Water conditioning
Dechlorination is not about how tough your fish are. It is about the filter.
A reasonable idea circulates in the hobby and it needs unpicking, because it is wrong for an interesting reason. The idea is that conditioning the water may not be necessary if you pick hardy enough fish, and the example usually given is something like a pleco or a discus. There are two separate things going on there and they are not the same size.
The smaller point, and it is a good example
Plecos are genuinely hardy. Discus are close to the opposite.
The pleco half is right, and better than right, because there is a published mechanism behind it: a sailfin pleco keeps its thermal tolerance in low oxygen as long as it can reach the surface to breathe air1.
Discus go the other way. They come from ion-poor Amazonian blackwater and their gills are specialised for it. Duarte and colleagues measured sodium uptake in discus and angelfish and found discus running a low-affinity, low-permeability strategy suited to very soft acidic water2. Their temperature requirement is high and narrow: Mattos and colleagues incubated discus eggs at 24, 26, 28, 30 and 32 C and found that 24 C produced body malformations and circulatory anomalies in the embryos3. Discus are kept warm because they need to be, soft because they are built for it, and very clean because a large fish in a warm tank fouls it quickly.
So the two fish in the example sit at opposite ends of the demand scale. That is not a criticism of anybody, it is a genuinely useful illustration: the way to tell a hardy fish from a demanding one is the biology, not the reputation.
The larger point, and it is the whole answer
Chlorine is not aimed at your fish. It is aimed at bacteria, and it hits yours.
Chlorine and chloramine are in tap water because they kill bacteria. That is their entire purpose and they are good at it. The nitrifying population in your filter is bacteria and archaea, and it is the part of the tank least able to absorb an insult.
Our own research on filter communities found ammonia oxidisers making up only 1.6, 3.1 and 3.8 percent of the prokaryotes on three home biofilter sponges after twelve weeks4. It also found the growth rates: nitrifiers turn a gram of nitrogen into about a seventh of a gram of new cells and grow at well under one doubling a day, while heterotrophs turn most of a gram of food into new cells and grow at three to six4.
Put those two together and the consequence is exact. A dose of chlorine knocks back the whole biofilm. The heterotrophs are back in a day. The nitrifiers take weeks. So the damage from putting untreated tap water into a tank does not appear as a dead fish. It appears as an ammonia reading three days later, in a tank that was cycled and stable, and the keeper has no idea what changed.
A hardy fish does not protect the filter. That is the answer, and it is why dechlorination is the one step that has nothing at all to do with which species you chose.
There is a second reason and it does concern the fish, but not through hardiness. The United States Environmental Protection Agency's recommended freshwater criteria for total residual chlorine are 19 micrograms per liter acute and 11 chronic5. Drinking water normally carries a residual of a few hundred to about a thousand micrograms per liter. That is roughly sixteen times the acute criterion at the low end and over fifty times it at the top. Chlorine is a gill irritant far below the concentration at which a person would notice it in a glass of water, and tolerating warmth or poor water quality is a completely different axis of toughness.
What is actually in tap water
Chlorine gas or sodium hypochlorite, present in the water as hypochlorous acid and hypochlorite. Volatile, reactive, and it decays as it travels through the pipes.
Usually monochloramine, made deliberately by dosing ammonia along with the chlorine. Far more stable, which is exactly why some utilities choose it. It survives a long journey through a network.
Mostly from the plumbing rather than the source. Copper from pipe, lead or zinc from older fittings. Copper is acutely toxic to invertebrates at levels that do not trouble fish, which is why shrimp keepers care about it more than anyone.
General hardness and carbonate hardness. Not a contaminant. This is the number that decides which species and which plants will thrive, and it is the one a keeper should know before buying anything at all.
Present in some supplies drawn from agricultural catchments. Rarely a problem and occasionally decisive.
Cold water drawn fast from a pressurised main can be supersaturated. It comes out as fine bubbles on the glass and on the fish. A minor hazard on a large rapid change, and it disappears if the water stands or is stirred for an hour.
Chlorine gases off. Chloramine does not. That is the whole practical difference.
Free chlorine leaves standing water. It is volatile, it reacts with organic matter, and light breaks it down. In an open bucket at room temperature a normal tap residual falls to near nothing within about a day, and much faster with an airstone in the bucket or the water splashed in from a height, because the mechanism is the same surface exchange described earlier on this page.
Monochloramine does not do this in any useful way. Utilities that use it choose it precisely because the residual persists for days to weeks across a distribution network6. A bucket of chloraminated water left overnight is a bucket of chloraminated water in the morning.
