The number that started this.
We keep a live tracker of Australian energy, oil and gas, infrastructure and data centre projects. As at this week it holds 2,510 projects, 1,504 of them in energy. Inside that energy set the split is the interesting part.
- 304 approved. Planning consent in hand, not yet building.
- 35 committed. Financial close reached.
- 123 under construction.
- 30 in commissioning.
Approved outnumbers under construction by more than two to one. Some of that is ordinary sequencing, because an approval is granted years before a shovel moves. But the ratio has been widening, and when we ask owners what is actually holding a specific project, the answer is rarely planning and increasingly rarely money.
It is the connection.
What the connection process actually demands.
Connecting a generator or a battery to the National Electricity Market is a negotiation, a modelling exercise and a commissioning programme rolled together. It runs roughly like this.
- Connection enquiry and application. The proponent approaches the network service provider, scopes the point of connection and starts the technical conversation.
- Generator performance standards. Every plant negotiates its own performance standards against the automatic and minimum access standards in the rules. This is where months disappear. Each standard is argued on evidence, and the evidence is modelling.
- Model acceptance. The plant has to be represented in both load flow and electromagnetic transient models that the network and AEMO will accept. For inverter based resources in a weak part of the grid, the electromagnetic transient work is the whole ballgame.
- Hold point releases. Energisation, then staged output, each released against measured performance that has to match the accepted model. A plant can be physically complete and still be months from full output.
None of this is new. What is new is the volume. Our tracker lists 610 solar projects, 344 batteries and 166 onshore wind projects across the country. Every one of them queues through the same small number of people.
Why the pool is so small.
The discipline sits at the intersection of power systems engineering, modelling software and regulatory negotiation. There is no degree that produces it. People arrive after years inside one of five places, and all five are hiring.
The network service providers. They assess the applications, so they train the deepest assessors. They are also the hardest employers to recruit out of, because the work is interesting and the pension and conditions are good.
AEMO. Small team, enormous influence, and anyone who has done time there is instantly credible on the proponent side.
Specialist consultancies. The firms that sell connection studies hold most of the contract capacity in the country. They are also where most proponents go first, which means the same engineers are working both sides of a busy market.
Inverter and turbine manufacturers. They own the models, and model behaviour is where disputes land. Manufacturer engineers understand the plant at a level nobody else does, and are usually tied to a global roster.
Owner side connection teams. Still rare. A developer with three or four projects has historically outsourced this, then discovered on project two that the knowledge walked out with the consultant.
In our experience there are a few hundred people in Australia who can genuinely run a connection from enquiry through to full output release. The number of projects that need one is in the hundreds as well.
What owners get wrong when they hire for it.
Four patterns come up often enough to be worth naming.
- The role is written as a project engineer role. It is advertised to a construction audience, the salary is benchmarked against construction, and the people who could do it never see it. Connections and compliance is a distinct market with its own pricing.
- The hire is made too late. Bringing a connections lead in after approval means they inherit a point of connection, a plant design and a set of assumptions they would have argued with. The value of the role is highest before the application goes in.
- Modelling is treated as a deliverable, not a capability. A consultant produces a model, the model is accepted, and then the plant behaves differently in commissioning and nobody in the owner's team can interrogate why. The hold point sits there for a quarter.
- Commissioning is staffed on the construction schedule. The electrical and compliance side of commissioning carries on long after the civil and mechanical teams have demobilised, and the budget rarely reflects it.
What we would do about it.
For owners and developers, the honest position is that you cannot hire your way out of a national shortage, so the plan has to be structural.
Hire one permanent connections and compliance lead early, well before the connection application, and give them authority over the technical negotiation. Buy the modelling depth on contract, because that market genuinely works and the day rates, while high, are cheaper than a quarter of delayed output. Keep the model in house after acceptance so somebody on your payroll can defend it. And staff commissioning as its own phase with its own people, rather than assuming the construction team absorbs it.
For engineers, this is the clearest value signal in the market right now. If you are a power systems engineer who can run electromagnetic transient studies and sit in a performance standards negotiation without flinching, you are in the smallest and best paid pool in Australian energy, and the queue of approved projects says that will hold for years.
Approved is not building. The people who close that gap are worth finding properly.