Pools
Pool salt calculator
Pounds equals gallons times 8.34 times the change in ppm, over a million. Every salt calculator does that sum and none of them ask where your numbers came from — which matters, because a test strip resolves to about 400 ppm, and on a 20,000 gallon pool that is 1.7 bags of uncertainty before anybody calculates anything. Salt is also the one pool chemical with no way back out, so this doses for the wrong end of your reading on purpose.
Written and edited by Pablo Ruiz Quintero
Last reviewed 2026-09-22 · Nothing you type leaves your browser
The arithmetic is one line. The reading is the problem.
Pounds equals gallons times 8.34 times the change in ppm, over a million. Every calculator does that. None of them ask whether the two numbers you fed in mean anything — and one of them almost certainly does not.
The number every dose is scaled by. If it is wrong, every chemical you ever add is wrong by the same proportion.
Measured off the shape of the floor, or read from a water meter on a refill.
The reading
How you measured matters as much as what you measured, and this is the input no other salt calculator asks for.
Zero if the pool is fresh and you are starting it up.
The colour blocks on the bottle are printed about 400 ppm apart, so that is the finest distinction the strip can make — whatever the packet claims. It tells you whether you are roughly in the right country, and that is a genuinely useful thing to know. It is not a number to dose from.
The most common system in American back gardens, and the one whose numbers everyone else quotes as though they were universal.
Evaporated or solar salt, 99.8 per cent sodium chloride or better, with no iodine and no anti-caking agent. It dissolves fast because it is milled for it, and it is what every manual asks for.
What to add today
Nothing
reading 2,400–3,200 ppm · band 2,700–3,400
Your reading cannot justify a dose
Dose to 3,200 ppm from the reading as it stands and a calculator would tell you 1.7 bags. But the reading could be as high as 3,200, and adding on that basis risks landing past the top of the band — which is the one mistake here with no way back. Get a finer reading before buying anything.
The margin is wider than the target
A salt test strip resolves to about 400 ppm and this cell’s whole band is 700 ppm wide. The measurement cannot see the band it is supposed to be aiming at — which is worth knowing before trusting any number computed from it, here or anywhere else.
The band, and what your reading actually pins down
Both on the same parts-per-million scale.
| If you measured with | Margin | Worth in bags | Buy today |
|---|---|---|---|
| A salt test strip · yours | ± 400 ppm | 1.7 | nothing |
| The salt cell's own display | ± 300 ppm | 1.3 | 1 bags |
| A liquid drop test | ± 200 ppm | 0.8 | 1 bags |
| A pool shop photometer | ± 100 ppm | 0.4 | 1 bags |
Same pool, same water, same day — four different answers, and the difference between them is entirely the instrument. A drop test costs about fifteen dollars and it is the cheapest genuine improvement available on this page.
What to buy
- Pool salt, 40 lb bags
Evaporated or solar salt, 99.8 per cent sodium chloride or better, with no iodine and no anti-caking agent. It dissolves fast because it is milled for it, and it is what every manual asks for.none - Where that lands the pool
From 2,800 ppm. One more bag would take it to 3,040.2,800 ppm - What the reading alone is worth
A salt test strip resolves to about 400 ppm. On this pool that is what the measurement is worth in salt before any arithmetic happens.1.7 bags
How to add it
The order matters more than people expect, and step three is where the over-dosing happens.
- 01Switch the cell off. It should not run while the salt is dissolving.
- 02Pour it into the shallow end and brush it across the floor. A heap dissolves from the outside and takes days.
- 03Run the pump for 24 hours before testing again. Testing early gives a low reading, and a low reading is what makes somebody add a second dose on top of a first that was already right.
- 04Re-test, then top up if it is still short. This is the step that gets skipped and it is the whole reason to aim low.
- 05Turn the cell back on.
If the volume above is a guess, that is the thing to settle first — working it out from the shape of the floor rather than an average depth changes this dose by more than any of these steps.
Worth knowing
- A salt test strip resolves to about 400 ppm, and the whole band this cell runs in is 700 ppm wide. The measurement cannot see the band. Any dose computed from it is arithmetic performed on a number that does not carry that much information.
