Planning data for the GB energy system in May 2026: FES, SSEP, CSNP, the Energy Digitalisation Framework and the DSI roadmap

Where the planning data stack stands in May 2026

The picture in May 2026 has shifted enough in the last twelve months that a short orientation is the right place to begin. Three moves reshape the stack. The first is the Centralised Strategic Network Plan methodology approval by Ofgem in April 2026. The methodology document sets out how NESO will compute the network investment envelope across transmission and distribution, how it reconciles inputs from the three transmission owners and the six distribution licence groups, and how the resulting plan feeds into the price-control framework. The first transitional CSNP, the T-CSNP, is due in June 2026 and the first full CSNP delivery is due by the end of 2028.^{10 6}

The second move is the Strategic Spatial Energy Plan timetable. NESO is developing SSEP with DESNZ under section 33A of the Energy Act 2023; the methodology was approved in May 2025, the first SSEP iteration is due in Q4 2026, public consultation runs in early 2027 and the final SSEP is due in Autumn 2027.⁵ SSEP is the first GB-wide statutory plan that locates where new electricity generation, storage, hydrogen production and CCUS infrastructure should sit at the regional level. It is the spatial counterpart to FES, and it is the input that CSNP reads to compute the network reinforcements that follow each spatial placement.

The third move is the FES cadence change. NESO has shifted FES from an annual publication to a multi-year cadence, with the next full FES due in 2028 and an Economics Annex added in 2025 to carry forward-looking cost projections between full editions. The annex extends the FES envelope without requiring a complete pathway rebuild every twelve months. The cadence change matches the move of the centre of gravity from FES (national pathway envelopes) to SSEP and CSNP (spatial placement and network investment), which now do more of the year-on-year work.³

Two further moves sit alongside these three. The Energy Digitalisation Framework, published jointly by DESNZ and Ofgem on 23 March 2026, sets a five-layer model that any data and digital programme in the sector has to align with: applications, governance, standards, communications, and physical assets. The framework is non-statutory but every commitment in it is implemented through statutory levers that already exist (Ofgem licence conditions, the Data (Use and Access) Act 2025, and the strategic policy statement to NESO).⁷ The Data Sharing Infrastructure first-draft architectural reference framework is due in August 2026; NESO holds the interim coordinator role under the OFG1164 Governance Decision of March 2025.^{8 9}

The May 2026 inflection is that every artefact in the stack now has a published, dated cadence that the rest of the regime can plan against. Before April 2026 the CSNP methodology was still in consultation and the SSEP iteration date was a working assumption. With the methodology approved and the iteration timetable confirmed, the strategic-planning data stack is for the first time a synchronised production line rather than a set of separately drifting clocks.

The five artefacts and how they depend on each other

The planning data stack divides into five named artefacts on three substantive layers. The pathway envelope (FES) is the national projection layer. The spatial and network plans (SSEP and CSNP) are the placement and investment layer; SSEP locates the new capacity that the FES envelope describes, and CSNP reinforces the network so that the SSEP placement can be delivered. The data-and-digital substrate (EDF and DSI) is the layer that lets the licence regime exchange inputs and outputs between the other three artefacts. The substrate matters because without a shared data model and a federated exchange, every CSNP iteration would have to negotiate bilateral data-sharing agreements with each transmission owner and each DNO group, and the cadence would not hold.

Two dependencies are load-bearing. SSEP depends on FES; the spatial placement uses the national envelope as the right-hand side of its allocation problem. CSNP depends on SSEP; the network investment plan reinforces the routes between the SSEP-placed assets and the demand centres they serve. Both dependencies are reflected in the published cadence: FES Economics Annex 2025, SSEP iteration 1 Q4 2026, CSNP T-CSNP June 2026, CSNP delivery end-2028. The CSNP transitional iteration in June 2026 runs before the SSEP iteration in Q4 2026 because the transitional plan is allowed to use the existing FES and the published methodology without waiting for the first SSEP cycle; the full CSNP in end-2028 reads the final SSEP of Autumn 2027 as one of its inputs.

The data-and-digital substrate has its own internal cadence. The Energy Digitalisation Framework of 23 March 2026 sets the five-layer model that every programme in the substrate aligns to. The DSI architectural reference framework first draft is due in August 2026 and is the document that turns the five-layer model into named integration patterns, semantic models and trust-spine commitments. The OFG1164 Governance Decision of March 2025 holds NESO as the interim coordinator for the substrate; the framework sets out a path to a permanent Digitalisation Coordination Function with a DESNZ consultation by end-2026.^{9 7}

The five-artefact partition is operational, not statutory. FES, SSEP and CSNP each have a different statutory or regulatory base. FES is a NESO publication under the Strategic Priorities document and the licence; SSEP is mandated under section 33A of the Energy Act 2023; CSNP is mandated through the Ofgem methodology approval and the price-control framework. The Energy Digitalisation Framework is a joint DESNZ and Ofgem framework with no single statutory base of its own; each commitment in it is implemented through an existing statutory lever (typically a licence condition).

The reader who arrives for the cadence reference uses the table above. The reader who arrives wanting to know how an individual artefact reads its inputs and writes its outputs uses the sections below. Each section sets out the artefact's purpose, its current methodology, the inputs it reads and the consumers that read its outputs.

The Future Energy Scenarios on a multi-year cadence with the 2025 Economics Annex

The Future Energy Scenarios are the long-running NESO scenario set that describes possible system futures across demand, generation, storage, hydrogen, heat and transport to 2050. FES has been the canonical projection stack for the GB electricity system since 1995 and has been published as FES (rather than as the predecessor Seven Year Statement) since 2011. The 2025 edition was the first published as NESO rather than as the National Grid System Operator, following NESO's establishment under the Energy Act 2023.³

The cadence change in May 2026 is the move from annual full publication to a multi-year cadence with an Economics Annex added between full editions. The 2025 Economics Annex carries the forward-looking cost projections (capital cost trajectories for offshore wind, onshore wind, solar PV, batteries, electrolysers, nuclear and hydrogen-ready gas plant; fuel-price trajectories for natural gas, coal and biomass; carbon-price trajectories under the UK Emissions Trading Scheme) that previously refreshed in each full FES. The annex extends the FES envelope by twelve months without requiring a complete pathway rebuild. The next full FES is due in 2028, which lets the next pathway rebuild absorb the first SSEP iteration of Q4 2026 and the T-CSNP of June 2026 as inputs rather than parallel publications.³

The 2025 FES kept the four-pathway structure (Holistic Transition, Electric Engagement, Hydrogen Evolution and Counterfactual in the 2024 names; the names evolve between editions). Holistic Transition treats the Clean Power 2030 target as the binding constraint. Electric Engagement assumes deeper electrification of heat and transport. Hydrogen Evolution leans on hydrogen for heavy industry and dispatchable generation. Counterfactual is the baseline that shows what the system would look like if 2024 policy ambition stayed put. Each pathway is run at hourly resolution to 2050 and consistent with one of the Climate Change Committee carbon budgets.

