Shared Filtration - Could This Work?

[daizeUK]

I've been wondering if this sort of setup could work (not that I'm thinking of getting two more tanks, oh no )
 
Two tanks, 35 and 120L
External filter, e.g. Eheim
External Hydor heater
Two Y junctions on the intake and return pipes to run both tanks through one filter
Single valve on return pipe to small tank to throttle flow (if necessary)
 
Could this work and what are the potential problems of doing it?
 
Click on diagram to see what I mean

 

[techen]

On paper it would work, I don't see much of an issue. But you'll need a large filter for it and you will want to avoid reducing the flow as it will build up the pressure on the turbine and kill it

 

[daizeUK]

As long as the filter was rated for 120+35 = 155L then that would be big enough?
 
If the throttle is only reducing flow to the small tank then I thought the remaining flow would simply be redirected to the big tank thus maintaining the same total flow rate with no strain on the pump, am I wrong?

 

[laurac94]

It could work, my concern would be disease. If one fish gets white spot..all of your fish in both tanks will be exposed too :'(

 

[ian_m]

Nope wouldn't work it's a recipe for a very very very wet floor. You will only need the slightest flow inbalance, dirt stuck to filter inlet in one tank for instance and you would end up emptying your 120litre tank into your 35litre tank (or other way round).
 
Two filters it is the safe "dry floor" solution.

 

[Chad]

I have done this.
 
One thing I did was put a valve on each line for the smaller tank as I did not want the flow in and out to be as great. This was easily done. The only issue with this is that the valves must be set equally or you end up either draining or filling the tank which is horrible!
 
Laurac is right, whatever happens in one tank happens in the other so parameters should be the same for each tank.
 
Better off not doing it really.

 

[Shaddex]

I have a 220l and a 110l using 1 filter.

Intake for filter goes in my 110l and outputs into my 220l. I have a homemade HOB drain to move the water from the 220l to the 110l.
Works a treat

Plus, only have to do a big water change in 1 tank rather than 2 lol

 

[DrRob]

I've done this on several different set ups. You'll see it more and more on complex breeder sets and in your LFS's you'll often see areas filtered in sections. It's also not vastly dissimilar to a reef refugium set up.
 
The 2 most important lessons that I learned from it were that you need to use an overflow system to sump system and put the filter intake in there, nothing else will ever quite be right for getting levels right. Once you're doing that then you may as well put the filter in the sump and just use a return pump.
 
The second lesson was to always use 2 overflows, one above the other as a redundancy, and to fix any tank where the second overflow was running.
 
I also tend to cheat and use more than 1 return pump as well, but that's just my overengineering things.
 
As above, the comment about disease spreading is valid, and a problem in one tank is a problem in both.

 

[Shadders]

Nope wouldn't work it's a recipe for a very very very wet floor. You will only need the slightest flow inbalance, dirt stuck to filter inlet in one tank for instance and you would end up emptying your 120litre tank into your 35litre tank (or other way round).
 
Two filters it is the safe "dry floor" solution.

I have to say I totally agree, even if you get the flow rates of both pipes exact how long will they stay the same, your going to get build up in and around both pipes, what are the chances of it being the same amount in each pipe, for example lets say you set you pipes running at 50lph in and out, then at some point there is some debris in the outlet pipe reducing that flow to 49lph, overnight lets say 10 hours that's 10l more going into tank than being taken out, where does that water go, on the floor, where has it come from, your other tank, imagine what would happen if the blockage was bigger, lets say 40lph, so 10lph x 10 hours = 100l ,  that's an empty tank and a very wet floor.
 
Like DrRob says the only real way to do it is by using a sump

 

[dgwebster]

you need to use an overflow system to sump system and put the filter intake in there, nothing else will ever quite be right for getting levels right. Once you're doing that then you may as well put the filter in the sump and just use a return pump.
 

Agreed. Any setup I have seen uses overflow with safety cutoff to a sump, then pump out and plumbing divide it. If the amount of water going in decreases, so does the water out. If the water going in increases, so does the water out, until you overload the outpipe, where the safety cutoff kicks in, stops water going in until more has emptied out.

 

[TwoTankAmin]

From the macro point of view this would be similar to central filtration which we all know is very doable. However when transferred to a non central setup such as a canister filter, the technical details become onerous. First, one should never do anything that inpedes the intake side of pumps and filters, only the outflow can be blocked without worrying.
 
A much easier method would work if you can have the two tanks side by side with their tops level with each other. On one tank is the intake and on the other is the return, the two tanks are then interconnected with a large U-tube which will always keeps the two water levels even. The draw back to this is you can not adjust the flow rates independently for each tank.
 
As noted above central systems use overflows which makes it impossible to have a tank get too full and put water onto the floor. The control over the rate in any tank adjusted by a control on the return flow for each tank out of the universal return pipes that service the entire system. The only amount of water that flows out of any tank is due to new water being added via the return.
 
The key to multiple tanks on a single filtration system isn't so much the pumps, sumps etc.- it is the method of using an overflow and return system that allows for variability in the rates for each tank without compromising the integrity of the equipment itself. Just splitting the in and out parts of a canister doesn't do that because it is a sealed (closed) system.

 

[daizeUK]

Many thanks for all the input everyone!  This is clearly not as simple as I thought and I think I'd be biting off more than I could chew at this stage.  Your advice is appreciated!