- At 2,800 ppm this pool is already inside the 2,700 to 3,400 ppm band, and once the reading's own margin is allowed for there is no dose that is clearly needed. The honest answer is to add nothing and re-test with something finer — a drop test or a shop sample.
- Salt is the one pool chemical with no way back: nothing consumes it and it does not evaporate. Under-dosing costs a bag and ten minutes; over-dosing costs a hose, a day and a re-balance. Aim low and creep up — and expect to add some every spring anyway, because backwashing, splash-out and rain all take salt out over a season.
On this page — 4 sections
The sum is one line. Your reading is the problem.
The short version
- A test strip resolves to about 400 ppm — worth 1.7 bags on a 20,000 gallon pool, before any arithmetic.
- Dose for the top of your reading’s range, because over-salting is the one error here with no way back.
- A scaled cell reads low and asks for salt it does not need. Clean it before believing it.
- Cells do not agree. Pentair IntelliChlor wants 3,400 ppm where Intex wants 3,000.
Every salt calculator on the internet performs the same calculation, and it is not a hard one: pounds equals gallons times 8.34 times the change in parts per million, divided by a million. You could do it on the back of the bag.
Which means the calculator is not where the error lives. It lives in the two numbers you typed in — and one of them is much worse than anybody admits.
A salt test strip resolves to about 400 parts per million. Not because it is a bad strip, but because the colour blocks printed on the bottle are that far apart; there is no finer distinction available to make. So a strip reading 2,800 means the pool is somewhere between 2,400 and 3,200.
On a 20,000 gallon pool that spread is worth 1.7 bags of salt. The instrument cannot tell the difference between needing two bags and needing none — and the whole operating band of a Hayward cell is only 700 ppm wide, which is narrower than the strip’s own margin.
| Measured with | Resolution | So 2,800 means | Worth in bags | Buy today |
|---|---|---|---|---|
| A salt test strip | ± 400 ppm | 2,400 – 3,200 | 1.7 | nothing |
| The salt cell's own display | ± 300 ppm | 2,500 – 3,100 | 1.3 | 1 bags |
| A liquid drop test | ± 200 ppm | 2,600 – 3,000 | 0.8 | 1 bags |
| A pool shop photometer | ± 100 ppm | 2,700 – 2,900 | 0.4 | 1 bags |
The same 20,000 gallon pool, the same water, the same day. The only thing that changes down the table is the instrument — and the answer changes from 1 bags to none.
MaterialsEstimator, How precisely pool salinity can actually be measured · checked 2026-09-22
Aim low, because only one direction is reversible
Here is the fact that ought to govern how anybody doses salt, and which no calculator acts on: the two directions are not the same.
Add too little and you add another bag. Ten minutes, ten dollars, and nobody ever knows. Add too much and there is no chemical that takes it back out. Nothing in a pool consumes salt — the cell borrows chloride to make chlorine, and the chlorine reverts to chloride — and salt does not evaporate with the water. The only remedy is dilution: salt water out, fresh water in, and then the fresh water has to be balanced for hardness and pH because it is not pool water.
An asymmetry like that has an obvious consequence and almost everybody ignores it. When your reading has a margin, dose for the top of that margin, not its middle. If a strip says 2,800, plan on the assumption that the pool might already be at 3,200 — because if it is, the dose that would have been correct at 2,800 puts you over.
| Dosed to the target | Dosed for the top of the range | |
|---|---|---|
| What the reading says | 2,800 ppm | 2,800 ppm |
| What it assumes | The number is exact | It could be 3,000 |
| Bags, on a drop test | 1.7 | 1 |
| If the pool was really low | Correct | A bag short — add one more |
| If the pool was really high | Over the band | Still inside it |
| Cost of being wrong | A hose and a day | A bag and ten minutes |
1.7 bags
What a strip's margin is worth
± 400 ppm on 20,000 gallons
13
Bags for a startup
From fresh to 3,114 ppm
3,590 gal
To come back from 3,900 ppm
The only remedy there is
700 ppm
The whole Hayward band
Narrower than a test strip's margin
A cell asking for salt is usually asking to be cleaned
This is the single most useful thing on the page, and it is not in any dosing chart.
