Editors Note:
The team is back from our summer break to inform you of moves in commodity markets and important news shaping our industry. This newsletter is slightly longer than our average run, but there is a lot to catch up on!
We cover the lasting impacts of the Iran conflict, the rapidly evolving regulatory landscape for data centers and large loads, changes to federal GHG standards, challenges facing NECEC, and DASI reforms
Source: EIA, Veolia
Source: EIA
Source: EIA
Source: Veolia
Source: NOAA
Oil and refined fuels
Source: TradingTools
Natural Gas
Source: Claude
For businesses, these conditions create greater uncertainty around future energy costs. The price of heating oil and refined products is under the most stress in the short term, but with geopolitical risk, growing LNG exports, demand growth, and winter weather all now influencing the market, staying flexible and monitoring price movements is crucial to managing energy budgets and finding value in the market.
The rapid growth of AI is driving a surge in U.S. data-center development and electricity demand. The large and highly concentrated loads are placing new demands on the power grid and raising questions around interconnection, reliability, infrastructure investment, and cost allocation. With no uniform national framework for addressing these loads, federal regulators, RTOs/ISOs, states, utilities, and local governments are taking different approaches to managing this growth.
What are Large Loads?
In the power planning world, “large load” typically refers to electric demand significantly large enough to materially affect the electric grid, requiring special interconnection, planning, reliability, and cost-allocation considerations. The term has entered the common vernacular with the increasing prevalence of datacenters, however there is no single national definition or regulatory framework for large loads. Thresholds vary by RTO/ISO, utility, and state, with 25 MW being increasingly used as a practical regulatory threshold, though some frameworks use 50 MW, 75 MW, or 100 MW+.
Source: Interconnection.fyi
The gaps are wide: SPP’s and Georgia’s definitions, for example, differ by a factor of ten. FirstEnergy in Ohio has asked to apply its data center tariff to any data center of any size at all. A 40 MW facility could simultaneously be a regulated data center under AEP Ohio’s tariff, an ordinary commercial customer in Georgia, Texas and Oklahoma, and, if interconnected at 69 kV or below in SPP - High Impact Large Load.
How large loads are regulated?
Like the ambiguity surrounding the definition of large loads, there is no single regulatory framework governing their development. Instead, regulation occurs through overlapping layers of federal, RTO/ISO, state, utility, and local authority, with each addressing different aspects of a project from wholesale market access and grid reliability to retail rates, infrastructure costs, permitting, and community impacts.
|
Jurisdiction |
Regulatory Body |
Primary Role |
|
Federal |
FERC / Federal Government |
Wholesale/transmission rules; interconnection policy; reliability and cost-shifting oversight |
|
RTO / ISO |
PJM, MISO, SPP, CAISO, NYISO, ISO-NE |
Interconnection, transmission planning, reliability, forecasting and market participation |
|
State |
PUCs / Legislatures |
Retail tariffs, cost recovery, ratepayer protection and environmental/energy requirements |
|
Utility |
IOUs & Public Utilities |
Customer interconnection, service agreements, deposits and infrastructure upgrades |
|
Local |
Cities, Counties & Authorities |
Zoning, permitting, land use and community requirements |
Source: Veolia
This layered structure means a single large load can be subject to multiple definitions, requirements, timelines, and approval processes, creating a regulatory landscape that varies significantly by market and state.
Without a unifying framework, jurisdictions are adopting different approaches and solutions to grapple with the growth of large loads. Federal policy prioritizes faster data-center development through streamlined permitting and environmental reviews, but it does not override state and local authority over zoning, utilities, and environmental approvals. Meanwhile, many RTOs/ISOs, states, utilities and local governments are pursuing large-load legislation focused on system reliability, cost responsibility, energy and water-use reporting, financial commitments, ratepayer protection, and community & environmental impacts. Some states are even considering construction moratoriums. The table shows how these approaches are evolving across key jurisdictions.
