VFD

Housedoc

Greetings,

I was called on a project that using a Grundfos 26-116 circulator to move fluid through a ground loop heat exchanger. The pump can easily provide the required 8 gpm at the lower 1/3 of the curve. I adjusted the circuit setter to limit the flow to 9 gpm. But the pump is extremely noisy. Does anyone know of an affordable variable frequency drive or Triac circuit controller that will slow this pump? It's 240 volt, 1 phase.

icesailor

Noisy Pump:

It sounds like an over sized pump, sized to work at the top of the curve but run at the bottom of the curve. I found that that "noise" that you hear when you slow the pump down is a sign of cavitation. In my opinion (worthless) a very bad thing. Cavitation will do very bad things in the system. What is the head pressure that the pump is delivering?

With submersible well pumps, it is often misunderstood that if you have a well that will only yield 8 GPM, you need a 8 GPM pump. If the water in the well is only 20' to water, and the bottom of the screen is 50', you have 30' of water to use. The pump only has to deliver enough pressure to push the water to the top of the well plus however much pressure you need for the usage, say 60#. So, the smallest pump you usually use is a 12 GPM, 1/2 Hp pump, that may only be 6 stages (6 impellors. The pump may only be rated to deliver enough pressure to "push" the water to a 200' head. However, if it is 200' to water, you need a pump that will "push" the water up 200' plus enough for the pressure. So, you need more impellors or "stages to do this. So, the more stages the pump has, the higher the pressure it is able to develop. It will deliver this pressure regardless of where it is located. If you throttle it down, it still tries to deliver the same pressure. Same as the smaller pump. So, putting a seemingly smaller rated GPM pump on a system, can deliver more pressure than the pump that is rated to deliver more water. It is the "head" that counts.

Well pumps. heating circulator pumps and water pumps work on the same physical principles. The pump has no idea what it is expected to do when running.

In my opinion, you are seeing a prime example of what I call "over-pumping" in a heating system. If the heating loop was highly restricted and the circuit setter said 9 GPM with no back pressure on the loop, the pump would still be cavitating. If the loop was able to handle 20 GPM. it probably wouldn't be making noise. But when you throttle it down to 9 GPM, the pump is trying to overcome the resistance of the loop. If it was a multi-speed pump set on low speed, it probably wouldn't be making noises.

Some don't really understand the concept of pump head and pressure head. It's the pumps ability to "push" water up an open pipe that runs vertically, 500' in the air. The "head" is how high the liquid is up the open pipe. If the water was captive, and you opened it into the 500' open pipe, it would still reach the same level.

Remember what we plumbers are taught. 1# PSIG will support a column of water 2.31' high. 1' of water in a pipe will exert .434# pressure at the base of a stack 100. high. Therefore, if you close the top of the pipe and pump the water to 10# PSIG, and you release the pressure, the water will be 4.34' up the pipe. It will also be up the pipe the same distance with the 10# of pressure. The air is compressed. You can't compress water.

If you open up and let that sucker run, and the noise goes away, you need a smaller pump. The head pressure and GPM's of the pump is too high.

IMO.

Subject to change by anyone that can convince me that my theory is wrong.

Housedoc

Pump too big

Icesailor I think you are describing the problem accurately. If I one the circuit setter fully, the pump jumps to 16 gpm and the noise goes away. That might be the best answer. I just don't know the consequences of pushing over 2x the flow rate through the loop. I have been told that a simple rheostat will slow it down if the pump is capacitor run style. Even so, I don't know if it will save any wattage.