How were these figures calculated?
Rule version 1.0.0, observation date 2026-09-25. 8673 model pages, 365 authority pages and 2863 district pages.
The caveat, kept off the first screen of a result
A noise result on a model page is not a measured level, proof of permitted development, or a guarantee against nuisance. It is an estimate from the MCS method for one outdoor unit at the stated directivity, using a manufacturer-declared sound power. The legal result is only pass, fail or unknown. Being within 2.0 dB of the limit is a precision marker, not a third legal result.
What is calculated
- Permitted development, noise and the grant call
scripts/contract_engine.py. Empty inputs stay unknown. - Sound pressure uses Q=2 (one reflecting surface) and, in the table, Q=4. No barrier. Distances 1, 3, 5, 8 and 12 metres. Scotland's legal point is 1 metre.
- Electrical draw is the registered cooling load divided by SEER (or EER), marked estimated. £ per summer multiplies that draw by 200, 400 and 800 hours and the Ofgem price-cap unit rate in data/build/tariff.csv. The hours are a scenario, not a measured summer.
- Conservation-area figures on a district are the share of sampled postcode points, not a share of land area, and not a yes for one address.
- Offices within 25 km are a haversine from the centroid to the registered office. That is not a service area.
What is pending_data
- Hot nights come from ERA5 reanalysis (method hn-era5arco-1). The grid is about 30 km and reads lower than street temperatures in cities. A zero means no night in that cell reached the threshold.
- Outdoor-unit volume, unless a later datasheet row carries it.
- Room area. No UK floor-area rule is applied.
- MCS air-to-air status in the catalogue is unknown, not a confirmed zero.
Author of the pages: Igor Ivitskiy, PhD in mathematical modelling, and a UK homeowner. See About.