So ageing water is a real technique with a narrow and specific scope, and knowing which disinfectant is in your tap decides whether it is a valid method or a placebo.
| It handles | It does nothing for |
|---|---|
| Free chlorine, especially with agitation | Chloramine, at all |
| Temperature matching | Heavy metals |
| Dissolved gas supersaturation | Ammonia |
| Nitrate, hardness, anything dissolved |
Which one does Armenia use? We went looking, and here is exactly how far we got.
What the operator publishes. Veolia Jur, the national water operator, states on its own drinking water page that its system comprises 132 water sources, 93 chlorination stations, 8 drinking water treatment plants, 35 laboratories and 11,582 km of pipe, and names its principal sources as springs including Aparan, Gyumush, Arzni, Shor-Shor, Arzakan, Katnaghbyur and Garni7.
What that supports. The operator names chlorination stations and no ammoniation or chloramination plant. Chloramination requires ammonia dosing as a separate deliberate step with its own equipment, and an operator listing its infrastructure would normally list it. The supply is predominantly spring-fed rather than surface water, which is the situation with the least disinfection by-product pressure and therefore the least reason to switch to chloramine in the first place.
What we could not establish, and we are not going to pretend otherwise. We could not find a published statement from Veolia Jur or from the Public Services Regulatory Commission naming the disinfectant or its target residual. The evidence points to free chlorine. We have not seen it stated.
You can settle it for your own address in five minutes. A chlorine test kit of the kind sold for swimming pools measures free chlorine and total chlorine separately. Draw a glass from the tap and test both.
If free is close to total
Free chlorine
Ageing water in an open container is a valid method, and an airstone in the bucket makes it fast. A conditioner is still simpler and still recommended, but you now know the fallback works.
If total is clearly higher than free
Combined chlorine, meaning chloramine
Ageing water is useless and a conditioner is not optional. Prefer one that also deals with the ammonia released when the chloramine bond is broken.
Either way
Do it once and it is settled
This will not change from month to month. One test answers the question for your building for good. And in the meantime the safe default is simple: a modern conditioner handles both, so use one.
What a dechlorinator actually is
The active ingredient in almost every aquarium water conditioner is a reducing agent, and in most of them it is sodium thiosulfate It reduces hypochlorite to chloride, which is harmless and already present in any water. The reaction is essentially instant at tank temperatures and the thiosulfate itself is not toxic at the doses used.
With chloramine there is an extra step, and it is the reason cheap and modern conditioners differ. Thiosulfate breaks the chlorine to nitrogen bond and takes the chlorine away, but the nitrogen stays, and it is released as ammonia Monochloramine is NH2Cl, so about 0.27 milligrams of ammonia nitrogen for every milligram of monochloramine on a molecular weight basis8.
Is that a lot? For a 30 percent change on a 100 liter tank with 1 mg/L of chloramine in the tap, that is 30 liters at 0.27 mg/L, which is about 8 milligrams of ammonia nitrogen released into 100 liters: roughly 0.08 mg/L in the tank8. A healthy filter clears that within hours. In a tank that is still cycling, or at a high pH where more of it is present as the toxic un-ionised form, it is not nothing.
This is why several conditioners advertise that they also bind or detoxify ammonia. Seachem states that Prime converts ammonia into a form the biofilter can still use but the fish are not harmed by; API states that its conditioners neutralise chloramine9. Following this site's rule on manufacturers, those are the makers' own statements about their own products, reported as claims. We did not find independent peer-reviewed testing of the ammonia-binding chemistry under aquarium conditions.
Three practical points that follow from the chemistry. Dose for the whole tank volume rather than the volume of new water, if you are adding water directly to the tank rather than treating it in a bucket, because thiosulfate is consumed by whatever it reacts with. A modest overdose is not harmful the way most overdoses are, but a very large one consumes dissolved oxygen, because a reducing agent in a tank full of oxygen will find it; do not pour it in by eye on a hot day. And a conditioner does nothing whatever for hardness or nitrate. It handles the disinfectant, and in some products heavy metals and ammonia. That is the entire list.
Reverse osmosis
When RO is genuinely needed, and when it is fashion.
Reverse osmosis pushes water through a membrane and produces water with almost nothing in it. It also wastes several liters for every liter it makes, it is slow, and the output has to be remineralised before use, because water with nothing in it is not a habitat.
Genuinely needed, in four cases. First, when the tap water makes the target impossible. Soft water specialists, wild caught tetras and Apistogramma, discus for the reasons above, and Caridina shrimp such as crystal reds all need hardness lower than hard tap water can reach by any amount of conditioning. You cannot remove hardness with a bottle. Second, when breeding those species, because many soft water fish will live in harder water and will not spawn in it. Third, when the tap carries something no conditioner touches, and high nitrate from an agricultural catchment is the usual case.
