Skip to main content

Water change calculator

Ammonia or nitrite spike: how much water to change

A water change replaces a share of the tank with new water. If the new water has no ammonia or nitrite, each change multiplies the reading by what is left: reading × (1 − share)changes. That is my arithmetic, and it assumes the new water mixes fully and nothing adds more ammonia while you work.

After 1 change
2 ppm
After 2 changes
1 ppm
After 3 changes
0.5 ppm

Each result is for 50% changes starting from 4 ppm. Retest after the change, because the arithmetic cannot see what the fish add in the meantime.

Reading after 25%, 30% and 50% changes

Reading in ppm after one, two or three water changes, starting at 4 ppm, with the share of the original left
Each change After 1After 2After 3
25% 3 ppm (75% left)2.25 ppm (56.3% left)1.69 ppm (42.2% left)
30% 2.8 ppm (70% left)1.96 ppm (49% left)1.37 ppm (34.3% left)
50% 2 ppm (50% left)1 ppm (25% left)0.5 ppm (12.5% left)

The percentages left do not depend on the starting reading. Two 50% changes leave 25%, three leave 12.5%.

A worked example

The test shows 4 ppm. A 50% change leaves 4 × 0.5 = 2 ppm. A second 50% change leaves 2 × 0.5 = 1 ppm, which is 4 × 0.5 × 0.5. Three 25% changes leave 4 × 0.75 × 0.75 × 0.75 = 1.69 ppm. To turn a share into liters or gallons, use water change by tank size.

What the extension source says

The University of Florida extension publication Ammonia in Aquatic Systems (UF/IFAS FA031, opened 2026-10-17) says that "A 25% to 50% water change will help remove some ammonia, assuming the incoming water source does not contain ammonia." It also puts feeding first: "The first thing to do when ammonia is present in a pond or tank is to reduce or stop feeding."

On toxicity, the publication says the portion of un-ionized ammonia "increases with increasing pH and/or temperature" (it also decreases slightly with increasing salinity). It adds that un-ionized ammonia "is approximately 100 times more toxic to fish than the ionized (NH4+) form." Test kits report total ammonia nitrogen (TAN). To get the un-ionized part, you multiply TAN by a factor from the publication's Table 1, which depends on pH and temperature.

Two factors from that table, at 24 °C: 0.0052 at pH 7.0 and 0.0502 at pH 8.0. With a TAN reading of 4 ppm, my arithmetic gives 0.0208 mg/L un-ionized ammonia at pH 7.0 and 0.2008 mg/L at pH 8.0. The same reading is about ten times more un-ionized ammonia at the higher pH.

The publication states: "Anytime the UIA is higher than 0.05 mg/L, damage to fish tissues can occur." It also says: "In healthy ponds and tanks, ammonia levels should always be zero." For nitrite, it says: "Nitrite is toxic to fish at levels as low as 0.10 mg/L." Those are the source's statements, and this page does not set a limit of its own.

My arithmetic: a water change also dilutes un-ionized ammonia, as long as pH and temperature stay the same. The publication notes that a water change helps in the short term and that a properly sized biofilter is the long-term fix.

Before the new water goes in

My suggestion: match the new water's temperature to the tank, and treat tap water with a conditioner first, since chlorine and chloramine are a problem of their own. The dechlorinator calculator gives the dose for the volume you are adding.

My suggestion: retest right after the change and again later the same day, and write the numbers down. If the reading climbs back, the cause is still feeding the problem, and another change only buys time. For nitrate instead, see how much water to change to lower nitrate.

This page is arithmetic and a summary of one extension publication. It does not replace advice from an aquatic vet or a local fish store about your particular fish.

Was this page useful?