What the FES dataset contains

An FES dataset for a single pathway contains the demand projection (total, peak, by sector, by region, by hourly profile), the generation mix (capacity and output by technology, by region), the flexibility and storage stack (battery capacity, pumped hydro, demand-side response, vehicle-to-grid, interconnectors), the hydrogen system (production split between electrolysis and reformation with carbon capture, storage, pipeline infrastructure, demand by sector), the heat sector (heat pump deployment, district heating, hydrogen boilers, hybrid systems, by region), the transport sector (EV uptake curves, charging infrastructure needs, electricity demand from transport), and the carbon-emission trajectory at the sector level. The Economics Annex carries the cost and price trajectories that feed every pathway. The dataset is published under the NESO Open Licence and is the reference projection set for almost every market-facing analysis in GB.

The Electricity Ten Year Statement (ETYS) is the network-impact counterpart to FES. ETYS reads the FES pathways and computes the transmission-network reinforcements that each pathway requires; the 2024 ETYS was the most recent edition.⁴ Under the post-REMA architecture, CSNP absorbs the ETYS role for the new build-out across both transmission and distribution, but ETYS continues to publish for the transmission-only operational view.

How DNOs translate FES at the distribution level

Each of the fourteen distribution licence areas across the six DNO groups runs a Distribution Future Energy Scenarios cycle that decomposes the FES national pathways by primary substation, by Bulk Supply Point and by year. The DFES feeds the DNO's Network Development Plan and its capacity heatmap, which in turn drives connection offers and reinforcement decisions. Where DFES quality is high, applications are routed to fast-track processes; where DFES quality is lower, DNOs default to conservative assumptions and connection cost estimates inflate. The DFES cadence is annual but staggered: each DNO publishes on its own calendar across the year.

The DFES output is the input to two downstream artefacts. The Embedded Capacity Register, published per DNO under Standard Licence Condition 50 with monthly updates within ten working days of month-end, records every generation and storage plant connected at the distribution level above one megawatt (with smaller schemes down to fifty kilowatts carried for visibility where the DNO holds data). The Long Term Development Statement, published under Standard Licence Condition 25 with the third Ofgem derogation letter of 13 May 2026 holding the Stage 2 publication date at 29 May 2026 in a validated CIM data model, records the asset state, the contracted load and generation, and the planned reinforcement at every substation in the DNO's area.

The shift from PDF to a CIM data model in the LTDS is the precondition for the strategic-planning artefacts above to read distribution-level inputs without bespoke parsing. Without a shared data model the spatial placement work in SSEP would have to take fourteen different file formats and reconcile them by hand for every iteration; with the LTDS in CIM format the placement work can run against a single schema across the six DNO groups. The same data-model substrate matters even more for CSNP because the network reinforcement plan has to compute thermal-loading and voltage-rise budgets at the asset level across every licence area simultaneously.

The FES feedback loop into the LTDS and back

The FES pathway envelope is not a single forecast; each pathway is an internally consistent world built around a different set of assumptions about consumer adoption, policy speed and technology cost. Each pathway therefore reads as a sensitivity envelope rather than a point projection. A market-facing analyst at a supplier reading the Holistic Transition pathway alongside the Counterfactual sees the range of plausible system states the planning data stack has to accommodate, and the FES dataset is the canonical artefact the analyst reads against. The pathway envelope feeds back into the LTDS through two routes. The first route is the planning-horizon assumptions that each DNO uses in the LTDS Stage 3 publication of 30 November 2026 (which adds the planning data and the connection-stage data to the Stage 2 topology); each DNO reads the FES pathway envelope against its own DFES decomposition and produces a Stage 3 publication that the next CSNP cycle then reads.

The second route is the connection-pipeline data that feeds the next FES cycle. A connection offer issued under the post-Gate-2 process becomes a committed pipeline entry; the committed pipeline is one of the inputs that the next FES pathway rebuild uses to calibrate the demand and the generation projection. The cadence below FES therefore reads upward and downward at once: the FES informs the connection process through the SSEP and the CSNP; the connection process informs the FES through the committed pipeline. The Economics Annex of 2025 carries the cost trajectories that bridge between full FES editions while the feedback loop continues to operate.

The multi-year cadence is the change that makes the feedback loop sustainable. Before the cadence change the FES rebuild ran every twelve months, which left the next iteration absorbing inputs from a network state that had moved underneath it; the SSEP iteration in Q4 2026 and the T-CSNP in June 2026 each take the FES envelope as a published reference, and the next FES rebuild in 2028 absorbs the final SSEP of Autumn 2027 and the operational outturn from the post-Gate-2 connections as substantive inputs. The cadence below the FES is therefore not a slower production line; it is a production line that has been re-paced to match the substantive cycle of the spatial and network plans.

The Strategic Spatial Energy Plan from Q4 2026 iteration to Autumn 2027 final

The Strategic Spatial Energy Plan is the post-REMA spatial counterpart to FES. NESO is developing SSEP with DESNZ under section 33A of the Energy Act 2023; the section obliges NESO to prepare a Strategic Spatial Energy Plan, to keep it under review and to revise it as the Authority directs, and obliges the Secretary of State to lay the SSEP before Parliament. The statutory hook is stronger than for any previous planning artefact in GB energy: FES, ETYS and the Holistic Network Design are all NESO publications under licence, but the SSEP is a statutory plan whose status as a material consideration for planning consents is anchored in primary legislation.⁵

The timetable is set: the methodology was approved in May 2025; the first SSEP iteration is due in Q4 2026; the public consultation runs in early 2027; the final SSEP is due in Autumn 2027. The methodology document sets out how NESO will compute the spatial placement across the eleven Regional Energy Strategic Plan administrative regions (one for Scotland, one for Wales and nine for England, broadly aligned with the Government Office regions), how it will reconcile against the FES pathway envelope, how it will treat environmental and land-use constraints, and how it will publish the result as an input to CSNP and to the connections queue reform.⁵

What SSEP iteration 1 will contain

The first SSEP iteration covers electricity generation, electricity storage, hydrogen production and CCUS infrastructure across GB. The output is a spatial map of optimal locations for each technology, with sensitivity to the FES pathways and to the eleven RESP regions. The most data-intensive single artefact in the planning stack is SSEP because the placement work reconciles generation capacity and pipeline (from the connections queue and the FES projections), network capacity and constraints (from the LTDS CIM models and the DNO heatmaps), demand profiles by zone, land-use and environmental constraints, hydrogen production potential, economic and demographic data, and offshore wind resource data. Every input has to read against the same map projection and the same temporal resolution.