A salt cell does not measure salt. It measures conductivity — how easily current passes between its plates — and infers salinity from that. It is a good proxy, right up until something else changes the conductivity.
Calcium scale does exactly that. It builds on the plates as a white crust, it insulates them, and less current passes. To the cell, that is indistinguishable from weaker brine. So a scaling cell reads low and asks for salt — and the owner, reasonably enough, adds some.
Do that two or three times and the pool is genuinely over-salted, the cell is still scaled, and it is still reporting low. The reading that prompted every dose was the symptom of a different problem the whole way through.
Before adding salt because the unit asked for it
- Take the cell out and look through it. Clean plates look like clean plates; scale is an obvious white crust.
- Clean it if there is any. A soak in dilute acid, or the manufacturer's own cleaning solution, and never a screwdriver — scraping ruins the coating.
- Test the water with something that is not the cell. A strip is coarse but it is independent, and independence is the point here.
- Check the water temperature. Cells read low in cold water too, and a spring reading on an unheated pool can be genuinely misleading rather than wrong.
- Look at how old the cell is. They last three to seven years, and a tired one reports oddly before it fails outright.
- Only then consider the pool short of salt.
| Salt system | Target | Band | Stops at | Bags from 2,800 |
|---|---|---|---|---|
| Hayward AquaRite | 3,200 | 2,700 – 3,400 | 3,800 | 1 |
| Pentair IntelliChlor | 3,400 | 3,000 – 4,500 | 4,500 | 2 |
| Jandy or Zodiac | 3,500 | 3,000 – 4,000 | 4,000 | 2 |
| Intex and other above-ground systems | 3,000 | 2,500 – 3,500 | 3,500 | none |
The same pool and the same drop-test reading, against four systems. There is no universal “correct salt level”, and the figure everyone quotes is simply Hayward’s.
Hayward, Pentair and Jandy, Salt chlorine generator operating specifications · checked 2026-09-22
What goes wrong
01Dosing from a test strip as though it were exact
Strips resolve to about 400 ppm, which on a 20,000 gallon pool is 1.7 bags. The reading carries less information than the dose calculated from it implies.
Instead — Use the strip to know roughly where you are, and a drop test or a shop sample before actually buying salt.
02Adding salt because the cell said low salt
A scaled cell reads low — calcium on the plates looks exactly like weak brine. Repeat the dose a few times and the pool is genuinely over-salted while the display still says low.
Instead — Take the cell out and look at it before buying anything. A pool that seems to want salt every few weeks has a cleaning problem, not a salt problem.
03Testing before the salt has dissolved
It comes back low, which invites a second dose on top of a first that was already right. This is the mechanism behind a large share of over-salted pools.
Instead — Brush it in, run the pump for 24 hours with the cell off, then test.
04Using one 'correct' salt level for every system
The figure quoted everywhere is Hayward's. Pentair IntelliChlor wants 3,400 ppm and Intex wants 3,000 — dose one to the other's number and you are at the bottom of a band or at a cut-out.
Instead — Read the manual for your own unit once, and write the target on the pump housing next to the volume.
05Scaling the dose off a volume from the average-depth formula
On a pool with a hopper bottom that formula runs about 17 per cent high, and the error goes into the salt in the direction that needs a hose to fix.
Instead — Settle the volume from the shape of the floor first. It is the single input that every dose you will ever make is multiplied by.
06Filling with softened water and adding a full startup dose
A softener exchanges hardness for sodium, so softened water arrives at around 350 ppm before anything is added.
Instead — Test the fill water before the first dose, even on a brand new pool. It is the one time the starting point is assumed rather than measured.
07Buying pellets instead of crystals
Pellets are the same salt pressed hard, and they sit on the floor for days. The pool tests low, somebody adds more, and then both doses arrive at once.
Instead — Crystals — pool salt or softener crystals, they are the same chemical. If pellets are what you have, expect several days and do not re-test early.
