Signs of Scale Inside a Tankless Water Heater

Scale in tankless water heater systems announces itself long before it fails the unit. The first sign is not a leak or an error code — it is a widening gap between the temperature you set and the temperature that arrives, worst at low flow. By the time a code appears, the heat exchanger has usually been coating for a year or more.

That gap is measurable with a thermometer and thirty seconds a month. This article covers what to watch, what your manufacturer actually publishes as a hardness limit, and how to turn that number into a flush interval.

The Five Signs, in the Order They Appear

Tankless units scale from the inside of a narrow copper or stainless heat exchanger outward. Because the passages are small by design, a deposit that would be invisible in a storage tank changes behaviour quickly. The symptoms arrive in a predictable sequence.

1. Longer time-to-hot at the tap. The first and least noticed. Scale is an insulator between the burner and the water, so the exchanger takes longer to bring the passing stream up to setpoint. You will notice it as an extra two or three seconds at the kitchen sink before anyone calls it a problem.

2. Temperature swings at low flow. A scaled exchanger holds temperature well at 3 gallons per minute and badly at 0.7. Bathroom faucets and trickling shower valves go lukewarm-hot-lukewarm while the tub runs perfectly. This is the signature symptom, and it is the one homeowners misattribute to the "cold water sandwich" that all tankless units produce on restart.

3. Reduced hot flow with normal cold flow. Scale narrows the exchanger's internal channels. If the hot side of a fixture is visibly weaker than the cold side across the whole house, the restriction is at the heater. This differs from whole-house pressure loss from scale in the supply piping, which affects hot and cold together.

4. Noise during firing. Popping, crackling or a low rumble as the burner runs. Water trapped underneath a scale layer flash-boils into steam bubbles that collapse against the metal. It is the same acoustic mechanism as a sediment-lined storage tank, in a component that costs a great deal more.

5. Error codes. Overheat and flow-rate faults. On Rinnai equipment the LC or LC0 code is an explicit scale warning; other brands throw generic overheat and ignition faults. This is the last stage, not the first, and the exchanger is usually well coated by the time it appears.

Manufacturer Hardness Limits and What the Warranty Says

Most articles on this topic assert that hard water "voids your warranty" without quoting anything. The manufacturers do publish numbers, they differ, and the differences matter. Here is what four major brands put in writing.

Three things follow from that table.

The clustering around 200 mg/L is not a coincidence. Roughly 11.7 to 12 grains per gallon is where the industry has settled on "this exchanger will not survive untreated." If your measured hardness is above that, you are outside the design envelope of most residential tankless equipment.

The exclusion targets the heat exchanger specifically. Nobody voids an entire warranty over water chemistry. They decline the one claim that represents most of the unit's replacement cost. A control board failure will still be covered; the exchanger will not.

Hardness is one row in a longer table. Rheem's water quality specification also limits TDS and pH; Rinnai's covers iron and manganese. Well owners in particular should read the whole table rather than the hardness line, because iron and manganese have their own limits and their own treatment that a softener alone may not meet.

You can pull your own model's numbers from the manufacturer directly — Rinnai publishes its warranty documents and Navien hosts its installation manuals — and the figure that governs your claim is the one in your model's manual, not the one in this table.

The Delta-T Log: Catching Scale in a Tankless Water Heater Early

This is the part almost nobody does, and it is the reason most tankless units are diagnosed late.

Delta-T is simply the temperature rise across the unit: outlet temperature minus inlet temperature. A clean exchanger transfers heat efficiently and holds a consistent rise at a given flow. A scaling exchanger loses that consistency at low flow first, because there is less thermal mass moving through to buffer the insulating layer.

What you need: an instant-read kitchen thermometer, a one-gallon jug, and a stopwatch.

The monthly routine, five minutes:

1. Note the incoming water temperature. If there is a cold tap upstream of the heater, use that; otherwise run the cold at the fixture nearest where the service line enters and let it stabilise for a minute.
2. Set the unit to a fixed temperature and leave it there permanently. 120°F is the usual choice.
3. Open a hot tap to a low flow — a bathroom faucet about a third open. Let it run 45 seconds, then measure the temperature in a cup under the stream.
4. Measure the flow at that same setting: time how long the stream takes to fill a one-gallon jug, then divide 60 by the seconds to get gallons per minute.
5. Write down four numbers: date, inlet temp, outlet temp, GPM.

Reading the log. What matters is not any single reading — it is the trend at comparable flow. On a clean unit, the outlet temperature at 0.8 GPM sits within a couple of degrees of setpoint month after month. Scale shows up as a slow, one-directional drift: the low-flow outlet temperature falls, or becomes unstable across the 45 seconds, while the high-flow reading still looks fine.

A drop of 4 to 5°F at low flow, sustained across two consecutive months, is worth a flush. That typically arrives six to twelve months before an error code does, which is the entire value of keeping the log.

Two things that will corrupt your readings: taking measurements at different flow rates and comparing them, and seasonal inlet temperature change. Winter groundwater can be 25°F colder than summer, which pushes the unit harder and changes the achievable rise. Always compare against the same month last year, or record the inlet temperature and compare deltas rather than outlet temperatures.

Flush Interval by Measured Hardness

The manufacturer baseline is annual. Your water decides whether that is generous or optimistic.

The awkward truth in the bottom half of that table: at high hardness, flushing is not maintenance, it is triage. A six-month flush cycle costs roughly what a year of softener salt does, and the exchanger is still accumulating deposit between flushes in the places circulation reaches least.

If you are in that band, the decision is not "flush or soften" — it is whether the tankless unit was the right choice for this water at all. Whether your hardness number justifies whole-house treatment is worth working out before you spend another year flushing.

What a Flush Cannot Fix

Descaling circulates a mild acid — white vinegar or a proprietary solution — through the exchanger for 45 to 60 minutes. It dissolves carbonate deposit that the solution can physically reach. That is a real and worthwhile intervention, and it has limits.

It does not reach dead legs. Circulation follows the path of least resistance. Passages that are already substantially blocked receive the least flow and therefore the least solution, which is the opposite of what you want.

It does not restore a pitted exchanger. Where scale has caused sustained local overheating, the metal underneath has been thermally stressed. Removing the deposit reveals the damage rather than repairing it.

It does not clean the inlet filter screen. Every tankless unit has one, most owners have never seen it, and a partly blocked screen produces the same low-flow symptoms as a scaled exchanger. Check it before you conclude anything — it is a two-minute job and it is a common false positive.

It does not change your water. A flushed exchanger on 18-grain water starts re-scaling the moment it comes back online, at exactly the rate it did before.

For a well-maintained unit on moderate water, the payoff is genuine. ENERGY STAR's data on tankless gas water heaters rests on the unit performing at its rated efficiency, and a scaled exchanger does not. The efficiency you paid a premium for is the first thing scale takes.

What to Do Next

1. Measure your hardness properly before doing anything else. Every decision on this page is keyed to that number, and a test strip costs less than a coffee.
2. Find your model's water quality table — it is in the installation or user manual, usually within two pages of the maintenance section. Compare your number to it and note the iron, pH and TDS limits while you are there.
3. Start the delta-T log this month. Four numbers, once a month, at a consistent low flow. It is the only early warning you will get.
4. Check for isolation valves at the unit. If they are there, a DIY flush is a Saturday morning. If they are not, price fitting them now rather than during a failure.
5. If you are above 12 grains, treat the water. Every manufacturer in the table above has told you in writing where their equipment stops, and flushing more often does not move that line.