The first iteration is presented as a working draft for consultation; the consultation in early 2027 lets stakeholders engage with the spatial choices the methodology produces; the final SSEP of Autumn 2027 absorbs the consultation responses. Subsequent SSEP iterations follow on a three-year cycle aligned with the FES pathway update window. The three-year cycle is the cadence at which SSEP outputs cascade into the Holistic Network Design iterations, the RESP refinements and the connection-window reforms.

What SSEP is not

SSEP is not a planning consent. The SSEP locates optimal regions; the planning consent for each individual asset still sits with the local planning authority under the Town and Country Planning Act 1990 (for England and Wales) or the Planning Etc. (Scotland) Act 2006, with Nationally Significant Infrastructure Projects routed under the Planning Act 2008 and the National Policy Statement framework. SSEP gives the local planning authority a material consideration that they have to weigh in their decision; it does not replace the local consent.

SSEP is also not a connection offer. A developer who reads SSEP and identifies a placement region still has to apply for a connection under the post-Gate-2 reform process; the connection offer is computed against the CSNP-reinforced network rather than against the queue position alone. The connections queue reform of April 2026, which progressed 283 gigawatts of generation and storage and 99 gigawatts of demand across Phase 1 to 2030 and Phase 2 to 2035, set the queue rule book that SSEP-aligned projects are now being routed through.¹¹

The risk of late SSEP delivery

If SSEP iteration 1 slips past Q4 2026, the Clean Power 2030 capacity envelope continues to act as the proxy spatial plan, but the post-Gate-2 connections process operates on weaker spatial signal, and the next full FES has to work harder to reconcile to the eventual SSEP. The most likely consequence of slip would be a long tail of SSEP-misaligned projects that have to be unwound after the fact rather than placed at first connection offer. The cadence below SSEP, especially CSNP T-CSNP in June 2026 and full CSNP delivery in end-2028, is dimensioned around the SSEP iteration arriving on time; the SSEP timetable holds the network reinforcement timetable together.

The eleven RESP regions and how SSEP reads them

The Regional Energy Strategic Plans split GB into eleven administrative regions: one for Scotland, one for Wales and nine for England (broadly aligned with the Government Office regions: North East, North West, Yorkshire and Humber, East Midlands, West Midlands, East of England, London, South East and South West). The geography follows administrative boundaries rather than DNO licence areas, which forces an additional reconciliation step but lets the RESP engage local-authority planning teams and Combined Authority strategy teams in their familiar geography. SSEP reads the eleven RESP regions as the canonical regional decomposition for the spatial placement, and the SSEP iteration 1 of Q4 2026 publishes spatial placement at the RESP level.

Each RESP region carries a distinct energy character that SSEP reads into the placement work. Scotland carries major onshore and offshore wind, subsea island connections, hydrogen potential and the B6 boundary transmission constraint. Wales carries Celtic Sea floating wind, the Wylfa nuclear site and the industrial South Wales cluster. The North East carries offshore-wind connections and industrial decarbonisation. The North West carries the HyNet hydrogen cluster, the Heysham nuclear cluster and high urban demand. Yorkshire and Humber carries the Humber CCS cluster, offshore wind and industrial demand. The East Midlands carries EV manufacturing, logistics corridors and rural solar. The West Midlands carries the urban demand centre, transport electrification and battery manufacturing. The East of England carries the East Anglia offshore wind cluster, the Bacton gas terminal and agricultural solar. London carries the highest demand density, heat networks and transport electrification. The South East carries the IFA interconnector, the Dungeness nuclear site and high housing growth. The South West carries solar resource, Celtic Sea floating wind and rural-network constraints.

The transitional RESP was published on 30 January 2026 as a first-of-kind baseline; the RESP methodology consultation ran from November 2025 to January 2026; the final RESP methodology is due in Summer 2026, with the full eleven RESPs following the methodology. The RESP cycle runs alongside the SSEP cycle and the two artefacts will live alongside each other for at least one cycle; the reconciliation between RESP geographies (which follow administrative boundaries) and DNO licence areas (which do not) is the next substantive governance question for the planning data stack.

The Centralised Strategic Network Plan from April 2026 methodology to end-2028 delivery

The Centralised Strategic Network Plan is the network-investment counterpart to SSEP. NESO is developing CSNP with Ofgem under the post-REMA architecture; the methodology was approved by Ofgem in April 2026, the transitional T-CSNP is due in June 2026 and the first full CSNP delivery is due by the end of 2028.^{10 6} The CSNP replaces the Network Options Assessment as the canonical network-investment plan; it sets out the network reinforcements needed to deliver the assets that SSEP places, and it sets the investment envelope that Ofgem then approves through the price-control framework.

The April 2026 Ofgem methodology approval document is the load-bearing decision behind the rest of the CSNP cadence. The document sets out how NESO will compute the network investment envelope across transmission and distribution, how it reconciles inputs from the three transmission owners (National Grid Electricity Transmission, SP Transmission, SSEN Transmission) and the six distribution licence groups, how it incorporates the SSEP spatial placement, how it allocates reinforcement to nodal versus zonal solutions, and how the resulting plan feeds into the price-control framework. The methodology is the rulebook that the T-CSNP and the full CSNP each iterate against.¹⁰

The T-CSNP transitional iteration of June 2026

The transitional T-CSNP is due in June 2026, which is two months after the methodology approval. The transitional plan is allowed to read the existing FES pathway envelope and the published methodology without waiting for the first SSEP iteration of Q4 2026. The T-CSNP covers the network reinforcements that the existing connection queue (post-Gate-2) and the Clean Power 2030 envelope together imply, and it sets the investment envelope that Ofgem then approves for the first wave of post-methodology projects. The T-CSNP is a working transitional artefact rather than the full delivery; the full CSNP at end-2028 is the cycle that integrates the final SSEP of Autumn 2027 as a substantive input.

