Troubleshooting
Why Your Tankless Water Heater Runs Cold in Winter
Your tankless unit did not get worse in December. The water got colder, and flow rate and temperature rise trade against each other. Here is the arithmetic, and the three real causes people mistake for it.
11 min readUpdated
Quick Answer
Why is my tankless water heater not hot enough in winter?
Because incoming water in North Texas drops to about 50-55F in winter, forcing your unit to add roughly 55 degrees instead of 25. Flow rate and temperature rise trade against each other, so the same unit that delivered 9 GPM in July delivers about 5.5 GPM in January. Frisco Tankless Water Heater Pros diagnoses winter output problems at (972) 430-9024.
- North Texas winter inlet temperature: roughly 50-55F
- Summer inlet temperature here: roughly 78-82F
- GPM = BTU input x efficiency / (temperature rise x 500)
- A 199,000 BTU condensing unit: ~9 GPM at 25F rise, ~5.5 GPM at 55F rise
- Scale, a clogged inlet screen and an undersized gas line all mimic the same symptom
- Correct winter sizing is the permanent fix; setpoint changes only shift the trade-off
The formula that explains the whole problem
There is one equation behind every winter tankless complaint in North Texas, and once you have seen it the behavior stops being mysterious. Flow rate in gallons per minute equals burner input in BTU per hour times thermal efficiency, divided by temperature rise in degrees Fahrenheit times 500.
- GPM = (BTU input x efficiency) / (temperature rise x 500)
- The 500 is a constant derived from water's specific heat: raising one gallon per minute by one degree Fahrenheit for an hour requires about 500 BTU. Everything else in the equation is a property of your unit or your weather.
Look at what is fixed and what is not. Burner input is fixed by the unit you bought. Efficiency is fixed by its design. The 500 is physics. The only variable that moves through the year is temperature rise — and it is in the denominator. Double the rise and you halve the flow. That is not a defect; that is the machine working exactly as designed.
| Month | Inlet temp | Rise to 105F | Available flow |
|---|---|---|---|
| July | 80F | 25F | About 14.8 GPM (limited by unit max) |
| September | 74F | 31F | About 11.9 GPM |
| November | 62F | 43F | About 8.6 GPM |
| January | 52F | 53F | About 7.0 GPM |
| February, cold snap | 48F | 57F | About 6.5 GPM |
What North Texas inlet water actually does through the year
Frisco's supply comes largely from surface reservoirs rather than deep wells, which matters: surface water tracks air temperature far more closely than groundwater does. A deep well might vary five degrees across a year. Reservoir-fed municipal supply here swings close to thirty.
- Peak summer, July through early September: roughly 78-82F at the meter
- Shoulder seasons, May and October: roughly 65-72F
- Late fall and early spring: roughly 58-64F
- Deep winter, January and February: roughly 50-55F
- Extended hard freeze events: can dip to 45-48F for several days
The distance the pipe travels through cold soil matters too. A home at the end of a long service run on a north-facing lot will see colder water than a home nearer the main. This is why two neighbors with identical units can have genuinely different winter experiences and both be telling the truth.
Diagnosing it yourself in ten minutes
Before assuming the unit is faulty, measure. Three numbers separate a correctly functioning unit doing physics from a unit that has an actual problem.
Measure inlet temperature
Run a cold-only tap for two minutes to clear the line, then hold a thermometer in the stream. Anything from 48 to 58F is normal for a Frisco winter.
Measure outlet temperature at the fixture
Run one hot fixture only, wide open, for two minutes. Measure the stream. If your setpoint is 120F and you are reading 118-122F at the tap with one fixture running, the unit is producing correctly.
Measure actual flow
Time how long the same fixture takes to fill a one-gallon container. Sixty seconds is 1 GPM; thirty seconds is 2 GPM. Repeat at each fixture and add the ones you actually run together.
Compare against the rise chart
Take your measured simultaneous flow and your measured rise, and check them against the unit's published flow chart. If your demand exceeds the chart, the unit is undersized for winter and nothing is broken. If your demand is well under the chart and the temperature still sags, something is wrong.
Cause one: the unit was sized for summer
This is the most common cause and it is an installation error rather than a failure. A unit selected against the headline flow rate on the box — those figures assume a 35-degree rise or less — will be roughly 40 percent short of expectation in a Frisco January.
| Unit BTU input | Advertised max GPM | Real GPM at 55F rise | Realistic winter use |
|---|---|---|---|
| 140,000 | About 7.0 | About 3.8 - 4.2 | One shower plus a sink |
| 160,000 | About 8.0 | About 4.4 - 4.8 | One shower comfortably |
| 180,000 | About 9.0 | About 5.0 - 5.4 | Two modest showers |
| 199,000 | About 9.8 | About 5.5 - 6.0 | Two showers comfortably |
| Two 199,000 manifolded | About 19.6 | About 11 - 12 | Four showers, or a soaking tub plus showers |
If your household genuinely needs more than one unit delivers at a 55-degree rise, the honest fixes are a second unit manifolded to the first, or a buffer tank on the outlet to absorb short peaks. Setpoint tricks and flow restrictors are palliatives, not solutions.
Cause two: scale on the heat exchanger
A scaled exchanger produces exactly the winter symptom people attribute to sizing, and it is the reason a unit that was fine last January is not fine this January. Nothing about the weather changed; the exchanger did.
