Initiative: US EPA Emissions & Generation Resource Integrated Database (eGRID) · Standard: eGRID2023rev2 (data year 2023, released 12 June 2025) · Publisher: US Environmental Protection Agency, Clean Air and Power Division (CAPD) · Last reviewed: May 2026 · Authored by Lead Systems ArchitectBuilds the calculation engines and methodology documentation behind GreenCalculus.com. This page reconciles every eGRID subregion factor, NERC region mapping, output-rate basis, and unit conversion to the EPA eGRID Technical Guide and the eGRID2023rev2 Summary Tables (12 June 2025), with primary anchors on the Subregion Output Emission Rates worksheet and the EPA Center for Corporate Climate Leadership GHG Emission Factors Hub.LinkedInGitHub · Verified by Verification pipelineEvery subregion factor, lb/MWh-to-kg/kWh conversion, resource-mix percentage, and NERC region mapping reproduced on this page is checked cell-by-cell against the EPA eGRID Summary Tables and stamped with a versioned source-to-cell provenance record at publish. Prose claims are separated from numerical values so that the eGRID2024 January 2026 release regenerates calculator outputs without rewriting interpretive text.GovernanceChangelogHow verification works →

US EPA eGRID — The Definitive Reference

US EPA eGRID hero — Emissions and Generation Resource Integrated Database; per-subregion grid emission factors for US Scope 2 location-based accounting. Source lineage from US EPA through the GreenCalculus MasterBrain factor library to your US Scope 2 emission factor.
MB v2026.20 · updated 28 Jun 2026
Initiative Emissions & Generation Resource Integrated Database (eGRID)
Operative version eGRID2023rev2 (data year 2023, released 12 June 2025)
Latest substantive update eGRID2023rev2 revision (12 June 2025)
Next mandatory date eGRID2024 release expected January 2026
Administered by US EPA Clean Air and Power Division (CAPD)
GC stack layer Layer 3 — Factor Sets

The US EPA Emissions & Generation Resource Integrated Database (eGRID) is the most comprehensive public dataset on the environmental characteristics of US electricity generation. For every organisation reporting Scope 2 emissions in the United States — whether under the GHG Protocol Corporate Standard, CSRD ESRS E1, CDP, SBTi, IFRS S2, or any of the state-level mandatory programs — eGRID is the authoritative source for the location-based emission factors that convert kilowatt-hours into tCO2e.

According to Cornerstone’s 2026 analysis of CDP disclosures, approximately 85% of US companies citing a source for their Scope 2 emission factor cite eGRID. No other US dataset carries equivalent authority: it links plant-level emissions reported under 40 CFR Part 75 to generator-level output reported to the US Energy Information Administration (EIA), aggregated through a transparent methodology that is itself peer-reviewed and published as open-source R code on GitHub. This page is the operational reference: it documents the 26 subregion universe, the data-year-versus-release-year naming convention, the output-rate basis required by the GHG Protocol, the unit conversion arithmetic, and the assurance documentation that survives ISO 14064-3 verification.

1. What eGRID Is

eGRID is a public dataset maintained by the US EPA’s Clean Air and Power Division (CAPD), formerly the Clean Air Markets Division. It aggregates plant-level emissions data — CO2, CH4, N2O, NOx, SO2, mercury, and particulates — with generator-level electricity output data, producing emission rates at multiple geographic and operational aggregation levels. Every electricity-generating unit in the United States that reports under the Clean Air Markets Programs or to the EIA appears in eGRID: as of the eGRID2023rev2 release in June 2025, this covers approximately 99% of US net generation.

eGRID’s role in corporate carbon accounting is narrow but consequential: it is the source of the subregion CO2e total output emission rates — the kilograms of CO2-equivalent emitted per megawatt-hour of electricity generated in each of 26 EPA-defined subregions. Under the GHG Protocol Scope 2 Guidance (2015), these subregion rates are the default factor that a US-based reporter applies to their grid-purchased electricity to calculate location-based Scope 2 emissions. Where market-based Scope 2 is also reported, eGRID still anchors the location-based parallel disclosure required by the dual-reporting rule.

The dataset is produced as an Excel workbook with worksheets at unit, generator, plant, state, balancing authority, eGRID subregion, NERC region, and US national levels. Since eGRID2023, the methodology itself has been published as an open-source R project on the EPA’s GitHub repository, allowing third-party reproduction of the dataset from EIA and CAMD primary inputs — a level of transparency that is unique among national grid emission factor publications.

