calinb

I stumbled across a link to a white paper. It's near the bottom of this page:
http://www.geocities.com/Tokyo/4468/nitrogen.html

I found the white paper to be a very interesting and valuable read! Apparently, in many typical water change scenarios, nitrates continue to increase for weeks or even MONTHS before equilibrium is reached! Many people think their nitrate levels are stable, when in fact, they are increasing so slowly that they don't notice it over a week or a month and they drop their guard. Months later, long after the aquarist grows weary of checking water for nitrates, the levels are sky high, despite frequent water changes and reasonable fish loads.

The paper gets a little mathematical but the fundamental truth about water changing is, to maintain equilibrium below any given nitrate level, the nitrate removed through water changes must equal the net gain in nitrate produced by the tank. The two other cases result in moving the equilibrium nitrate level point; if more nitrate it removed than the net nitrate produced by the tank, the nitrate level will decrease to a new equilibrium point. If less nitrate is removed than the net nitrate produced by the tank, the nitrate level will increase to a new equilibrium level.

For example, say I measure my nitrate increase over a week period and find that it's increasing by 10 ppm / week -- say it went from 10 ppm to 20 ppm. If I wish to maintain 20 ppm or less, I could do a weekly 50% water change, which would remove 10 ppm (half the tank's nitrate) at the end of each week and the nitrate level would swing between 10 ppm and 20 ppm between changes. If I change less than 50% of the water on this weekly basis, the nitrate level will increase to a new equilibrium level higher than 20 ppm maximum. If I change more than 50%, the eventual equilibrium realized will be less than 20 ppm maximum nitrates.

But let's say I wish to do smaller and more frequent water changes. I could do a 25% water change when the nitrate reached the 20 ppm limit and the nitrate would immediately drop to 15 ppm (1/4 of the 20 ppm nitrate was removed, leaving 15ppm in the tank.) Given that it would only take 1/2 a week to regain that 5ppm removed in the water change, I'd need to do a 25% water change twice per week and the nitrate level would swing between 15ppm and 20ppm between changes, which results in a higher average nitrate level than the less frequent, 50% water change performed every week produced.

For a 75% water change schedule, the nitrate levels would swing between 5ppm and 20ppm and the water would require changing every 1.5 weeks. Contrary to what many people might intuitively think, when nitrate levels are in equilibrium, larger, less frequent water changes actually result in lower average nitrate levels than small frequent water changes.

A daily water change would require changing 1/7 * 100% = 14,3% of the water each day. So, in this fictitious 10ppm / week nitrate accumulation and 20 ppm nitrate limit example, we have

Freq vs. % changed
Daily 14.3%
2x /wk 25%
1x /wk 50%
1.5 weeks 75%

Let's look at the extreme limits necessary to maintain 20 ppm maximum nitrate-- 100% water change every other week and continuous flow water change. The 100% change results in nitrate swings between 0 ppm and 20 ppm and has the lowest average nitrate level of all the water change schedules

At the other extreme, a continuous flow water exchange system would need to remove 10 ppm of nitrate per week or half the tank's water every week. A 50 gallon tank would require a water exchange flow of 25 gallons / week = 3.57 gal / day = 0.149 gal / hour = .0025 gal / minute = 0.317 ounces / minute--about 1/3 oz per minute in and out of the tank. The nitrate level would not swing at all but would remain constant and in equilibrium at 20ppm with this continuous flow scenario.

Well..read the white paper, keep up with testing until you know how fast nitrates build in your tank. And even then, test occasionally.

...and check my arithmetic. I'm pretty tired this early A.M.

eepruls

Would removing so much water eliminate too many ammonia and nitrite loving bacteria or are there enough in the filter, gravel, etc. to keep the tank from a mini-cycle?

calinb

Possibly, I guess, but I've heard that, once a tank is fully cycled, the bacteria reside on the surfaces of stuff in the water rather than just floating around in the water itself. Biological filters are simply designed to have a lot of surface area to increase the bacteria density in any given volume of filter, but the bacteria are also stuck to all the plants, walls, gravel, hoses, decorations, etc. I know that water changes can block or impede tank cycling after an agent like BIO-Spira is used because, until the bacteria get established and the tank is fully cycled, they DO reside in the water. I "canceled" my first BIO-Spira purchase and treatment by doing daily or 2x daily 50% water changes.

On the other hand, I changed as much water as possible in a recently cycled tank that I'd cycled fishless with ammonia. It was the only way to get the final nitrAte level down to an acceptable level after processing 5 ppm ammonia for a week! The tank remained cycled after I removed all but the water in the gravel and HOB filter. So, based on that experience, I think that, once a tank cycles, you may do a massive water change without damaging your biological filtering capability. It's probably best not to stir up the gravel or filter and also best to add the new water immediately after draining out the old water. I think bacteria start to die out pretty quickly (or they become inactive) if the objects they're stuck to dry up. I'm not even close to a bacteriologist so this is just speculation.

