Foundations stage summary

8 min 6 concepts 5 figures

The Foundations stage builds the vocabulary that every later module depends on. It sets out what energy system data actually is, in Ofgem's terms and at 2026 scale, who holds it across the six groups of organisations that make up the GB ecosystem, the identifiers and registers that key every record, the physical network of wires and pipes that produces the raw measurements, a working taxonomy for classifying what flows, and the national instrument that produces most of it: the smart metering estate and the DCC network behind it.

One argument runs through the stage. In GB you find energy data by knowing the institution, and you read energy data by knowing the identifier. There is no single national hub: registration truth lives in the CSS, settlement truth with Elexon, gas truth with Xoserve, system truth with NESO, and the rules with Ofgem and the code bodies. So the professional skill this stage teaches is routing: given any dataset or question, name the identifier it is keyed by, the register that holds it, the organisation that operates that register, and the instrument that governs access.

The sections follow the stage's teaching order, so you can read straight through to rebuild the stage in your head, or jump to the concept you need. Each section links back to its module for the full treatment.

What you carry out of this stage

  • Define energy system data in Ofgem's post-October 2024 terms and explain why the broadened definition changed what the Data Best Practice Guidance covers
  • Name the six groups of the GB energy data ecosystem and route any major dataset to its institutional home
  • Decode an MPAN and an MPRN, explain why import and export are separate meter points, and keep the roughly 350 physical GSPs distinct from the 14 GSP Groups
  • Locate the four main telemetry sources on the network map, from transmission SCADA to gas chromatographs
  • Classify an unseen dataset with the course's 13-category taxonomy, place it on the five-rung sensitivity ladder, and name its producer, consumer and governing instrument
  • Describe the smart metering estate accurately: the SMETS generations, the three WAN regions and the 4G transition, the DCC2 handover dates, and which DUIS read family serves which use case

The route through GB energy system data, from meters to certification

Each stage answers one question and hands the next what the pill under the arrow names, so the exam at stage four is asked of the vocabulary, the lifecycle and the judgement together rather than of the last stage alone.

Four stages run from vocabulary to proof: foundations, the data lifecycle, markets in motion, then exam and certification, with codes, Ofgem and data policy governing the route from above. Source: this course's syllabus, 31 modules across four stages.

The route through GB energy system data, from meters to certification Four stage nodes run left to right inside a thin frame: stage 1 Foundations, 6 modules, asking what the system and its data are; stage 2 The data lifecycle, 8 modules, asking how a reading becomes money; stage 3 Markets in motion and strategy, 8 modules, asking where the market is going; stage 4 Exam and certification, 3 exam surfaces, asking whether you can prove it. Labelled arrows between the stages read builds the vocabulary, follows the data and applies the judgement. A governance band above stages 1 to 3, labelled codes, Ofgem and data policy, drops into stage 2 (rules) and stage 3 (strategy). The Foundations stage is highlighted as the current stage. CODES · OFGEM · DATA POLICY stages 1 to 3 rules live in stage 2 strategy lives in stage 3 STAGE 1 Foundations 6 modules What is the systemand its data? STAGE 2 The data lifecycle 8 modules How does a readingbecome money? STAGE 3 Markets in motionand strategy 8 modules Where is the marketgoing? STAGE 4 Exam andcertification 3 exam surfaces Can you prove it? builds the vocabulary follows the data applies the judgement

Energy system data has a legal definition, and Ofgem widened it in October 2024

Energy system data means any data relating to the energy system, whether produced by, for, or about the energy system. That is Ofgem's definition, adopted in its decision on the Data Best Practice Guidance on 8 October 2024, and the wording matters because it replaced a narrower test that covered only data necessary for the operation of the system. The broadened definition pulls planning forecasts, asset registers, consumer consent records and innovation datasets into the scope of the guidance, so datasets that once sat outside the rules are now inside them.

