UKenergy

What it is

UKenergy answers three practical questions for battery-storage development in Great Britain: where there is land a battery could go, which battery projects are already in the pipeline and how far along they are, and which commercial and industrial (C&I) sites could use a battery behind the meter. Kacper Saks built most of it, working with Claude Code as a coding agent; co-founder Paweł Kulak contributed the second version of the C&I module.

• Land (England and Wales): 807,954 freehold parcels of at least 0.5 ha near the electricity network, with distance to each voltage class, nearest substation, published spare capacity, land cover, project-size fit (up to 2 MW, 2–20 MW, 20 MW+) and nearby storage, solar and wind projects
• BESS projects (Great Britain): 2,625 projects from the Renewable Energy Planning Database with planning status, connection-queue position (Gate 1 / Gate 2, dates) and matching entries from network operators’ embedded-capacity registers
• C&I sites (the Midlands and Greater London): 2,133 and 1,110 buildings with a screening estimate of battery savings on published 2026/27 network tariffs; where no tariff area matches, the tool shows “not computed” rather than a false zero
• Investor view: low, base and high scenarios across three project structures, with revenue and CAPEX assumptions tied to published benchmarks and trading terms marked as hypothetical

Data pipeline

Everything starts from open data: HM Land Registry INSPIRE polygons, OpenStreetMap lines, substations and land cover, the Renewable Energy Planning Database, the grid operator’s connection-queue data and the spare-capacity maps that network operators publish. A Python pipeline (shapely, pyproj, NumPy, osmium) cuts parcels from the raw polygons, extracts the network, computes distances and land cover per parcel and packs the result into gzip map tiles. The Leaflet web app loads a national top list at start-up and streams parcels and lines once you zoom in. The interface works in Polish and English, on desktop and on a 390 px phone.

Ranking and how it was checked

The first ranking is a transparent 0–100 score built from distance to the medium- and high-voltage network, parcel area, spare capacity, land cover and site advantages. Against 2,098 storage projects in England and Wales, 73.4% had a selected parcel within 300 m and 99.2% within 1 km. The check as first defined was mis-specified, so validation was redefined with temporal and spatial hold-outs and confidence intervals. Version 2 is an L2-regularised logistic regression on parcel features only, with anything derived from nearby projects excluded as label leakage.

• Cut-off 31 December 2022: version 1 AUC 0.674 (without its storage penalty), version 2 AUC 0.798 (95% CI 0.781–0.814)
• Cut-off 31 December 2023: version 1 AUC 0.660 (without its storage penalty), version 2 AUC 0.795 (95% CI 0.771–0.815)
• Version 2’s hit rate in the top 1,000 parcels is 2.6 times that of version 1
• The intervals straddle the AUC > 0.8 target, so the criterion is not counted as met; the network data is from 2026, and the label measures where developers apply, not where projects succeed

What the tool does not claim

The tool is designed to keep what is confirmed apart from what is assumed, and to say so on screen.

• Source, date and scope of each piece of evidence are recorded as entered; the tool does not verify document contents or whether a connection is live
• Open data does not name landowners; an importer for HM Land Registry corporate-ownership data is built but awaits data access; a draft contact process for private individuals awaits UK legal review, and no outreach channel is enabled until it is approved
• The ranking only orders manual site checks; it is not a probability of success
• CSV exports leave out notes and are protected against spreadsheet formula injection

How it was built

88 commits and 30 merged pull requests between 23 and 30 September 2026, tracked against a written implementation plan whose statuses separate “confirmed”, “partial”, “to do”, “unverified” and “blocked”. Python unit tests, model tests and Playwright acceptance tests, including the 390 px phone layout, run in CI on every push. On an emulated phone with throttled 4G, the first screen appears in 0.7–0.9 s; the 3 s target for full-map readiness is not yet met (3.88 s, median of five runs).

Status

The phase plan runs into early 2027. Phase 1 exit criteria (50 field-verified parcels and AUC above 0.8) are not yet met, and no pilot has started. Version 2 of the ranking is frozen as a candidate; the app still ranks with version 1 until a deployment decision. The last commit was on 30 September 2026.