limestone slurry pump 800 psi

Modified Pump Curves

WWSME

I am reviewing a document for a duplex submersible sewage lift station and it makes reference to a modified pump head capacity curve where the friction losses in the suction and discharge piping of the individual pumps are omitted from the system head-capacity curve. It says "instead, these losses are subtracted from the head-capacity curves of the individual pumps to obtain modified pump head-capacity curves, which represent the head-capacity capabilities of the pump and its individual valves and piping combined." Apparently, this is a reference from Metcalf and Eddy Wastewater Engineering, Collection and Pumping of Wastewater, but I cannot seem to track down our copy. If there is a wet well and an associated valve chamber with a plug valve and check valve, are they suggesting that the friction loss of all the piping and valves between the pump and the point where the two pumps flows combine is to be subtrated from the individual pump curves? Why would this be necessary? It sounds as though they are suggesting you do a system-head curve for the station over the flow range of the pump and subtract the resultant losses from the pump curve and then generate a system head curve for the common forcemain. Why not just use the original pump curves and generate a system head curve to determine the system performance?

LittleInch

From your description it sounds as if they are doing exactly that. Why? At a guess they have established a "standard" set of piping system which they either supply as a package or believe is common and maybe they got tired of people not doing proper system curve calcualtions starting from the pump and then blaming the pump for not providing enough head. A bit odd alright but so long as you know where in the system the pump curve starts at you can then design away from that.

77JQX

Using the pump curve modified in the way suggested simplifies the analysis of parallel pumping arrangements. It transfers part of the system curve that is unique and specific to each pump (the individual pump's suction and discharge losses) onto the pump curve. That way you can develop the system curve for the common piping system (supply, suction header, discharge header, distribution, etc.) based on total pump station flow, without reworking the system curve for different pumping scenarios (Pump A vs Pump B vs Pumps A & B etc.). The approach assumes constant speed pumps, or at least that the friction losses in the pump's unique suction and discharge piping does not significantly vary as the total pump station flow varies.

WWSME

Thanks....makes perfect sense!

RJSH

Is there another way to do this so that don't have to mofify the pump curve that a pump manufacturers provided? I hesistate to mofiy what other has supplied.