Pump Efficiency Data

[mas985]

I came accross this data the other day and thought I would share it with the forum. The California Energy Commission in conjunction with PG&E, a power company in California, has collected energy efficiency data on pool pumps from various suppliers. The methodology used allows pumps from different manufactures to be compared directly. Each pump is tested under the same effective plumbing configuration. Currently, there are two curves which represent two different plumbing configurations and by my estimates, they represent the following:

Curve-A is equivalent to a 2.0" suction & 1.5" return plumbing system with 60' runs (high head loss)
Curve-B is equivalent to a 1.5" plumbing system with 250' foot runs (really high head loss)

Because the current curves represent very poor plumbing designs, pump manufactures have requested a third plumbing system which reflects a more reasonable head loss for better designed pools. I will keep an eye out for that version.

Anyway, Curve-A is probably reasonable to compare pumps and the column of most interest is labeled as Curve-A Energy Factor (CAEF). This is simply the gallons pumped per Watt-Hour so the higher the number the better and it provides a direct efficiency comparison of various pumps.

In addition, you can use the CAEF to determine the watts per turnover for any pool although keep in mind that it will be somewhat pessimistic because of the high head loss assumption but at least it will be an upper bound.

This should make it much easier to compare the energy efficiency of pumps.

[EDIT] One disturbing point on this data is that some of the test points do not lie on the published head curves which could mean that either the test data is inaccurate or the published head curves are inaccurate.

[EDIT2] The CEC recently removed the ability to directly download the Excel spreadsheet. You now have to do a web query and then export that to a spreadsheet. From what I understand, they will be adding Curve C which is more inline with a typical 2" plumbing system.

 

[JasonLion]

Fascinating. By sorting on column M you can really see the huge advantage that variable speed pumps have on low speeds, numbers around 10. Dual speed pumps on low speed are getting 5 to 6, while hardly any single speed pumps do much better than 3.

One thing that surprised me was that the Pentair WFE-3 does worse than the WFE-4. That doesn't make any sense to me since the WFE-4 is higher HP and otherwise ought to be the same.

 

[mas985]

Fascinating. By sorting on column M you can really see the huge advantage that variable speed pumps have on low speeds, numbers around 10. Dual speed pumps on low speed are getting 5 to 6, while hardly any single speed pumps do much better than 3.

And those factors relate directly to cost. So a 10:3 EF ratio is a 10:3 cost savings.

One thing that surprised me was that the Pentair WFE-3 does worse than the WFE-4. That doesn't make any sense to me since the WFE-4 is higher HP and otherwise ought to be the same.

The biggest difference between the WFE-3 and -4 is the motor efficiency (does not include wet end). For some reason the electric motor used on the 3 has much lower efficiency than the 4. Not something they usually publish. Lower HP motors tend to have worse efficiency but not by that much.

 

[JasonLion]

The biggest difference between the WFE-3 and -4 is the motor efficiency (does not include wet end). For some reason the electric motor used on the 3 has much lower efficiency than the 4. Not something they usually publish. Lower HP motors tend to have worse efficiency but not by that much.

Curiously, the WFE-2 is the best of the three of them (according to the PG&E data) and the lowest HP (with the highest efficiency of the entire WhisperFlo series). I always dismissed the WFE-2 because by the specs it draws just as much power as the WFE-3 but appeared to move less water (based on the pump curve). Based on the PG&E data there is something weird going on with the Pentair specs (or the PG&E data is wrong).

{EDIT}Pentair updated the specs. Now the WFE-2 is shown as using less power than the previous listing.{/EDIT}

 

[mas985]

I am a bit suspect about the data as well since some of the operation points do not fall on the published head curves. However, that could also indicate that the head curves are incorrect. It is too bad that the pool industry does not have the same level of documentation as the commercial pump industry. They publish a complete set of data on each pump using detailed laboratory measurements. The pool industry publishes specs more like cartoons then actual data. I guess they really don't want you to know how the pumps perform. The most forth comming data is from the Intelliflo where they actually show the energy usage for various operating points. But then again, they are trying to sell the energy efficiency of the pump.

 

[deanrinehart]

Yikes...disappointed in the ePump numbers. The Intelliflo is ahead by a stunning margin.

 

[taekwondodo]

Isn't the real metric here the "GPM/Watts"? IOW, how many watts it takes to move a gallon of water?

I added a few cells on my own (GPM/W), and the Intelliflo still wins by an outstanding margin - but it gives a more meaningful number.

I currently have a WFE-8 (2hp WhisperFlo) - it moves 0.022 gallons of water in one minute per watt (or, takes 45 watts to move a gallon of water per minute) of water in a higher head configuration.

