Published live (2026-07-02). This is the first definite treatment: not a principle and not a search residue, but a specified, replicable design — put through the bridge test as its gate, and published with the verdict on its face. The deliverable is the template; the Ogwen valley is the worked instance, not the point. Process: a fourteen-agent run — four evidence researchers (regulation, existence proofs, economics, ownership forms; ~40 sourced findings), four expansion lenses, a walls pass with classification, synthesis, then three independent adversaries attacking the result against the bridge test. Verdict, unanimous: passes-with-conditions — and the conditions are the most valuable part, so they are printed here, not managed away.
The design constraint
The bridge test forbids any design that needs a campaign, a founding meeting, or a heroic act of collective will. So the constraint, set before generation began: a no-faith ladder — every rung individually rational for its adopter on day-one terms, with the cell emerging as a by-product of self-interested rung-climbing, never as a prerequisite. No rung may require belief in the whole.
The template — the ratchet cell
Nine rungs. No rung requires the next; every rung makes the next cheaper. The name is honest: the mechanism is a ratchet, not a leap.
Rung 0 — the paid originator. A pre-existing local development entity (the Partneriaeth Ogwen shape), grant-funded (Great British Energy Community Fund; Ynni Cymru — the latter election-contingent, flagged). Its output is the artefact kit: one-page bilateral supply agreement, exemption self-assessment checklist, air-space lease and PPA templates, share-offer skeleton, metering runbooks. Origination is a paid job, never volunteer faith — and no later rung's rationale cites grants.
Rung 1 — the hermit and the cylinder. Any household, fully unilateral: rooftop solar (7–11 year payback at ~26p/kWh retail, 0% VAT to March 2027) and — the cheapest entry anywhere on the ladder — a £50–150 smart immersion diverter converting ~13p export into ~26p displaced water-heating. The value basis is self-consumption, not the retail/export spread — so the rung survives the spread compression that success itself causes. Standing rule inherited by every rung above: share electrons, never heat — since 27 January 2026 even a single shared heat vessel across dwellings trips full Ofgem heat-network authorisation, with no de-minimis.
Rung 2 — the party-wall dyad. One exporting household, one adjacent neighbour, a private wire, a sub-meter, a standing order: seller earns ~19–20p instead of ~13p, buyer pays ~19–20p instead of 26p. Both strictly better off from the first traded kWh. Regulatory footing: the Class C on-site/private-wire supply exemption — self-assessed, no pre-approval. Explicitly non-gating: the install cost (£500–2,000) is unverified, and where adjacency fails adopters skip to rung 5. An accelerant, not a gate.
Rung 3 — the free-roof anchor. The strongest single offer in UK community energy, already deployed at scale by Egni Co-op: the mutual leases the air-space above a large daytime-loaded roof (farm shed, school, chapel hall, workshop), installs at its own cost, and sells the host power behind their own meter at ~18p vs ~26p — zero capex, zero liability for the host, kit transfers free at year 21. One decision-maker, no community required. This also dissolves the split incentive by construction: the mutual owns the asset, so the roof-owner and the benefit-capturer no longer need to be the same person.
Rung 4 — the saver and the hedger. The capital layer, and where the institution materialises: a Community Benefit Society issuing community shares (FCA-exempt, one-member-one-vote, asset-locked) at the sector-standard 4–7% — a channel this exact population has already used (Ynni Ogwen: £450–500k raised in about two months, ~85% from the LL57 postcode). The society arrives late, as the wrapper the finance legally requires — governance after value, ownership as the residue of accumulated trades. The gate's sharpest finding lands here: the ~85%-local raise pattern means this capital is bought on a blended civic-financial motive, not pure yield — see the verdict below.
Rung 5 — the match-tariff member. Through-grid sharing without a licence: the Energy Local club model, born in this valley — half-hourly matching of local demand to local generation at a member rate between the generator's raw price and retail, with a licensed partner supplier handling top-up and billing. Bethesda members report ~24% average savings, and the club is closed with a waitlist — demand pre-aggregated with no campaign, because the campaign already happened and hit a supply wall. (An in-flight settlement reform, P441, would clean up the plumbing; it is treated strictly as a windfall — no rung depends on it.)
Rung 6 — the flexibility member and the anchor battery. Batteries and smart thermal stores enrolled where they already pay solo (time-of-use arbitrage stands alone); mutual-fronted, savings-split batteries behind anchor-business meters. The club's match percentage rises mechanically (Bethesda ran ~52%); the "community battery" exists only as an accounting fact emerging from selfish placements.
Rung 7 — the winter mutual. The one leg no household solves alone, mutualised honestly: first call on the shared winter-peaking asset (in Ogwen, the existing 100kW hydro) in exactly the half-hours home kit fails, plus pooled flexibility and — proposal — aggregated purchasing of the residual. No rung may promise grid independence: winter import and standing charges are bought, not designed away.
