If you only do one thing: before you shop for a bigger water heater, wait until April and see whether the problem is still there. In most of these houses it will not be, and you will have saved yourself several thousand dollars.
January and February are our busiest months for water heater calls, and a good share of those calls are a water heater that is working correctly. The complaint is always some version of the same thing: it used to be fine, and now the second person in the shower gets cold water.
The water heater did not change. The water going into it did.
Where your cold water comes from, and what it does in January
Most of Northern Virginia drinks the Potomac. The USGS has a gauge on it at Point of Rocks, upstream of the intakes that feed Fairfax Water’s Corbalis plant and Loudoun Water, and it has been logging water temperature every day for years.
The July average is about 83 degrees. The January average is about 37.
That is not what arrives at your tap, because the water warms up traveling through buried mains toward the temperature of the soil. The standard model the Department of Energy uses for this, run against thirty years of Dulles weather normals, puts the water actually entering a house here in the low 70s at the end of summer and right around 50 in early February.
Call it the mid 40s to low 50s in January against the low 70s in August, and note that during a long draw like a shower you are pulling fresh water through the main faster than it can warm up, so the colder end of that is the honest one.
Your water heater now has to do meaningfully more work, with the same burner, to fill the same tank.
The arithmetic, which is the whole article
A 105 degree shower is not 105 degree water from your tank. It is 120 degree tank water blended with cold water at the valve. So the colder the cold is, the more tank water each minute of shower consumes.
The fraction of your shower that has to come from the tank is the difference between the shower and the cold, divided by the difference between the tank and the cold. On a 50 gallon tank set to 120, for a 105 degree shower:
| Cold water coming in | Share that must come from the tank | Gallons of shower you get |
|---|---|---|
| 70 degrees (August) | 70% | 71 gallons |
| 50 degrees (a normal January) | 79% | 64 gallons |
| 45 degrees (a cold snap) | 80% | 62 gallons |
That is 12.5 percent less shower out of the identical tank, at the identical setting, with nothing wrong. At a normal 2 gallon per minute showerhead it is about 36 minutes of hot water in August and 31 in January.
One honest caveat: you never actually get every gallon at full temperature, because cold water mixing in at the top of the tank as it refills drags the outlet temperature down before the tank is empty. So the real numbers are lower than the table. The summer to winter ratio is not.
The half that actually causes the cold shower
Five minutes lost is annoying. It is not what generates the phone call. The recovery is.
Heating 50 gallons from 70 degrees up to 120 takes about 20,800 BTU. Heating the same 50 gallons from 45 degrees takes about 31,200. Exactly 50 percent more energy, for the same tankful.
Your burner does not get bigger in the winter. So a common 40,000 BTU gas water heater that can regenerate about 104 gallons of usable shower water an hour in August manages about 61 in a cold snap. That is a 42 percent drop in how fast it comes back. Electric is worse, because a single 4,500 watt element has far less to work with than a gas burner.
Now the household version. Person one showers and notices nothing. Twenty minutes later, person two gets what the tank has had time to recover, and in August that was plenty and in January it is not. Nothing broke. The queue just got longer.
Why the number on your label did not warn you
This is the part we have never seen anybody write down, and it is worth a paragraph.
Every storage water heater carries a first hour rating on its EnergyGuide label, which is the federal estimate of how much hot water it can deliver in an hour starting full. The federal test procedure specifies the conditions: supply water held at 58 degrees, delivering at 125.
Run the blending arithmetic at those label conditions and you get a hot fraction of 0.701. Run it for a Northern Virginia August, 70 degree inlet and a 120 degree setpoint, and you get 0.700.
The same number to three decimal places. For shower purposes, the rating on your water heater is a summer rating. Nothing on that label is wrong or dishonest. It was simply never measured in January, and the shortfall you are experiencing is the gap between a test condition and a Virginia winter.
Tankless owners: this is bigger for you, not smaller
A tankless unit has no reservoir. It has a fixed burner and it spends that heat raising water from whatever comes in to whatever you asked for. Colder in means a bigger lift means fewer gallons per minute. There is no tank of already-hot water to hide the difference.
Take Rinnai’s own published table for one of their bigger units. Same box, same gas, only the incoming water changed:
| Condition | Temperature rise | Gallons per minute |
|---|---|---|
| The brochure number | 35 degrees | 11.1 |
| Northern Virginia, August | 50 degrees | 7.8 |
| Northern Virginia, January | 70 degrees | 5.5 |
| Cold snap | 75 degrees | 5.2 |
Roughly a third less than August, and less than half the number on the brochure. Rheem is quietly candid about this and labels its own 35 degree row “MAX,” with a footnote saying flow rates vary with incoming water temperature and set temperature. Older non-condensing units are worse still, landing closer to 4.5 gallons a minute in a January.