08Pre-loading extra salt against future losses
Losses come from backwashing, splash-out and rain, and they arrive gradually and are trivially corrected. An overshoot arrives at once and is not.
Instead — Dose to where you want to be now. Expect to add a little most springs, and treat that as normal rather than as a failure to plan.
- 01Settle the volume firstEvery dose scales by it. A volume from the average-depth formula runs high on most inground pools, and that error goes into the salt in the direction that cannot be undone.
- 02Find out what your cell actually wantsThe manual, not the internet. A Pentair wants 3,400 ppm and an Intex 3,000, and dosing one to the other's figure is how a cell ends up at its cut-out.
- 03Test with the best method you haveA drop test costs about fifteen dollars and has twice the resolution of a strip. Before a large dose, a shop sample is free and better than both.
- 04Clean the cell before believing its displayA scaled cell reads low and asks for salt. If the reading came from the unit itself, take it out and look at it first.
- 05Dose for the top of your reading's range, not its middleIf a strip says 2,800 it could mean 3,200. The dose has to be safe at 3,200, because that is the end you cannot come back from.
- 06Brush it in with the cell offA heap on the floor dissolves from the outside for days. Spread it, and leave the cell off while it goes.
- 07Wait 24 hours before testing againTesting early gives a low reading, and a low reading is what makes somebody add a second dose on top of a first that was already correct.
- 08Top up if it is still shortThis is the step that makes aiming low free. Skipping it is the only reason aiming high ever seemed sensible.
Every dose on this page is multiplied by the pool’s volume, and that number is wrong more often than the salt reading is. The average-depth formula everyone uses runs about 17 per cent high on a pool with a hopper bottom — which is most inground pools in America, and about two extra bags on the calculation above.
Settle the volume first- ppm
- Parts per million by weight, the same as milligrams per litre. Because a gallon of water weighs 8.34 lb, the salt needed to move a pool by a given ppm follows from the volume and nothing else.
- Salt cell
- The chlorine generator on a saltwater pool. It passes current through the water to split chloride into chlorine, and the chlorine reverts to chloride afterwards — which is why salt is not consumed.
- Operating band
- The salinity range a cell is built to run in. It is not one number across the trade: a Hayward wants about 3,200 ppm and a Pentair 3,400, with bands of very different widths.
- Cut-out
- The salinity at which a cell stops generating to protect itself. Above it the pool quietly makes no chlorine at all, which is usually discovered when the water goes green.
- Resolution
- The finest distinction a test can actually make — for a strip, the spacing of the printed colour blocks. It is not the same as the accuracy claimed on the packet, and it is what bounds how precisely anybody can dose.
- Titration
- A drop test, where reagent is added a drop at a time until the colour turns and the count gives the reading. Twice the resolution of a strip for a few dollars more.
- Scale
- Calcium deposited on the plates of a cell. It insulates them, so the cell conducts less — and since the cell infers salinity from conductivity, it reads low and asks for salt it does not need.
- Solar salt
- Salt made by evaporating brine in the sun. It is what pool salt and softener salt both are: the same chemical at the same purity, in different bags at different prices.
- Anti-caking agent
- An additive that stops table salt clumping. It has no place in a pool, which along with the iodine is why table salt is on the do-not-use list.
- Startup dose
- The salt added to a freshly filled pool. It is the largest single dose the pool will ever get and the one most worth measuring for, because the fill water is not always at zero.
- Softened water
- Mains water run through a water softener, which works by exchanging hardness for sodium. It arrives with salt already in it — enough to matter when filling a pool.
- Dilution
- Removing salt water and replacing it with fresh. It is the only way to lower salinity, because nothing consumes salt and it does not evaporate.
Common questions
- How much salt does my pool need?
- For a 20,000 gallon pool starting from fresh water and aiming at a Hayward's 3,200 ppm: about 520 lb, or 13 bags of 40 lb, which lands it at 3,114 ppm. The arithmetic is gallons times 8.34 times the change in ppm, divided by a million. The hard part is not the sum — it is that both numbers you feed into it are less certain than they look.
- How many bags of salt for a 20,000 gallon pool?
