Pump & filter operation and testing

[stev32k]

It's a long story, but I found that my main pump is over sized for my filter which is a Hayward S-244T sand filter rated for 62 gpm. The pump is a Sta-Rie single speed 1-1/2 H.P. with a maximum flow of almost 120 gpm. When the filter is clean the inlet pressure is 24 psig which according to the pump curve corresponds to a flow of 82 gpm. I'm pretty sure I had intermittent episodes of flow channeling because I've seen the inlet pressure drop by 3 or 4 psi when it should have been going up, not down.

I added a valve to the pump discharge and put a pressure gauge on the filter outlet so I could limit the inlet flow from the pump and determine the pressure drop across the filter. What I found was surprising. At full pump flow with an inlet pressure of 24 psig the filter outlet pressure was 6 psig. That's a differential of 18 psi and that is way too much for a sand filter. I varied the inlet pressure and recorded both inlet and outlet pressure and estimated the flow from the manufacturer's pump curve (after back washing). The results are:

Inlet, psi--Outlet, psi---Diff.psi------Flow, gpm
24-----------6--------------18------------ 82
25----------11--------------14------------78
26----------18--------------8--------------73
27----------20 -------------7--------------68
28----------23--------------5--------------63
29----------26--------------3--------------60
30----------27--------------3--------------58

The flow numbers are very rough because it's hard to read the very small graph the Mfg. sent, but the pressures are good numbers. I repeated the test several times and got good repeatability. What is most interesting is the sharp rise in differential pressure between an inlet of 26 to 24 psi. The filter has been running with an inlet pressure of 28 psig for the past week and the differential has only increased 1 psi. I'm not sure at what point on the pressure curve the filter should be back washed but think I'll watch flow from the returns and compare that to the outlet pressure try to make a correlation.

 

[mas985]

Some backwash valves have quite a bit of head loss so those numbers are not totally out of the question. Add that to the loss in the sand and it is possible to have that much pressure loss across the filter.

Also, when determining flow rates, you need to take several things into account:

If you use the filter gauge for a return head measurement, since it is about 3' above the pump, you need to add three feet of head to compensate. The best place to measure return head is at the pump drain plug so it includes the backwash valve and no other corrections are required.

The suction head needs to be added to the return head. If it is small, it can be neglected but most of the time it is 5-10' of head.

 

[stev32k]

My goal in this is to measure differential pressure across the filter, avoid channeling, and learn how long the filter can run between back washes. I'm not so much interested in absolute accuracy as I am in repeatable measurements. I'm surprised that Hayward does not provide an instrument connection on the filter discharge like they do on the inlet. Inlet pressure is important, but provides only half the information needed to determine how the filter is operating. The other half is the outlet pressure and the difference between the two is the most important measurement.