Scale from Frisco's 8-11 grain water insulates the water side of the exchanger. The burner fires at full rate, but less of that heat reaches the water. Simultaneously, scale narrows the passage, so flow through the unit drops. Both effects push the same direction, and both are worst in winter when the unit is already working hardest.
- Rumbling or kettling while the unit fires — steam pockets forming under scale
- Outlet temperature that oscillates during a long draw instead of holding steady
- Flow noticeably lower than it was when new, measured at the same fixture
- Occasional overheat or flow-sensor codes that clear on a power cycle
- Gas usage trending upward with no change in household habits
Cause three: the gas line cannot feed full fire
This one hides beautifully because the unit is fine and the plumbing is fine. The problem is that the burner cannot reach its rated input, so every number in the flow equation collapses. It shows up specifically in winter because that is the only time the unit is asked to run at full fire for extended periods.
It also shows up specifically on cold mornings, which is the tell. On a January morning the furnace is running, and if the furnace and the water heater share a marginal gas branch, the water heater loses. Manifold pressure sags below the minimum the unit needs and it either derates or faults out.
| Observation | Suggests |
|---|---|
| Performs fine at night, poorly on cold mornings | Furnace and water heater competing for gas |
| Worse when the gas range or a gas dryer is running | Undersized shared branch or manifold |
| Ignition or flame-failure codes only at full demand | Manifold pressure dropping below minimum |
| Fine at half flow, faults at full flow | Pipe cannot carry rated BTU at the developed length |
| Started after adding a pool heater or generator | Total connected load now exceeds meter capacity |
The measurement that settles it is manifold pressure at the unit while it runs at full fire with the furnace also firing. If it drops below the manufacturer's minimum, the pipe or the meter is the problem, and no amount of work on the water heater will fix it.
Cause four: a clogged inlet screen or a failing sensor
The cheapest fix on the list and the one most often overlooked. Every tankless unit has a fine mesh screen on the cold water inlet. It catches sediment, pipe scale and construction debris, and when it clogs it restricts flow into the unit — which the flow sensor reads as low demand, which causes the unit to modulate down or refuse to fire at all on small draws.
- Shut off the cold isolation valve and relieve pressure at a hot fixture
- Remove the inlet screen — usually a small plug or union at the cold connection
- Rinse it in a vinegar solution and brush off scale and debris
- Reinstall, restore water, and purge air from the hot lines before firing the unit
The other cheap failure is a drifting outlet thermistor. If the sensor reads high, the unit throttles the burner because it believes it has already reached setpoint. The symptom is water that is consistently 8 to 15 degrees below the display value, at all flow rates, in all seasons — which is how you distinguish it from a rise problem.
Fixes that genuinely work, ranked by cost
Clean the inlet screen — under $50, sometimes free
Always do this first. A restricted screen mimics almost every other fault on this list and takes fifteen minutes.
Descale the heat exchanger — $189 to $349
If it has been more than a year, this is the highest-value step. In Frisco water it frequently restores several tenths of a GPM and eliminates temperature instability outright.
Raise the setpoint from 120F to 125-130F — free
This does not create capacity, but it buys usable comfort: more cold water gets mixed at the valve, so a given comfortable shower consumes less hot flow. Do not exceed 140F, and be aware of scald risk and accelerated scaling at higher setpoints.
Install lower-flow showerheads — $40 to $150
Dropping two showers from 2.5 to 1.75 GPM cuts simultaneous demand by 1.5 GPM. In a household that is 1 GPM short in January, this alone can close the gap.
Correct the gas supply — $400 to $1,500
If manifold pressure sags at full fire, upsizing the branch is the only real fix, and it usually resolves several apparently unrelated complaints at once.
Add a second unit or a buffer tank — $2,000 to $4,500
The correct answer when peak simultaneous demand genuinely exceeds one unit's winter capacity. Manifolded units share load and give redundancy; a buffer tank smooths short peaks more cheaply but does not raise sustained capacity.
What about the cold water sandwich?
- Cold water sandwich
- A slug of cool water arriving mid-draw, caused by hot water left in the pipe from a previous use reaching the fixture first, followed by unheated water that entered while the burner was ramping up, followed by properly heated water.
This is a distinct problem from winter capacity, though winter makes it more noticeable because the cool slug is colder. It is characteristic of short, repeated draws — someone washes their hands, then two minutes later someone starts a shower.
- A small buffer tank on the outlet, typically 2 to 6 gallons, absorbs the slug almost completely
- A recirculation loop keeps the branch line hot and largely eliminates it
- Modern units with faster ignition and better modulation reduce it substantially compared with older models
- It is unrelated to unit sizing — a larger unit does not fix it
If it is still cold after all of that
A unit that is descaled, has a clean inlet screen, adequate gas pressure at full fire, and still cannot hold setpoint at moderate flow has an internal fault — a failing exchanger, a drifting sensor, a gas valve not opening fully, or a control board that has lost its calibration. That is a diagnostic visit, not a checklist.
Frisco Tankless Water Heater Pros can measure inlet and outlet temperature, actual flow, manifold pressure under load and combustion characteristics, and tell you whether you are looking at a repair, a maintenance issue or a sizing decision. Call (972) 430-9024.