After seeing several questions on pool pump efficiency and energy cost comparisons, I thought it might be helpful to put together a simple spreadsheet that allows a pool owner to compare the costs of several pool pumps. This spreadsheet uses the data from the California Energy Commission web site which shows the energy efficiency of many different pool pumps.



Documentation is now in the spreadsheet and can be downloaded here.

Hopefully, you will find this useful and please let me know if something isn't clear.

Hi Mark this is great data, unfortunately I do not see my pump model listed. My pump is a StaRite model P2RA5F, 1.5 hp. Do you know which model I could substitute this for????

The JWP5FL-6A1 should be close enough.

Also, I found an error on the Scenario2 tab so delete that tab and duplicate Scenario1 if you need to.

Pretty nifty table for a quick comparison of pumps. I have a couple of questions:

I.) The "Pluming (sic) Curve - Curve A" is selectable across row 2 --- should the drop downs for columns C, D & E also be used for input? They seem to work independently: though they are greyed out, one can select Curve A or Curve B in any of the columns, changing the output.

II.) In the post you've documented both Curve A and the vaporware-like Curve C, so what does Curve B apply to?

Thanks ---polyvue

Currently, all the columns are tied to column B for the plumbing input to force the same plumbing for all pumps. You can select a different plumbing for each column but I saw no reason to do that. If you are comparing pumps, they should all be on the same plumbing.

Curve A is approximately equivalent to a pool with 2" suction/pad lines (or 2x1.5") and 1.5" return lines with 55' runs. Not too unreasonable but still not a great design.

Curve B is approximately equivalent to a pool with 1.5" suction/pad lines and 1.5" return lines with 190' runs. I'm not sure why on earth they even put Curve B in that table but is a worst case.

Curve C is approximately equivalent to a pool with 2.5" suction/pad lines (or 2x2") and 2" return lines with 100' runs. More reasonable but may take some time to collect the data so it could be a while before they have that.

OK. But the data for Curve C doesn't yet exist, right? So, little help there. And what makes those yokels at CEC believe that a typical home run is 100'?! I think 20' is pushing it on my system. Would think they'd have at least halved this figure to a more reasonable 50'...

For precision, I think I'm going to have to go back to some of your original calculations -- will mull these over and get back to you. Thanks!

You could just bump up the efficiency some to approximate Curve C. I estimate that the efficiency should increase by about 15% for Curve C over Curve A. Something better than Curve C might be 20% over A but I doubt it would get much better than that.

I will update the SS to allow you to add or subtract efficiency.

Edit - Updated

Updated the spreadsheet to use Curve C which is extrapolated from Curves A and B.

I also added a way to change speeds for a variable speed pump.

Mark, this grid is great - thanks for sharing your work with us!

Regarding this statement, if I wish to compare my current pump (Pentair WF28; ss up rated 2 hp, unlisted in your ss) to the two speed version (WFDS28), would it be a reasonable to simply compare the WFDS28 high speed to the WFDS28 low speed? Further, should the relative pump speed be left at 100% for both high and low?

Nominal speed is what ever the profile dictates. So 100% for high speed profiles is 3450 RPM and 100% for low speed profiles is 1725. You can set 50% on a high speed profile and it will approximate the low speed profile but using 100% on the low speed profile will be more accurate. The speed parameter is mainly for the variable speed pumps.

In your case, since there is no profile for a WF28, you can use the WFDS28 high speed profile as a proxy. However, you might still want to have three separate profiles for comparison. For two speed pumps, it is likely you will split the run time between the speeds so for example, you could have the following profiles:

Profile 1: WFDS-28 (HIGH SPEED) @ 1 turnover per day (reference pump)
Profile 2: WFDS-28 (HIGH SPEED) @ 0.5 turnover per day
Profile 3: WFDS-28 (LOW SPEED) @ 0.5 turnover per day

You then add the cost of 2 & 3 and compare it to 1. Also, you can split the run time anyway you want for profiles 2 & 3 but it is more realistic to model it this way. Also, if you split the pump cost between the run times proportional to the run time, then you can just add each of the cost parameters for the two speeds and the totals will be correct.

I updated the spreadsheet with a two speed summary to show you how to set it up (see Scenario #1). Also, I had to fix the run time calculation as it didn't take into account the turnovers per day.

Because of the interest on ROI form this post and others, I added a section to do the ROI calculation although it doesn't include any probability of failure calculations as sc23 did. Thanks for the idea.

Key Assumptions:

20k pool
1 Turnover per day (25% at higher speed, 75% at lower speed)
120 month study period

Existing Pump is the WFE-28 (shown as two profiles for consistency)
2 Speed Pump is the WFDS-24 at a cost of $600
Intellifo VS with an interface box at a cost of $1100 (enable at least a two speed timer)
Intellifo VF with no interface at a cost of $1500 (uses internal timer for two speeds)
The Intelliflo at high speed is set to the same flow rate as the 2 speed at high speed but the low speed setting is 1000 RPM which tends to be the most efficient.

Scenario1 - $0.12 per kwh, the national average
Scenario2 - $0.24 per kwh, 2x the national average

In scenario 1 the two speed has a better NPV than the VS while in scenario 2 the VS has a better NPV but the VF is still not as good as the two speed. As you can see the outcomes are primarily dependent on energy costs, the total pump implementation costs and how the pump is going to be used.

What was most surprising to me is that NPV spread between the different pumps was only a few hundred dollars.