Rung 8 — cell-in-a-box. The cell packages itself: the rung-0 artefact kit, a pre-negotiated supplier framework, shared back-office — sold to the next community's development trust, collapsing setup from years to weeks. Replication pays the coordination bill: the levy funds the paid officer that keeps every cell off volunteer burnout.
The worked instance — Bethesda enters mid-climb
The strongest evidence that the ladder is real: the Ogwen valley did not start it — it is already standing on rungs 0, 4, 5, and 7's asset (Partneriaeth Ogwen; Ynni Ogwen CBS and its share raise; the first Energy Local club in the UK, now waitlisted; the winter-peaking hydro). The verified binding constraint is generation supply, not demand — so the valley's next moves are precisely the missing rungs: a roof survey then Egni-pattern leases on the frozen roofs; a doorstep cylinder survey then a bulk diverter rollout to lift the match percentage; each new roof mechanically reopening waitlist places; then the whole thing documented as the first cell-in-a-box master copy. First-week homework, before anything else: obtain Ynni Ogwen's monthly generation data — the hydro-covers-winter claim currently rests on general UK run-of-river seasonality, not this hydro's meters, and it is not allowed to bear load until it does.
The gate verdict — printed, not managed
Three independent adversaries, three verdicts of passes-with-conditions. The narrowed claim that survives, stated exactly:
Bridge-clean and coordination-free for the median adopter at rungs 1–4; coordination-light (paid, budgeted, never volunteered) at rungs 5–8; and false for the acute fuel-poverty cohort, whose access requires funded outreach and a community-benefit fund that actually exists and performs. Bethesda's own history proves both the need and the mechanism.
The conditions that must be discharged before any rung's template ships as "ready":
- The capital layer must drop the "plain financial product" claim. The evidence says community-share capital is civic-blended (~85% local), and the 2026 economics (whether an ~18p PPA at today's install costs still clears 4–7% after debt and the developer fee) have not been re-modelled. This is the weakest load-bearing rung; if the re-model clears only 2–3%, the capital rung is faith-dependent and the template must say so.
- A residual saint survives in the design: the CBS legally requires unpaid volunteer directors carrying fiduciary liability. The paid officer does not cover this. Either board support is budgeted too, or the template admits one rung of civic contribution it cannot pay for.
- No projection may price any rung off today's spreads. Ofgem's live shift of network-cost recovery onto standing charges shrinks the trading rungs' fuel; economics must be based on self-consumption, avoided-network and flexibility value — the parts that persist.
- A solicitor must eyeball the primary exemption text (SI 2001/3270 Schedule 4 and the January 2026 guidance) — the class caps and private-wire reading currently rest on secondary triangulation.
- Rung 2's install cost and rung 6's second-life battery assumptions are unverified; each rung stays flagged until its number is real.
The walls ledger
| Wall | Class | Residue |
|---|---|---|
| Seasonal storage | Physical floor | Summer-to-winter storage is ~two orders of magnitude beyond daily cycling; winter import is bought, with its standing charges. |
| Finite asset pool | Physical floor | Viable roofs, adjacencies, waitlist depth and the ~85%-tapped local investor base are surveyed quantities, not designables. |
| Network-cost shifting | Political contest | Cell savings shift fixed-cost recovery onto roofless, low-usage non-participants — regressively, and growing with success. The in-design answer: cells voluntarily pay a fair capacity-based backbone share. The honest answer: this is a distributive contest no template dissolves. |
| Revocable exemption footing | Political contest | Every rung below 5 stands on self-assessed class exemptions inside a live Ofgem review — and energy licensing is reserved, so Wales holds no devolved lever if Westminster tightens. Mitigation is posture: move while the window is open. |
| Split incentive; fuel-poor access; coordination labour; the originator; supplier counterparty; spread compression | Frame-artefacts | Each dissolved into the design (mutual-owned assets; funded outreach from ring-fenced surplus; paid coordination; grant-funded origination; partner-supplier model with a licensed fallback; self-consumption basing) — with the residue of each named above rather than hidden. |
The optional layer, kept optional
A local currency/credit layer can sit on top of a working cell — paying for the Δω-positive work the market ignores — and an idea-stage proposal for a delivery surface already exists in these notes. It is deliberately not part of this template: the cell must stand on kilowatt-hours and pounds alone. A template that needs its own money is not a bridge.
What this establishes
The treatment gives the fitness-function note its second real case, and the gate its first formal run: the bridge test, applied adversarially, neither rubber-stamped the design nor killed it — it narrowed the claim to what the evidence carries, which is exactly what a gate is for. And it surfaced the general lesson early: in this domain the bridge exists for the median adopter, while the cohort that needs it most requires the one thing the test forbids — which is not a reason to abandon the test, but the honest shape of the equity residue, now priced (a ring-fenced surplus and funded outreach) instead of wished away.
Not yet promoted into propositions.ts. Forming register; scaffold name. The template ships as "ready" only when its five conditions are discharged; the Ogwen instance starts with a roof survey, a cylinder survey, and one request for the hydro's monthly meter data.