Your battery is paid at one node. Stop planning it at the zone.

A hub price is an average, and a battery never gets paid an average. It settles at one point on the network, and the difference between that point and the hub is the whole game. Plan the asset at the zone and you are optimizing against a number it is never actually paid.

Averages hide the congestion you’re paid for

A zonal or hub view smooths across dozens of nodes. That’s reasonable for a gas desk thinking in regional terms; it’s actively misleading for storage, because the basis between the settlement node and the hub is the congestion that determines the return. Score every node independently and the ranking reshuffles — the nodes that actually pay are frequently not the well-known ones. A view that can only see the hub can’t see the asset.

Structural versus transient: which spread is worth a strategy

Not all node congestion is equal. Some is a persistent feature of the network — a constraint that shows up most afternoons and will still be there next quarter. Some is a weather or outage event that won’t repeat. The distinction is what tells you whether a spread is worth building a strategy and a siting decision around, or just worth trading once. A signal that can’t separate the two hands you an average of both and calls it an opportunity.

Capture rate is the honest measure of the edge

The temptation is to argue value in headline revenue. But revenue depends on how volatile the year happened to be, which no tool controls. What a forward, node-level signal actually changes is the share of whatever spread exists that you capture — a merchant unit dispatched on a zonal view leaves node-level basis on the table every day. Report the capture rate directly and an uplift claim can be argued from something real, rather than borrowed from a good year.

Plan a battery at the zone and you optimize against a price it is never paid. The unit settles at a node, the money is in the node-to-hub basis, and the only defensible way to talk about edge is the share of that spread you actually capture. Everything upstream — the forecast, the ranking, the schedule — should be built at node resolution, because that’s where the asset lives.