The Architecture of a Charge No One Voted For
In April 2026, an independent report revealed an uncomfortable figure: between $658.6 million and $685.6 million in charges that California electric customers may be paying without solid operational justification. The origin lies in Diablo Canyon, the only remaining nuclear plant in the state, and in the legislative agreement that extended its operational life until 2030 in 2022.
The mechanism is more technical than it appears, but it is worth unraveling. When the California Legislature approved a $1.4 billion state loan to finance the plant's extension, it did so with the explicit promise that PG&E would recoup that money from a federal grant from the Department of Energy. The deal was: the state lends, the federal government reimburses, the customer does not pay more. However, the Department of Energy determined that the plant needed only $741.4 million to operate through 2030. On the same day the Legislature approved the $1.4 billion loan, PG&E had requested only $1.1 billion from the federal government. The gap between what was requested, what was granted, and what was loaned created a hole that someone has to fill.
The report indicates that that "someone" is consumers, through four structures of charges established directly by law, bypassing the ordinary process of the California Public Utilities Commission, which typically involves hearings, technical reviews, and regulatory oversight. By skipping that process, the charges became shielded from usual scrutiny.
What the Numbers Reveal When Added Up
The arithmetic underpinning the report is the most revealing part of the case. The charges are not limited to sustaining the operation of the plant: more than $100 million annually is allocated to shareholder compensation, and between $260 million and $270 million annually finances grid projects that need not be related to Diablo Canyon. PG&E customers bear $190.8 million a year solely for the volumetric performance charge, set at $13 per megawatt-hour.
If all these charges were eliminated between 2027 and 2030, the report projects a total savings of $1.84 billion for consumers across the state. In individual terms: about $250 annually for PG&E customers, $80 for Southern California Edison customers, and $60 for San Diego Gas & Electric customers. This is no small figure for a family already facing one of the highest electricity rates in the country.
What makes this analytically interesting is not the accusation itself, which PG&E denied, and its spokesperson claims that the loan money will be fully allocated to eligible projects with previous audits not finding irregularities. What is relevant is the architecture of the financing model: when the costs of critical public infrastructure are set by law rather than by technical regulation, the adjustment mechanism disappears. There is no body that can recalibrate the charges if conditions change, and conditions have changed: the federal government granted less money than expected.
The Consumer as the Residual Financier of Infrastructure
This is where the story stops being just a matter of California energy politics and becomes a business behavior pattern with implications for any regulated industry.
Diablo Canyon generates more than 8% of California's total electricity and about 17% of its carbon-free energy. A report from the Public Utilities Commission projected in July 2025 that closing the plant would create a shortfall of about 1,500 megawatts in the grid. This dependence is not accidental: it is the result of years of energy policy that did not build enough alternative capacity. And this dependence is precisely what makes it politically challenging to question the costs associated with keeping the plant operational.
The emerging pattern is this: when an infrastructure becomes too important to shut down, those who operate it gain a negotiating position that transforms the consumer into the ultimate financier. Not because a contract says so, but because the alternative—shutting down the plant—is politically unacceptable. The customer did not consciously contract that risk. They inherited it.
This has a direct reading for companies operating in sectors where the end user has no real exit options: the rate becomes a mechanism for transferring corporate risk onto the captive consumer. In July 2025, the Public Utilities Commission already ordered PG&E to pay $43.2 million to its customers for mismanagement of maintenance on one of its generators. This precedent suggests that the regulator has the capacity to act, but also that it acts reactively, case by case, rather than structurally.
The Job the User Did Not Know They Were Contracting
For SMEs operating in California and for any business relying on regulated supplies, this case offers a concrete operational reading. Charges set by law without periodic review are not rates; they are long-term commitments disguised as line items on the bill. The difference matters when constructing cost projections for three or five years.
The report calculates that without the additional charges from 2023 to 2030, Diablo Canyon would cost over one-third less than it costs under the current scheme, and power sales would exceed operating costs by $164 million. In other words: the plant would be profitable on its own. The extra charges do not fund the operation; they fund other things that customers did not choose to finance.
That is the model failure that this case precisely exposes. The work that the electric consumer believes they are contracting is reliable and clean energy supply at a price reflecting the true cost of producing it. What they end up contracting, in systems where rate-setting escapes the ordinary regulatory process, is something broader and more opaque: the financial stability of an operator whose exit from the market is unacceptable to anyone.
