A published queue position counts applications, and most applications never build. The effective wait for a project that will actually connect is shorter than the nominal queue implies and far less predictable, because withdrawal rates vary by zone and by developer quality.
What does a queue position mean?
That an application was filed and accepted, nothing more. Filing is inexpensive relative to development, so queues fill with speculative positions held as options rather than as committed projects.
A position therefore describes the paperwork ahead of you rather than the projects ahead of you, and the two differ by a wide and unpublished margin.
Filing behavior also differs by developer in a way that is legible to people in the market. A developer with land control, an offtake conversation under way and a financing route is holding a position they intend to build; one with none of those is holding an option. Both occupy the same line in the register, and the register does not distinguish them.
| Nominal queue position | Effective position | |
|---|---|---|
| Counts | Applications filed and accepted | Projects that will actually build |
| Published | Yes, by the network operator | No — estimable only from people with adjacent positions |
| Useful for | Establishing the upper bound | Underwriting a commercial operation date |
How many projects actually withdraw?
Enough to change the answer materially, and the rate varies by zone, by developer and by how the network operator sequences studies. It is not published in a form that supports a project-level estimate.
Withdrawals also cluster. A network upgrade allocation that lands badly can remove several positions at once, which makes the queue move in steps rather than smoothly — and pushes the build program with it.
The clustering matters more than the average rate for underwriting purposes. A queue that loses a third of its positions steadily is a manageable assumption; one that loses the same third in two events, triggered by upgrade allocations nobody could time, produces a date that is either much earlier or much later than the central case. Averages hide the shape.
What is the effective wait?
Shorter than nominal and far less predictable. That combination is awkward for underwriting, because it means the central case improves and the variance widens at the same time.
The practical route to an estimate is asking developers with adjacent positions how the queue ahead of them has actually behaved, which is observable to them and to nobody else.
Widening variance with an improving central case is an awkward combination to present, and it is usually resolved by presenting only the central case. The more honest treatment is to carry both — a shorter expected wait and a wider band — and to say which observable events would collapse the band, because those are the ones worth monitoring after close.
Who can tell you the truth?
Former transmission planning staff and developers with projects in nearby positions. Both see withdrawal patterns and study sequencing that no public register describes, and both will describe them on a screened call.
Network operator staff are the better source on sequencing; developers are better on withdrawal behavior, because they see who is actually building.
Combining the two sources is what makes the estimate defensible rather than anecdotal. A single developer describes their own experience and generalizes from it; three developers in different positions in the same zone, read against a former planner's account of how studies were sequenced, produce something closer to a distribution than to a story.
How should a model treat it?
As a distribution rather than a date. A single commercial operation date drawn from a nominal queue position is the assumption most likely to be wrong in a renewables model, and it is usually the one presented with least qualification — with the offtake price a close second.
Modeling a range and stating what would narrow it is both more honest and more useful to an investment committee than a point estimate nobody believes.