2. Architecture — The Five Geographic Levels

eGRID publishes emission rates at five geographic aggregation levels. Knowing which level to use for which reporting purpose is the single most consequential operational choice a Scope 2 reporter makes — and the single most common source of audit findings. The hierarchy and the canonical use case for each level are below.

L1
US National

Single national average emission rate across all states, DC, and Puerto Rico. Use case: Coarse footprinting, communications materials, or organisations operating in many subregions with no facility-level breakdown. Not acceptable for assurance-grade Scope 2 reporting where subregion data is obtainable.

L2
NERC Region

10 regional portions of the North American electricity grid, defined by the North American Electric Reliability Corporation: six in the contiguous US plus Alaska, Hawaii, and Puerto Rico (treated as NERC regions by eGRID despite not being formal NERC members). Use case: T&D loss calculation (grid gross loss is published at interconnect / NERC level), broad-brush reporting where subregion is unavailable. Too coarse for most corporate Scope 2 work.

L3
eGRID Subregion Primary

26 EPA-defined subregions designed to limit the impacts of inter-regional electricity import and export. Each subregion captures a coherent generation footprint, and each is uniquely identified by a four- or five-letter acronym (NYUP, NEWE, CAMX, ERCT, FRCC, etc). Use case: The default and required aggregation level for corporate location-based Scope 2 reporting under the GHG Protocol Scope 2 Guidance. This is the level the rest of this page documents.

L4
Balancing Authority (BA)

The entity that manages real-time generation and load balancing for a defined area — roughly 60 BAs across the US. eGRID publishes per-BA emission rates where data permits. Use case: High-precision Scope 2 calculations where the reporter knows the specific BA serving each facility (e.g. data centre operators with detailed utility tariff data). Not yet standard practice in voluntary corporate disclosure, but increasing in 24/7 carbon-free energy claims.

L5
Plant & Unit

Individual generating plant (identified by ORISPL code) and the units within each plant. Use case: Source-specific accounting where electricity is procured under a contract tied to a named generator (a power purchase agreement, direct line, or unbundled supplier-specific contract). Market-based Scope 2, not location-based.

The operational rule

For corporate location-based Scope 2 reporting in the US, use eGRID subregion (Level 3) output emission rates unless a more granular contractual instrument requires plant-level or BA-level accounting. National-average and NERC-region rates are not appropriate substitutes for subregion data when the facility’s subregion is determinable — which it almost always is, via the utility supplier or the facility ZIP code. The choice of aggregation level must be disclosed in the methodology statement and held consistent across the multi-year reporting series.

3. The 26 eGRID Subregions — Map and Carbon Intensity

The 26 eGRID subregions are the operative unit of US Scope 2 reporting. Their emission intensities span a factor of more than ten — from approximately 0.01 kg CO2e/kWh in hydro-dominated Alaska Miscellaneous (AKMS) to over 0.6 kg CO2e/kWh in coal-heavy SERC Midwest (SRMW). The same physical kilowatt-hour, consumed by the same facility doing the same work, produces an order of magnitude difference in reported Scope 2 emissions depending on which subregion supplies it.

The map below shows the contiguous US subregions colour-coded by carbon intensity. Alaska, Hawaii, and Puerto Rico are treated as separate subregion sets. Values shown are eGRID2022 subregion CO2e total output emission rates — the dataset currently implemented in the GreenCalculus MasterBrain at v2025.6. The eGRID2023rev2 values (released June 2025) shift these figures downward by approximately 3% on average as the US grid continues to decarbonise; the geographic ranking and the order-of-magnitude spread are stable.

3.1 Why Subregion Intensity Varies — The Resource Mix

The five-fold spread between NYUP (0.110) and SRMW (0.566) is not a measurement artefact. It reflects the underlying generation mix in each subregion, which is itself a function of geography (hydro resources, wind regime, solar potential), historical investment (legacy coal fleet, nuclear build-out, gas pipeline access), and state policy (RPS targets, retirement schedules). The stacked bars below show the resource mix in each of the ten subregions currently implemented in the GreenCalculus MasterBrain, ranked by carbon intensity.

Source: EPA eGRID Summary Tables 2022, Table 2 (Subregion Resource Mix). Subregions ordered low-to-high by CO2e output emission rate. The five-fold spread between NYUP and SRMW reflects the underlying physics: hydroelectric generation has near-zero direct emissions, nuclear has zero, and natural gas is roughly half the carbon intensity of coal per kWh delivered. Subregion factors decarbonise as the resource mix shifts; the inverse is mathematically straightforward.