|
Jurisdiction |
Regulatory Body |
Large-Load / Data Center Policy |
|---|---|---|
|
FERC |
|
|
|
|
|
Directing 6 RTOs/ISOs to speed large-load interconnection while preventing cost shifting; federal bill would require states to “consider” adoption of standards to cover incremental grid costs caused by >100 MW loads but Congress has not implemented anything to date.. |
|
RTO/ISO |
|
|
|
|
MISO |
Proposing enhanced reliability, monitoring, forecasting and ride-through requirements for loads >50 MW, including ≥25 MW computational loads. |
|
|
NYISO |
Paused new hyperscale data centers for up to 1 year while developing standards for grid, energy, water, environmental and community impacts. |
|
|
CAISO |
Developing large-load definitions and flexible interconnection options to improve grid integration while protecting reliability and existing customers. |
|
|
ERCOT |
Paused/audited data-center queue; new rules add study fees, financial commitments and energization requirements while improving queue discipline. |
|
|
SPP |
FERC-approved CHILLS allows large loads to use available non-firm transmission for up to 7 years while securing firm generation/upgrades. |
|
|
PJM |
Developing IRAS and a Large Load Registry for ≥50 MW loads, with requirements for new supply, potential curtailment during capacity shortages, and greater load visibility; new ride-through standards for data centers. |
|
State |
|
|
|
|
Oklahoma |
Proposed tariff allows >75 MW loads to self-supply generation while retaining the utility as retail supplier, with different commitments for utility-supplied power. |
|
|
PA |
Pennsylvania requires >25 MW data centers to secure new/clean power, pay full incremental costs and obtain local approval, with preferential treatment for compliant projects. |
|
|
VA |
Virginia is tightening permitting, transparency and environmental requirements and considering/upholding upfront payment for dedicated transmission costs. |
|
|
OH |
Ohio is debating 25 vs. 50 MW thresholds and cost allocation; AEP’s tariff adds minimum-demand, ramp and financial requirements for >25 MW loads. |
|
|
MA |
>25 MW projects must demonstrate incremental clean energy and address grid costs, with interconnection deposits, community benefits and environmental requirements. |
|
|
GA |
Large-load contracts are expanding rapidly, with stronger ratepayer protections and utility responsibility for early termination costs. |
|
|
UT |
≥100 MW loads can bypass utility generation service and procure from non-utility generators while continuing to use utility transmission; cost/backup obligations remain under review. |
Source: FERC, RTOs, Utilitydive, Interconnection.fyi, PUCs, Others
States are converging on one goal to make large loads more accountable for the grid costs and risks they create: Texas is tightening queue and financial requirements, Georgia is using large-load contracts and ratepayer protections, Oklahoma is enabling self-generation, and Utah allows large loads to bypass utility generation service while retaining transmission access.
PJM is similarly accommodating large load growth, but with conditions. It is developing a mechanism that increasingly requires new loads to bring their own power supplies, curtail during system stress, and avoid imposing incremental reliability costs on existing customers. Its one-time backstop procurement to address the 2028-2029 capacity auction shortfall is also closely linked to rapid large load growth and would add capacity while working on longer-term reforms. However, whenever it clears, PJM assigns the cost to each utility zone, leaving states to decide how it gets distributed among retail customers. For ratepayers, and for large loads themselves, the same procurement could therefore mean very different bill impacts depending on the state in PJM.
Therefore, the overall trend is clear: Federal → enable growth; RTOs → manage reliability/interconnection; States → control costs, siting, and community/environmental impacts. This division is creating a growing gap between federal efforts and those of RTOs, state, and localities.
Public sentiment is widening that gap. A poll from Aug ‘26 shows a sharp shift in public opinion against data centers, with 71% of Americans now opposing a new facility near their home, up from 42% in September 2025.
Source: Heatmap
Public opposition is also translating into local policy activity and is speculated to play a substantial role in the midterm elections. This November, residents in 28 localities across six states will vote on data center ballot measures, with proposals ranging from construction bans to permanent moratoriums on new facilities. This growing local activity adds another layer of uncertainty for developers and highlights the increasingly complex environment in which large-load projects are being planned and permitted.
Source: Multistate.US
Large load thresholds vary widely, so the same facility can be regulated in one jurisdiction and treated as likely any other industrial customer in another. At the same time, costs are shifting toward large loads, which are increasingly expected to fully fund their energy demand and any grid buildout or expansion and to include ratepayer protection provisions. This is a dramatic shift from the conversation two years ago when co-location and bypass of the transmission system altogether seemed to have momentum. Finally, bring-your-own-power is spreading, with self-supply, new-generation requirements, and curtailment provisions increasingly serving as the price of faster interconnection.
