small daily vrs large monthly water changes

[beaslbob]

walkerbrody on other forum

Was asking what the effect of automating water changes to do 1g/day vrs ~20%per month.

Could someone please pass on these ideas.

What happens with water changes is that the just before water change values converge on:

just before water change values=values in replacement water+(rise in value between water changes)/(fraction of water changed)


for this example assume we are talking about a 1%/day change or a 20%/20 days change. (and assume you have some constant daily change in some parameter).

With 1%/day:

just before water change=(replacment water)+(daily change)*(1 day)/(1/100)=
_____________________=(replacement water)+100* daily change


with 20%/20 days:

just before water change=(replacement water)+(daily change)*(20 days)/(1/5)
_____________________=(replacement water)+(daily change)*20 days*5
_____________________=(replacement water)+100*daily change


the same exact values as the 1%/day

but the after water change value will:

%____before____after___buildup___nextbefore
1%___100________99_______1_______100
20%__100________80______20_______100

Of course (as most of you who know me expect me to say) the key to maintaining anything in the tank is to limit the daily change in conditions. When you acheive that then water changes have no effect and are unnecessary.


my .02

 

[H@rry]

- - - - - - - whoosh! - - - - - - -

I might be able to follow that early in the morning but not at this time of day!

 

[beaslbob]

- - - - - - - whoosh! - - - - - - -

I might be able to follow that early in the morning but not at this time of day!

:xd::xd::bigsmile:

Aw come on H@rry. It only took me 35 years to figure all that out.


:squigglemouth:

 

[hawkinsrgk]

I feel so much better that it is not just me. I have read this thread 5 times trying to figure it out.

 

[beaslbob]

I feel so much better that it is not just me. I have read this thread 5 times trying to figure it out.

I don't blame ya.

I have edited the equations.

:cry:

What happens is lets say you have a constant rate of tank changes and are doing a certain water change schedule with constant replacement water.

Say you change 10% of the water.

After awhile the conditions just before the water changes are the same. (takes a little while).

In order for the conditions to be constant you have to remove the build up between changes by the water change.

so if you are changing 1/10 of the water, the conditions before a water changes grow to 10 times the change between water changes. down to 9 (1/10) back up to 10.

if you change out 1/5 the water then the the conditions grow to 5 times. down to 4 (1/5 change) back up to 5.

1/3-------- 3 times 2 back up to 3

And so on

By "tieing" to water change level to the days between changes and using the daily tank change rate resulted in the 100 times daily tank change above. Small frequent changes or large infrequent changes result in the same before water change levels. The larger changes results in lower after change values but there are more days for the change to build up. In fact a constant flow through of 1% per day the tank would result in a constant 100 times the daily changes in tank conditions. So if the tank had 1ppm/day nitrates increasing a constant flow through the tank of 1%/day with 0 nitrate water would result in a constant nitrate level of 100pm. If you are consuming calcium a 1ppm/day, using 400ppm calcium replacement water, and doing the 1%/day constant flow, the calcium will be at a constant 300ppm.

Now H@rry may have to study that in the morning. :xd:

 

[Dasani]

I say do a 15-25% weekly for <50g and 20% monthly for >50g and screw all the math
XD

 

[walkerbrody]

Beaslbob, Thank you for your help. I tutored linear algebra and differential equasions at UAH many years ago, maybe its because its 3am, but I'm having a hard time understanding. I think you are saying eventhough the change is less, because the changes are more frequent the end result is the same. I can increase the size of the water change container, and use a weekly timer instead of a 24 hour one to change 5% or more of the water once per week instead of 0.7% per day if you think this would be better. Thanks again for your help

 

[beaslbob]

Beaslbob, Thank you for your help. I tutored linear algebra and differential equasions at UAH many years ago, maybe its because its 3am, but I'm having a hard time understanding. I think you are saying eventhough the change is less, because the changes are more frequent the end result is the same. I can increase the size of the water change container, and use a weekly timer instead of a 24 hour one to change 5% or more of the water once per week instead of 0.7% per day if you think this would be better. Thanks again for your help

interesting. and you probably have a much better math background then this old engineer whose last math course was in the '60's.

I got a pm and sent this reply. Hopefully this will give more detail but when read will help clarify my KISS analysis and comments above:

1. 
   
