Methodology and sources
Every figure on this site is either a field copied from a published record or arithmetic on such fields using an assumption stated here. There are three sources and four assumptions.
1. Pump specifications: ENERGY STAR Certified Pool Pumps
| Dataset | ENERGY STAR Certified Pool Pumps (U.S. EPA) |
|---|---|
| Access method | Socrata JSON endpoint https://data.energystar.gov/resource/m8cf-pkii.json?$limit=500&$order=pd_id |
| Retrieved | 2026-10-17 |
| Records held | 423 certified records in summary; 103 in full detail, grouped into 63 model pages |
| Licence | ENERGY STAR certified-product data are published by the U.S. Environmental Protection Agency as open government data. The ENERGY STAR name and mark belong to the EPA; this site is not affiliated with or endorsed by the EPA or any manufacturer. |
| Missing values | Fields the source leaves blank are stored as null and render as "not published". No blank is filled with an estimate. |
| Extraction note | Socrata JSON endpoint read in batches in a browser session (the build shell has no route to data.energystar.gov). Full-detail records: the Pentair, STA-RITE, Hayward, Jandy, SPECK, Waterco, Black and Decker, Blue Torrent, Orka and SharkJet rows, plus representative single-speed and two-speed rows so every speed class is represented. Summary fields (brand, model number, type, speed, hhp, WEF, minimum WEF) were taken for all 423 certified records. |
| Verification | Field names and values as returned by the dataset on 2026-10-17. Blank fields in the source stay null here and render as 'not published'. Two-speed, multi-speed and single-speed records carry only one measured operating point, so their pages show one. Model names are reproduced as published, including the record whose model name reads 'TirStar VS' where the manufacturer's product is TriStar VS, and including names published entirely in capitals. Two presentational exceptions are applied: the 'PMP ' order prefix on some partner records is dropped from the displayed name, and where a record's model name is only the brand name the model number is displayed instead. One brand name arrived from the source with a damaged non-ASCII character and is recorded in the catalog as 'Trevi' without its accent. Cross-check performed on retrieval: the published minimum_wef of all 423 records reproduces the Version 2.0 Table 1 threshold formula for its sub-type and size class to within 0.01, so the whole list is certified against Version 2.0 thresholds rather than Version 3.0. One record (Orka 94028963, above-ground two-speed, 0.761 hhp) reports a WEF of 4.100 against its own published minimum of 4.12; the values are reproduced as published and not corrected. |
The records as held are published at /data/pumps.json, provenance block included.
What each field means
| wef | reported_wef_kgal_kwh — Weighted Energy Factor in thousands of gallons per kilowatt-hour, from the DOE test procedure the ENERGY STAR specification cites (10 CFR 431.464(b)). Higher is better. |
|---|---|
| minWef | minimum_wef — the ENERGY STAR minimum WEF for this pump's sub-type and size class, as published with the record. |
| hhp | hydraulic_horsepower_hhp — rated hydraulic horsepower: the pump's measured water output power, not the motor's nameplate rating. |
| ratedHp | rated_horsepower_hp — the motor's nameplate horsepower. It is not comparable between pumps and does not set efficiency. |
| low | flow_gpm / head_fth20 / rpm_at / power_w at the low-speed measurement point. |
| high | the high-speed measurement point. Published for variable-, multi- and two-speed records that have one; null otherwise. |
| curves | curve_a / curve_b / curve_c flow and input power: the three reference system curves in the test procedure. |
| standbyWatts | standby_average_power_w — average power in standby mode, where published. |
| wefMargin | DERIVED by this build: wef − minWef. Not a source field. |
Why one page can carry several model numbers
Manufacturers certify the same tested pump under several model numbers — a different control board, a touchscreen, a
retailer's own part number, a saltwater trim. Those records carry identical test data. Giving each one its own page
would be the same numbers repeated, so records that share a brand, sub-type, speed class, hydraulic horsepower,
reported WEF and measured input power are treated as one pump, and the page lists every model number and marketing
name on the group. The grouping rule is in scripts/build-pumps.mjs and the pd_id of every
record in a group is printed on its page.