- From zero, 13 bags. From a reading of 2,800 ppm it depends entirely on how that reading was taken: 1 bags on a shop photometer, 1 on a drop test, and on a test strip the honest answer is none at all, because the strip cannot tell whether the pool is already where it should be. Same pool, same water, same day.
- Why does my salt calculator give a different answer to yours?
- Because most of them dose straight to the target from the number you typed in, as though that number were exact. A test strip resolves to about 400 ppm — its colour blocks are printed that far apart — so a reading of 2,800 means somewhere between 2,400 and 3,200. On this pool that spread is worth 1.7 bags. Dosing from the middle of it risks overshooting, and overshooting is the one error here that costs a day.
- Can you remove salt from a pool?
- Not chemically. Nothing in a pool consumes salt — the cell borrows chloride to make chlorine and the chlorine reverts to chloride — and it does not evaporate with the water. The only way down is dilution: take salt water out and put fresh water in. A pool at 3,900 ppm needs about 3,590 gallons replaced, and the fresh water then needs balancing for hardness and pH.
- Why does my salt cell keep saying low salt?
- Very often because it is scaled, not because the pool is short. A cell works out salinity from how well the water conducts between its plates, and calcium on those plates insulates them — which looks identical to weak brine. So a scaling cell reads low and asks for salt it does not need. If a pool seems to want salt every few weeks, take the cell out and look at it before buying another bag.
- What kind of salt should I use in a pool?
- Evaporated or solar salt at 99.8 per cent sodium chloride or better, with no iodine and no anti-caking agent. Water softener salt crystals are the same chemical at the same purity and usually cheaper — about $2 a bag less — so they are a genuine saving. Crystals, though, not pellets: pellets are the same salt pressed hard and they sit on the floor for days. Rock salt carries insolubles that stain plaster, and table salt is iodised, full of additives and absurd money by the pound.
- How long does pool salt take to dissolve?
- About 24 hours with the pump running, if it has been brushed across the floor rather than left in a heap. Test before that and the reading comes back low — which is exactly how people end up adding a second dose on top of a first that was already right. Leave the cell switched off while it dissolves.
- What salt level should a saltwater pool be?
- Whatever your cell's manual says, and that varies more than people expect: Hayward AquaRite about 3,200 ppm, Pentair IntelliChlor about 3,400 ppm, Jandy or Zodiac about 3,500 ppm, Intex and other above-ground systems about 3,000 ppm. Salting a Pentair to a Hayward's figure leaves it at the bottom of its band where cells under-produce, and salting an Intex to a Pentair's figure puts it at its cut-out, where it makes no chlorine at all.
- Do I need to add salt every year?
- Usually some, yes. Salinity drifts down over a season because backwashing, splash-out and rain all remove salt water or dilute it — evaporation is the only process that concentrates it, and topping up afterwards cancels that out. So expect to be adding rather than removing, which is another reason there is no case for pre-loading a pool with extra.
- Does the fill water already have salt in it?
- Sometimes enough to matter. Mains water runs around 50 ppm, A well runs around 200 ppm, Softened water runs around 350 ppm. Mains water is negligible against a 3,200 ppm target. Softened water is not — a softener works by exchanging hardness for sodium, so it arrives salty, and filling a pool that way and then adding a full startup dose is a reliable route to an over-salted pool. Test before the first dose, even on a brand new pool.
- What happens if the salt is too high?
- Above the top of the band the cell works harder than it was designed to, and above its cut-out it simply stops generating — so the pool quietly makes no chlorine, which is usually discovered when the water turns. Very high salinity is also hard on stone coping, on some pool furniture and on a heater's internals. None of it is catastrophic and all of it is tedious, because the only remedy is a hose and a day.
Send this to whoever is paying
The link carries your measurements, so whoever opens it sees the same numbers you are looking at — and you can come back to it yourself tomorrow with the delivery note in front of you.
Plain links, no tracking scripts. Nothing here reports back that you shared it.
About the author
Pablo Ruiz Quintero — Editor
Editor of MaterialsEstimator. Decides which documents count as a source, what every default value is, and where the line sits between what the reader parses and what the engines compute.
More about how this site is edited