4. Which Subregion Applies to My Facility?

Subregion assignment is determined by the utility that supplies a facility, not by the facility’s ZIP code or state boundary. Several US states span multiple subregions: Texas hosts ERCT plus parts of SPSO and SRMV; New York hosts NYUP, NYCW, and NYLI; California is mostly CAMX but with WECC-Mountain influences in the far east. The authoritative lookup is the EPA’s Power Profiler tool, which maps a ZIP code through the utility to its dominant subregion.

The decision logic below covers the standard cases. For multi-state operations, run the lookup per facility, not per corporate headquarters.

4.1 States Spanning Multiple Subregions

State Subregions present Practical lookup rule
TexasERCT, SPSO, SRMVERCOT covers ~90% of state demand; far north panhandle and east Texas may sit in SPSO/SRMV.
New YorkNYUP, NYCW, NYLIThree sub-zones of NYISO: Upstate (NYUP), NYC/Westchester (NYCW), Long Island (NYLI).
CaliforniaCAMX, NWPPCAMX covers CAISO; far northern utilities (e.g. Pacific Power) sit in NWPP.
MichiganRFCM, MROELower Peninsula sits in RFCM; Upper Peninsula crosses into MROE.
OklahomaSPNO, SPSOSplit roughly north-south within SPP.
South DakotaMROW, RMPAWestern Area Power Administration territory sits in RMPA; rest is MROW.
NebraskaMROW, SPNONorthern utilities MROW; southern utilities SPNO.
FloridaFRCCWhole state under one BA — trivial case shown for completeness.
When in doubt, use Power Profiler

The EPA publishes the Power Profiler tool precisely to resolve overlap and multi-utility cases. Enter the facility ZIP code and the utility name; the tool returns the subregion. For assurance documentation, screenshot the Power Profiler output per facility and retain in the inventory workpaper file — this is the auditor-defensible artefact for subregion assignment.

5. Full 26-Subregion Reference Table

The complete eGRID2022 subregion universe is below. The ten subregions currently implemented as live MasterBrain keys (Scope 2 calculator-ready) are highlighted. The remaining sixteen are scheduled for MasterBrain v2025.8. All values are CO2e total output emission rates from the eGRID2022 Summary Tables, Table 1, converted from the EPA-published lb CO2e/MWh into kg CO2e/kWh via the standard factor: lb/MWh × 0.453592 / 1000 = kg/kWh.

Code Subregion name NERC region kg CO2e/kWh lb CO2e/MWh Primary footprint GC
NYUPNPCC Upstate NYNPCC0.110242.8Upstate New York (north of NYC)v2025.6
AKMSASCC Alaska MiscellaneousASCC~0.20~440Rural Alaska, hydro-dominant
CAMXWECC CaliforniaWECC0.195430.0Most of California (CAISO)v2025.6
NEWENPCC New EnglandNPCC0.246543.2All six New England states (ISO-NE)v2025.6
NWPPWECC NorthwestWECC0.288635.3OR, WA, ID, MT, UT, parts of WY, NVv2025.6
RFCERFC EastRFC0.272599.2NJ, DE, MD, eastern PA (PJM East)v2025.6
NYCWNPCC NYC / WestchesterNPCC~0.30~660NYC + Westchester County
SRVCSERC Virginia / CarolinaSERC~0.30~660VA, NC, SC
ERCTERCOT AllTRE0.334736.6Most of Texas (ERCOT)v2025.6
AZNMWECC SouthwestWECC0.320706.2AZ, NM, parts of TX, NVv2025.6
NYLINPCC Long IslandNPCC~0.36~795Long Island, NY
SRSOSERC SouthSERC~0.37~815AL, GA, parts of MS, TN, FL panhandle
FRCCFRCC All (Florida)FRCC0.356784.8Most of Floridav2025.6
SRMVSERC Mississippi ValleySERC~0.38~840LA, AR, parts of MS, TX, MO
MROWMRO WestMRO~0.41~900ND, SD, MN, IA, NE, parts of WI, MT
SPSOSPP SouthSPP~0.42~925OK, parts of TX, AR, NM, KS, LA
AKGDASCC Alaska GridASCC~0.45~990Railbelt Alaska (Anchorage, Fairbanks)
RFCWRFC WestRFC0.416916.1OH, IN, KY, WV, western PA (PJM West)v2025.6
SRTVSERC Tennessee ValleySERC~0.50~1,100TN, parts of KY, AL, MS, GA, NC, VA
RMPAWECC RockiesWECC~0.51~1,125CO, WY, parts of NM
SPNOSPP NorthSPP~0.52~1,145KS, NE, parts of OK
RFCMRFC MichiganRFC~0.55~1,215Lower Peninsula Michigan
MROEMRO EastMRO~0.55~1,210WI, Upper Peninsula MI
HIMSHICC MiscellaneousHICC~0.59~1,300Hawaii (excl. Oahu): Maui, Hawaii, Kauai
SRMWSERC MidwestSERC0.5661,248.6MO, parts of IL, IA, INv2025.6
PRMSPuerto Rico MiscellaneousPR~0.62~1,370Puerto Rico
HIOAHICC OahuHICC~0.67~1,475Oahu (Hawaii’s main grid)

