct from the input rating, which is why I trust it more.
The link between them holds in one line. First hour rating is roughly the usable volume of a full tank plus one hour of recovery on top. Capacity and recovery both sit inside it, which is why an FHR quoted on its own says nothing about hour two.
Side by side
| criteria | 50 GAL GASatmospheric, 40,000 BTU/h | 50 GAL ELECTRICresistance, 4,500 W |
|---|---|---|
| Heat actually reaching the water | 30,400 BTU/h40,000 input at about 76% recovery efficiencydouble | 15,400 BTU/h4,500 W, one element at a time |
| Recovery, 90 °F rise | 40 GPHholds this pace indefinitelywin | 20 GPHone shower every 45 minutes |
| First hour rating on the label | 84 galmost 50s land between 80 and 90win | 63 galmost 50s land between 62 and 66 |
| Reheating a tank drained to cold | ~75 min37,500 BTU of work at 30,400 an hourwin | ~2 h 30 msame work, half the heat |
| DOE draw-pattern bin for the UEF test | HighFHR of 75 gal or more, 84 gal/day test draw | MediumFHR 51 to 74 gal, 55 gal/day test draw |
| Hot water still arriving in hour three | 40 galburner keeps up with two showerswin | 20 galone shower, then wait |
| Reaching a high-80s FHR | 50 galalready there at nominal size | 80 galbuy volume to replace missing recoverythe usual fix |
One rush and out the door? First hour rating settles it, and either fuel can be specced to cover you. A second wave ninety minutes later? Recovery settles it, and at identical capacity the gas tank is twice the appliance.
The arithmetic that links the two
Water takes 8.34 BTU to lift one gallon by one degree Fahrenheit. So gallons per hour equals the heat reaching the water divided by 8.34 times the temperature rise. A 40,000 BTU/h gas burner at a recovery efficiency of 0.76 puts 30,400 BTU into the water each hour. Divide that by 8.34 times 90 and you get 40.5 gallons per hour. A 4,500 watt element is 15,354 BTU/h before losses and almost all of it lands in the water, so the same division gives 20.5. That is the whole reason two tanks of identical volume behave nothing alike.
Now stack the capacity on top. A 50-gallon tank surrenders roughly 45 usable gallons before the outlet temperature drops out of the test window, because cold inlet water mixes in as the hot leaves. Add one hour of recovery: 45 plus 40 is 85 on gas, and published labels for 50-gallon gas tanks sit between 80 and 90. On electric, 45 plus 20 is 65, and labels sit between 62 and 66. The sum is an approximation, since thermostat behaviour and stratification move it a few gallons either way. It is close enough to sanity-check any spec sheet in the aisle.
Turn that into something you can picture. A ten-minute shower at a mixed 2 gallons a minute pulls roughly 15 gallons of stored hot water. The gas tank replaces it in about 22 minutes. The electric tank needs 45. One supports a rolling queue of showers, the other does not.
Why the same fifty gallons recovers at two different speeds
A gas burner under a residential tank is a 30,000 to 40,000 BTU/h device. A standard electric tank carries two 4,500 watt elements and a thermostat interlock that lets only one run at a time, so it is a 15,400 BTU/h device whichever element is calling. Raw input differs by about two and a half times. After recovery efficiency, which flatters electric because resistance heat has nowhere to escape, delivered heat differs by roughly two. Two is what shows up in the recovery column.
Some electric models can be configured to fire both elements together, which roughly doubles recovery and roughly doubles the electrical demand. That is a question for whoever is responsible for your panel, and this site does not size circuits. The point is that 20 GPH is a design decision rather than a law of physics, and it is the decision behind most of the 80-gallon electric tanks sold here. The 40, 50 and 80 gallon comparison is really a question of how much recovery a household is willing to buy as stored volume.
First hour rating quietly changes the efficiency number you are comparing
This one catches people out. Under the federal uniform test procedure a storage heater is assigned a draw pattern by its first hour rating: very small below 18 gallons, low from 18 to 50, medium from 51 to 74, high at 75 and above. Each pattern uses a different daily draw volume, roughly 10, 38, 55 and 84 gallons. The UEF printed on a unit is what it achieved under its own pattern.
Two heaters with the same UEF were therefore not necessarily measured against the same workload. A tank in the medium bin was asked for 55 gallons a day; a tank in the high bin was asked for 84. Standby losses spread over a larger draw look smaller. The 50-gallon pair in the table above already straddles that line: the gas tank sits in the high bin at 84 gallons, the electric sits in medium at 63. Across two fuels that barely matters, because 0.65 against 0.92 is a gap no draw pattern can explain. Within one fuel it matters a great deal, and a 50-gallon electric held up against an 80-gallon electric is exactly that case, so read the draw pattern out of the AHRI listing before you make anything of a few hundredths of UEF. This site’s model assumes 64 gallons a day, which lands between the medium and high patterns, and I would rather say so than pretend the label was built for your house.
What the first hour rating cannot tell you
The honest limitation is that FHR is measured against a laboratory draw sequence, and that sequence is not your morning. The rig draws at a steady rate with defined pauses. Your house runs two overlapping showers, a shave, a dishwasher on the hot fill and a washing machine somebody started without telling anyone. FHR is a repeatable comparison between products, not a prediction about a household. Read 84 gallons as “outruns that 63-gallon tank by about a third”, never as “we get 84 gallons”.
Recovery carries its own asterisk. It scales inversely with the rise, so a tank listed at 40 GPH on a 90 °F rise delivers about 45 on an 80 °F rise and about 36 if the sheet quoted 100. That never changes a purchase. It does explain why the tank felt different in July.
One more thing worth pricing before you solve a recovery problem with gallons. Running cost tracks the fuel and the efficiency, not the size of the vessel. Take the site’s delivered-heat anchor of 4,000 kWh a year, divide by a resistance tank’s UEF of 0.92, and you buy 4,348 kWh, or $739 a year. Divide the same heat by a heat pump’s 3.5 and you buy 1,143 kWh, or $194. That $545 a year is the largest running-cost gap in the house, and it is the reason to read a hybrid’s first hour rating carefully rather than write off the category. The heat pump against resistance comparison covers where the gap holds.
I keep meeting people who were sold an 80-gallon electric tank on a rule of thumb and now pay to keep 80 gallons hot for a household of two. I would not buy stored volume to fix a problem a recovery number already describes to the gallon. Decide which of the two questions your household actually asks. Two houses with the same number of people in them can sit a long way apart on that question. Once capacity is settled the badge matters less than it looks, which the Rheem and A.O. Smith comparison explains.
Sources
- FTC EnergyGuide label for storage water heaters, which prints first hour rating in gallons.
- DOE Energy Saver, sizing guidance and the definition of first hour rating.
- DOE uniform efficiency descriptor test procedure, for the four draw patterns and the FHR thresholds that assign them.
- AHRI Directory of Certified Product Performance, for per-model UEF, FHR and draw pattern.
- Manufacturer specification sheets and data plates, for recovery in GPH at a stated rise, burner input and element wattage.
- US EIA residential energy prices, behind the $1.50 per therm and $0.17 per kWh used above.
Every figure above comes from a published rating or from arithmetic you can repeat with a calculator, and where a number rests on an assumption the assumption is printed beside it. The method page sets out how the running-cost models are built and which constants they hold fixed.