   This analysis of mine was one of those oh no wonder moments for me.
   ​

1. 
   
   ​

1. 
   
   There are several assumptions which may not be realistic but simplify the analysis to consider only the effects of the water changes.
   
   

   1) we are talking about concentration: effect size of the tank is irrelevant
   2) the concentration is linear: effect: things addup. so concentration is replacement water and tank increases can be considered seperately then added together
   3) everything is constant: effect tank build up, replacement water are constant
   4) I am considering where the tank winds up not the intermediate effects. effect: what are the values just before water changes and after and at every point in between.​

   
   

   replacement water effects: The tank eventually closes in on the values of the replacement water. ​

   
   

   tank build up efects: In order for #4 to be true the water change must remove the build up between changes. So if you change 1/10 of the water the tank will build up to 10 times the build up between changes.
   Then the water change will remove 1/10 the build up so down to 9 to rise to 10 times before the next change. 1/5 results in 5 times. 1/3 3 times, 1/100 100 times and so on.​

   
   

   What I did with the example was to specify the build up and size of water changes in days. So water changes were say 1%/day or multiples of that. Like 1%/day, 10%/10 days, 33%/33 days, 50%/50 days etc etc etc. Because there were specified that way the just before water change values were the same but after water change value shows more variance. Consider 100%/100 days. You have 100 times the daily build up then drop down to 0 by the water change. Tthen 100 days later you are back up to 100 times the daily build up. 50% every 50 days you have 50 times the daily build up between changes. But you are only removing 50% of the water. So the just before water change values must be twice that 50 or 100. Then down to 50 and back up to 100.​

   
   

   To that you add the concentration on the replacement water. So say your replacement water has 30 then the before and after values are incresed by 30. So 50%/50 days is 130 down to 80 up to 130.​

   
   

   consider this:​

   
   

   (1%/day) ​

   
   

   days____%change___buildup___before change____after_____next
   100______100_______100_______100____________000___ ___100
   050______050_______050_______100____________050___ ___100
   030______030_______030_______100____________070___ ___100
   020______020_______020_______100____________080___ ___100
   010______010_______010_______100____________090___ ___100
   005______005_______005_______100____________095___ ___100
   001______001_______001_______100____________099___ ___100
   0.5______0.5________0.5_______100____________99.5_ ____100
   constant 1%/day change________100____________100______100​

   
   

   So in a tank the is increasing nitrates at 1ppm/day or consuming calcium at 1ppm per day in order to limit the change to 5 ppm you would have to do water changes of 100% every 5 days or greater water changes. And with water that has 0 nitrates and perfect calcium. 20%/day would do it also.​

   
   

   Hopefully this should point to that fact that water changes which are convienent to hobbists will not maintain parameters. Tthe key is to get the tank balanced out to nitrates are being constantly maintained at 0 and calcium/alk/mag are being dosed. Then water changes have no effect and can only degrade the system.​

   
   

   Always always ask question when this old school nerd confuses you. Better to stop the ignorance, force me to be clearer and correct my errors, then for both of us to remain ignorant.​

   
   
   

   my .02 ​

   
   ​

1. 
   

 

[beaslbob]

From that to me it is obvious to me that in order to maintain tank parameters you must limit the build ups. Consume the ammonia/nitrItes/NitrAtes/phosphates are they are being generated. (I like macro algaes) and dose calcium/alk/mag as they are being consumed. and do that in a stable way so shocks to the system are rapidily recovered and parameters return to acceptable values.

 

[walkerbrody]

Im installing a calcium reactor this week, and I have a 30gal fuge full of chaetomorpha algae. So that should help keep everything stable. Tomoko linked this artical Water Changes in Reef Aquaria by Randy Holmes-Farley - Reefkeeping.com on the other forum. I says pretty much what you are saying. Small daily water changes aren't quite as good as large monthly ones for removing impurities and replenishing nutrents but it is close enough that the stability of the overall system makes it worth the effort. Thanks for your help. I think I'm going to start on my 1 gal a day change system soon.

 

[beaslbob]

That is a good article Tomoko posted.

One word of warning--Dr. Holmes-Farley is much better then beaslbob. So you may ruin your reputation with any mention that anything I said was about the same. :bigsmile: