What does a water change actually fix?
A water change is not a treatment. It is a blend: you remove some percentage of the tank and put back the same volume of new salt water, so every dissolved thing in the tank moves a fraction of the way from where it is to whatever the new water holds. Change 25% and each parameter closes a quarter of that gap. That is the entire mechanism, and two useful consequences fall straight out of it.
The first is a hard ceiling. A water change can never take a parameter past your salt mix's own value, because at a 100% change the tank is the new water. If your alkalinity is 7 dKH and your mix comes out at 7 dKH, no water change of any size raises your alkalinity — you would be diluting toward a number lower than the one you already have.
The second is an asymmetry nobody seems to print: water changes are excellent at removing and poor at adding. Fresh salt water contains essentially no nitrate or phosphate, so for those the new water is at zero and every single change helps. For alkalinity, calcium and magnesium you are pulling toward a number only slightly better than yours, and progress stops there.
A tank at 40 ppm nitrate and 7 dKH, changed 25% with Red Sea Coral Pro (12 dKH):
- Nitrate 40 → 30 ppm. A quarter gone, and repeating it keeps working: 40 → 30 → 23 → 17 → 13 → 9 ppm.
- Alkalinity 7 → 8.25 dKH. Barely more than a test kit can resolve.
Same water, same effort. The nitrate job is nearly done and the alkalinity job has barely started.
Can a water change raise my alkalinity?
Only if your mix is higher than your tank, and only up to the mix. Here is a tank sitting at 7 dKH that wants the bottom of the mixed-reef range, 8 dKH, against every mix with a published figure:
| Salt mix | Mixed alk | After a 25% change | Change needed for 8 dKH |
|---|---|---|---|
| Red Sea Coral Pro | 12 dKH | 8.25 dKH | 20% of the system |
| Red Sea Salt (blue) | 8 dKH | 7.25 dKH | 100% of the system |
| Instant Ocean Sea Salt reported | 11.5 dKH | 8.13 dKH | 22% of the system |
| Instant Ocean Reef Crystals reported | 11 dKH | 8 dKH | 25% of the system |
| Tropic Marin Pro-Reef | 7 dKH | 7 dKH | not reachable — mix tops out at 7 |
| Fritz RPM | 8.5 dKH | 7.38 dKH | 67% of the system |
| Brightwell NeoMarine | 7.5 dKH | 7.13 dKH | not reachable — mix tops out at 7.5 |
Tank at 7 dKH. Values are the mixes' published figures at 35 ppt, midpoints where a range is published; "reported" marks the two whose makers publish no parameter spec. Computed with this site's engine.
2 of the 7 mixes cannot get this tank to 8 dKH at all, at any volume, because they mix at or below it. Of the ones that can, the cheapest is a 20% change. At the other end, Red Sea Salt mixes at exactly 8 dKH, so the arithmetic returns 100% — which is not a water change, it is replacing every drop in the system and landing exactly on the number with nothing to spare. Treat any figure over about 50% as the tool telling you this is the wrong job for a water change.
Every number in that table is the manufacturer's, and your bucket will not match it. Mixes settle and separate in shipping, absorb moisture, and get scooped inconsistently; the same product measured by different people routinely lands a full dKH apart. Since this figure is the ceiling on what any water change can do, it is worth knowing: mix a batch, let it circulate overnight, and test it once. Then type that number into the calculator instead of picking a brand.
Water change or dose it? The honest comparison
Take a 300 L (79 gal) system at 7 dKH that wants 8, with Red Sea Coral Pro in the bucket. The water change route needs 20% of the system — about 60 L mixed, heated, circulated and hauled. The dosing route is 5.7 g of soda ash, which is about a level teaspoon.
That is not an argument against water changes. It is an argument against using them for the wrong job. Water changes are for export and replenishment — nutrients out, trace elements in. Correcting a specific parameter deficit is what dosing is for, and it is cheaper, faster, more precise and gentler on the tank. If you find yourself planning a 40% water change to move alkalinity a point, you have reached for the wrong tool.
The reverse also holds. If nitrate is 60 ppm, no additive fixes that — dilution and export do. And when a parameter is too high, a water change is often the only gentle way down, because there is no such thing as un-dosing.
Why the last bit always costs the most
Each change closes the same fraction of the remaining gap, not the same amount, so progress slows as you approach the new water's value. Going 40 → 30 ppm nitrate took one 25% change; going from 12.7 to 9.5 took another, and moved a tenth as far. This is why "just do more water changes" stops working near the end, and why chasing the last few ppm of nitrate by dilution alone is a poor use of salt.
It also means the tool will sometimes tell you a target needs a change larger than your whole system. That is not a rounding artifact — it is the honest answer that the parameter is on the wrong side of your mix, and the result panel says so rather than printing an absurd volume.
What about the parameters you didn't ask about?
A water change moves everything at once, which is the argument for it and the trap in it. Raising alkalinity with a big change of a high-alk mix also drags calcium and magnesium toward that mix's numbers, and drops nitrate and phosphate — which in a low-nutrient tank can take you from "slightly high" to "bottomed out and bleaching". The result table shows every parameter for exactly this reason: check the ones you weren't trying to change.
Rate matters too. Alkalinity is the parameter corals notice fastest; moving it more than about 1 dKH in a day is asking for tissue recession in SPS, no matter how you move it. A water change large enough to fix alkalinity in one go is usually large enough to be worth splitting over two days, and the salinity of the new water should match the tank before it goes in.