"My pool won't hold chlorine" covers several different problems, and treating the wrong one is why people spend two weeks adding chlorine into what feels like a hole in the ground.
The main three are sunlight, demand from something in the water, and ammonia. They need different responses, and a little testing separates them.
Key takeaways
- Without stabilizer, PHTA says bright sun can destroy most of a pool's chlorine within an hour. Test CYA first.
- Chlorine above 10 ppm can bleach a DPD test to a false low or zero, Taylor says. Dilute and retest before chasing a problem.
- Testing at dusk and again at dawn takes sunlight out of the picture, so whatever is lost overnight was used by something in the water.
- Ammonia takes about 7.6 ppm of chlorine per ppm of ammonia nitrogen to break through, by EPA's figure, and more in practice.
- Chlorine held by high stabilizer still tests as free chlorine. CDC says the reading should rise in proportion to CYA.
First, rule out sunlight
PHTA's stabilizer bulletin says that without cyanuric acid, bright sunlight can decompose most of the chlorine in a pool within an hour. The chart it reproduces, from Monsanto's tests, shows about half gone in 30 minutes and about 90% gone in two hours. With 25 ppm of cyanuric acid, about two thirds was still there after three hours.
So test CYA before anything else. PHTA gives 30 to 50 ppm as the ideal range. An outdoor pool well below that is losing chlorine to the sun, and the fix is stabilizer rather than diagnosis.
At the other end, high CYA produces a different illusion. PHTA says chlorine bound to cyanuric acid still tests as free chlorine, and CDC's pool code says free chlorine should rise in proportion as CYA rises. A reading that would be plenty at 30 ppm of stabilizer may not be at 90. That is chlorine working slowly, not disappearing, and the answer is either more chlorine, in proportion, or less stabilizer, by dilution. A 2019 peer-reviewed risk study recommends keeping CYA no more than 20 times the free chlorine. CDC caps CYA at 90 ppm and closes a public pool past 45 times.
Two answers that are not demand at all
Before committing a weekend to breakpoint arithmetic, rule out the cases where the chlorine was never actually gone.
The test is wrong. Taylor lists three ways a DPD chlorine test reads low:
- High chlorine bleaches it. Above 10 ppm, the color can bleach out to a false low or zero. Dilute the sample with distilled water and retest.
- Old reagent fails. Taylor's DPD #2 darkens to brown or pink as it ages, and once it does it no longer works.
- The sample sat. Taylor says not to let a sample sit for any length of time, and to take it mid-pool, away from returns and feeders.
A pool that "will not hold chlorine" and turns out to be holding it perfectly well is a happy outcome.
Metals are using it. Orenda says metals are the first thing chlorine oxidizes, and EPA's water treatment manual puts the cost at about 0.62 ppm of chlorine for every ppm of iron. Water from a well can carry enough to matter, and the usual tell is staining or a tint to the water. Testing for metals settles it. A sequestrant does not take the metal out: PHTA says it slows the metal's reaction with chlorine to limit staining, and that the sanitizer itself breaks the sequestrant down.
There is a gentler version of the first case, too: a pool where nothing is wrong and the owner is testing at four in the afternoon. Taylor recommends testing at the same time each day, and says morning is ideal because sunlight has not started breaking the chlorine down yet.
The overnight loss test
No maker or agency publishes this test, but the logic holds: take the sun out of the picture, and whatever chlorine disappears was used by something in the water.
- 1
Test at dusk
5 minutesOnce the sun is off the water. Record free chlorine precisely, because this test depends on the difference between two numbers. A drop kit reads in steps of 0.2 or 0.5 ppm, depending on the sample size; strips cannot resolve that.
Bring chlorine to a normal level for your CYA if it is low. You need something there to lose.
- 2
Leave the pump running
overnightCirculation matters, because you want the whole pool represented rather than the top foot. Do not add anything else.
- 3
Test at first light
5 minutesBefore the sun reaches the water. Same kit, same technique.
- 4
Read the difference
—There is no published pass mark, so here is the scale. Wojtowicz measured a clean, stabilized pool in Phoenix, with no swimmers, losing about 14% of its chlorine a day, and losing it at night as well as in daylight. That loss came from the chlorine slowly reacting with its own stabilizer in warm water.
So a small overnight loss is normal. A loss well beyond that, night after night, means something in the water is using chlorine.
Demand from something in the water
BioGuard puts chlorine demand down mostly to contaminants, organic and inorganic. On a pool, that is usually algae, debris, or what swimmers bring in.