The cadence below the T-CSNP is the price-control allowance process. Each CSNP iteration sits inside the RIIO-T3 transmission price control (which begins April 2026) and the RIIO-ED3 distribution price control (which begins April 2028). The price-control framework converts the CSNP investment envelope into allowed expenditure across the price-control period; the licensees deliver against the allowed expenditure and report quarterly to Ofgem under the Regulatory Information Guidelines.

What the full CSNP at end-2028 contains

The full CSNP at end-2028 reads the final SSEP of Autumn 2027 as the spatial-placement input, reads the next full FES of 2028 as the pathway-envelope input, and reads the LTDS CIM publications from each DNO and the ETYS from NESO as the network-state inputs. The output is a single integrated investment plan across transmission and distribution, with each reinforcement project named by location, voltage, technology, justification (which SSEP placement it serves and which FES pathway it is consistent with), commissioning date and cost. The plan is the canonical reference that the price-control framework and the connection offer regime both read.

The CSNP integration of transmission and distribution is the post-REMA architectural shift the methodology approval enables. Before April 2026 the network-investment work split across the Network Options Assessment for transmission (run by NESO with the three transmission owners) and the bilateral DNO-by-DNO planning at distribution (under each DNO's own Network Development Plan). The CSNP folds both into a single integrated planning cycle, which means the network investment for a new offshore wind farm in the North Sea can be planned alongside the distribution reinforcements in the South East that absorb the export, in one document with one methodology and one price-control envelope.

How CSNP reads its inputs

The CSNP cycle has four substantive input streams. The first is the SSEP spatial placement, which sets where the new generation, storage and hydrogen production sits. The second is the FES pathway envelope, which sets how much capacity each technology has at each year out to 2050. The third is the network state at each transmission and distribution licence area, which the LTDS CIM publications carry. The fourth is the demand-side projection from the DFES and the supplier-aggregated half-hourly settlement data after the Market-wide Half Hourly Settlement programme cuts over in July 2027. Without GC0139 the licence regime would have to negotiate bilateral data-sharing agreements for each stream; with GC0139 in place each stream is a licence-mandated publication.¹

The price-control envelope that CSNP sits inside

The CSNP investment envelope sits inside the RIIO price-control framework. RIIO-T3 is the transmission price control that began in April 2026 and runs to March 2031; RIIO-ED3 is the distribution price control that begins in April 2028 and runs to March 2033. Each price-control period sets allowed expenditure for each licensee against the outputs the licensee is contracted to deliver; CSNP is the document the price-control framework reads to compute the network investment that each licensee is allowed to recover from consumers. The CSNP cadence is therefore not a free-standing planning exercise; it is the input to a regulatory mechanism that converts the plan into allowed expenditure and then holds the licensees to delivery against the allowance.

The price-control framework runs on a five-year cycle; CSNP runs on a longer iteration cycle. The T-CSNP of June 2026 lands inside the RIIO-T3 period at the second month of the cycle, which lets the first wave of post-methodology projects enter the price-control allowance window without waiting for the next periodic review. The full CSNP at end-2028 lands inside the RIIO-ED3 period at the eighth month of the cycle, which lets the integrated transmission-and-distribution investment plan enter the next round of price-control allowance windows. The cadence is calibrated so that each CSNP iteration arrives at a point in the price-control cycle where the framework can convert the plan into allowed expenditure without a separate code change.

The reinforcement type space that CSNP allocates over

The reinforcement work that CSNP allocates over splits into three substantive types. The first type is the bulk reinforcement (new transmission lines, new substations, new high-voltage cables) that the post-Gate-2 connection pipeline and the SSEP spatial placement together require; the NESO Beyond 2030 work scoped roughly eighty gigawatts of new transmission capacity by 2030, and the CSNP reads that scoping into specific transmission projects. The second type is the targeted reinforcement (uprated transformers, additional bays at existing substations, dynamic-rating reconductoring) that addresses thermal or voltage-rise constraints that the LTDS publications and the DNO heatmaps surface. The third type is the network-management investment (Active Network Management schemes, capacitor and reactor switching, STATCOM and Static Var Compensator devices) that defers or avoids reinforcement by using existing capacity more efficiently. CSNP allocates across all three types, and the methodology document sets out how the allocation is computed for each substation and each transmission boundary.

The cadence below the reinforcement allocation is the procurement cycle. Each named reinforcement project enters a procurement window under the price-control framework; the licensee procures the equipment, the engineering and the construction; the project is commissioned at the indicative date in the CSNP plan. The lead time from CSNP publication to commissioning is typically five to seven years for bulk transmission reinforcement, eighteen months to three years for targeted distribution reinforcement, and twelve months to two years for network-management investment. The lead times are the reason the cadence of the planning data stack matters: a five-year-out CSNP iteration that misses the next price-control window slips the commissioning by the full lead time, which means the connection offers that depend on the reinforcement slip by the same amount.

The Energy Digitalisation Framework five-layer model

The Energy Digitalisation Framework was published jointly by DESNZ and Ofgem on 23 March 2026. It is the first time the two bodies have produced a single shared framework for the data and digital track. The full title is "Energy Digitalisation Framework: a shared vision for digitalising Great Britain's energy system" and it was issued under joint signatures by the Secretary of State for Energy Security and Net Zero and the Chief Executive of Ofgem.⁷ The framework is non-statutory in itself, but every commitment it contains is implemented through statutory levers that already exist: Ofgem licence conditions, the Data (Use and Access) Act 2025, and the strategic policy statement to NESO.

The framework partitions the data and digital landscape into five layers. The five layers form a stack from the user-facing top down to the asset-facing bottom: applications at the top, governance just below them, standards in the middle, communications below standards and physical assets at the foundation. Each layer has a distinct purpose and a distinct set of lead bodies; together they describe the substrate the rest of the planning data stack reads its inputs and writes its outputs through.

