Indiana Regulators Reopen AES Indiana’s $71 Million Rate Case Over a Google Data Center and a BlackRock-Led Buyout

On September 2 the Indiana Utility Regulatory Commission voted to reopen AES Indiana’s $71 million rate case. The commission had already acted on the case. The vote returns the increase to active review rather than letting the earlier outcome stand.

The commission cited two reasons: the utility’s pending acquisition by a BlackRock-led consortium, and its plans to serve a Google data center.

The reopening. A state commission revisiting a commercial rate increase it has already decided, on the strength of a single large-load customer and a change of control, is a procedural development worth recording on its own terms. The narrow question is what a commission does when the facts underlying a rate order change after the order issues. The broader fact is that an approved commercial rate schedule moved from settled to contested.

What a savings model assumes. Every behind-the-meter storage proposal in a commercial building rests on a demand charge stated in dollars per kilowatt on a published tariff. Load shape, battery degradation, and dispatch accuracy are all modeled as uncertain. The tariff is the input treated as fixed, because a rate case is supposed to produce a number that holds until the next one. Indiana is a case where that input was reopened.

Two other developments this week show large-load rate structure being set through channels that sit outside the general rate case.

The riders. An RMI analysis reported by Latitude Media on September 2 modeled what a 300 MW data center running at a 100% load factor would pay across several utilities. Portland General Electric collects above $300 million a year. Dominion collects roughly $241 million, Evergy Kansas $184 million, and Xcel Colorado $144.8 million as proposed.

The composition is where the spread lives. According to the analysis, the difference across utilities is driven mostly by riders and adjustments: PGE collects $114 million more than Evergy through those mechanisms alone, including a $130 million fuel cost adjustment rider and a $26.28 million community benefit surcharge.

Riders reset on their own schedules, through trackers and annual filings rather than through the multi-year general rate case. On PGE’s numbers, the mechanism carrying much of the gap between two utilities’ bills for identical load is the one that adjusts most often and is litigated least.

The contract terms. The protective provisions utilities are writing into these large-load tariffs run long: 15-year minimum contracts, exit fees, and minimum demand charges set at 80 to 90% of contracted capacity.

A minimum charge set as a percentage of contracted capacity bills the contract rather than the meter, which is the one determinant a peak-shaving battery cannot move. A battery can clip a measured peak. It cannot clip a number written into a service agreement. Colorado’s Utility Consumer Advocate warned in the Xcel proceeding that shortfalls default to spreading across all customer classes. These terms currently apply only to negotiated large-load service, well above any commercial building threshold, and there is no evidence in the record that the determinant is migrating to standard commercial classes.

California’s two bills. AB 2383 (Zbur) and SB 886 (Padilla and McNerney) cleared both houses on August 31, the final day of the 2026 session, and now sit with Governor Newsom, who has until the end of September to sign or veto. Roughly $1.8 billion in contested ratepayer cost allocation sits behind the fight.

The package directs the California Public Utilities Commission to create special rates and updated rules for data-center electricity use, including assigning qualifying data centers responsibility for the grid upgrades their interconnection triggers. That is a new rate structure created by statute and implemented through a commission process the legislature opened, running alongside the general rate case cycle rather than inside it.

The shared property. In each of the three jurisdictions, a rate outcome is being set or reset outside the proceeding that was supposed to fix it. In Indiana the level of an already-decided increase is back under review. At PGE the bulk of the gap between two utilities’ bills for identical load sits in riders that reset outside the base rate proceeding. In California the framework itself is being written by legislative direction on a September deadline.

A commercial demand charge remains structurally distinct from these mechanisms. It is billed on measured peak demand and settled by a published schedule, with no event notice and no enrollment window. Duration is the property that distinguishes it from an incentive program or a capacity payment: a published tariff schedule settles without a program cycle behind it.

The revision interval is what today’s filings speak to. A savings projection quoting a current tariff over a ten-year term carries an implicit assumption that the schedule is stable across that term. The Indiana reopening is a data point on how firm that assumption is, and proposals carrying multi-year savings projections have reason to state a tariff-revision assumption explicitly rather than leave it implied.

Two contrary readings deserve equal weight. Reopening cuts against the tariff assumptions behind any savings model, and it also signals that commissions will revisit cost allocation when large loads arrive mid-case. Every proceeding that assigns upgrade cost to the load that caused it reduces the share of that cost landing in ordinary commercial delivery charges. The same regulatory attention that makes tariffs less predictable in the short term is also the mechanism that keeps data-center-driven investment out of the commercial classes over the longer one.

Sources