The scale is worth holding in numbers. At the end of March 2026, 72 percent of all meters operated in smart or advanced mode, on DESNZ Q1 2026 statistics, with more than 41 million smart and advanced meters installed. Elexon expects MHHS processing to handle up to 500 billion half-hourly meter readings a year from May 2027. Thousands of SCADA-monitored substations are sampled every 2 to 10 seconds, producing around 69 million data points an hour for a large DNO, and about 20 gas chromatographs at National Transmission System entry points continuously measure the calorific value that prices every gas bill in GB.

The stage's reading frame is five unresolved tensions: open data against security risk, centralisation against fragmentation, consumer consent against system need, innovation speed against regulatory pace, and national standards against local variation. None of them has a clean answer. Every decision in energy data trades one legitimate goal against another, which is why the course keeps returning to who decides, under what instrument, rather than to any single right answer.

Energy system data: the definition Ofgem broadened in October 2024

Before October 2024 the Data Best Practice covered three categories, and those three cross unchanged under the KEPT arrows; planning data, consent records, asset registers and innovation datasets join them, so a dataset out of scope then can be in scope now.

Ofgem's 8 October 2024 decision moved planning data, consent records, asset registers and innovation datasets into scope. Source: Ofgem Decision on Data Best Practice Guidance.

Ofgem's 2024 broadening of the Energy System Data definition Two parallel stacks. The left stack shows the narrow pre-2024 scope of Energy System Data: operational meter data, network operations and settlement reads, drawn with dashed borders. The right stack shows the broad post-October-2024 scope with the same three categories kept plus four new categories highlighted in red soft: planning data, consumer consent records, asset registers and innovation datasets. A brand-red band below states the decision date 8 October 2024. BEFORE · NARROW SCOPE Operational data only AFTER · BROAD SCOPE Produced by, for, or about Operational meter data DCC reads, telemetry Network operations SCADA, switching Settlement reads MHHS half hours out of scope before 2024 out of scope before 2024 out of scope before 2024 out of scope before 2024 Operational meter data DCC reads, telemetry Network operations SCADA, switching Settlement reads MHHS half hours Planning data Capacity scenarios, FES NEW Consumer consent records CCS, opt-in registers NEW Asset registers Substation, cable, transformer NEW Innovation datasets Network Innovation, SIF NEW KEPT KEPT KEPT DECISION · 8 OCTOBER 2024 Ofgem broadened the Energy System Data definition

Six groups hold the data, so the institutional map is a data catalogue

More than 40 organisations run the GB energy data estate, and they sort into six groups: government and regulators (DESNZ and Ofgem), system operation (NESO, publicly owned since 1 October 2024 under Energy Act 2023 powers), networks (three electricity transmission owners, six DNO groups, four gas distribution networks and National Gas), the code bodies (Elexon, RECCo, DCC and ElectraLink), industry bodies, and market participants. Each group produces, consumes or governs a distinct slice of the estate.

Data custody follows the institutions. NESO holds system data, the Future Energy Scenarios, the Data Portal and coordination of the Data Sharing Infrastructure. Elexon runs the BSC, settlement, the Insights Solution, the flexibility market facilitator role and the Smart Data Repository arriving in autumn 2026. The DCC operates the smart meter WAN and the Central Switching Service. RECCo governs the Retail Energy Code, the enquiry services and the Consumer Consent Solution. Xoserve is the gas central data services provider, running UK Link and Gemini, and the DNOs publish the LTDS, DFES, the Embedded Capacity Register and capacity heatmaps.

The everyday consequence is that a data request is routed, not searched. Whoever operates the register you need also answers to a specific code: the CSS is operated by the DCC but governed under the REC, settlement data flows under the BSC, and gas retail data lives under the UNC. One dated fact belongs in this map: the Gas and Electricity Enquiry Services replaced the older Data Enquiry Service on 18 July 2022, so pointing anyone at the DES today is pointing them at a retired service.

Six groups, three roles: the GB energy data ecosystem

Five organisations in the framework column set the rules for GB energy data, but the fifteen that operate the system, the four code bodies and the twenty-plus market participants have to carry them, putting a rule in far more hands than the number that wrote it.

