Why regulators reject grid modernization spending

Read enough rate-case decisions and a pattern emerges in what regulators strike from grid modernization asks. The disallowed spending is rarely rejected because the panel doubted the technology. Regulators do not rule that a distribution management system is unproven or that sensors do not work. They rule that the applicant failed to make the case — and applicants fail it in the same handful of ways, decision after decision, utility after utility. That consistency is bad news for the utilities that keep making the same arguments, and good news for anyone willing to read the decisions before writing the next application.

The pattern

The first failure is benefits unmoored from need. The application describes what the platform will do — improve reliability, enable DERs, empower customers — without demonstrating that the system currently suffers from the problem the platform solves. A reliability investment on a network whose reliability indices are stable and mid-pack invites the obvious question, and applications routinely fail to anticipate it. Benefit claims that cannot be traced to a demonstrated deficiency read as aspiration, and regulators do not fund aspiration with ratepayer money.

The second is alternatives never priced. Regulatory tests almost universally require that the proposed investment be compared against other ways of solving the problem, including solving it later or not at all. Applications routinely treat this as a formality — a paragraph asserting that alternatives were considered — rather than an analysis with numbers in it. A panel that cannot see what the do-nothing case costs cannot conclude the investment is prudent, and will say so.

The third is pilots that never scaled — in either direction. Utilities run pilots, report success, and then file for full deployment with a business case that bears no arithmetic relationship to the pilot results. Or worse: the pilot's measured benefits, extrapolated honestly, do not support the full program, and the application quietly substitutes vendor estimates. Regulators notice. A pilot that cannot be traced into the deployment case is worse than no pilot, because it demonstrates the utility had the evidence and chose not to use it.

The fourth is asks that outrun demonstrated readiness. A utility with a troubled meter-data-management implementation in its recent past, asking to fund an enterprise analytics platform, is asking the regulator to bet against the record. Panels weigh organizational capability whether or not the application invites them to, and a program whose scale exceeds anything the utility has delivered lands as risk, not ambition.

What survives

The spending that survives scrutiny has a recognizable shape. It is traced to named constraints: this station is at capacity, this feeder class drives a measured share of outage minutes, this connection queue has doubled. Named, located, quantified — deficiencies the regulator can verify from the utility's own filed data, not gestured at through industry trends.

It is quantified against a do-nothing baseline. The strongest applications spend as much effort costing the status quo as costing the proposal, because the increment between them is the entire case. When deferral is genuinely cheaper, the strong application says so and asks for less — which is precisely what makes the panel trust the parts where it asks for more.

And it is staged with off-ramps the regulator can see. Tranches tied to demonstrated outcomes, decision points where the program can be redirected, reporting that lets the panel verify progress before the next tranche. Staging converts a request for trust into a structure for accountability, and panels approve structures far more readily than they extend trust.

Writing the filing backwards

Most grid modernization filings are written forwards: the utility selects a platform, assembles the benefits that platform is said to deliver, and then searches its system for problems those benefits might address. The result reads exactly like what it is — a purchase in search of a justification — and experienced panels can smell it in the executive summary.

The filing should be written backwards. Start from the deficiency the regulator can verify: the measured constraint, the deteriorating index, the queue that is growing on the record. Establish it, price its trajectory, price the conventional fix. Only then introduce the investment, sized to the deficiency rather than to the vendor's reference architecture. The discipline is not rhetorical. Writing backwards forces the utility to discover, before filing, which parts of its wish list actually have a case — and to cut the parts that do not, rather than have the panel cut them publicly.

A filing is an argument, not a catalogue. Every element that cannot carry its own weight in that argument weakens the elements that can.

A test

There is a simple test worth applying to any grid modernization business case before it goes anywhere near a regulator. Remove the vendor's name and the product's name, and read it again. If the case still works — if the deficiency, the baseline, the alternatives, and the quantified benefits stand on their own, and the named product is merely the currently preferred way to close the gap — then it is an investment, and it will read like one. If the case collapses without the product, because the benefits were the product's brochure restated in the utility's voice, then it is a purchase. Regulators approve investments. Purchases they disallow, and they are right to.