Which state pays for interconnectors?

This article was originally posted on the Australian Energy Council website.

Paul Keating said “Never get between a premier and a bucket of money”. When it comes to paying for interconnectors, the reverse holds equally true.

The Integrated System Plan (ISP) for as much as $23 billion to be spent over the next 20 years to develop a much deeper National Electricity Market (NEM) transmission grid has opened up an inevitable, and rather unedifying, fight between the states over how to divide the cost pie: each wanting the smallest slice. The Energy Security Board has been given the fraught task of sorting out the squabble¹. And then to add further confusion to the mix, last week’s federal technology roadmap has opened the door to some Commonwealth funding².

We take a look at how major transmission paths are presently funded, and some options.

The Warring Tribes

At the time of federation, there was a transcontinental telegraph system, so the founding fathers sensibly allocated telecommunications to the Commonwealth in the Constitution. The states have no part in how this document was delivered to you.

However, the concept of electrical power being moved between the colonies was then beyond imagination. Energy was not mentioned in the Constitution and by default went to the states.

Thus, for most of the twentieth century, each state developed its own self-reliant power system and shunned submitting themselves to a national architecture. This led to plenty of amusing anomalies, such as the Hume Power Station on the Murray River having two equal sized units, one connected to New South Wales and another to Victoria, each continuously running at equal output, even when the river was below half-flow.  The late Robert Booth captured the theme in the title to his book, Warring Tribes ³.

Interconnection between Victoria and New South Wales was unavoidable when the Snowy Scheme was constructed. Belated interconnections to South Australia (1989) and Queensland (2000) eventually followed, however state utilities saw these purely to share opportunistic flows between them, and never really as a “national grid”.

How is transmission funded now?

The “shared” network is the vast majority of poles and wires: everything that is not part of a dedicated connection to a single generator or consumer. Shared costs are recovered through complex formulae that ultimately result in it mostly being paid by consumers through smeared energy levies. There is some geographical cost reflectivity, but it’s not particularly pointed.

This sharing occurs intra-state. Shared transmission costs are pooled into one state kitty recovered within that state – even in Victoria where there are multiple transmission owners. Assets supporting interconnection similarly fall along geographical lines, sometimes right up to the last tower on either side of the border.

Thus, although interconnections provide national benefits, their funding is largely local, opening inevitable fairness complaints – why should we be paying for this when it helps someone else?

Whilst states largely pay for everything within their borders, there is, at the margin at least, a small degree of national sharing:

• When the original interconnection between the Electricity Trust of South Australia (ETSA) and the State Electricity Commission of Victoria (SECV) was built, a now expired deal was negotiated that saw ETSA pay a share of the SECV assets. Bilateral deals remain possible, but seem unlikely with planning moving into the national sphere under the Australian Energy Market Operator (AEMO).
• When AEMO sells Settlement Residue Instruments – a NEM surplus created by interconnectors – the profits go to defraying some transmission costs on either side of the interconnector. This tends to be small compared to the cost of the assets.
• The “Inter-Regional Transmission Use of System” (IR-TUOS) charge leads to some sharing. At the end of the year, the net energy flow across an interconnector is tallied up and the net exporting state’s transmission agency charges the other as if the net flow was one of their own customers near the border. Because interconnectors have two-way flows, and there is no attempt to assess the value of the flows, IR-TUOS charges tend to be also very small compared to the true costs and benefits of interconnection.

How was Basslink built?

The Basslink cable between Victoria and Tasmania built in 2005 is peculiar on many fronts:

• Most of its assets are in Commonwealth waters – so there is no state transmission agency to take on the job by default.
• It is Direct Current (DC) and therefore independently controllable. Unlike an alternating current interconnector it is not an intrinsic part of the broader network: technically it behaves more like a controllable generator than a line.
• It was built as a merchant asset, and not underwritten by captive transmission customers.

Just after the NEM started, we followed a North American experiment into merchant transmission. This is where entrepreneurs build DC lines purely to profit from the price difference at each end, i.e. the entrepreneurs rather than AEMO retain the settlement residue. This is how Basslink was built, subject to a long-term agreement with Hydro Tasmania. Costs are not recovered from customers so it avoided arguments about which state should pay.

Two other smaller merchant interconnectors were built at a time of enthusiasm for merchant transmission: Directlink and Murraylink. However the experiment is now considered a failure as:

• DC technology is expensive except in situations like undersea cables where there is no alternative.
• The concept was more suited to nodal markets: Directlink and Murraylink are connected to the edges of their respective states and were so heavily affected by congestion elsewhere they proved unprofitable.

Ultimately Directlink and Murraylink were reclassified as part of the shared network at a written-down value that is now being recovered from customers like conventional network assets.

It seems most unlikely that any of the very expensive interconnectors being proposed in the ISP will be merchant, so the question of who pays is therefore unavoidable.

What is an interconnector anyway?

When thinking about an “interconnector”, most people envisage a discrete transmission line crossing a state border. But, except possibly in the merchant cases described above, this is quite wrong. The Alternating Current (AC) transmission network is a highly intricate spiderweb operating collectively where every link affects every other link. An “interconnector” is in fact only a mathematical construct that exists within AEMO’s dispatch algorithm and settlement systems used to subdivide regional pricing and settlement. And those subdivisions are arbitrary – they have no basis in physics or economics and come about purely through the historical accident of colonisation.

