Framework 02 · Pipeline

Horizon Pipeline.

Every renewable project — from feasibility through commissioning — linked to its developer, financier, equipment suppliers and grid constraint. Per-project gate-by-gate connection probability, calibrated against actual NESO and equivalent outcomes.

Request a demo Read the queue analysis →

The problem

700 GW in the queue. Most will not connect on time.

Over seven hundred gigawatts of nominal contracted capacity sits in the UK transmission-entry-capacity queue. Many projects will not connect on their original date. Many will never connect at all. The old project-finance model — discount the queue date by twelve months — has stopped working. Horizon Pipeline replaces it with a gate-by-gate hazard function.

What changed · TM04+ and the gating reforms Recent NESO reforms convert the queue from a first-come-first-served list into a milestone-gated cohort. Projects that miss milestone gates are pushed back, sometimes by years. The new risk model treats every gate as a hazard event, with a probability and a calibrated interval. Horizon Pipeline scores every active project against every gate.

The graph

11,800 projects. Five entity classes. One graph.

Nodes

Projects · Developers · Financiers · OEMs · Grid nodes

Every active or recently-active project in the eleven jurisdictions we cover at developer-grade detail. Each project linked to the developer of record, the most recent financing round, the named equipment suppliers, and the relevant grid constraint.

Edges

Ownership · finance · supply · constraint

Edges carry attributes: ownership share, finance instrument and tenor, supply contract status, queue position and gate history. The graph is queryable through a standard interface; outputs include subgraphs scoped to a portfolio.

Gates

Per-project hazard function

For each project, the hazard at each future gate is modelled from queue position, milestone evidence in the documentary record, comparable project history, and policy probability from Horizon Policy. Calibrated quarterly against closed outcomes.

Outputs

Connection probability + interval per gate

For every project, every gate: P(pass), 90% interval, calibration class. For every portfolio: weighted IRR and capacity outturn distribution. For every developer: queue exposure ranked.

Sample output

Gate-by-gate scoring on a real-shape portfolio.

Project (anonymised)	Tech · MW	Original TEC	P(connect by TEC)	Most likely slip	Driver
SE-OFFSH-04	Offshore wind · 720 MW	2029	0.61	+9 mo	Consenting backlog
NE-SOLAR-22	Solar · 180 MW	2027	0.47	+14 mo	TM04+ gate slip
SW-WIND-09	Onshore wind · 64 MW	2026	0.72	+3 mo	Distribution co-ordination
EM-BESS-17	BESS · 220 MW	2026	0.81	+2 mo	Commissioning slack
WALES-SOLAR-11	Solar · 96 MW	2028	0.34	+19 mo	Queue cleansing
SCOT-OFFSH-02	Offshore wind · 1.2 GW	2031	0.43	+22 mo	Transmission build
NW-SOLAR-31	Solar · 240 MW	2027	0.55	+11 mo	Distribution upgrade
SE-BESS-08	BESS · 100 MW	2026	0.78	+3 mo	EPC contracting

Illustrative. Tenant outputs are tied to identified projects, with full documentary trail and quarterly calibration against actual gate outcomes.

What it changes

Three places money has been mispriced.

Late-stage acquisitions of ROC-era assets

Where a TEC date is treated as a fact rather than a probabilistic forecast, transactions are routinely overpaid by hundreds of basis points of IRR. Pipeline outputs the IRR distribution under the gate-by-gate hazard function.

Sleeve PPAs priced off original TEC dates

Sleeve and corporate PPAs increasingly carry connection-slip clauses. Pricing them off the calibrated probability rather than the headline date is now table stakes for serious offtakers.

Equipment forward orders

OEMs forward-order against developer pipelines that systematically over-state on-time connection. Pipeline outputs let OEMs hedge slot allocation against probabilistic commissioning curves.

Coverage

Eleven jurisdictions, deepening.

Pipeline launched on the UK, where the queue dynamics force the issue. We have since added Germany, Ireland, the Netherlands, Belgium, France, Spain, Italy, ERCOT, MISO and CAISO. Each new jurisdiction is added only when we have eighteen months of closed outcome data to calibrate against; we publish each calibration class as it goes live.

The next coverage release — under quarterly review — covers PJM, the Nordics, and the Iberian peninsula at distribution-network resolution.