PJM's January 2026 load forecast projects unusually rapid peak-load growth in the Dominion zone and explicitly adjusts that zone for data-center growth. The record matters because it is used for regional planning, but its categories are planning inputs and forecasts: they do not establish that every identified project was built, energized, or consuming the forecast megawatts.
What the January 2026 record is
PJM posted its 2026 Load Forecast Report on January 14, 2026. The report says it is an independent forecast prepared by PJM staff and provides a 20-year outlook for peak load, net energy and several related demand components across PJM zones and the RTO. Its model inputs include economic assumptions, weather scenarios and sector data, with additional adjustments where PJM expects large changes that are not adequately represented by the base model.
For the Dominion zone, PJM identifies two explicit adjustments: growth in data-center load and a voltage-optimization program. That is a planning statement about how the forecast was constructed. It is different from a meter record, an interconnection completion record, a certificate that a particular facility has opened, or an hourly operating history for a named data center.
Sources for this section2026 PJM Load Forecast Report ↗
What PJM projects for the Dominion zone
The Dominion dashboard in the 2026 report shows summer peak load growing at an annualized 5.4% over the next 10 years, 4.9% over 15 years and 4.1% over 20 years. The corresponding winter figures are 5.1%, 4.7% and 3.9%. For comparison, PJM's RTO-wide summer peak is projected to grow 3.6% per year over 10 years and 2.4% over 20 years.
Those percentages describe a zonal planning forecast. The Dominion zone includes a broad Virginia service territory and multiple load-serving entities and customers. A zonal growth rate cannot be converted into the operating demand of one campus without a project-specific record tying that facility to an actual service date, meter history or other verified load data.
Sources for this sectionDominion dashboard in the 2026 PJM Load Forecast Report ↗
How the data-center adjustment enters the forecast
PJM's 2026 Long-Term Load Forecast Supplement provides the methodological detail behind the Dominion adjustment. It says Dominion Energy requested an adjustment for data-center growth and that an adjustment of this type has existed in some form since the 2014 Load Forecast Report. Dominion, Northern Virginia Electric Cooperative, Old Dominion Electric Cooperative and Rappahannock Electric Cooperative supplied historical hourly information and forecast peak information for the next 20 years.
The supplement also describes an effort to avoid double counting. PJM says it removed monthly data-center load from commercial EIA data and removed hourly data-center load from the historical loads used in model estimation before applying the separate large-load treatment. It further distinguishes 'firm' and 'non-firm' data-center additions. That classification is part of forecast construction: it expresses the degree of commitment PJM assigns to proposed load for planning purposes, not a finding that every listed project has entered service.
- The adjustment is added to a broader load-forecast model.
- Historical data and forecast additions are treated separately.
- Firm and non-firm are forecast categories, not operating-status certificates.
Sources for this section2026 PJM Long-Term Load Forecast Supplement ↗
Observed load and forecast load are different evidence layers
PJM's May 29, 2026 Load Forecast Accuracy Report makes this distinction visible. Its Dominion pages plot historical monthly data-center load peaks and actual July peaks separately from successive annual large-load-adjustment forecasts. The chart therefore contains both observations and projections, but they are not interchangeable series.
An observed monthly peak can support a statement about measured aggregate load for the covered group and period. A future large-load adjustment supports a statement about PJM's planning assumption. Neither one, without additional project-level data, establishes the demand of a specific facility. The same boundary applies when a forecast line is revised between forecast vintages: revision is evidence that the planning estimate changed, not that physical demand rose or fell by the same amount.
Sources for this section2026 PJM Load Forecast Accuracy Report ↗
The 2026 document timeline and current status
The controlling annual forecast was posted January 14, 2026. PJM posted the detailed Long-Term Load Forecast Supplement on February 6 and the 2026 Load Forecast Accuracy Report on May 29. PJM's load-forecast development page also lists a July 24, 2026 mid-year update and labels it 'Informational Only.'
As checked on October 4, 2026, PJM's public forecast-development page still identifies the January 2026 report as the 2026 Long-Term Load Forecast and separately identifies the July material as an informational mid-year update. That sequence matters when citing dates: a later update or meeting material does not change the publication date of the underlying January report, and an informational update should not be described as a new full annual forecast unless PJM says so.
Sources for this sectionPJM Load Forecast Development Process ↗2026 PJM Load Forecast Report ↗
What this record does not establish
The forecast does not establish that every data-center project represented in a utility request was constructed, energized or operating. It does not show that a named facility is drawing its requested capacity, that a non-firm project will become firm, or that the full zonal forecast will materialize on the schedule shown. A planning adjustment is not an operating meter.
The record also does not establish environmental impacts, legal compliance or health effects at a particular site. Electricity demand can be relevant to generation and transmission planning, but claims about generator use, emissions, air quality, water consumption, wetlands or land-use effects require the corresponding permits, operating reports, monitoring data, utility records or local decisions. Nor does a projected load request by itself identify who should bear a particular transmission or generation cost; that question requires tariff, cost-allocation and commission records.
- Requested or forecast load is not actual operating load.
- A firm planning category is not proof of construction or energization.
- A zonal forecast does not identify the demand of one facility.
- Grid-demand evidence does not substitute for environmental or land-use records.
What additional record would support a stronger claim
To move from the zonal forecast to a facility-level statement, the strongest record would connect the project to a specific utility service or interconnection status and then to observed demand. Depending on the question, useful records could include an executed service agreement, energization or construction-completion documentation, hourly or monthly meter data, a utility filing that identifies the project, or an official load-service status report.
For a system-cost or reliability claim, additional evidence may include PJM transmission-planning studies, utility rate filings, Virginia State Corporation Commission orders, cost-allocation records and NERC or PJM disturbance reports. Virginia's State Corporation Commission has already documented the broader planning context in its 2025 Combined Annual Report, but that statewide regulatory summary still does not replace the project-specific records needed to say what one facility is actually using.
Sources for this sectionVirginia SCC 2025 Combined Annual Report ↗
The source controls over this summary. Corrections should identify the disputed sentence and the strongest available primary record.
Questions about framing and source selection: [email protected]. A named editor has not yet been published; no cited researcher is represented as responsible for this page.
Send the page URL, disputed claim and supporting source. Review does not require endorsement.
Request a correction →