The large-load landscape is still evolving, and regulatory requirements will likely remain dynamic across federal, RTO/ISO, state, utility, and local jurisdictions. Our team will continue to track these policy and market developments, assess their implications for energy costs, reliability and procurement strategies, and help clients navigate the increasingly complex and fragmented regulatory framework surrounding large loads and data centers.
The Trump Administration eliminated GHG standards for power plants and weakened standards for automobiles in September. The electric power and transportation sectors are responsible for over half of the country’s total GHG emissions, with ~28% from transportation and ~25% from power generation.
Source: Claude & EIA GHG Inventory
On September 14th, the Environmental Protection Agency (EPA) announced the repeal of carbon emissions limits for coal and gas-fired power plants that were previously approved during the Biden Administration. The EPA also proposed removing every remaining greenhouse gas emission standard for the power sector, which is currently being held for public hearing.
In 2024, the EPA implemented strict GHG emissions limits for new and existing fossil fuel fired power plants under Section 111 of the Clean Air Act. The regulations applied to existing coal-fired plants and new gas plants that operate more than 40% of the time and plan to continue operating through 2039. Such facilities were required to eliminate 90% of their carbon dioxide emissions via carbon capture and storage (CCS) technologies or co-fire with natural gas by 2032.
The current administration argues the Biden-era regulations exceeded EPA’s authority under Section 111 of the Clean Air Act and created inefficient use of natural gas, driving up electricity prices. The EPA argues that repealing CO₂ standards for power plants will save $310 billion by reducing compliance costs, which will trickle down to ratepayers. EPA also argues that GHG emissions from power plants have “no material impact on global climate change” and any analysis of associated public health harms are too convoluted to fall under the scope of power sector regulation. (The Biden administration had estimated that by 2035 its regulations would prevent 1,200 premature deaths, 360,000 cases of asthma attacks, avoid 48,000 school absences and 57,000 lost workdays.)
This week the Trump administration also announced changes to the Corporate Average Fuel Economy (CAFE) standards, which sets average fuel economy regulations for car manufacturers. The new rules require most vehicles to average 34.9 miles per gallon by 2031, a ~30% drop from the previous standards of 50.4 miles per gallon. Affordability was again the central rationale behind the decision. However, analysis from the USDOT shows that average cost per vehicle will decline by $1,289 but increase fuel costs by over $1,600 over the vehicle’s lifespan.
The New England Clean Energy Connect (NECEC) was envisioned as a cornerstone of Massachusetts' clean energy strategy. Under 20 year contracts, Massachusetts utilities agreed to purchase approximately 9.55 TWh of Hydro-Québec hydropower annually, equivalent to roughly 20% of Massachusetts electricity demand. The project was intended to provide low cost, low carbon electricity, displace fossil fueled generation, reduce wholesale power costs, and reduce the need for additional renewable generation and associated credits.
Map of the New England Clean Energy Connector
Source: Iberdola - New England Clean Energy Connector
Those benefits have been tested almost immediately. After a series of interruptions during its first year of operation, Massachusetts utilities and Hydro-Québec (HQ) have filed dueling lawsuits over their contractual obligations.
What Happened?
Source: ISONE External Interface Metered Data
While low flows have put the NECEC on pace to underprovide its annual contracted volumes, the value of its’ imported power changes dramatically depending on the New England power market’s current conditions.
Our analysis of NECEC’s 2026 underperformance periods highlights the difference:
Source - ISONE External Interface Metered Data
What to Watch in the Current Litigation
The lawsuits focus on who pays for NECEC's failures, but the real issue is more fundamental: the project's value is heavily dependent on its timely supply during grid stress events. NECEC drives power prices down the most when New England faces constrained power supply volumes coinciding with spikes in electricity prices, exactly when customers need it most. When NECEC fails during these critical periods, utilities must scramble to buy expensive replacement power, which severely takes away the savings the project was supposed to deliver. In other words, NECEC works best when it's needed most, but that's also when its failures hurt the most.
Source: NEISO data, Veolia analysis
Market data disclaimer: Data provided in the "Market Data" section is for the newsletter recipient only, and should not be shared with outside parties.