Layer 1: Applications

The applications layer is the layer where a user (a planner, a supplier, a connection applicant, a regulator, a consumer-facing service) reads or writes data. The layer covers the connection portals, the planning analytics tools, the market-facing dashboards, the consumer-facing services and the regulator-facing compliance tools. The applications layer reads against the standards layer below it and depends on the communications layer for transport. The framework sets the expectation that the applications layer will be open and discoverable through the DSI catalogue, which makes new applications cheaper to onboard and reduces lock-in to a single platform.

Layer 2: Governance

The governance layer holds the rules, the licence conditions and the codes that bind every data flow. The layer covers Standard Licence Condition 25 of the Electricity Distribution Licence (which produces the LTDS), Standard Licence Condition 50 of the same licence (which produces the Embedded Capacity Register), the Distribution Code and the Grid Code, the Balancing and Settlement Code (which Elexon runs), the Connection and Use of System Code (which NESO runs), the Smart Energy Code and the Retail Energy Code, the Data (Use and Access) Act 2025, the OFG1164 DSI Governance Decision of March 2025 and the Energy Smart Data and Privacy Framework. The governance layer is the layer through which any data publication obligation lands on a licensee; the framework does not displace the existing governance instruments but rather points at them as the implementation levers for each of its commitments.

Layer 3: Standards

The standards layer holds the data models, the schemas and the validation rules that every layer above reads and writes against. The layer covers the Common Information Model (IEC 61970-301 Edition 7.0 with Amendment 1:2022 for the EMS-API base, IEC 61968-13 Edition 2.0 (BS EN IEC 61968-13:2021) for the distribution profiles, ENTSO-E CGMES 3.0 with the Application Profiles Library v1.1.1 patch of October 2025 for the cross-border exchange), the SHACL validation that the BSI engagement portal hosts at https://cim.bsigroup.com/, the master resource identifiers (mRIDs) that tie each topology element to a stable identifier across publication cycles, and the standard data dictionaries that the DSI catalogue uses. The standards layer is where the LTDS Stage 2 publication of 29 May 2026 in CIM format sits; it is also the layer that the SSEP and CSNP cycles read for the network-state inputs.

Layer 4: Communications

The communications layer holds the transport infrastructure that carries data between the producers and the consumers. The layer covers the Smart Meter Communications Licence and the DCC infrastructure that runs the SMETS2 meter messaging, the DSI federated exchange that the OFG1164 Governance Decision sets out, the Balancing Mechanism Reporting Service (BMRS) that Elexon publishes, the NESO Data Portal, the DCC Data Service Provider interface and the Wide Area Network that the regional Communications Service Providers operate, and the Application Programming Interfaces (APIs) that each licensee publishes for its own datasets. The communications layer is the layer through which a planner reading the LTDS CIM dataset for the DNO area they are placing a project in actually receives the bytes.

Layer 5: Physical assets

The physical-assets layer holds the meters, the telemetry devices, the substations, the transformers, the lines and the cables that produce the data and that the data describes. The layer covers the 35.9 million domestic and small non-domestic electricity and gas meters in the field as of early 2026, the substation telemetry from the three transmission owners and the six DNO groups, the Phasor Measurement Units at the transmission level, and the asset register that each licensee maintains under its licence. The physical-assets layer is the foundation of the stack because every byte that flows through the four layers above it ultimately resolves to a measurement, a state change or a configuration of a physical asset that one of the licensees owns and operates.

The five-layer model reads upward and downward at once. A planner reading the SSEP iteration 1 of Q4 2026 receives the result through the applications layer; the application reads against the standards layer (CIM and the DSI catalogue) for the model and the metadata; the standards layer reads against the communications layer (DSI federated exchange) for the transport; the communications layer reads against the governance layer (OFG1164 and the licence conditions) for the rules; and the rules ultimately bind the licensees that own the physical assets at the bottom. The framework is the document that makes this reading explicit; every existing instrument continues to operate on its own terms.

What the framework does not displace

The framework is the strategic articulation; it does not displace any existing instrument. Data Best Practice Guidance v3.5 (June 2025) remains the operative Ofgem guidance for data publication. The Digitalisation Strategy and Action Plan licence condition (under which each licensee publishes its DSAP) remains the operative obligation for each licensee. The OFG1164 DSI Governance Decision of March 2025 remains the operative governance for the DSI. The framework points at each of these as the implementation lever for the commitments in the framework document, which means a reader who wants to know how a framework commitment will be enforced reads through to the underlying instrument.⁸

The seven design principles the framework asks every programme to align to

Alongside the five-layer model the framework holds seven design principles that every delivery programme must align to. The principles are the test that the EDF reader applies when reading any new data publication or any new platform proposal; each principle reads against one or more of the five layers and against one or more of the underlying instruments. The seven principles are: trusted and secure (datasets pass the DSI trust-framework checks and the cybersecurity controls aligned to the NIS Regulations 2018 and the Cyber Assessment Framework); interoperable by design (common standards adopted from the start, CIM where applicable, published API contracts); simple and accessible (discovery through the DSI catalogue, low barriers for small and medium enterprises and consumers, documented licensing terms); responsive and future-proof (adaptable to hydrogen, CCUS, heat networks and emerging storage; modular publication patterns); efficient and cost-effective (avoiding duplication, building on the DSI, DIP and CCS rather than parallel platforms, gating through the Strategic Innovation Fund against re-invention); innovative and competition-driven (enabling new services and business models through data access, removing gatekeeping where unjustified); and deliverable (phased onboarding, coordinators stepping up by 2027, the Digitalisation Coordination Function operational from 2028 onward, milestones tied to RIIO-3).

The principles do real work. Principle five (efficiency) is the basis on which the Strategic Innovation Fund refuses to fund a project that re-invents a function already delivered by the DSI or the DIP. Principle one (trust) is the basis on which the trust framework rejects a dataset publication that does not show evidence of provenance and quality control. Principle six (innovation) is the route by which the Smart Data scheme under the Data (Use and Access) Act 2025 connects energy data to the cross-sector economy. The first time the principles were used to push back on a proposal in a code panel meeting was the BSC Panel meeting of November 2025, which referenced principle five to reject a proposed parallel data exchange route that would have duplicated the Data Integration Platform. The minute of that meeting is the cleanest precedent for the principles having teeth.

