Pump Performance in Real Systems

[Lawrence of PA]

I have read several threads here and in other forums related to pump performance.  There seems to be some confusion about how pumps work, and what we need to know to size a pump correctly.  I am a retired engineer, and I spent most of my career working  with pumps in fluid systems.

Pond pumps are centrifugal pumps.  Manufacturers love to quote us numbers that sound very appealing: 4500 gph; max head 22 ft.  Well, we can't get 4500 gph AND 22 ft at the same time.  The 22 ft head is measured when the pump is working into a shutoff line -- this is zero flow.  The 4500 gph is measured when the discharge hose is essentially lying on the ground -- typically about 1 ft of head.

A centrifugal pump has a "characteristic curve," which describes the relationship between head and flow over the entire range of the pump's operation.  For illustration, I'm using pump curves for Sequence 1000 pumps, since I plan to install one this summer.  The manufacturer, MDM, provides very good application information on these pumps, including the head-flow curve and the power consumption curve.

In a plumbing system, as flow increases, the flow resistance (aka flow friction, or "friction head") increases too.  But it increases roughly as the square of the flow.  That is, if flow doubles, the friction head goes up by a factor of 4.  A great reason to minimize flow friction, by reducing fittings, length, elbows, curves, etc.; and by increasing pipe diameter.

Since the piping system resistance goes up as flow increases, and since the pump's ability to move water decreases as resistance (friction head) increases, the most reliable way to predict what will happen is to plot the system and the pump on the same graph.

For my system, I determined that all the fittings, piping, etc., were equivalent to 165 ft of 2" pipe.  Using a "plumbing primer" from another forum, I plotted friction head vs flow for this piping configuration.  Then I plotted the pump curve for a Sequence 4500 on the same graph.  The result is the attached jpg.

This shows that as pump head decreases, flow increases.  Because the pump can deliver more water if there is less resistance.  This also shows that as flow through the system increases, the friction head increases.  The point where these two graphs cross is the "operating point" for this pump in this system.  This graph predicts that I will get about 3000 gph from my 4500 gph pump in this configuration.  The original of this graph is in pretty colors, but it's too large to meet the file size for upload.

Since this a sort of marginal flow, I will re-do this with a 5100 gph pump and recheck the result.  In the next post.

[Lawrence of PA]

Here's what I got using a Sequence 5100.  The flow increased to about 3400 gph.

The moral of the story is:  we don't always get what we intend.  I had to move up one pump size, and even that didn't provide a whopping increase.  The good news is that this external pump uses less power and moves more water than my submersible.

[reddad35]

I wish I was more of a researcher than a wanter ponder.

 I built my pond, with a design in mind. I purchased submesibles until I got the effect I wanted. I cut the output in half on every pump I puchased and never took into account the actual flow rate and/ or power it took to run these things. I look back and have found I got lucky in my purchases. I get the wanted flow and wanted effects as well as having economic pumps as far as power costs. WEE I could replace the 4 pumps I have with one but I would have to use alot of plumbing to create the effect I have now and the actual filter flow I want. I do hope some day I will take more of an interest in actual power vs power usage. Effects vs. wattage and the  like.

Nice post I hope to gain the ability to know about and to worry about it some day.