So, when the ISP recommends a new “interconnector”, it is in fact recommending a complicated series of investments strengthening the overall grid – most of which are deep within the states and inseparable from the rest of the network. For example, the transformers at the South Morang Terminal station, in Melbourne’s northern suburbs, are no less a part of the New South Wales to Victorian interconnector as the lines that actually cross the Murray River.

Who pays: An inane issue with a lot at stake

So, following the default expectation that those assets associated with an “interconnector” are paid for by the customers who happen to live in the states in which they are built, anomalies are immediately apparent. For example, the majority of expense associated with Project EnergyConnect, between New South Wales and South Australia, falls in New South Wales. New South Wales customer groups don’t oppose the project, but consider this unfair. Yet on the other hand, if costs were shared according to the relative sizes of the customer bases, it would fall even more strongly upon them.

New South Wales’ geographical shape will always result in the majority of the assets required for an interconnector with any of their three neighbours naturally falling within their borders. So it’s unsurprising it would seek to change this default, whilst the other states would resist.

We could dream of a “one Australia” attitude: AEMO once proposed nationally pooling all transmission costs⁴. But that utopian vision has problems too. State governments are imposing ever greater parochial variations on the NEM. For example, Victoria recently passed legislation to empower the state minister to direct the construction of transmission assets outside of the NEM’s cost-benefit based planning framework⁵. It would clearly be unreasonable, and a moral hazard, for customers in other states to pay for such follies.

There is no perfect answer here, except to go back in time and alert the founding fathers to the NEM!

Beneficiaries pay

The ISP has recommended a future construction of the next Bass Strait cable, Marinus Link, at a cost of around $1.8b (Stage I, with a further $3.2b for Stage II). There is no intention to build it merchant like Basslink, and, being mostly outside state territory, there is not even a default sharing arrangement. The Tasmanian Government’s view is clear, that it will be for the benefit of the NEM as a whole, and as the smallest state, Tasmanian customers should pay much less than half.

TasNetworks has articulated an interesting approach⁶ that includes some good economic rationale married with a great pile of pragmatism. The approach intentionally doesn’t open the pandora’s box of who pays for existing assets – what’s done is done. It also leaves alone the allocation of all shared transmission costs to customers and largely retains the “each state looks after itself” regime. However for new assets that are part of interconnector projects, it proposes costs be shared in a new way.

The NEM’s transmission planning cost-benefit framework⁷ is already supported by economic modelling which is necessary to show that the total market benefits of an interconnector project exceed its cost (with the notable exception of projects directed by the Victorian minister). It does not matter where those benefits fall – it could be customers anywhere, or even generators who benefit from the interconnector.

TasNetworks proposes that the modelling can determine the incidence of the customer benefits, i.e. it can identify which states’ customers benefit and by how much. The relative proportions are then used to determine what shares of the new project will be allocated to each states’ customers.

This seems fair, but it comes with some complexities to be aware of:

• Beneficiary customers can be quite remote, i.e. even customers in Cairns will pay for some of a Bass Strait asset.
• A cost-benefit test’s beneficiaries are not just customers. It is quite possible, and indeed very sensible, for much of an interconnector’s beneficiaries to be generators – say windfarms presently being constrained. TasNetworks accepts this, but rather than open another pandora’s box about generators paying for transmission, proposes only looking at the ratios of the customers’ benefits across the states. But, if the customers’ side of the benefits is small, it could result in billions of transmission costs being allocated to customers in a state whom the modelling suggests only benefit by millions.
• It opens a new concern about the boundary of interconnector projects and planning for intra-state purposes. As noted previously, they are not really separable and this will require judgement and regulatory scrutiny.
• Billions of dollars of cost allocation will hinge on the vagaries of a market modeller’s results. Anyone who has ever engaged a modeller will naturally feel some discomfort about that, and disputants are sure to magically calculate quite different results.

The Commonwealth to the rescue?

Many of the intractable difficulties of the Australian federation have been resolved with buckets of Commonwealth money. There are signs of that in last week’s announcements, with federal taxpayers’ money being suggested in order to advance interconnectors. Its early days, but we should be wary of such a white knight:

• The money being spoken about is only in the order of hundreds of millions, when the projects are in the billions. Thus it seems intended to be used to grease state-federal negotiations rather than a way to achieve good objective planning.
• There are moral hazards, in that states lose the incentive to negotiate agreements maturely between themselves when they can go crying to their parent.
• Every interconnection project unavoidably affects every competitive activity in the NEM, be it generation, storage, retailing etc., so to reduce investor risk it is critical that the planning process is predictable and transparent. A federal funding process risks becoming haphazard and increasing rather than reducing risks across the NEM.
• There is a new problem of fairness in that Western Australian and the Northern Territory will miss out on this largesse as they have no interconnectors.
• Finally, there is an underlying question of whether it is appropriate for taxpayers money to be spent on electricity networks which is effectively subsidising energy consumption.