The four functional data domains under the framework

Alongside the five-layer model and the seven principles, the framework also partitions the data landscape into four functional domains and assigns coordination responsibility to a named organisation in each. The four domains are: Core energy system service (coordinator NESO, confirmed, covering operational data including frequency and balancing-mechanism actions and curtailment, asset and topology data, planning data including FES, ETYS and RESPs, and the connections queue); Behind-the-meter asset (coordinator Elexon, provisional, covering static and dynamic data on solar PV, batteries, EV chargers, heat pumps and other distributed energy resources, plus aggregator data); Consumer (coordinator the Retail Energy Code Company, confirmed, covering switching, consent, tariff, vulnerability and usage-related data routed through the consumer consent path); and Metering data (coordinator Elexon, provisional, covering half-hourly and non-half-hourly electricity meter reads, gas meter reads and export meter data for Smart Export Guarantee payments).

Two of the four coordinator designations are provisional. Elexon's appointment as the Behind-the-Meter coordinator depends on the outcome of the Ofgem Enhancing Asset Visibility consultation that Elexon responded to in February 2026; Elexon's appointment as the Metering Data coordinator depends on the integration decision between the Smart Meter Data Repository and the Settlement Data Repository. Either consultation can produce a different organisational arrangement, and the framework explicitly permits the coordinator to change without amending the framework document itself. The DESNZ consultation on the permanent Digitalisation Coordination Function by end-2026 is the cycle in which the provisional designations are confirmed or revised.

The DSI roadmap and the architectural reference framework due August 2026

The Data Sharing Infrastructure is the operational backbone of the EDF's presumed-open data principle. The DSI is a federated discovery, identity, consent and exchange layer that lets any participant in the GB energy system (a supplier, a DNO, NESO, a third-party flexibility provider, an academic researcher, the regulator) request and receive data from any other participant under common identity, consent and schema rules. The DSI does not store the underlying data; it is the wiring rather than the warehouse. The data continues to live with whoever owns it (settlement data with Elexon, network connection data with the DNOs, smart-meter consumption data with the suppliers under the DCC arrangements); the DSI publishes the discovery catalogue, runs the identity and consent checks, and routes the request to the data owner.⁸

The DSI has been in delivery since the Digital Spine Feasibility Study of 2023 and the DSI Pilot of September 2024. The OFG1164 DSI Governance Decision of March 2025 set the governance model and named NESO as the Interim Coordinator with a defined transition path to a permanent operator.⁹ The MVP is in delivery against a small set of high-priority data products through the NESO Interim Coordinator function in 2026; the Public Beta runs from 2028 with the product catalogue expanded and the long tail of participants onboarded; the permanent operator is expected to take over from NESO after the consultation on the Digitalisation Coordination Function closes (DESNZ to consult on the DCF form and powers by end-2026; mobilisation two to three years after consultation closes).

The architectural reference framework first draft due August 2026

The architectural reference framework (ARF) is the document that turns the five-layer model in the EDF into named integration patterns, semantic models and trust-spine commitments. The first draft is due in August 2026; NESO is the lead body and publishes the ARF under the licence direction the EDF references. The ARF covers the L1-L5 activity model (which sets out the activities the participants perform on the DSI), the five-layer digital stack alignment to the EDF (applications, governance, standards, communications, physical assets), the semantic-model lineage from CIM into the DSI catalogue, and the integration patterns that the DSI participants will implement against. The ARF is the document a participant reads to know how to onboard against the DSI rather than against a bilateral data-sharing agreement.⁷

The ARF first draft is a working draft for sector engagement; the consultation runs through the rest of 2026; the finalised ARF lands in 2027 and is the substrate for the DSI Public Beta in 2028. The cadence below the ARF is the participant onboarding programme; each transmission owner, each DNO group, the suppliers, the third-party flexibility providers and the regulator each onboard against the ARF in sequence.

What the DSI is not

The DSI is not a data warehouse. The DSI is not a single platform that any participant has to migrate their data into; participants retain ownership of their own data and publish through their own APIs against the standards layer. The DSI is not a replacement for any existing data flow; the Balancing Mechanism Reporting Service continues to publish under the BSC, the DCC continues to carry smart-meter messages under the Smart Energy Code, the LTDS continues to publish in CIM format under the third Ofgem derogation letter. The DSI is the discovery, identity, consent and exchange layer that sits over all of these; it is the answer to the question "how does a planner who needs the LTDS Stage 2 from each of the fourteen distribution licence areas and the ETYS from NESO and the half-hourly settlement aggregates from Elexon all in the same place actually receive them?"

How the DSI reads the planning artefacts

Each of the three planning artefacts (FES, SSEP, CSNP) publishes to the DSI catalogue under the ARF integration patterns. A consumer of the FES dataset (a market-facing analyst at a supplier, a planning team at a DNO, an investment desk at a developer) discovers the dataset through the catalogue, authenticates against the identity layer, checks the licence (NESO Open Licence for FES), and receives the dataset through the federated exchange. A consumer of the SSEP iteration 1 of Q4 2026 follows the same pattern; a consumer of the T-CSNP of June 2026 follows the same pattern. The DSI is the layer that makes the three artefacts reachable from a single onboarding rather than from three separate bilateral agreements.

The Behind-the-Meter asset coordination decision (Elexon provisional under the EDF's four-domain partition, contingent on the outcome of Ofgem's Enhancing Asset Visibility consultation to which Elexon responded in February 2026) is the next governance decision to land in the DSI workstream. The Behind-the-Meter assets (the solar PV, the batteries, the EV chargers, the heat pumps, the aggregator data) are the operational substrate that the SSEP and CSNP cycles read for the demand-side projection. The Flexibility, Markets and Assets Register (FMAR) is the named platform that the EDF references as the anchor for the Behind-the-Meter coordination; FMAR is in build for a 2027 launch.

The DSI MVP catalogue: what is in scope for 2026

The DSI MVP catalogue is the first published list of data products that the DSI exchange supports. The catalogue scope for 2026 covers a small set of high-priority data products that the planning data stack reads at the SSEP and CSNP cadence. The catalogue entries are: the FES dataset under the NESO Open Licence; the SSEP iteration 1 of Q4 2026 (when published) under the NESO Open Licence; the T-CSNP of June 2026 (when published) under the NESO Open Licence; the LTDS Stage 2 of 29 May 2026 and Stage 3 of 30 November 2026 in CIM format under each DNO's data publication terms; the Embedded Capacity Register from each DNO under SLC 50; the Distribution Future Energy Scenarios from each DNO under each DNO's open data terms; and the connection queue position data under the post-Gate-2 process. Each catalogue entry carries the licence, the custodian, the publication cadence, the schema reference (CIM where applicable), the data quality indicators (provenance, completeness, timeliness) and the API contract that the consumer reads against.

