The era of flat demand is over
For roughly 15 years, U.S. electricity demand was essentially flat. Energy efficiency gains offset population and economic growth almost exactly. Grid planners got comfortable with that pattern. Procurement strategies got built around it.
That period ended around 2023. The EIA projects U.S. electricity consumption to reach about 4,189 billion kWh in 2025 and 4,278 billion kWh in 2026, up from roughly 4,097 billion kWh in 2024 — the first sustained three-year growth streak since 2007. Several drivers compound:
- AI and data centers. Hyperscale data center growth in PJM's Virginia load zone is the most visible example, but the dynamic is regional, not just one corridor. NERC estimates data centers could account for up to 12% of all U.S. power demand by 2028.
- Electric vehicles. EIA data shows electricity use from light-duty EVs has grown nearly fivefold since 2018. Large-scale charging hubs draw load comparable to small factories.
- Industrial electrification. Shifts away from fossil-fuel heating and process energy are adding gigawatts of new demand in key markets.
- Reshoring and regional growth. Advanced manufacturing, logistics, and clean-energy production are all heavier loads than the offices and warehouses they often replace.
The supply side is constrained
Meanwhile, supply isn't moving at the same pace. Three forces tighten what's available:
Generator retirements
Coal plants continue retiring on schedule. Older gas plants are retiring earlier than originally planned as economics shift. New solar, wind, and battery storage get built, but provide capacity differently than thermal generation — and the pace of new build isn't filling the gap.
Interconnection queue backlogs
New generation projects sit for 4–5 years (often longer) before getting through interconnection studies and agreements. PJM's queue had over 200 GW of pending projects as of late 2024. Reforms like PJM's Fast-Track (RRI) are accelerating subsets of the queue, but the structural backlog persists.
Reserve margin tightening
FERC-approved rule changes have tightened reserve margin requirements across multiple ISOs. Same available supply, higher required quantity, higher capacity prices. This is most visible in PJM's capacity auction outcomes — but it's a pattern, not an isolated event.
What grid stress means for buyers
The supply-demand gap shows up on commercial bills and operations in four ways:
| Exposure | How it shows up | Direct cost impact |
|---|---|---|
| Price volatility | Spot prices spike sharply during heat waves, cold snaps, and grid stress events | Indexed-pricing portfolios; unhedged exposure |
| Capacity cost increases | Peak Load Contribution (PLC) events set next year's capacity tag — and tags are getting more expensive | Annual capacity bill (often 15–25% of total cost) |
| Curtailment risk | Large commercial loads are first to be asked (or required) to reduce consumption during emergencies | Operational disruption; production losses |
| Infrastructure bottlenecks | New facility interconnections face delays in congested zones; capacity allocation slows | Project timeline risk; possible relocation considerations |
To make this concrete: during PJM's June 2025 heat event, demand peaked at 160,560 MW — surpassing PJM's seasonal forecast of 154,000 MW. Real-time wholesale prices climbed to $1,334 per MWh at 7 PM, compared with $52/MWh just a week earlier. Petroleum-fired generation tripled from the prior day to meet demand. PJM had ~179,000 MW of generation capacity plus ~8,000 MW of demand response on the books — the event still confirmed how quickly conditions can tighten.
25× spike
PJM real-time prices, week-over-week, during June 2025 heat event ($52/MWh → $1,334/MWh) — exactly the kind of scarcity exposure index-heavy buyers face
What buyers should actually do
Grid reliability isn't a buyer problem to solve. It's a buyer reality to plan around. Four moves matter:
1. Treat demand response as procurement
DR programs aren't just an optional revenue stream — they're a procurement tool. Enrolled load reduces your capacity tag (cutting next year's capacity bill) and gives you advance notice and operational protection during curtailment events. Load management is the demand-side half of energy strategy. In a tight grid, it's no longer optional.
2. Forecast and manage capacity tags
Your capacity tag — your peak demand contribution during ISO-defined system peaks — sets next year's capacity charges. In PJM, those are the 5 highest summer afternoon peaks (5CP). Tag management means knowing when those hours are likely, having operational responses ready, and tracking your actual peak contribution through the season. Reductions of 15–25% year over year are achievable for facilities with operational flexibility.
3. Layer hedges to manage spike exposure
Index-heavy portfolios benefit on average — spot prices average below forward curves most of the time. But scarcity events can erase a year of savings. Layered hedging covering 50–75% of load smooths the volatility while preserving most of the upside. The discipline of hedging on schedule outperforms hoping for stable spot pricing.
4. Plan facility interconnections early
For new facility builds or expansions in congested zones (Virginia data center corridor, ERCOT North Texas, parts of NY and NJ), interconnection timelines have grown materially. Start the conversation 18–24 months before you need power. Treat capacity allocation as a constraint that can shape facility siting decisions, not as a checkbox at the end of the planning process.
The longer view
Multiple reforms are stacking on top of each other to address the supply side: PJM Fast-Track (RRI), surplus interconnection processes, cluster study reforms, and capacity market rule changes. New generation is being added — solar, batteries, gas peakers, and even nuclear additions are in motion. But the supply response will take years to fully catch up to demand growth that's already happening.
The buyers who plan around the gap — building capacity awareness, hedging discipline, demand response, and earlier interconnection conversations into operations — are the ones who absorb the volatility instead of being defined by it.
Bottom Line
The grid is responding to two simultaneous changes: demand climbing for the first time in a generation, and supply slow to catch up. The gap is real, and it's reflected in capacity prices, spot volatility, and operational risk. Don't plan as if reliability stress is temporary. Plan as if it's the new operating environment — because it is. Demand response, capacity tag management, layered hedging, and early interconnection planning are how disciplined buyers navigate it.
