My salt cell says low salt but the water tests fine - which do I trust?
Trust the water. Then verify it twice, because the answer decides whether you carry a bag of salt to the pool or leave it in the garage.
Here is the rule this whole article exists to plant: never add salt on the cell's word alone. Salt is the one thing in your pool you cannot dose your way out of. Chlorine burns off, acid gets consumed, but salt only leaves the water when the water itself leaves, through splash-out, backwash, and rain overflow. Add two bags on a false reading and you will be living with them, or draining them, for a long time.
Your cell does not actually measure salt
This is the part that makes the whole conflict make sense. There is no little tongue in the salt cell tasting the water. The cell passes a current between metal plates and estimates salt from conductivity, how easily that current flows. More dissolved salt, better conduction, higher reading.
The problem is that salt is not the only thing that changes conductivity. Three other things do, and every one of them pushes the reading in the same direction: down.
Cold water reads low. Conductivity falls as water cools, roughly a quarter between a swimming-warm 85 °F (29 °C) and a spring-startup 60 °F (16 °C). Cells with poor or aging temperature compensation translate that straight into a low salt number, which is why "low salt" warnings cluster in April and October while a test strip says nothing changed. Cold water also genuinely slows chlorine production, and many systems throttle back or shut down entirely once the water drops into the 50s °F (10–15 °C). So a cold-weather warning usually means "it is cold," not "add salt."
Scale on the plates reads low. Making chlorine electrolytically leaves the water at the plate surface locally alkaline, and calcium scale loves to form exactly there. A film of scale insulates the plates, current drops, and the cell reports the missing current as missing salt. The salt is fine; the plates just cannot see it. This is worth checking first because it is the one cause you can fix in an afternoon: pull the cell, look for white crust between the plates, and clean it the way your manufacturer specifies.
Worn plates read low. Cells are consumables. The catalytic coating on the plates erodes with every hour of production, and three to seven years is a normal lifespan. As the coating thins, the cell passes less current at the same salt level, and the reading sags no matter what you add. A cell in year six that has started chronically crying low salt, in warm water, with clean plates, is most likely telling you about itself.
Notice what all three have in common: the water is fine and the reading is not. Which is why the next step is never a bag of salt.
Test the water independently first
Before anything goes in the pool, get a salt number from something that is not the cell:
- Salt test strips. The titrator style you dip and read where the band stops climbing. Not laboratory-grade, roughly ±200–400 ppm, but honest and independent.
- A drop kit. Silver nitrate titration, the same chemistry pool stores use. Tighter than strips, and the best home option if you own a salt pool for the long haul.
- A pool store test. Free, and useful as a second opinion when the strip and the cell disagree by a lot.
Now compare. Your cell's display against an independent test is not a tie to split down the middle. If the strip says 3,100 ppm and the store says 3,150 and the cell says 2,400, the water is at ~3,100 and your cell has a problem, cold, scaled, or old, in roughly that order of likelihood by season and age. Two independent measurements that agree beat one electronic estimate every time.
If the independent test also comes back low, good news: the cell was right, the fix is cheap, and it is just arithmetic.
The bag math
A saltwater pool wants roughly 2,700 to 3,400 ppm, and you should aim for the middle, around 3,000, never the top. The formula:
8.3 lb of pool salt raises 10,000 gallons by 100 ppm.
The practical version: one 40 lb bag raises 10,000 gallons by about 480 ppm. Scale it to your pool by dividing, so in 15,000 gallons a bag gives you about 320 ppm, and in 20,000 gallons about 240 ppm.
Worked through, for a 15,000 gallon pool that independently tests at 2,600 ppm:
- Gap to target: 3,000 − 2,600 = 400 ppm
- Pounds needed: 8.3 × 4 × 1.5 = 49.8 lb, call it one bag now
- One 40 lb bag ≈ +320 ppm → lands you near 2,920 ppm
Pour it into the shallow end, brush until it dissolves, run the pump for a day, then retest before deciding about the second half-bag. Salt dissolves fully but mixes slowly, and because it never leaves on its own, the whole game is add in stages and retest instead of aiming high. Undershooting costs you another trip to the pool with a bucket. Overshooting costs you water.
The failure mode that ruins equipment
Here is how it goes wrong, and it goes wrong slowly enough that nobody notices until the damage is done.
The cell reads 2,500. You add a bag. The cell still reads 2,500, because the problem is a six-year-old cell, not the salt. A few weeks later you add another bag. The display barely moves. By the end of the season you have fed it five or six bags, and an independent test would show the actual water at 5,500 or 6,000+ ppm, double the target.
Above its range the cell gains nothing, and the water slowly turns corrosive: heat exchangers, stainless ladder rails and light rings, and soft coping stone all sit in it every day. And because salt only exits with the water, the fix at that point is a partial drain and refill, in stages, retesting as you go. To drop 6,000 ppm to 3,000 you are replacing roughly half the pool's water. That is the true cost of trusting the display, and it is why the independent test comes first, every single time.
When it really is the cell
Run the checklist in this order: salt confirmed in range by an independent test, water reasonably warm, output percentage up, pump run hours long enough to matter. If free chlorine still lags after all of that, stop adding things to the water and look at the hardware, scale on the plates first, then age. A cell past year five with fading output is doing exactly what cells do, and no amount of salt resurrects worn plates. Budget for the replacement rather than chasing the reading. One more thing to rule out before condemning the hardware: chlorine that is being actively consumed - demand climbing day after day, big overnight losses - is a water problem no cell can outrun, and it has its own diagnosis path in why a pool won't hold chlorine.
The short version
- Never add salt because the cell said so
- Cold water, scaled plates, and worn plates all read falsely low
- Test independently: strips, a drop kit, or the pool store
- If truly low: 8.3 lb per 100 ppm per 10,000 gallons, one 40 lb bag ≈ +480 ppm in 10k gal
- Add in stages, brush, retest, and aim for ~3,000 ppm, never the top of the range
- If salt is in range and chlorine still lags, check the cell for scale and check its age
The pattern that catches a dying cell early is divergence: your logged salt tests holding flat while the display drifts down. Basinly makes that visible, log each independent salt test and it charts against your 2,700–3,400 ppm band, and if you add a custom parameter for the cell's displayed number, the two traces pulling apart is the cell aging in plain sight. When salt genuinely is low, the dose plan sizes the pounds of pool salt to your pool's actual volume and aims for the middle of the band, so a bag never goes in on a guess.