Continuous (automatic) water changes?

[shawn]

I read a post a while back on RC about how fewer numbers of larger volume water changes were more efficient than the same volume of water changed in larger numbers of lesser volume. They had some explanation that made good sense, and I pretty much bought into the idea.

Then I was talking with a fellow reefer (Bidadari's husband, Adrianus) about continuous water changes, and the subject of efficiency came up. Seemed like the efficiency would be very low, since you would essentially be doing very many very small volume water changes. However, we wondered just how bad the efficiency would be.

I'll leave out the assumptions and the math (I can follow up if people care), but the efficiency turns out to be about 63% - pretty good compared to my expectations.

Some numerical analysis gives this chart of efficiency:

Water change -- # of changes -- efficiency
100% -- 1 -- 100%
1/2 -- 2 -- 75%
1/4 -- 4 -- 68%
1/8 -- 8 -- 65%
1/16 -- 16 -- 64%

Turns out the limit of the efficiency as the size of the water change goes to 0 and the # of changes goes to infinity (continuous water change) is: (1 - 1/e) = .63 = 63%

So, it really isn't significantly worse from an efficiency standpoint than doing 25% water changes, plus, you reduce the potential shock to livestock as the changes won't cause large swings in ph, salinity, temperature, etc...

Seems like all one needs is a two-channel peristaltic pump that adds salt water on one channel and removes it on the other. Set it at a rate that equals the weekly water change you want to do and keep one container of new salt water and empty the other container of waste water. Since the rate would be low you probably don't even need to aerate or heat the new water.

Any one here doing something like this? I was thinking of removing the water just after the return to the sump, and putting new water in just downstream from where I removed it.

What issues do people see with this approach?

 

[Denise Goodheart]

I''m interested in this one... when I got my liter meter and read it can do this a light bulb blinded me... I thought this was a great way to do the changes... at least it would be regularly done.... one the other hand I heard that 2 - 5% changes (give or take) a week are good and it gets you looking onto your tank more (if that is possible) and likely to notice problems or changes...????
Denise

 

[~Flighty~]

We do this, but it is called a leak. Not exactly what you mean

 

[shawn]

~Flighty~, good one!

Denise, I know exactly what you mean. I often do pump & sump cleaning, etc., when I do water changes. I'm making the assumption that these tasks will still get done at the proper frequency...

 

[wrassefan]

Wouldn't your supply of mixed saltwater evaporate some waiting to be supplied to the tank?? Potentially, wouldn't the sg be higher in some cases depending on evap rate?? Just some thoughts.

 

[KevK]

Could you drill a hole in a closed container, run the hose through the hole to minimize evaporation? I agree you would have to pay attention to the SG, but it seems like a good solution to large water changes.

 

[shawn]

I'd use a container with a lid for the replacement water so there wouldn't be any significant evaporation. I do that now when I prepare SW for changes while heating and aerating.

 

[dz6t]

wait a minute, I think there must be some flaw in the "science".

In flesh water world, the continue water change is the MOST efficient way to do water change. It is what happen in river and lake.

Also, contiune water change stress the fish the least.

If you ever visit goldfish and discus farm in southeast asia, you will see that is what they do there (water is cheap down there).

 

[shawn]

I defined efficiency as how effectively a given volume of replacement water is at removing substance dissolved in the pre-change water.

For example, a 100% water change is 100% efficient since all the original water and all the original substance dissolved in it is removed, and then replaced with new, clean water.

For two 50% water changes, 50% of the original water and dissolved substance is left after the first change. The second 50% water change removes half of that, leaving 25% of the original. Thus, with 25% of the original water and dissolved substance left, the "100%" of water changed only removed 75% of the original water, and is only 75% efficient.

Note that this model assumes perfect mixing of existing water and replacement water between "changes". In most cases, this is probably a reasonable approximation to reality.

As the volume of each water change gets smaller and the total number of changes gets larger, the efficiency gets lower. However, the lowest it will ever get is about 63%. Pretty strange, huh?

Note that this deals with efficiency, not effectiveness. Barring 100% water changes, which are pretty much impractical, the most effective method is to change more total volume in a given period of time. It seems clear that changing 100% of the volume over the period of a day is better (considering only dilution) than changing 1% of the volume over the same period of time. However, it also costs about 100 times more, so unless your replacement water is really cheap...

 

[~Flighty~]

It is the least efficient in terms of getting the old water out and new water in.

Take a few 50% water changes on a tank
First you have 100% old and 0% new in the tank
you take out half of the old water and add new now you have a 50% 50%mix
now you take out half of the 50% mix and add new to get 75% new 25% old and so on.

If there was red dye spilled in the water, 50% water changes would dilute it, but never remove it. A continuous change would remove it even slower.

 

[~Flighty~]

Darn, I hate posting the same thing at the same time.