Six groups cover every organisation that touches GB energy data. Source: Ofgem Energy System Data Best Practice Guidance, Annex A.

Six groups, three roles: the GB energy data ecosystem A three-column grid. The framework column on the left holds two cards for the rule-setting groups: government and regulators (DESNZ, Ofgem, ICO) and industry bodies (ENA, BSI). The operating column in the middle holds two cards for the data-moving groups: system operation (NESO) and networks (14 organisations). The enabling column on the right holds two cards for the groups that build platforms and trade in the market: code bodies (Elexon, RECCo, DCC, ElectraLink) and market participants. Each card states the count of organisations, the primary data role, and the key organisations. FRAMEWORK GOVERNMENT AND REGULATORS 3 ORGANISATIONS ROLE Set policy and legal basis KEY ORGANISATIONS DESNZ, Ofgem, ICO INDUSTRY BODIES 2 ORGANISATIONS ROLE Coordinate standards KEY ORGANISATIONS ENA, BSI CIM Advisory Group OPERATING SYSTEM OPERATION 1 ORGANISATION ROLE Balance the system live KEY ORGANISATIONS NESO NETWORKS 14 ORGANISATIONS ROLE Carry power and SCADA data KEY ORGANISATIONS 3 TOs, 6 DNOs, NGT, 4 GDNs ENABLING CODE BODIES 4 ORGANISATIONS ROLE Run the data platforms KEY ORGANISATIONS Elexon, RECCo, DCC, ElectraLink MARKET PARTICIPANTS 20+ ORGANISATIONS ROLE Trade, supply, flex KEY ORGANISATIONS Suppliers, generators, aggregators

An MPAN is a bundle of market decisions wearing a serial number

Every record in the market is keyed by a small set of identifiers. The electricity MPAN's 21 digits encode settlement treatment, not just identity: the top line carries the profile class, meter time-switch code and line loss factor class, and the core carries the distributor id, the unique number and a check digit. The gas MPRN does the same job for the other fuel, and MPXN is the collective term. Reading an identifier is therefore reading a set of market decisions about how that meter point settles.

Import and export are separate meter points, not signs on one number. A solar home holds an import MPAN and a separate export MPAN, each with its own register, which is why an import register never runs backwards. And the physical network must be kept distinct from its settlement geography: about 350 physical grid supply points roll up into just 14 GSP Groups, and it is the Groups that settlement and most regional data products use. Mixing the two up is how central settlement services get wrongly described as regional ones.

The registers hold the market's memory. The Central Switching Service has been the registration source of truth for both fuels since July 2022, with the Retail Energy Location record carrying the address; UK Link is the gas-side backbone; and profile classes with their standard settlement configurations are legacy structures now dying under MHHS. Identifier quality is a market-wide cost, because a wrong address or a mislinked meter point fails switches and pollutes settlement downstream.

The wires and pipes are the data producers, tier by tier

The electricity system is a voltage cascade, and each tier produces its own category of data: generation connects at 11 to 25 kV, the 400 and 275 kV supergrid moves bulk power, grid and primary substations step down through 132 and 33 kV, high-voltage distribution runs at 11 kV, and the low-voltage network delivers 400 and 230 V to premises where the smart meter sits. Gas mirrors this with the National Transmission System and the local distribution zones, and gas billing is a two-input measurement: metered volume priced by continuously measured calorific value.

The telemetry lives in four main places: transmission SCADA, distribution monitoring, boundary metering at the grid supply points, and NTS and LDZ telemetry on the gas side. These feeds are operational technology data, carried on networks physically separated from the internet, and they are exactly the datasets where the open-data presumption meets its security limit.

The changing grid is why this module matters beyond background. Electric vehicles, heat pumps and rooftop solar all connect at the low-voltage edge, the layer the DNOs historically monitored least, so the fastest-growing operational question sits where the historic data is thinnest. That gap explains much of what stages 2 and 3 cover: half-hourly settlement, flexibility registers, and the monitoring and estimation products that stand in where meters do not exist.