Compared to the IntelliFlo moving 0.188 gallons of water in one minute per watt (or, takes 5.3 watts to move a gallon of water in a minute).

In other words, the IntelliFlo is moves 8.5x the amount of water that the WhisperFlo does for the same money (wattage)! (Someone check my logic, as this seems INSANE)

OUCH! Time to upgrade (And take advantage of PG&E's $100 rebate).

- Jeff

 

[JasonLion]

{EDIT}I don't appear to be able to write this down correctly. Read what mas985 says in the next post.{/EDIT}

In any case, the IntelliFlo, run on low speed, is very dramatically better than any single speed pump.

 

[mas985]

"GPM/Watt" or "Gallons/Watt-hour" are the metrics one should use. The spreadsheet uses "Gallons/Watt-hour" which is simply "GPM/Watt" * 60 but either one is fine to use.

 

[taekwondodo]

In any case, the IntelliFlo, run on low speed, is very dramatically better than any single speed pump.

Agreed - and I tried that logic on the wife, and she says to me, "Just where do you think you're going to come up with $1200 (for the pump and controller)?" !!!

(Note - had to replace my front transfer case last week in my Avalanche, for $2.2K. So much for upgrading the pump this year).

 

[lochman]

The spreadsheet is from the California Energy Commission, although the data may have been collected by PG&E. California adopted amended appliance efficiency regulations last year that requires various appliances, including pool pumps, to be tested in a standardized way. The spreadsheet shows some, but not all, of the data required to be collected under the new pool pump testing standard. The Curve A data is from the point on the pump performance curve where total system head = 0.0167 * flow^2. The 0.0167 number is an estimate of system performance in a "typical" pool, but it is not necessarily the best efficiency point on the pump performance curve.

I think the low speed performance data for the variable speed pumps in the spreadsheet gives something of an apples and oranges comparison. The Intelliflo pumps were tested at 1000 and 1100 rpm, and the Sta-Rite pumps were tested at 800 rpm. The spreadsheet indicates only that the two Jandy ePump models were test at low speed, but other information at the Energy Commission suggests that the Jandy ePumps were tested at half of full speed, or 1725 rpm. A variable speed pump has three variables: flow, head, and speed. The definition of Curve A in the testing method establishes a fixed ratio between head and flow at any given speed, but it is impossible to have a meaningful comparison of pump efficiencies unless speed is held constant across all test subjects. At full speed (3450 rpm), the Intelliflo pumps are not as efficient as the ePumps. A variable speed pump is always more efficient at a lower speed than at a higher speed. For example, compare rows 122 and 123 to see how an increase of 100 rpm causes efficiency to drop in the Inteliflo pump. We can't know whether the Intelliflo or the ePump is more efficient at low speed unless both pumps are tested at the same low speed. I hope the Energy Commission releases more useful comparison data when it becomes available.

 

[BobF]

what does column G mean?

 

[mas985]

I believe that it simply indicates if the pump has the capability for more than one speed.

 

[JasonLion]

I have an update on the WFE-2 vs. WFE-3 comparison. Pentair updated their specs and now lists the WFE-2 as taking 8.8/4.5-4.4 amps, while the WFE-3 remains at 11.2/6.0-5.6 amps. This makes the PG&E data much more plausible.

 

[mas985]

I think the problem with 1/2 HP pumps, and the Hayward Northstars exhibit the same issue, is that it is very difficult to make a low HP induction motor so while the impeller might be for a 1/2 HP motor, they can only find a 3/4 HP full rated motor which might have a SFHP of over 1. So the motors tend to be rated well over the required HP and amps for the impeller. If you actually measure the amps on the 1/2 HP motor, it is well below the rating of the motor. So the motor may be running at 2/3rds or less of full rating.

This is one of the reasons you can take almost any pump and downsize it by replacing the impeller (and diffuser in some cases). The loss of efficiency for underloading the motor is usually made up by the higher efficiency of larger rated motors and the reduced energy consumption. Theoretically, an induction motors can be underloaded to any extent. This is the primary reason a two speed motor loses efficiency at low speed but is still more efficient than high speed. The affinity laws ensure that the overall efficiency improves even though motor efficiency my decline. The same is true for a reduced impeller. Theoretically, you can put a 1/2 HP impeller on a 2 HP motor and it will be more efficient than the 2 HP impeller with the same motor. In fact, it might only be slightly less efficient than a 1/2 HP impeller on a 1/2 HP motor.

I wanted to add that this is one of the reasons the nameplate information is of limited use. The rating of the motor only tells you so much. It really doesn't tell you what the design point of the impeller is in relation to the motor. This is what makes the CEC data so valuable. It is a direct comparison of the energy use between pumps under the same conditions. Hopefully over time, it will become more reliable.