Practical translation: two showers and a dishwasher is comfortable in August and marginal in January, from a unit that has not changed. And there is a second trap. When the water feels weak people push the tap further toward hot, which cuts the total flow through the unit, and if it drops near the minimum the burner needs the unit modulates oddly or drops out. That is the cold water sandwich, and it gets worse in winter for exactly this reason.
When it really is the water heater
Everything above is a water heater working correctly in the cold. Here is how to tell it apart, and this list is not ours: it is close to Bradford White’s own troubleshooting table, which lists “very cold inlet water to heater” as a named cause of not enough hot water right alongside the genuine failures.
It got worse suddenly rather than gradually. Weather does not change overnight in October. Parts do.
It is also bad in July. If you had this complaint in August, the cold water is not your answer.
Electric tank, and you get a short burst of hot then cold that never comes back. Classic failed lower heating element. The upper element heats the top of the tank and that is all you get.
Lukewarm everywhere, all the time, at every tap. Suspect a broken dip tube, the plastic tube that carries incoming cold water to the bottom of the tank. When it breaks, cold water dumps in at the top and mixes straight into the outlet. There was a genuinely defective run of them in heaters built between August 1993 and October 1996, which was settled in a class action rather than a recall, so nobody ever contacted the owners. If your heater is that old you have a bigger problem than the dip tube.
Rumbling or popping from the tank. Sediment on the bottom, boiling water trapped underneath it. Worth flushing and worth knowing that its effect on efficiency is smaller than the trade generally claims.
Tankless that was fine last winter and is not this one. Scale. Hard water here builds up in the heat exchanger and steals capacity from a unit that was already tight in January. A descale runs $79 to $365, most around $100 to $194, and it is the cheapest thing on this page.
And the one almost nobody checks: a crossover. If cold water is leaking into your hot water piping somewhere, you are diluting the hot at every tap in the house and no water heater on earth will fix it. Rheem’s own service bulletin says outright that a plumbing crossover “can cause a complaint such as ‘water not hot enough’,” and the usual culprits are a worn cartridge in a single-handle faucet, a washing machine valve, a failed mixing valve, or a recirculation line with a bad check valve.
The crossover test, because you can do it yourself
This is Rheem’s procedure and it takes ten minutes.
1. Close the cold water shutoff valve going into the water heater.
2. Open only the hot side of a two-handle faucet, ideally the highest one in the house. If you have a recirculation pump, turn it off first.
3. The water should stop after the pressure bleeds off. If it does, you have no crossover.
4. If it keeps running, feel it. If it keeps running and turns cold, you have a crossover. If it keeps running and stays hot, the shutoff valve you just closed is leaking through, and the test tells you nothing until that valve is replaced. That second branch is the trap, and it is why people get false answers off this test.
5. To find it, shut the hot and cold supply valves to the washing machine first, because its mixing valve is the most common offender and the easiest to isolate. If the flow stops, that is your answer.
6. If not, feel the two supply lines under each single-handle faucet. If both are equally cold on one of them, that faucet is letting cold across and needs a cartridge or a rebuild.
One more thing Rheem flags that nobody else does: in a house with only single-handle faucets, run the whole test twice using two different faucets, because the faucet you used for the test might be the one causing the problem.
What to do about it if nothing is broken
Stagger the showers. Unromantic, free, and it is the actual answer for most families. Give the tank forty minutes between people instead of twenty.
Run the dishwasher and the laundry at night.
Put in a low-flow showerhead, which stretches the same gallons across more minutes.
Turn the thermostat up, carefully. Storing hotter and mixing down at the fixtures buys back a meaningful amount of usable capacity, on the order of a quarter or more. It is a real tradeoff, so here are both sides honestly. Hotter water scalds faster: the CPSC puts 150 degree water at about two seconds to a serious burn, which is why 120 is the standard recommendation and why this is a bad idea in a house with small children or older adults unless you fit thermostatic mixing valves. And hotter storage scales faster; one national lab study puts a tank at 140 degrees at roughly four years to significant scaling against something closer to thirty at 120. Talk to somebody before you do it.
Do not buy a bigger water heater in January. That is the whole point of this page. If the tank is genuinely undersized for the house it will still be undersized in June, and you will know because the problem will still be there. If it disappears in April, the tank was fine and you would have spent the money on the weather.
If it turns out you do need one, our installed ranges are $2,000 to $4,800 for a tank and $5,500 to $7,500 for tankless or a hybrid, and the water heater calculator will narrow it for your house. Tank versus tankless, honestly, is here.
Do this week
If the hot water got worse this month, write down the date, and write down whether it was also a problem last July. Those two facts answer this question better than a service call will.
Then do the crossover test, because it is free, it takes ten minutes, and when it finds something it is usually a faucet cartridge rather than a water heater.
This is one of ten in the Cold Season Field Guide. If you have no hot water at all rather than not enough, start here instead.
SwiftPro does water heater service and replacement across Northern Virginia. (703) 997-9757, or book a time online.