Six voltage tiers, six data categories: the GB electricity cascade

What happens on the LV tier can only be inferred from the tiers above and below it: five of the six carry a monitored or partial chip, and the one gap falls between the secondary substation and the meter, so the tier nearest the home is the one nobody sees.

Six voltage tiers each generate a distinct data category. Source: Distribution Licence Condition 25 (LTDS), Ofgem RIIO-ED2 monitoring scope.

Six voltage tiers, six data categories: the GB electricity cascade A vertical stack of six rows from generation at 11 to 25 kV down to your home at 230 V. Each row shows the tier number, the voltage range, the physical role, the data category pill (GEN, E1, E3, E5, D1, D6, D5, A1, B1) and a monitoring visibility chip. The Level 5 low-voltage row is emphasised and tagged VISIBILITY GAP because DNOs historically have no monitoring between secondary substations and homes. TIER VOLTAGE ROLE AND OWNER DATA CATEGORY VISIBILITY LEVEL 1 11 to 25 kV Generation GENERATORS, NESO GEN MONITORED LEVEL 2 400 / 275 kV Supergrid transmission NGET, SPT, SHET E1, E3, E5 MONITORED LEVEL 3 132 / 33 kV Grid and primary substations DNO GROUPS (14) D1 MONITORED LEVEL 4 11 kV Distribution feeders DNO GROUPS D6 PARTIAL LEVEL 5 400 / 230 V Low voltage to premises DNO GROUPS (RIIO-ED2) D5 VISIBILITY GAP LEVEL 6 230 V at meter Smart meter at your home DCC, SUPPLIERS A1, B1 MONITORED

Thirteen categories and five sensitivity rungs, honestly attributed

The course groups the estate into 13 categories covering 87 distinct data types, each with a defined producer, consumer, governing instrument and sensitivity classification. The attribution matters: this is the course's teaching map, not an industry standard, and quoting it as an external authority is one of the stage's named pitfalls. Its value is practical rather than official, because it turns the shapeless phrase energy data into a set of types you can route.

Sensitivity does the legal work. The five-rung ladder runs from open data to restricted, and the personal against operational distinction carries most of the weight: half-hourly consumption data is personal data when it can be linked to an identifiable household, because it can reveal routines and occupancy, while SCADA telemetry is operational data whose risk is national security rather than privacy. Ofgem's Data Best Practice Guidance presumes openness, and the sensitivity classification is what a holder must argue from when it withholds.

The taxonomy earns its keep only when it routes real decisions. Worked end to end, a half-hourly import reading and a DFES scenario workbook land in different categories, on different rungs, with different producers, consumers and governing instruments, which is the whole point: two datasets that both count as energy system data obey entirely different rules. Profile classes appear in the taxonomy flagged as legacy, because MHHS replaces profiling with actual half-hourly data.

Thirteen categories cover eighty-seven GB energy data types

The type counts on the cards run from three to ten rather than splitting the eighty-seven types evenly across the thirteen categories, making a rule written for one category reach far more of the estate than the same rule written for another.

Thirteen categories, eighty-seven types. Source: Ofgem Data Best Practice Guidance, BSC, REC, SEC, UNC, DCUSA, CUSC, Grid Code.