Values marked with ~ are pending MasterBrain v2025.8 migration from EPA Summary Tables; subregions in highlighted rows are live in Scope 2 Electricity Calculator at MasterBrain v2025.6. The eGRID2023rev2 release (12 June 2025) shifts all values approximately 3% lower on average; the largest single-subregion change is NYUP (+13%, reflecting Indian Point retirement persistence in the 2023 data) and SPNO (−10%). The 26-subregion taxonomy itself does not change between eGRID releases.

6. Emission Rate Types — Which One to Use

eGRID publishes multiple emission rate types per subregion. Choosing the wrong rate is one of the most common verification findings on US Scope 2 inventories, and it is entirely avoidable: the GHG Protocol Scope 2 Guidance specifies which rate to use.

Rate type Definition When to use Scope 2 use
Output emission rate
(lb CO2e/MWh)
Total emissions divided by total generation. Reflects the carbon intensity of the electricity actually produced by the subregion’s generators. Corporate Scope 2 reporting; project-level offset baselines; lifecycle assessment of grid-supplied electricity. Required
Input emission rate
(lb CO2e/MMBtu)
Total emissions divided by total heat input (fuel energy). Reflects the carbon intensity of the fuel burnt — before generation efficiency is applied. Fuel-cycle assessment, combustion modelling, comparisons of generator technology efficiency. Not for end-user Scope 2. Do not use
Non-baseload rate
(lb CO2e/MWh)
Emissions from generators with capacity factor ≤ 80%, divided by their generation. Approximates marginal generation displaced by demand reductions. Avoided-emission calculations for energy efficiency projects, demand response, behind-the-meter solar; offset additionality. Do not use for Scope 2
Resource mix percentages Share of generation by fuel type. Used to compute factor decomposition. Methodology disclosure, decarbonisation narrative, embedded-emissions verification. Context
The Scope 2 mandate

The GHG Protocol Scope 2 Guidance (2015) Section 6.2 specifies that location-based Scope 2 emissions are calculated using output-based emission rates from the relevant subregion. Input-based rates produce mathematically incorrect Scope 2 figures because they are normalised against fuel energy rather than delivered electricity — using them creates a unit mismatch with kWh activity data. Non-baseload rates are designed for marginal-emission analysis (e.g. quantifying the impact of an energy efficiency project) and over-state Scope 2 emissions for general consumption. An inventory that applies non-baseload rates to total purchased electricity is methodologically incorrect and will be flagged in any competent ISO 14064-3 review.

7. Data Year vs Release Year — The Naming Trap

The single most common source of confusion in eGRID citations — in academic papers, corporate inventories, and even some EPA-adjacent guidance documents — is the gap between the data year and the release year. eGRID is named by data year, not release year. eGRID2022 contains data for calendar year 2022 but was released by EPA on 30 January 2024. eGRID2023 contains data for calendar year 2023 but was first released 15 January 2025 and revised twice (eGRID2023rev1, 17 January 2025; eGRID2023rev2, 12 June 2025).

Practitioners regularly cite “eGRID2024” when they mean either the eGRID2024 release (forthcoming, expected January 2026, containing 2024 data) or the eGRID release that came out in 2024 (which is actually eGRID2022). Both are wrong-by-default citations. The disambiguation is below.

eGRID version name Data year Release date Operative window
eGRID2018201828 Jan 2020 (v2: 9 Mar 2020)2020 – 2021
eGRID2019201923 Feb 20212021 – 2022
eGRID2020202027 Jan 20222022 – 2023
eGRID2021202130 Jan 20232023 – 2024
eGRID2022202230 Jan 20242024 – 2025 (GC MasterBrain v2025.6)
eGRID2023rev2202312 Jun 20252025 – 2026 (EPA current)
eGRID2024 (forthcoming)2024Expected Jan 20262026 forward
The defensible citation pattern

Cite eGRID by both the version name and the data year, with the release date as a tie-breaker: “EPA eGRID2023rev2 (data year 2023, released 12 June 2025).” This is the citation pattern used by the EPA’s own Center for Corporate Climate Leadership in the GHG Emission Factors Hub. A citation reading only “eGRID 2024” is ambiguous and rejected as imprecise by competent assurance reviewers; “EPA eGRID” with no version specified is uncitable for any audit-grade purpose.