 

[Scuba_Dave]

You need to live on the ocean, and just keep pumping in a small amount, and let the overflow drain back to sea

 

[mikestod]

The point is really ease of maintenance. A 100% water change is only really necessary in a system with major issues. On a healthy system, water changes are performed to reduce the accumulated organic materials that remain in our system, and recharge levels of organic materials used by the system. The best way to phrase it is "Out with the bad, in with the good."

A dynamic ecosystem in a static environment (our reef system)needs a way to reduce the constantly building nitrates, phosphorus, silicates, etc., while replenishing the chemicals used for growth such as calcium, magnesium, stontium, etc. as well as keeping the aquatic conditions like pH, slainity, temp, light, etc as close to the norm as possible.

That being said...Our systems are unique, as are our systems' needs. Just as we dose 2 part additives or use calcium reactors to keep calcium and alkalinity stable, we can use other ways to keep our other parameters in check. I'd have to guess that constant partial water changes would work fine for a majority of the systems out there. In my opinion, it is only necessary to do massive water changes when water parameters are out of whack. Starting with good water quality, it should be easily maintained at acceptable levels (never perfect in a sealed environment, but as close to perfect as we can maintain it). The trick is to figure out the proper ratio of water to change in the appropriate time frame (like adjusting a calcium reactor or using a proper dosing method)

If I had the space for the equpiment, I'd be automating my water changes. If I lived on the ocean, I'd take Dave's advice!

 

[dz6t]

The model only works under the condition that there is no organic and inorganic (Ca, Mg etc) being produced or consumed between water changes. Which means you have no living thing in the tank.

 

[shawn]

dz6t,

Well, I wouldn't go so far as to say it doesn't "work", it is just not as accurate as is could be. All models are a tradeoff between complexity and accuracy. Adding time to the picture makes things harder in a number of respects. For example, how does one measure efficiency in a system where things are constantly changing? For larger water changes, when are the changes done relative to the points in time where we make measurements of contamination? Are we talking about peak levels of contamination, average, or minimum? What is the appropriate rate of new contamination vs. rate of water change? By dropping time, all these issues go away, and I bet the end result, from a reef-keepers point of view, is pretty similar (but of course, I may be wrong).

Not that these aren't all potentially interesting and important effects, but I think that they probably serve to cloud the issue more than clarify it.

Maybe I'll put together a numerical model that includes time - an analytical model is out of the question (for me). Like I said though, interpreting the results may be a little more difficult.

Based on your comments, I gather you have specific objections with the conclusion of continous water changes that I posted. Is your objection that the "efficiency" should be higher or lower, or with the concept of "efficiency" altogether?

The bottom line of my post was that continous water changes seemed like a good thing:

So, it [continous water change] really isn't significantly worse from an efficiency standpoint than doing 25% water changes, plus, you reduce the potential shock to livestock as the changes won't cause large swings in ph, salinity, temperature, etc...

 

[KAS]

I'm somewhat with Wrassefan. If you do the constant water change, you are essentially doing a water change AND a topoff (for tank evaporation). Because you are topping off with saltwater, I believe your salinity with increase over time.

 

[MickyB]

Had a LitermeterIII setup on a 150 gallon tank with 3 pumps Inlandreef (Matt) helped me with the settup. Had one pump putting in Ro/DI from a 55 gallon barrel for evaporation about 3 gallons a day. Another pump Taking out 1 gallon a day of tank water and sending it outside the house. Another pump putting in 1 Gallon a day from a 55 gallon drum with premix Ro/DI and Instant ocean. The litermeter did it in very small amounts thru a 24 hour period.

 

[~Flighty~]

I'm somewhat with Wrassefan. If you do the constant water change, you are essentially doing a water change AND a topoff (for tank evaporation). Because you are topping off with saltwater, I believe your salinity with increase over time.

Well, you just need to adjust the salinity of your new water to balance out. If you are going to do 10 gallons a day and you usually need to do 1 gallon of topoff for evaporation a day, then add one gallon rodi to 9 gallons of SW to get the right mix to add all day.

 

[shawn]

All right, I spent way to much time on this and the data is too hard to explain, but I did a model that delt with the concentration of "contaminates" that resulted with various water change methods.

I linked to a pdf that describes what I did and the results, in case anyone is interested:

http://reef.mxpark.com/waterChange.pdf


Here is the summary:

1. The maximum "contaminate" concentration is only affected by the average rate of water changed. The size of the water changes does not affect the maximum "contaminate" concentration.
2. Increasing the size of the water change causes a lower minimum and average "contaminate" concentration.
3. For a "continuous" water change, the maximum, average, and minimum "contaminate" concentrations are essentially equal (most stable). While for 100% water changes, they are the most different (most unstable).



As far as water change vs. water topoff goes, I would do just like what MikeB said: 1 pump removes water change water and 1 pump adds water change water (both the same volume), and then another pump adds topoff water. Works the same as "normal" water changes as far as salinity.

 

[dz6t]

Actually, I believe continue water change is a good thing too. May be I did not explain myself clear enough.