PGE Isolated Data Center Costs, and Its Commercial Power Rates Fell 2.1 Percent
Portland General Electric will raise average rates on its largest data center customers by 29 percent. Under the same cost allocation, average commercial rates fall 2.1 percent. The two numbers come from a single order, and the relationship between them is the point.
The mechanism is a new rate class. Oregon’s POWER Act directed the state Public Utility Commission to stop socializing the cost of serving very large loads across everyone else. PGE’s answer is Schedule 96, approved May 7 and effective June 10, which pulls data centers and other loads above 20 megawatts into a dedicated class with terms no other customer faces.
The terms. Schedule 96 sets minimum demand charges at 90 percent of contracted system capacity, so a data center pays for the grid it books whether or not it draws the power. Large loads sign contracts ranging from 10 to 30 years. Projects above 100 megawatts pay an additional surcharge. Each provision does the same work: it ties the cost of new large-load infrastructure to the loads that create it, and it removes those costs from the pool that sets everyone else’s rates.
That is the arithmetic behind the 2.1 percent. When the cost of serving a fast-growing class is carved out, the rates for the remaining classes do not have to absorb it. In PGE’s case they fall.
The inversion. The working thesis across much of the storage sector has been that data center load lifts commercial and industrial bills. The logic is straightforward. Hyperscale demand forces utilities to build transmission, substations, and generation; cost-of-service ratemaking spreads that capital across all classes; commercial demand charges climb as a result. PGE is the counter-case. The same load-growth pressure produced the opposite outcome for commercial customers, because Oregon chose to isolate the cost rather than socialize it.
Two philosophies. One approach socializes large-load costs and lets commercial rates rise with them. The other, written into the POWER Act, walls the costs off and lets commercial rates fall. Both are now live in different jurisdictions. Which one a given territory adopts is becoming a first-order determinant of where commercial electricity bills go over the next several years.
That distinction is where behind-the-meter storage economics live or die. A commercial battery earns its return primarily by shaving the monthly demand peak that sets the demand charge. The size of that charge, and the direction it is heading, is the single largest input to payback. In a socialized territory, rising demand charges shorten payback periods and expand the set of buildings where storage pencils out. In a cost-isolating territory like PGE’s, the same load growth flattens or lowers commercial demand charges, and the arbitrage a battery is sold on gets thinner.
The measurable piece. The 2.1 percent is an average across the commercial class and across the whole bill, not a stated cut to the demand charge specifically. Schedule 96 reworks who pays for capacity; it does not publish a new commercial demand rate. Whether a given commercial account sees its demand charge fall, hold, or rise depends on the rate schedule it sits on and its load shape. What the order establishes is direction rather than a line item. The direction is down.
How far it spreads. Oregon put the cost-isolation principle into statute, which makes it more durable than a single commission order. Roughly 23 states have now finalized large-load tariffs of their own, and the terms vary. Virginia’s GS-5 schedule, which sets demand floors at 85 percent of transmission and 60 percent of generation capacity, takes effect January 1, 2027. Some of these tariffs borrow PGE’s minimum-demand and long-contract structure; others stop short of removing the costs from the general rate pool. The tariff language, not the headline rate change, decides which way commercial rates move.
For anyone modeling commercial storage returns across a multi-state footprint, the takeaway is that territory now matters as much as tariff. A demand-charge savings assumption carried over from a socialized market into a cost-isolating one will overstate the return. PGE has shown that a regulator facing 29 percent data center growth can hand commercial customers a rate reduction. Where that happens, the case for a battery has to rest on something other than a demand charge that is no longer climbing.
The data center is still the largest new load on the American grid. What PGE settled is who pays for it. That answer, repeated state by state, will shape the commercial storage map as much as any cell price or tax credit.
Sources