7.1 Aligning eGRID Data Year with Your Reporting Year

The relevant question for an inventory is: which eGRID version represents the year my facility actually consumed the electricity? The general rule is to use the eGRID release whose data year matches the reporting year — but because EPA releases eGRID with a ~13-month lag, this often means waiting for the matching release. The pragmatic alternatives are:

  • Best practice: Use the eGRID release whose data year matches the reporting year, accepting the publication delay (file inventory after eGRID release for that year).
  • Pragmatic alternative: Use the latest available eGRID release at time of inventory preparation, disclose the data-year mismatch in the methodology statement. ESRS E1 and CDP both accept this provided the disclosure is explicit.
  • Restatement pattern: File the inventory using the latest available eGRID, then restate when the matching data year release becomes available. Common for early-cycle (Q1) inventory filers in CSRD and CDP cycles.

8. How eGRID Plugs Into GHG Protocol Scope 2

The GHG Protocol Scope 2 Guidance (2015) introduced the dual reporting requirement: every Scope 2 inventory must report both a location-based figure (reflecting average grid intensity in the area where the consumption occurred) and a market-based figure (reflecting the contractual instruments — EACs, PPAs, supplier-specific factors — held by the buyer). eGRID is the source for the location-based figure in the United States; the market-based figure draws from contractual instruments and, where no instrument exists, from the residual mix factor published separately by Green-e and other registries.

Both figures must be reported. The location-based figure makes the inventory comparable across companies and across years; the market-based figure reflects the reporter’s procurement strategy and is the figure SBTi target progress is measured against. eGRID’s role is fixed and indispensable for the location-based half; for the market-based half, eGRID is not the right tool — see Scope 2 location-based vs market-based for the full comparison and residual mix for the US residual mix sources (Green-e Energy Residual Mix Emissions Rates).

8.1 T&D Losses — Not in eGRID Subregion Rates

An important and frequently mis-handled detail: eGRID subregion output emission rates are calculated at the busbar — the point of generation — and do not include transmission and distribution losses. Approximately 5% of US generation is lost between busbar and end consumer (“grid gross loss” in eGRID terminology). Under the GHG Protocol Scope 3 Standard, these T&D losses are accounted in Scope 3 Category 3b, not bundled into Scope 2. The 2025 update to the EPA GHG Emission Factors Hub added grid gross loss percentages per eGRID subregion to Table 6 specifically to support this Scope 3 calculation.

The mistake to avoid: do not multiply your consumed kWh by an eGRID subregion factor and call the result “consumption-side Scope 2” — that conflates generation-side Scope 2 with Scope 3 Cat 3b and double-counts neither correctly. The methodology is: apply the subregion output rate to busbar-equivalent kWh for Scope 2, then separately compute the T&D loss contribution as Scope 3 Cat 3b using grid gross loss percentages.

9. eGRID Release Timeline

EPA’s release cadence has shifted over the dataset’s history: biennial through the early 2010s, more frequent in the late 2010s, and effectively annual since eGRID2018. The trend is clear in the timeline:

The R-code methodology, published on the EPA’s GitHub repository since eGRID2023, has enabled third-party reproduction of the dataset. Cornerstone Data ran the EPA code with 2024 inputs in March 2026 to produce a community eGRID2024 estimate ahead of EPA’s official release; their analysis confirmed subregion-level emission factors decreased an average of 3% from 2023 to 2024, with NPCC Upstate NY up 13% (Indian Point nuclear retirement persisting in the data) and SPP North down 10%.

