09-08-2014, 07:46 AM
Join Date: Jul 2013
Location: McKinney, TX (Dallas/DFW)
| | Re: Pump for sump tank
I was thinking of doing something similar, but a friend warned me it was a bad idea for the reasons you listed.
Doing some research, it seems the problem is that the pump is having to suck water from the sump, and then push it up to the display. This rapid low to high pressure can cause cavitation. This is the process of the water effectively vaporizing into microbubbles that then rapidly collapse. This cause all sorts of weird problems from ultrasonic vibrations and localized extreme heat. In any case, it basically trashes your pump.
Many canisters, like Eheims, have the pump on top of the canister. These are designed to be below the intake not next to it (since your intake is in the sump). This setup can cause them to run dry if you power off. The return line will drain down to the sump's water level. If the the section of the sump you draw water from is below the pump, then your pump and some filter media will be sitting dry and unless it's self priming, will fail to restart (and will burn up).
There is also the problem of the return lines head pressure. Many canisters have pumps more equipped to be circulation powerheads, than real pumps. The return's highest point is typically only an inch or two above the tanks waterline, and requires very little pressure to push it over (<0.1psi). As a sump return pump, it's pushing past about 4ft of head pressure (1.73psi). That's a LOT more work.
When I set mine up, I'm going to try a few things to mitigate these problems. For one, I'll be using an Eheim 2262. They have a 1262 pump on them that are designed for head pressure. I'll also have my canister feed a tall Rexx Grigg CO2 reactor. This will cause a large loop to follow after the pump so when the pump is powered off, the return can't drain past the pump. The water level on the reactors output will match the level in the sump, and the pump will stay wet inside of a syphon between the sump and reactor. I'm also going to install 2 check valves in series after the CO2 reactor. This should help prevent the water from draining down in the first place. I'm also hoping that with this loop, the pump will run effectively as if submerged since being in a constant syphon, it doesn't have to do the work of pulling water up over the sumps wall, it's effectively just pushing water up from the reactor. It's a minor difference, but if cavitation is a problem, this may help with it.
Basically, the CO2 reactor you're looking to add, may actually help you if the base of it is below the sumps waterline.