Thirteen categories cover eighty-seven GB energy data types A three-column grid of thirteen category cards labelled A through M. Each card shows the category letter in a brand-red block on the left with the type count below it, then the category name and an example type to the right. Categories A (consumption and metering) and D (network infrastructure) are emphasised because they account for the largest share of regulated data volume. A brand-red total band beneath the grid states 87 data types across 13 categories with the Ofgem Data Best Practice Guidance annex as the source. A 7 TYPES Consumption and metering EXAMPLE Half-hourly reads (A1) B 6 TYPES Smart meter infrastructure EXAMPLE DUIS, PKI, IHD C 6 TYPES Registration and switching EXAMPLE MPAN, MPRN, CSS events D 10 TYPES Network infrastructure EXAMPLE Topology, SCADA, GIS E 8 TYPES System operations EXAMPLE Frequency, inertia F 8 TYPES Wholesale markets EXAMPLE BMRS, settlement, REGO G 5 TYPES Ancillary services EXAMPLE FFR, reserve, restart H 6 TYPES Flexibility and DER EXAMPLE FMAR, EV, storage I 6 TYPES Gas system EXAMPLE Linepack, CV, flows J 6 TYPES Low carbon EXAMPLE Carbon intensity, CCUS K 7 TYPES Planning and forecasting EXAMPLE FES, SSEP, connections L 3 TYPES Offshore and marine EXAMPLE Offshore wind SCADA M 9 TYPES Regulatory and compliance EXAMPLE RIIO RIGs, NIS, theft TOTAL · 87 DATA TYPES ACROSS 13 CATEGORIES Ofgem DBP Guidance Annex

Smart metering is the national instrument, and its network is three WAN paths, not one

The smart metering estate is the largest single data instrument in GB, and its generations still matter. SMETS1 meters, installed first and initially tied to individual suppliers, went through a national enrolment: about 15 million were in the programme's eligible scope and just over 11.1 million had migrated onto the shared DCC network by May 2026. SMETS2 meters are DCC-native from installation. Inside the home, the HAN links meter, in-home display and comms hub; the WAN carries data from the comms hub to the DCC.

The WAN is not one technology. The North region uses Arqiva long-range radio and covers Scotland and northern England only. The Central and South regions run on Telefonica and Virgin Media O2 cellular, and they include all of Wales. The estate is also mid-transition: 4G comms hubs for the North were agreed on 30 May 2025, the DCC's current licence expires in September 2027, Ofgem's decision on the successor licence came in April 2026, and the business transfer is expected in November 2026, so DCC2 is a data continuity question, not a distant procurement.

What the meter records is equally precise. Import and export are separate registers against separate MPANs, so an import register never goes negative, whatever the midday sun is doing. And reads come in families under the DUIS specification: service request 4.1.1 reads instantaneous import registers on demand, while scheduled half-hourly profile reads for billing use the 4.8 family. Matching the read type to the use case is the difference between a snapshot and a settlement-grade series.

Three WAN paths converge on the DCC backbone for the national smart-meter estate

VMO2 cellular, Arqiva radio mesh and the VWAN broadband fallback arriving in 2026 differ in technology and in the premises they reach, yet all terminate on the DCC backbone, so every read is authorised under SEC Section H whichever path carried it.

Three WAN technologies cover the national smart-meter estate. Source: DCC Annual Service Description; Smart Energy Code Section H.

Three WAN paths converge on the DCC backbone for the national smart-meter estate Three rows, each tracing a coverage regime from the Home Area Network on the left, through the DCC WAN technology in the middle, to the coverage characterisation on the right. Row one is Telefonica VMO2 cellular for the Central and South regions, which include all of Wales. Row two is Arqiva long-range radio at 868 MHz for the North region, covering Scotland and northern England. Row three is the planned VWAN broadband fallback for hard-to-reach premises in 2026. A callout band records the 4G transition: 4G comms hubs on the Vodafone network agreed for the North region from 30 May 2025. A red DCC backbone band closes the figure. HOME AREA ZigBee 2.4 GHz Meter, gas, IHD, CAD DCC WAN Telefonica VMO2 cellular 2G to 4G mobile CENTRAL AND SOUTH, WALES COVERAGE Strong urban and suburban; ~7m meters hit by 2G/3G sunset HOME AREA ZigBee 2.4 GHz Meter, gas, IHD, CAD DCC WAN Arqiva long-range radio 868 MHz long range low power SCOTLAND, NORTHERN ENGLAND COVERAGE Better rural and obstructed cover via meter-to-meter relay HOME AREA ZigBee 2.4 GHz Meter, gas, IHD, CAD DCC WAN VWAN broadband (planned 2026) Customer broadband fallback HARD-TO-REACH PREMISES COVERAGE Basements, deep buildings, persistent wireless dead spots 4G TRANSITION 4G comms hubs on the Vodafone network agreed for the North region from 30 May 2025 DCC BACKBONE All three WAN paths terminate here; SEC Section H authorises every read national estate

The traps this stage warns against

  • Placing Wales in the radio region of the smart meter WAN because the map shows radio in the North.

