Why Connecting a Power Project to the Grid Takes Years

Building a solar farm or a wind project is a construction task measured in months. Getting it connected to the grid and able to sell power is a regulatory and engineering task measured in years, and it is the binding constraint on how fast generation capacity actually grows.

The bottleneck is not manufacturing or finance. It is the queue.

What interconnection involves

A generator cannot simply plug into the network. Adding power at a point changes flows across the system, and the operator must establish that the addition will not overload equipment or destabilise the network under contingency conditions.

That requires studies: modelling the project's effect under a range of scenarios, identifying upgrades needed, and assigning who pays for them. Each stage takes months, and each depends on assumptions about which other projects in the queue will proceed.

The cascade problem

That last dependency is the source of much of the delay. Studies are conducted in sequence, and each assumes the projects ahead of it are built. When one withdraws, the assumptions underlying every study behind it change, and work may have to be redone.

Queues have therefore accumulated projects that were never likely to be built — speculative applications submitted cheaply to hold a position. Each one distorts the studies behind it. Reforms in various jurisdictions have moved toward studying projects in clusters and requiring deposits that make speculative entries costly, with results that are still emerging.

Transmission is the deeper constraint

Interconnection queues are a symptom. The underlying issue is that moving power from where it is cheap to generate to where it is consumed requires high-voltage transmission, and building transmission is harder than building generation.

The result is that the best generation sites — consistently windy, reliably sunny — are frequently far from demand and poorly connected to it.

Curtailment

When generation exceeds what the network can carry, output is curtailed: the operator instructs generators to reduce production despite available wind or sun.

This is an ordinary operational necessity and, at scale, a signal. Persistent curtailment in a region means the constraint has moved from generation to transmission, and further generation there adds less than the headline capacity suggests. It also damages project economics, because revenue assumptions were built on expected output.

What storage does and does not solve

Batteries shift power across hours, which addresses part of the problem: storing midday solar for the evening peak reduces curtailment and improves the value of generation.

They do not solve geography. Storage moves energy in time, not space. A region with more generation than transmission capacity still needs wires, and a region needing power for days rather than hours during a weather event needs something other than the durations most current battery installations provide.

Who pays for the upgrades

When a study finds that connecting a project requires network reinforcement, someone must fund it, and the allocation rules are a persistent source of dispute.

Under one approach the connecting generator bears the full cost of upgrades its project triggers. That is straightforward and produces a queue-position lottery: the project that happens to be studied when the network runs out of headroom carries costs that later projects then benefit from for free.

Under a shared approach, costs are spread across beneficiaries, which is fairer and considerably slower to agree, because identifying beneficiaries of a transmission line is itself contested.

Neither rule is obviously right, and the choice materially affects which projects get built — a cost assignment can render an otherwise viable project uneconomic overnight.

Capacity, energy, and being paid for two different things

Electricity markets typically pay generators for two distinct products, and conflating them is a common source of confusion.

Energy payments compensate for electricity actually produced. Capacity payments compensate for being available to produce when needed, whether or not the plant runs.

The distinction matters for comparing technologies. A resource producing cheaply but only when the weather cooperates earns well on energy and less on capacity. A plant that runs rarely but can be called on reliably may earn most of its revenue from capacity.

This is why the cost of generation alone is an incomplete comparison. A system needs both cheap energy and firm availability, and the market pays separately for each because they are separately valuable.

Behind the meter, and avoiding the queue entirely

One response to interconnection delay is to avoid the connection altogether by building generation on the same site as the load it serves — a data centre with dedicated generation, a factory with its own solar array.

This sidesteps the queue for the portion of output consumed on site, which is why it has become attractive for large industrial loads facing multi-year waits. It does not remove the need for a grid connection entirely, since most such sites still draw power when their own generation is unavailable.

It also raises questions regulators are still working through: whether large self-supplied loads should contribute to the fixed costs of a network they rely on as backup, and what happens to those costs if enough load departs.

What this means commercially

For anyone assessing an energy project, the queue position and transmission situation are as material as the resource quality and the equipment cost. A project with excellent sun and no path to market is not a project.

It also explains why announced capacity and operating capacity diverge so widely. Announcements reflect intentions; the queue determines outcomes, and a substantial share of what enters it never emerges.