PJM Queue — Note 008

Eight weeks, one dataset, and everything I got wrong.

Eight weeks, one dataset, and everything I got wrong.

Kenneth · 2026-08-12

Source: LBNL Queued Up 2026 Edition (CC BY 4.0), through end-2025. Analysis my own.

This is the last of the series. It is the accounting.

What the series claimed, and how it would fail

Every note rested on one public file that anyone can download. **Nothing here

required privileged data, a subscription, or a relationship.** That is the point

worth making, and it is uncomfortable: the numbers everyone quotes are wrong in

ways a weekend and a spreadsheet reveal.

The recurring pattern across all eight:

**A composite figure is quoted as though it were a base rate. It is an average

over populations that behave nothing like each other.**

Technology, vintage, operator, size — each splits the 68% into ranges wider than

the number itself. That is not a subtlety. **It means the headline is not merely

imprecise, it is the wrong shape.**

What I could not establish, listed plainly

Why it deteriorated. Cluster reform, upgrade cost allocation, speculative

filing and interconnection queue reform are all plausible. I tested none of them.

Every note that touched the question said so.

Why ERCOT differs. Connect-and-manage is the obvious candidate. Also untested.

Whether recent cohorts are fully resolved. They are largely, not entirely.

Some active requests may yet be built, which would lower the recent rates. That

cuts the same way everywhere, so the comparisons hold even if the levels move.

Battery's conversion rate is depressed by youth as well as failure. Stated in

Note 006 rather than buried.

What would change my mind

• A 2022–23 cohort that resolves materially better than 2020–21 would mean the

collapse was a cluster-reform artefact rather than a trend.

• ERCOT's advantage narrowing as its queue grows would suggest the difference was

volume, not process.

• Battery conversion above 5% in three years would mean the 0.5% was youth.

All three are checkable from the same file, next year, by anyone.

The standing offer

This is published research. Every figure is a count from a free dataset at emp.lbl.gov/queues — check it yourself rather than taking my word for it.

rates for a real position rather than a composite that fits nothing.

If you think a number in this series is wrong, tell me which and why. **Every

figure is a count from a free file**, so disagreement is settleable rather than

a matter of opinion, and I would rather be corrected than quoted.

Kenneth · kookyeus@yahoo.com

Commercial energy — supply contracts, bill audit, PJM queue research