My habit is to look where algae starts out of sight: behind a ladder, under the steps, in a light niche, on the shaded wall. Then brush everything, including the places you normally skip, and shock to a level that suits your CYA. Hold it until the overnight loss is back to normal. Holding it is the part people abandon early.
Filter cleanliness matters here too. A filter that cannot catch what you kill leaves dead material circulating, and the demand persists after the algae is gone.
Ammonia, and why it is different
The distinctive pattern is speed. You add a normal dose, test an hour later, and there is nothing there. Not reduced: gone. Repeat, same result.
CDC's pool code describes what is happening. Chlorine added to water containing ammonia reacts with it first, forming chloramines, and only when enough is added to destroy them does free chlorine remain. That point is called breakpoint. WHO says the reaction with ammonia is extremely rapid. CDC names urine, sweat and the environment as sources of ammonia. Hach adds that it is a byproduct of bacteria breaking down organic matter, so a pool left without chlorine has nothing removing it as it forms.
Three things make it different from ordinary demand:
The dose is large. EPA's guidance manual puts breakpoint between 5 and 7.6 parts of chlorine per part of ammonia nitrogen, by weight. Wojtowicz says it takes about 8.9 in practice, because some of the ammonia ends up as nitrate. CDC's code gives a simpler working rule: about 10 times the combined chlorine reading.
You cannot test for the ammonia directly. Taylor's pool kits do not measure it. LaMotte and Hach sell ammonia tests, but for aquariums, ponds and natural water. Once chlorine is in the pool, ammonia shows up as combined chlorine, which a FAS-DPD kit does measure.
Stopping partway wastes everything. Chlorine added below breakpoint is used up without leaving any free chlorine. Adding half of what is needed, twice, over two weekends, achieves less than adding all of it once.
The method is unglamorous: add a heavy dose, wait an hour, test. If it is gone, add more. Keep going until a reading finally holds, then keep it there until the overnight loss is normal.
The unglamorous alternative
Sometimes the arithmetic says otherwise.
If a pool has serious ammonia, high stabilizer and water that has been standing for months, the chemical route can cost more than replacing water. PHTA and WHO both say dilution is how stabilizer comes down, and a 2021 study of an enzyme method still described partial or complete draining as the standard approach.
Work out roughly what the chlorine will cost before committing. A partial drain and refill, within your pool maker's limit, resets stabilizer and hardness at the same time, and there is no prize for treating your way out of something.
Go deeperShocking properlyThe level that works for your stabilizer, and why holding it is the part that matters.Frequently asked questions
Why does my pool lose chlorine so fast?
Three broad causes. Sunlight destroys unstabilized chlorine quickly; PHTA says bright sun can take most of it within an hour. Contaminants use it up around the clock, and BioGuard and Orenda both list organic material and metals. And ammonia, from urine and sweat or from organic matter breaking down in an unchlorinated pool, reacts with chlorine before anything else does, so normal doses disappear.
What is the overnight chlorine loss test?
Test free chlorine at dusk, note it, and retest at first light before the sun is on the water. With sunlight out of the picture, chlorine that disappeared was used by something in the pool. No maker or agency publishes the test or a pass mark. For scale, Wojtowicz measured a clean, stabilized pool with no swimmers losing about 14% of its chlorine a day, at night as well as in daylight.
Why does chlorine disappear immediately in my pool?
Add a normal dose, test an hour later, and find nothing: that pattern points to ammonia. CDC's code describes chlorine reacting first with ammonia to form chloramines, and WHO says that reaction is extremely rapid. Ammonia comes from urine and sweat, and from organic matter breaking down in water left without chlorine. It does not respond to small doses.
How much chlorine does it take to clear ammonia?
EPA's guidance manual puts breakpoint between 5 and 7.6 parts of chlorine per part of ammonia nitrogen by weight, and Wojtowicz says about 8.9 in practice. Pool test kits do not measure ammonia, so the practical method is to add a heavy dose, test after an hour, and keep adding until a reading finally holds. Everything added short of breakpoint is used up without leaving any free chlorine.
Can a high CYA level make chlorine seem to disappear?
Not disappear, but work more slowly, which can look the same. PHTA says chlorine bound to cyanuric acid still tests as free chlorine, and CDC's code says free chlorine should rise in proportion as CYA rises. CDC caps CYA at 90 ppm and treats a CYA-to-chlorine ratio above 45 to 1 as a reason to close a public pool.