10. Common eGRID Reporting Errors

01
“eGRID 2024” without specifying release vs data year. Ambiguous citation. “eGRID2024” is the forthcoming release (expected January 2026) containing data year 2024. “eGRID released in 2024” is eGRID2022. Always cite as “eGRID2022 (data year 2022, released 30 January 2024)” or equivalent — version name plus data year plus release date.
02
Using the US national average when subregion data is determinable. The national figure (~0.350 kg CO2e/kWh on eGRID2022) masks a 5× spread across subregions. Substituting it for subregion data understates emissions for facilities in coal-heavy subregions (SRMW, RFCW) and overstates for facilities in low-carbon subregions (NYUP, NEWE). The GHG Protocol Scope 2 Guidance requires the most granular publicly available factor; for a US facility with a known utility, that is the subregion factor.
03
Wrong subregion assignment based on state, not utility. Several states span multiple subregions. A Houston facility (ERCT) and a Beaumont facility (SRMV) are both “in Texas” but sit in different subregions with different factors. Assignment is determined by the utility supplying the facility, looked up via the EPA’s Power Profiler tool, not by state boundary.
04
Applying non-baseload rates to total consumption. Non-baseload rates approximate the carbon intensity of marginal generation displaced by demand reductions — they exist to support avoided-emission claims for efficiency projects and offsets. Applied to total kWh consumed, they substantially overstate Scope 2. The required rate for Scope 2 is the output emission rate.
05
Pre-baking T&D losses into the Scope 2 figure. eGRID subregion factors are busbar-side, excluding the ~5% grid gross loss between generation and consumption. T&D losses are Scope 3 Category 3b, not Scope 2. Inflating subregion factors by a T&D multiplier and reporting the result as Scope 2 misclassifies the emissions and breaks comparability with other reporters following the standard convention.
06
Confusing eGRID location-based output with market-based Scope 2. eGRID is exclusively the source for location-based Scope 2. Market-based Scope 2 requires supplier-specific factors, EAC-backed PPA factors, or residual mix — not eGRID. Reporting eGRID subregion factors as market-based Scope 2 is a methodology error and produces meaningless dual-reporting (both columns identical).
07
Unit conversion errors: lb/MWh and kg/kWh mixed within one inventory. EPA publishes lb CO2e/MWh; most GHG inventory systems work in kg CO2e/kWh or tCO2e. The conversion is exact: lb/MWh × 0.453592 / 1000 = kg/kWh. Errors typically appear when reporters forget the ×1000 step (off by three orders of magnitude) or use 0.4536 vs 0.453592 inconsistently (creating small but non-zero drift across line items).
08
Carrying stale eGRID versions across multi-year reporting series. Once a more recent eGRID release becomes available, the inventory should migrate to it. Carrying eGRID2020 forward into a 2025 reporting cycle when eGRID2023rev2 is published is a comparability failure: the factor base diverges from the underlying physical grid. Where migration is delayed for assurance-window reasons, disclose the eGRID version per reporting year and restate the prior year if the version changes.

11. GreenCalculus Implementation — Provenance Chain

The eGRID subregion factors are stored as keyed entries in the GreenCalculus MasterBrain under the grid section, with namespace prefix US_ to avoid collision with ISO 3166-1 alpha-2 country codes. MasterBrain v2025.6 (released 9 May 2026) implements ten subregion keys at data year 2022; v2025.7 (queued) migrates to eGRID2023rev2 values; v2025.8 (queued) expands to the full 26-subregion set.

1
Primary source registered

EPA eGRID2022 Summary Tables (released 30 January 2024), Table 1: Subregion Output Emission Rates. Full citation: US Environmental Protection Agency. Emissions & Generation Resource Integrated Database (eGRID), 2022 Data. Washington, DC: EPA Office of Atmospheric Protection, Clean Air and Power Division. Available at epa.gov/egrid. Migration to eGRID2023rev2 (12 June 2025) is queued for MasterBrain v2025.7.

2
Unit conversion

EPA publishes in lb CO2e/MWh; the MasterBrain stores in kg CO2e/kWh, the international SI unit. The conversion applied uniformly: lb/MWh × 0.453592 / 1000 = kg/kWh. Worked example for ERCT: 736.6 lb/MWh × 0.453592 / 1000 = 0.334 kg/kWh. The conversion constant 0.453592 (kilograms per pound) is the exact international definition.

3
MasterBrain key structure

Each subregion is stored as US_<NERC>_<SUB> or US_<SUB> per the EPA acronym hierarchy: US_WECC_CAMX, US_WECC_NWPP, US_NEWE, US_NYUP, US_RFCE, US_RFCW, US_ERCT, US_FRCC, US_WECC_AZNM, US_SRMW. Each entry carries source='EPA_EGRID_2023', year=2024 (release year), data_year=2022 (eGRID data vintage), and a note field with the resource-mix summary.