The MVP catalogue is the first iteration; the Public Beta catalogue from 2028 extends the scope to cover the Behind-the-Meter asset data through FMAR, the consumer data through CCS, the metering data through the Data Integration Platform and the Settlement Data Repository, the wholesale market data through BMRS, and the asset register data from each licensee. The cadence below the catalogue is the data quality cycle; each catalogue entry is held to a published quality threshold (timeliness, completeness, conformance) and the DSI trust framework rejects publications that do not meet the threshold. The trust framework is the operational counterpart to the EDF principle one (trusted and secure).

How the ARF relates to the LTDS CIM model

The ARF first draft of August 2026 reads the LTDS CIM data model as one of its substantive inputs. The LTDS CIM is the canonical example of a regulator-mandated, machine-readable data publication in the GB regime, and the ARF integration patterns are designed so that any future regulator-mandated publication can follow the same publication pattern that the LTDS sets. The pattern is: standards layer (CIM with the appropriate profile, validated against a SHACL conformance report); communications layer (publication through a defined API, with metadata in the DSI catalogue); governance layer (a Direction or a Standard Licence Condition that mandates the publication at a defined cadence); applications layer (analyst-facing tooling that reads the published dataset). The ARF generalises the pattern beyond the LTDS so that the SSEP, the CSNP, the DFES, the ECR and the future asset registers under FMAR all publish under the same pattern.

The generalisation matters because the strategic-planning artefacts have until now followed their own publication patterns. FES publishes through the NESO Data Portal under the NESO Open Licence; ETYS publishes alongside it through a different page on the same portal; the LTDS publishes through each DNO's own portal under each DNO's data publication terms; the ECR publishes through each DNO's open data portal on its own monthly cadence. The ARF integration patterns absorb each artefact's existing publication route into a consistent pattern that the DSI catalogue can index, the DSI trust framework can validate, and the DSI federated exchange can route. The work the ARF is doing in August 2026 is the work of writing the integration pattern that lets the existing publication routes continue to operate while also being discoverable and validatable through the DSI.

GC0139 and the whole-system planning-data exchange

GC0139 is the Grid Code change that makes the whole-system planning-data exchange a licence obligation. The change was raised as a workgroup proposal in 2025 and progressed to a workgroup report on 3 December 2025; the proposal is for the Authority (Ofgem) to amend the Grid Code to mandate the network licensees to publish the planning data that SSEP and CSNP consume, in a format that NESO can ingest at the SSEP and CSNP cadence.¹ Without GC0139 the licence regime would have to negotiate bilateral data-sharing agreements with each of the three transmission owners and each of the six DNO groups for each SSEP and CSNP iteration; with GC0139 in place the data publication becomes a Grid Code obligation that the Authority can hold the licensees to.

GC0139 is the licence-layer counterpart to the standards-layer CIM work. CIM provides the shared data model that any participant reads and writes against; GC0139 provides the licence obligation that the publication actually happens at the cadence the strategic-planning cycle requires. Together the two close the loop that the EDF five-layer model describes: the governance layer (GC0139 in the Grid Code) lands the obligation on the licensee at the physical-assets layer; the licensee publishes the data through the communications layer (the DSI federated exchange); the data reads against the standards layer (CIM); the application layer (the SSEP modelling tool, the CSNP analysis tool) consumes the data through the DSI catalogue.²

What GC0139 changes for a transmission owner

A transmission owner reading the GC0139 obligation finds three new publication requirements landing on it. The first is the planning-stage asset data (the contracted load and generation at each substation, the planned reinforcement, the asset attributes at each connection point) at the cadence the SSEP cycle requires (annual for the iteration window, quarterly for the consultation window). The second is the network-state data (the operational asset attributes, the topology, the protection scheme) at the cadence the CSNP cycle requires (semi-annual for the transitional iteration, annual for the full delivery). The third is the validation data (the SHACL conformance report for each CIM publication) that lets the consumer trust the data without bespoke reconciliation. The cadence is set by the GC0139 obligation rather than by the bilateral agreement.

What GC0139 changes for a DNO group

A DNO group reading the same obligation finds the same three publication requirements landing on it, with the addition of the LTDS Stage 2 and Stage 3 obligations that the third Ofgem derogation letter of 13 May 2026 already holds the DNO to. The integration matters because the SSEP iteration 1 of Q4 2026 reads inputs from all fourteen distribution licence areas across the six DNO groups in one cycle; without a shared publication cadence the SSEP work has to choose a synchronisation window that wastes the data from licensees that have already published. GC0139 sets the synchronisation window as a licence obligation.

Where GC0139 sits in May 2026

The GC0139 workgroup report of 3 December 2025 is the most recent published document on the change. The proposal is currently in the Authority decision window; the Authority's decision will determine whether the obligation lands on the licensees in time for the T-CSNP of June 2026 (which can run on the existing bilateral agreements) or for the SSEP iteration 1 of Q4 2026 (which is the first cycle where the GC0139 cadence will materially reduce the bilateral negotiation burden). The Grid Code modification process runs through the Code Administrator (NESO) and the Authority decision; the workgroup report is the input to the Authority's decision.¹

How GC0139 reads against the CIM standards layer

The licence obligation that GC0139 lands is a publication obligation; the publication has to happen at the cadence and in the format that the SSEP and the CSNP cycles consume. The format requirement reads against the standards layer of the EDF five-layer model, and specifically against the CIM standards that the LTDS already publishes under. The workgroup report sets out that the GC0139 publications should follow the CIM data model (IEC 61970-301 for the EMS-API base, IEC 61968-13 for the distribution profiles, CGMES 3.0 for the cross-border alignment) with the GB-specific extensions that the BSI engagement portal at https://cim.bsigroup.com/ hosts. The validation step uses the SHACL Shapes Constraint Language that the W3C published as a recommendation in 2017 and that the CGMES community has adopted for profile validation; a publication that does not pass the SHACL validation against the published profile is not a compliant publication.