    Instead: The North region's long-range radio covers Scotland and northern England only. All of Wales sits in the Central and South regions, which run on Telefonica and Virgin Media O2 cellular.

  • Using GSP and GSP Group interchangeably, and placing central settlement services at physical substations.

    Instead: About 350 physical grid supply points roll up into 14 GSP Groups, and settlement works on the Groups. The SVAA is one central Elexon service, not a presence at each GSP.

  • Expecting a solar home's import register to run backwards around midday.

    Instead: Export is recorded on a separate export register against its own export MPAN, so the import register never goes negative. Smart Export Guarantee payment depends on half-hourly export data.

  • Citing the 13-category, 87-type taxonomy as an industry standard.

    Instead: The taxonomy is this course's teaching map. Use it to route producer, consumer, governing instrument and sensitivity questions, and attribute it as the course's grouping, never as an external authority.

  • Quoting a smart meter percentage without naming its denominator.

    Instead: Pin estate figures to the DESNZ quarterly statistics and state the base once: at the end of March 2026, 72 percent of all meters operated in smart or advanced mode. A percentage of homes, or of smart-capable meters, is a different number.

Core distinctions

  • An MPAN identifies an electricity meter point and its 21 digits encode settlement treatment; an MPRN identifies a gas meter point; MPXN is the collective term for both
  • A GSP is a physical grid supply point, one of about 350; a GSP Group is one of the 14 settlement regions they roll up into, and settlement and regional data products use the Groups
  • The import register records energy taken from the grid and the export register records energy sent to it; they sit on separate MPANs and the import register never goes negative
  • A smart meter is a SMETS device in homes and small businesses communicating through the DCC; an advanced meter is the earlier non-domestic category read remotely at half-hourly grain, and DESNZ's 72 percent counts both together
  • SMETS1 meters were installed supplier by supplier and needed enrolment onto the DCC, with just over 11.1 million migrated by May 2026; SMETS2 meters are DCC-native from installation
  • Personal data covers anything linkable to an identifiable household, including half-hourly consumption; operational data such as SCADA telemetry raises security rather than privacy risk, and the two follow different access rules
  • The HAN is the in-home network linking meter, display and comms hub; the WAN is the national network carrying data from the comms hub to the DCC, split across three regional paths

That is the Foundations stage in one place. A legal definition of energy system data and the scale behind it, six groups of institutions that each hold a known slice of the estate, the identifiers and registers that key every record, the network tiers that produce the raw telemetry, a working taxonomy with an honest attribution, and the smart metering instrument with its three WAN paths and its coming DCC2 handover. The foundations scenario practice now puts that vocabulary under pressure with realistic situations, so the common mistakes get caught before stage 2 follows a reading from the meter all the way to money.

Sources and further reading

  • DESNZ smart meter statisticsThe quarterly estate figures behind this stage, including 72 percent of all meters in smart or advanced mode at the end of March 2026.
  • DCC network updatesThe SMETS1 enrolment milestones, the WAN region technologies and the 4G transition behind the smart metering section.
  • Elexon Section S simple guideThe supplier volume allocation guide behind the GSP Group teaching and the identifier treatment of settlement.
  • Ofgem Smart Export GuaranteeThe export tariff scheme behind the import and export register teaching and its half-hourly data requirement.
  • NESO launch announcementThe 1 October 2024 creation of the publicly owned National Energy System Operator, the naming fact this course applies throughout.