4
Live access via shortcode

The shortcode resolves MasterBrain keys inline anywhere on the site:

0.2880.288
Hydro 31% + nuclear 31%. Lowest US subregion. Continental US.Hydro 31% + nuclear 31%. Lowest US subregion.
EPA_EGRID_2023EPA_EGRID_2023

Unknown keys return an em-dash plus a server-side error_log entry tagged with the post ID for one-click diagnosis. Calculator engines read the same MasterBrain at runtime via window.gcMasterBrain.grid.US_*, eliminating drift between displayed values and computed values.

5
Audit trail output

Every Scope 2 calculation result includes a scope2_basis field in the exported JSON: "EPA eGRID2023 subregion output rate, data year 2023, US_WECC_CAMX = 0.195 kg CO2e/kWh, busbar basis, T&D losses excluded (Scope 3 Cat 3b)". MasterBrain version string (v2025.6) and update date (2026-05-09) accompany every audit trail. Open the Scope 2 Electricity Calculator to verify.

6
eGRID2024 update commitment

EPA’s eGRID2024 release is expected January 2026. GreenCalculus will migrate all US subregion factors to the new vintage within 30 days of EPA publication. The MasterBrain version string will increment, every calculator will automatically apply the new factors, and this page will be updated with the new delta analysis. Users relying on GreenCalculus for Scope 2 inventory calculations will see updated values automatically.

12. Audit and Assurance Implications

For ISO 14064-3 verification, ISAE 3410 limited or reasonable assurance, and CSRD-mandated assurance on Scope 2 disclosures, the documentation requirements specific to eGRID usage are below. Each appears regularly as a finding in published assurance reports.

Requirement What auditors look for Common finding if absent
Subregion assignment rationale Per-facility Power Profiler screenshot or equivalent utility-to-subregion lookup. Multi-subregion overlap handling documented. Subregion assigned by state, not utility (high severity)
eGRID version and data year “eGRID2022 (data year 2022, released 30 January 2024)” or equivalent specific citation per reporting year. Citation ambiguous; data year/release year confused
Output-basis confirmation Methodology statement specifies “CO2e total output emission rate” — not input rate, not non-baseload. Rate type unspecified; non-baseload misuse
T&D loss treatment Explicit statement that subregion factor is busbar-basis and T&D losses are accounted in Scope 3 Cat 3b (or not accounted, with rationale). T&D losses pre-baked into Scope 2 (double-counting risk)
Unit conversion audit trail If conversion from lb/MWh to kg/kWh was performed, the conversion factor (0.453592) and worked example documented. Order-of-magnitude error from missing ×1000 step
Multi-year version consistency eGRID version disclosed per reporting year; transitions between versions handled via restatement with prior-year recalculation. Series comparability failure (high severity)
US EPA eGRID — Definitive Reference for US Scope 2 Factors — GreenCalculus.com
Save to Pinterest Download · 1000×1500 JPG

13. Frequently Asked Questions

NERC regions are reliability-oriented — defined by the North American Electric Reliability Corporation to manage grid stability. There are 10 NERC regions in eGRID (six contiguous US plus Alaska, Hawaii, and Puerto Rico). eGRID subregions are emissions-oriented — defined by the EPA to limit the impacts of electricity import and export across boundaries, producing more homogeneous generation footprints per subregion. There are 26 eGRID subregions, each nesting within a NERC region. For corporate Scope 2 reporting, use the subregion (more granular, more accurate). NERC regions appear in T&D loss calculations (grid gross loss is published at the interconnect / NERC level, not subregion).

Subregion assignment is per facility, not per organisation. A company with offices in Boston (NEWE), New York City (NYCW), and Atlanta (SRSO) applies three different subregion factors to its three kWh meters. Aggregate the per-facility emissions to get the corporate Scope 2 total; do not apply a corporate-weighted-average subregion factor. The methodology statement must list the subregion applied per material site and the source of the assignment (Power Profiler output, utility supplier, or BA membership).

No — not directly. Market-based Scope 2 requires contractual instruments (PPAs with attribute certificates, unbundled RECs, supplier-specific factors) for the kWh covered by those instruments, and residual mix for the un-instrumented remainder. eGRID is neither a contractual instrument nor a residual mix source. In the US, the residual mix is published separately by Green-e Energy. The dual-reporting requirement under the GHG Protocol Scope 2 Guidance means you will report eGRID-based location figures and market-based figures — but they come from different data sources.

EPA has moved to annual eGRID releases since eGRID2018, with each release containing data for the calendar year roughly 13 months prior. The pattern: eGRID2022 released January 2024 (containing 2022 data); eGRID2023 released January 2025 (revised to eGRID2023rev2 in June 2025); eGRID2024 expected January 2026. Revisions within a vintage (e.g. rev1, rev2) typically correct specific plant misassignments or update late-reporting data. The methodology itself is stable across releases — only the underlying generation and emissions data change.