The master resource identifiers (mRIDs) are the second standards-layer commitment that GC0139 makes operational. An mRID is a universally unique identifier assigned to each topology element (each substation, each transformer, each line section, each connection point) at the first publication and retained across every subsequent publication that includes that element. The mRID lets a SSEP or CSNP analysis tool compare publications from successive years and identify which elements have been added, removed or modified; without stable mRIDs the comparison would require structural matching across the topology and would be fragile to renaming. GC0139 makes the mRID retention a licence obligation alongside the publication itself.

How GC0139 reads against the BSC and the CUSC

GC0139 sits in the Grid Code, but it reads against the Balancing and Settlement Code and the Connection and Use of System Code as well. The reading is at the data-flow level rather than at the code-text level. The half-hourly settlement data that Elexon runs under the BSC (and that the MHHS programme makes universally half-hourly from July 2027) is one of the demand-side inputs that GC0139 expects the licensees to read against; the network connection data that NESO holds under the CUSC is one of the connection-state inputs that GC0139 expects the licensees to publish against. The cross-code reading is the post-REMA architectural shift the GC0139 workgroup report is designed to enable; the previous Grid Code amendments tended to read against the Grid Code alone, but GC0139 reads across the three operational codes simultaneously.

A worked example: tracing a 49 megavolt-ampere battery connection through the stack

A developer planning a 49 megavolt-ampere battery in the East Midlands in May 2026 reads each artefact in the stack in a specific order. The read order below makes the dependencies between the five artefacts concrete.

The read order above makes one point explicit. Each of the five artefacts is a different cycle with a different cadence and a different lead body; the developer's project case has to reconcile across all five. The cadence change in 2025 to 2026 (FES multi-year with Economics Annex, SSEP methodology approved May 2025, CSNP methodology approved April 2026, EDF published 23 March 2026, ARF first draft due August 2026) is the change that makes the cadence reconcilable for the first time. Before April 2026 the developer reading the stack found a synchronised methodology decision (CSNP), a published spatial timetable (SSEP), a refreshed pathway envelope (FES Economics Annex) and a substrate framework (EDF) within twelve months of each other.

The developer who reads each artefact at its published cadence rather than at the date the connection offer is needed finds that the project case strengthens at each iteration. SSEP iteration 1 in Q4 2026 confirms (or revises) the East Midlands signal; T-CSNP June 2026 confirms the reinforcement; LTDS Stage 3 on 30 November 2026 adds the planning data and the connection-stage data to the Stage 2 topology; the full CSNP at end-2028 integrates the final SSEP and the next FES. The cadence is the production line that the planning data stack delivers against.

What changes when the DSI MVP catalogue is in production

The seven-step read order above assumes that the developer goes to each data source independently. The DSI MVP catalogue (in delivery through 2026) changes the read order in two ways. The first change is the discovery step: the developer reads the DSI catalogue once to find every input the project case needs (FES, SSEP, T-CSNP, the NGED LTDS Stage 2 and Stage 3, the NGED Embedded Capacity Register, the NGED DFES, the connection-queue position for the relevant substation), rather than visiting seven different portals to find each input. The second change is the validation step: the developer reads the SHACL conformance report for each CIM publication through the same DSI catalogue, rather than running the validation against each publication independently.

The two changes together reduce the onboarding cost from days (visiting each portal, downloading each dataset, running each validation) to minutes (reading the catalogue, authenticating once, receiving each dataset through the federated exchange). The reduction matters because the developer is one of many consumers of the planning data stack; the same reduction applies to the planning team at the supplier, the analyst at the investment desk, the regulatory team at Ofgem and the modelling team at NESO. The DSI MVP is the moment the cost of reading the planning data stack drops below the cost of negotiating bilateral arrangements for each input.

What changes when GC0139 lands as a licence obligation

The Authority decision on GC0139 (currently in the decision window in May 2026) is the second change that the developer reading the stack should follow. When GC0139 lands as a licence obligation on the transmission owners and the DNO groups, the planning data inputs become a published obligation at a set cadence rather than a bilateral arrangement. The change reads downward into the project case: the developer who needs the NGED Stage 3 publication of 30 November 2026 finds it on the published cadence rather than negotiating its release; the developer who needs the SSEP iteration 1 of Q4 2026 finds it on the SSEP cadence; the developer who needs the T-CSNP of June 2026 finds it on the CSNP cadence. The cadence is no longer a working assumption; it is a licence-mandated calendar.

The combined effect of the DSI MVP catalogue and GC0139 is the operational shift the May 2026 inflection sets up. Before April 2026 the planning data stack read as five separate publications with five separate cadences and five separate access routes. After the May 2026 inflection the stack reads as a single synchronised production line with a single access route through the DSI catalogue and a single set of licence-mandated cadences. The shift does not change the substantive content of any of the five artefacts; it changes the cost of reading them as a single connected stack rather than as five separate publications.

Reading the stack as one production line, May 2026 onward

The five-artefact partition set out at the top is the partition the May 2026 inflection makes operationally readable for the first time. FES on the multi-year cadence with the 2025 Economics Annex carries the national pathway envelope; SSEP from Q4 2026 through Autumn 2027 carries the spatial placement; CSNP from the April 2026 methodology through the T-CSNP of June 2026 to the full delivery at end-2028 carries the network reinforcement; the Energy Digitalisation Framework of 23 March 2026 carries the five-layer model that ties the substrate together; the DSI architectural reference framework due August 2026 carries the integration patterns that make the substrate operational. Each artefact has its own lead body, its own cadence and its own statutory or regulatory base; together they describe a single connected production line that the licence regime can read against.

The cadence reference is the table in the orientation section above. The dependency reading is the lead diagram. The integration with the licence regime is the GC0139 section. The worked example for a 49 megavolt-ampere battery in the East Midlands shows the read order that a developer follows in May 2026 and how the read order changes when the DSI MVP catalogue is in production and when GC0139 lands as a licence obligation. The stack now shows the planning data inputs, for the first time, on a synchronised cadence that the licence regime can hold together as a single production line; the substantive content of each artefact continues to evolve, but the cadence is settled.

The work that remains in 2026 is the work of writing the SSEP iteration 1 against the methodology, publishing the T-CSNP against the methodology, finalising the ARF first draft against the EDF five-layer model and obtaining the Authority decision on GC0139 against the workgroup report. Each of these is a substantive deliverable with its own cycle, its own consultation and its own stakeholder engagement; none of them is a small piece of work. The cadence reference and the dependency reading set out above are the framing that makes each deliverable readable as part of a connected stack rather than as a free-standing publication.