Home / Methodology

Methodology & sources

Rule #1 at TonnageCheck: no fabricated prices or specs. Every equipment price, efficiency rating, tax credit statement and benchmark on this site comes from a public source we link to. Where a number is a modelled estimate rather than a listed price, it carries an estimate badge and an explanation.

Equipment prices

32 systems and 1 line set, 36 price points from HVACDirect and Alpine Home Air, captured October 24, 2026 by scripts/fetch-energystar-seed.mjs. The script reads each listing's published price, model numbers and spec table, identifies itself with an honest user agent, waits a second between requests, and only fetches pages the retailers' robots.txt allows. Listings it cannot parse are dropped and logged, never filled in by hand.

These are public online retail prices. Contractors usually buy from distributors at unpublished prices, typically lower. We label them "wholesale equipment value" on the scorecard because they are the closest verifiable public proxy for what the equipment costs, not because they are distributor prices.

Efficiency ratings

For each system we query the U.S. EPA ENERGY STAR, data.energystar.gov dataset (ENERGY STAR Certified Air-Source Heat Pumps dataset (w7cv-9xjt)) by outdoor model number, and accept a record only when its indoor model also matches the listing, or its SEER2 matches the listing exactly. 13 of 32 systems match; those pages show SEER2, EER2, HSPF2, COP at 5°F, capacity at 5°F and 47°F, and the AHRI reference number straight from the ENERGY STAR record, with a link to it. For the rest, ratings come from the retailer's spec table for that listing's own components and are labelled "listing". Systems below 15.2 SEER2 are below the ENERGY STAR minimum for split systems, which we note.

Installed-cost benchmarks

Fair installed ranges come from the National Residential Efficiency Measures Database (REMDB), published by the National Laboratory of the Rockies (formerly NREL), U.S. DOE. We download its data file (https://remdb.nlr.gov/db.json) and evaluate its regressions exactly as the REMDB web app does: installed cost = multiplier × (x1 × metric1 + x2 × metric2 + b) + adder (retrofit), evaluated at low / mid / high coefficients — the same calculation the REMDB web app performs.

REMDB uses SEER (SEER1). We convert with SEER ≈ SEER2 ÷ 0.95, the usual approximation for ducted split systems after the 2023 test procedure change. When we have matching listings in your tier we use their median SEER2; otherwise the midpoint of the tier.

REMDB publishes 2023 dollars. We restate them with CPI-U (CPIAUCSL (CPI-U, all items, seasonally adjusted)): 2023-12 = 308.741, 2026-09 = 336.298, factor 1.0893.

REMDB costs are national averages, divided by location multipliers in the underlying data. We do not apply a regional adjustment, because we do not have a public, citable one at ZIP level. Expect high-cost metros to run above the midpoint.

Grades

We compare the graded quote (your total minus any itemized ductwork) with the benchmark midpoint:

GradeQuote vs. benchmark midpointStatus
AAt or below the midpointFair
BUp to 15% aboveFair
C15–30% aboveOverpriced
DMore than 30% above, up to the top of the rangeOverpriced
FAbove the top of the benchmark rangeRed Flag

A quote below the bottom of the range is also flagged as a Red Flag, because very low prices often mean missing scope.

Tax credits and rebates

What we store

If you leave the box ticked, an anonymized copy of your audit is stored: the first three ZIP digits, system type, size, optional brand and model, quote, equipment value and grade. Second-opinion requests, newsletter sign-ups and contractor applications store the details you type, and are only readable by TonnageCheck staff. See privacy.

All sources