And fourth, the one that most Armenian keepers actually have. Top-up water in summer. The fan above evaporates about three percent of the tank a day, and that water has to be replaced with something mineral-free or the tank's dissolved solids climb every single day of the season. That is not fashion, it is arithmetic. Distilled water or clean rainwater does the same job; RO is simply the convenient version if you are producing it anyway.
Not needed for hardy community fish, and several of them actively prefer harder water and do worse in RO: guppies, platies, many barbs and rasboras, danios and most livebearers. Not needed for Neocaridina, the common cherry shrimp, which wants reasonable hardness rather than none. Not needed for almost any commonly sold plant. And not needed to make a tank cleaner, which is a common misunderstanding: RO removes minerals, a water change removes organic waste, and the second is not the first.
The honest position. RO is a tool for hitting a target the tap cannot hit. If you do not know your tap water's general and carbonate hardness, you do not yet know whether you need it. Buy a GH and KH test before buying a membrane. The test costs a small fraction of the unit and it answers the question.
- de Lima, M.C.M., Campos, D.F., Kochhann, D. and Val, A.L. (2025). Effects of oxygen level on thermal tolerance in Amazonian catfishes with bimodal respiration. The Journal of Experimental Biology. Paywalled; abstract free on Europe PMC. The sailfin pleco against bristlenose comparison, and the finding that air breathing preserves thermal tolerance only in the species equipped for it. DOI 10.1242/jeb.247610
- Duarte, R.M., Ferreira, M.S., Wood, C.M. and Val, A.L. (2013). Effect of low pH exposure on Na+ regulation in two cichlid fish species of the Amazon. Comparative Biochemistry and Physiology Part A 166(2): 411 to 418. Paywalled; abstract free on Europe PMC. The discus and angelfish sodium uptake kinetics, and the evidence that discus are ion-poor blackwater specialists rather than generalists. DOI 10.1016/j.cbpa.2013.07.022
- Mattos, D.D.C. and others (2024). Effect of temperature on the embryonic and larvae development of discus fish Symphysodon aequifasciatus and time of first feeding. Zygote. Paywalled; abstract free on Europe PMC. The finding that incubation at 24 C caused body malformation and circulatory anomalies. DOI 10.1017/s0967199424000236
- Our own research, W3, nitrifying bacteria versus waste bacteria. The source for the 1.6, 3.1 and 3.8 percent ammonia oxidiser figures on home biofilter sponges, the biomass yields of about 0.15 against 0.63 to 0.67, and the growth rate contrast of well under one against three to six per day. That file carries its own primary citations. The whole argument that dechlorination protects the filter rather than the fish rests on it. See also our cycling page.
- United States Environmental Protection Agency. National Recommended Water Quality Criteria: Aquatic Life Criteria Table. Free. Freshwater total residual chlorine: acute criterion 19 micrograms per litre, chronic criterion 11 micrograms per litre, published 1986. epa.gov
- Zhang, H. and others (2022). Review of nitrification monitoring and control strategies in drinking water systems. International Journal of Environmental Research and Public Health 19(7): 4003. Open access. Cited for why utilities use chloramine, namely that the residual persists across a distribution network, and for the nitrification problem that persistence then causes inside the mains. DOI 10.3390/ijerph19074003
- Veolia Jur (Veolia Armenia), drinking water service page. The company's own statement of its infrastructure: 132 water sources, 93 chlorination stations, 8 treatment plants, 35 laboratories, 11,582 km of pipeline, and the named spring sources. A self published primary source about itself, treated as a company statement and not as an independent finding. veolia.am
- Our own calculation, not a published finding. Monochloramine to ammonia. NH2Cl has a molecular weight of about 51.5 and its nitrogen 14.0, so 0.27 mg of ammonia nitrogen is released per mg of monochloramine broken. Worked here for a 30 percent change on a 100 litre tank at 1 mg/L in the tap. Molecular weights only; no measurement of what a particular conditioner actually releases. How we handle our own arithmetic
- Manufacturer product pages, quoted as manufacturer claims and nothing more. Read only for their own wording on chloramine and on ammonia binding. Seachem Prime and API water conditioners. No independent peer-reviewed evaluation of the ammonia binding chemistry under aquarium conditions was found.
- Our own research, TP2, siting, environment, electrics and water conditioning, which is the working file behind this page and carries the full arithmetic, the full source list and the complete record of what could not be verified.
Continue