Is now the time for DERs to scale?

The first reason it failed, and Lubershane counts himself among the people who got it wrong, is that the grid simply did not need these resources. His evidence is an occasion in 2017 when he gathered a group of utility people, many of them distribution engineers and several specifically tasked with considering distributed resources, and asked where on their systems they were finding non-wires alternatives. The answer was that they were not: one or two outliers, nothing systematic, and he came away convinced they were looking in good faith. The underlying reason was 15 to 20 years with very little load growth until very recently, which meant few places under pressure, upgrades that were not especially expensive when they were needed, and capacity markets with plenty of supply. Kann’s framing is that distributed resources were a vitamin rather than a painkiller, and Lubershane extends it usefully: a vitamin that was not fully approved, because a distributed resource does not deliver the same product as a wires upgrade. Build out the substation and you have capacity in that neighborhood every hour of the year. Add storage instead and you have a dispatchable resource with a limited amount of energy in it, which means the planner has to be sure how many hours a day and how many days a year the capacity is needed, and sure the battery will never be called, drained and called again before it can recharge.

The second reason was cost, and the interesting part is which half of the cost prediction came true. The optimists were right that hardware would get dramatically cheaper: lithium-ion cells and packs on the back of the electric vehicle industry, solar panels, and the cost of making any device connected and controllable over the internet. What they missed is that the soft costs would barely move. Customer acquisition stayed stubborn, and so did installation, because shrinking a large battery system into a box wired into a home’s breaker panel with all the required safety equipment takes skilled labor on site. So the installed price of a residential battery stayed high through a decade in which cell prices collapsed. Lubershane dates the first real movement in installed system cost to roughly the last two to three years.

What has changed is the need, unambiguously. He is confident that the same engineers convened today would name plenty of hotspots where load growth is about to make a feeder very expensive to upgrade. The cost question he leaves open, and his hope is a reinforcing loop rather than a technology trend: need drives utilities to run real procurement programs, programs create volume, volume drives installed costs down. Two things make him think the loop has started. Virtual power plants of genuinely useful size, hundreds of megawatts, are now being dispatched. And a chart from the chief executive of a large demand response aggregator, showing how many times a year its resources were actually called rather than merely enrolled, has climbed steeply in the last two years. He reads that as evidence of need and, as much, of grid operators growing comfortable using resources they used to treat as theoretical.

The strongest objection comes from the host, and Lubershane concedes most of it. Kann says he is not worried about cost, because cost is comparative and every alternative is getting more expensive, from gas turbines to the retail electricity rate that a customer-sited resource is measured against. His worry is that in electricity, “needed” has repeatedly failed to mean “happens.” Change runs through state-level regulation and utility programs that arrive late, or arrive small, or arrive with rules nobody can meet, and hundreds of megawatts of virtual power plants is very little against a problem measured in gigawatts. The whole thesis rests on whether that inflects within two or three years. Lubershane’s answer is not a rebuttal. He names his own bias, an economist’s habit of assuming that systems make rational decisions, and restates his position as a bet that the economic fundamentals finally win out against the friction. He then scopes the forecast down himself: not every utility running a gigawatt program, but every major utility plausibly running a several-hundred-megawatt program within five years, with gigawatt-plus scale in the more mature and more constrained markets. And he adds the limit the title invites you to forget. This does not replace the gas plants now being built and does not remove the need for the other resources. It makes a dent.