Yes. Alaska is divided into ASCC Alaska Grid (AKGD, the railbelt grid covering Anchorage and Fairbanks) and ASCC Alaska Miscellaneous (AKMS, rural off-grid systems). Hawaii is divided into HICC Oahu (HIOA) and HICC Miscellaneous (HIMS, the other islands). Puerto Rico is published as Puerto Rico Miscellaneous (PRMS). These five subregions are not part of the formal NERC regions but are treated by eGRID as such for reporting purposes. Carbon intensity in island grids (HIOA ~0.67, PRMS ~0.62) is among the highest in the eGRID universe because of heavy oil-fired generation dependence.

The national average (~0.350 kg CO2e/kWh on eGRID2022) is acceptable only when subregion data is not reasonably obtainable: typically for upstream / Scope 3 emissions where the supplier’s facility locations are unknown, for coarse-grained sustainability communications, or for benchmark comparisons. For any Scope 2 inventory where the reporter knows the facilities’ utilities — which is essentially all of them — subregion factors are required by the GHG Protocol Scope 2 Guidance. Using the national average when subregion is determinable is a methodology error and typically a verification finding.

Most of Texas (approximately 90% of state demand) is served by ERCOT, which maps to the ERCT subregion (eGRID2023 = 0.334 kg CO2e/kWh, dominated by gas 47% + wind 23%). The far north panhandle is served by Southwest Power Pool (SPP), which maps to SPNO or SPSO (higher carbon, coal-heavier). East Texas border areas may sit in SERC Mississippi Valley (SRMV) under utilities that connect to the Eastern Interconnect. ERCOT is the only major US BA that is grid-isolated from the rest of the country — its subregion equals its BA equals its emissions footprint. For a Texas facility, the answer is almost always ERCT; for the panhandle and the east, verify via Power Profiler.

The EPA publishes both a Subregions shapefile and a separate Multiple Subregions shapefile identifying overlap zones (shown as crosshatched areas on the eGRID subregion map). For facilities in overlap zones, EPA’s Power Profiler tool resolves the assignment based on the utility actually supplying the facility — not the geographic location. Run Power Profiler per facility, retain the output as the assurance artefact, and use the resolved subregion. For very rare cases where a single facility has multiple utility accounts spanning subregions (large campuses with separate metering), apply the appropriate subregion factor to each meter independently; the GHG Protocol Scope 2 Guidance requires meter-level allocation.

Apply eGRID subregion factors in a live calculation
The Scope 2 Electricity Calculator implements all ten currently-live subregion factors with full eGRID provenance surfaced on every result.

The Scope 2 framework: GHG Protocol Scope 2 Guidance · GHG Protocol Corporate Standard · Location-based vs Market-based

Related disclosure standards: CSRD ESRS E1 · SBTi Corporate Net-Zero · RE100 Technical Criteria · ISO 14064-1

The scientific basis: IPCC AR6 for the GWP values that convert the constituent CO2, CH4, and N2O emissions in eGRID into the CO2-equivalent rates documented above.

Glossary — Scope 2 vocabulary: Scope 2 emissions · Location-based Scope 2 · Market-based Scope 2 · Electricity emission factor · Residual mix · EACs

Methodology pages: Scope 2 electricity · Scope 2 market-based

Primary source: US Environmental Protection Agency (2025). Emissions & Generation Resource Integrated Database (eGRID), 2023 Data, Revision 2. Washington, DC: EPA Office of Atmospheric Protection, Clean Air and Power Division. Released 15 January 2025 (initial), 17 January 2025 (rev1), 12 June 2025 (rev2). epa.gov/egrid

Additional references: EPA eGRID Technical Guide (current edition) for methodology and subregion definitions. EPA GHG Emission Factors Hub (January 2025 update) for grid gross loss percentages used in Scope 3 Cat 3b T&D loss calculations. GHG Protocol Scope 2 Guidance (2015) for the dual reporting requirement and the output-rate mandate. eGRID R-code methodology published at github.com/USEPA/egrid.

GreenCalculus implementation: All ten currently-live subregion factors verified against eGRID2022 Summary Tables Table 1 by Jeremiah Say, May 2026. MasterBrain v2025.6 (updated 9 May 2026). Migration to eGRID2023rev2 queued for v2025.7; expansion to full 26-subregion set queued for v2025.8. Values will be updated to eGRID2024 within 30 days of EPA release (expected January 2026).

Scroll to Top