I've recently become interested in establishing a biotype aquarium with Chocolate Gouramis (I guess I'm looking for a new challenge already ) The problem with these fish is they like very soft water with low Gh, Kh and also low pH, and maybe some other stuff contained in Indian almond leaves. They are more susceptible to nitrates than other fish too, so very low nitAte levels are desirable.

My worry is not that large water changes would harm the bacteria but, rather, the large water changes would shift the water chemistry. It would be best to "pre-age" a supply of soft water with Indian almond leaves, peat, etc. in an uninhabited heated and filtered tank or container to match more closely match the water parameters of the biotype. My LFS says, with the right water, these fish are easy and not to worry! We are blessed with very soft water in Portland, Oregon. Doh ne' no stinkin' RO!


I'd prefer to get more experienced opinions on the risks associated with large water changes. I'm not recommending any particular water change regime here; I'm only doing the arithmetic and passing on some reading material that I found useful and interesting.

Butterfly

The large water changes wouldn't remove too many beneficial bacteria as they reside mostly on the gravel, decor and glass. The only danger I can see is that it might cause a ph fluctuation.
If water changes of 25% are done each week and the gravel vacuumed there is no need for large water changes to get rid of Nitrates. Live plants will also suck them up and keep them in control.
Carol

eepruls

That's true. The plants would help bring the nitrates down.

calinb

If we assume that plants don't process nitrate exponentially (where the amount processed is a function of the concentration) we can still do simple arithmetic on the "net" nitrate production of the tank.

For a 25% water change in any given water change period, I get the following for equilibrium:


So if a tank produces 10 ppm of net nitrate weekly, a weekly 25% change will keep nitrates under 40ppm. If we accept 40 ppm as the absolute highest limit for fish, 10 ppm would be associated with the "stocking capacity" of the tank, given a weekly 25% water change regiment.

If a tank produces 7.5 ppm of net nitrate weekly, the same water change regiment will keep the tank under 30 ppm nitrate. 20 ppm maximum nitrate is the result, given 5 ppm net nitrate production, etc. Just multiply the output produced in any given change interval by 1 / 0.25 (or 4 for 25%) to get the maximum nitrate level in that change interval.

Also, everything scales. If you double your change frequency, the maximum nitrate level is cut in half, once equilibrium is attained.

Again, simple arithmetic assumes that nitrate intake and out-take (by plants) processes are not a function of nitrate concentration. I suspect this is close to the truth at reasonable nitrate levels.

Richard Taylor had 3 goldfish (9 inches worth) in his 55 gallon and found his water change schedule to be insufficient. It's easy to test and determine the net nitrate output of a tank. With a little arithmetic, appropriate water change regiments (quantity and frequency of water changes) may be determined and selected.

Butterfly

Goldfish are naturally heavy waste producers, and so of course water changes and tank size would have to be adjusted for waste produced. The amount of plants can be adjusted to use the amount of Nitrates produced. Since Nitrates are plant food I think it would be simpler to just let them suck them up.
Carol

susitna-flower

I think you need to meet timandkaren...of DIY fame!

You are such an interesting person...in a very good way!

As Carol said plants are the key....TimandKaren have (hope still do) a 200 gallon tank set up on a DIY nitrate filter...house plants growing in a container with just slow circulating aquarium water going in and slowly back to the tank. The lights grow under artificial light, and grow FAST! They take most if not all the nitrate, and this tank exists without water changes, in an overstocked condition!

Calinb, since you haven't been on the forum long you don't know this chap, but he is a real thinker like you are, and in the time it has taken for the rest of us to set up a tank or two he has turned his whole life upside down building a "fishhouse", then having to move and re-build it. Soon you will see his fishhouse talked about again, and I'm sure you two will have lots of ideas to bounce back and forth!

calinb

Sounds like TimandKaren are all set up to do some quantitative experiments too! I don't have any confidence in my assertion that plant uptake of nitrates, within the range of interest to the aquarist, is roughly independent of nitrate concentration. In another post, I mentioned that I'm not bacteriologist. Well, I'm not a botanist either.

Maybe I find some time to do a little online research on nitrate consumption in aquatic plants.

Apart from the math, Russell W. Taylor's paper illustrates how water changes that almost, but not quite sufficient, are more likely to result in a growing water quality problem going unnoticed in the short or even "medium" run. If a water change program is grossly insufficient, short run water testing is likely to reveal it.

I almost forgot...that reminds me of my idea to heat a tank with a 24/7 personal computer. I have a multimedia PC (actually, two of them) in my entertainment center. At this time, one of them is cooled by water and a large radiator. It's a personal digital video recorder so I never turn it off. If I keep the CPU busy encoding video data and just spin the video graphics processing unit GPU) with a "power virus" when it's not decoding data, I know I can get 200 W of heat out of it or more! That's enough for a 50 gallon or maybe my new 125 for goldfish.

Thanks for the introduction, Susitna-flower!