2. The efficiency requirement: ENERGY STAR specification
The minimum_wef on each record is the ENERGY STAR threshold for that pump's sub-type and size class. The
thresholds come from
ENERGY STAR Program Requirements for Pool Pumps, Version 3.1 (final, effective 19 July 2021), Table 1, and
the sub-type and speed-class definitions quoted on this site come from its Section 1. The specification assigns WEF
measurement to the U.S. Department of Energy test procedure for dedicated-purpose pool pumps at 10 CFR 431.464(b); it
does not restate the procedure's load points, and this site does not either.
ENERGY STAR's key product criteria page
publishes the same threshold table in a shorter form. Both were retrieved 2026-10-17.
One consequence matters for reading any table here: the minimum is a function of size, and it differs between sub-types. An above-ground pump meets a lower bar than an in-ground pump of the same hydraulic horsepower, and a small pump meets a higher bar than a large one. A high WEF therefore does not mean "better pump for my pool" across sub-types — only within one.
3. The electricity price: EIA Electric Power Monthly
| Source | U.S. Energy Information Administration, Electric Power Monthly, Table 5.3, "Average Price of Electricity to Ultimate Customers: Total by End-Use Sector" |
|---|---|
| Value used as the default | 18.31¢/kWh — u.s. average retail price of electricity to residential customers, July 2026 |
| Twelve-month comparison | 17.99¢/kWh for the 12 months ending July 2026 |
| Retrieved | 2026-10-17 |
July 2026 is the latest month published on the retrieval date and sits inside the swimming season, which is when a pool pump actually runs. It is a national average: state averages range widely, so the calculators default to this figure and let you replace it. Dividing the dollar total on your own bill by its kWh total gives the number that applies to you, taxes and fixed charges included.
4. The four editorial assumptions
- The run schedule. Editorial default, not a source value: a 12-hour day split 10 hours at the low-speed measured point and 2 hours at the high-speed measured point, over a 180-day season. Every figure on the site built on it says so, and the calculator lets you change all three. Every cost figure outside the calculators uses it, and every page that shows one says so. It is not a recommendation, and it is not from any source.
- A month is 30 days and a year is 365. Monthly and annual costs are the daily figure multiplied, with no seasonal weighting beyond the season length you set.
- Standby fills the idle hours. Where a record publishes a standby figure, cost pages apply it to the hours the pump is not running. Where it does not, those hours are counted as zero, which understates the true figure slightly.
- Comparable pumps are within 0.2 hydraulic horsepower. The "comparable pumps" table on each model page is that band, restricted to the same sub-type, and ordered by reported WEF. It is a sizing heuristic for browsing, not a statement that any two pumps are interchangeable in your plumbing.
Category medians are the median reported WEF across the certified records of a sub-type, computed from the dataset as held. Category medians are the median of reported WEF across all catalog records of that product type (and of that type and speed class), computed by scripts/build-pumps.mjs. They describe the certified list only — not the market, which also contains uncertified pumps.
5. What this site does not do
- It does not test pumps. Nothing here has been measured by the author; every number is traceable to a published record or to the arithmetic above.
- It does not predict your pump's draw. Certified figures come from a standardised test rig at defined heads. Your plumbing sits elsewhere on the curve.
- It does not give a turnover requirement. Those are set by state and county pool codes for public pools and are not uniform.
- It does not advise on motor replacement compatibility. The dataset carries no mounting, frame, voltage-wiring or shaft data, and ENERGY STAR's own requirement for replacement motors is listed as "TBD" in the specification. Getting that wrong damages equipment, so the gap is left open rather than guessed at.
- It does not quote prices, rebates or payback periods. No source here publishes them.
6. Corrections
If a value on any page does not match its source record, that is a bug, not an interpretation. Each model page links
the exact dataset query for its own pd_id, and the data version (2026-10-17) and retrieval
date are printed in the footer of every page, so any discrepancy can be traced to a specific extract.