The first problem with the 2020 Integrated System Plan is that it was released by the Australian Energy Market Operator, and not by federal energy minister Angus Taylor.
Contrast this when the ill-fated National Energy Guarantee was announced – we had the head of the Energy Security Board, AEMO, the Australian Energy Market Commission and the Australian Energy Regulator all lined up behind the Prime Minister of the day. When the ISP is announced we have the AEMO on its own. Perhaps I am over reading this, but I doubt it.
When was the last time anyone heard Taylor talk about the ISP in public? Or even give a speech about it, either positive or negative.
To be fair, the ISP does point out that under the terms of the agreement between the Feds and the NSW arrangement over one of the key target projects in the ISP – the central west renewable energy zone – will result in any excess of costs over benefits of transmission being met by the Federal Government. But it needs more than just clauses in contracts.
How much legal force does the ISP have?
Always bear in mind that the ISP proceeds without the existence of any decarbonisation objective in the National Electricity Objective or National Electricity Law. (One of the great scandals of the country’s energy laws.
Yet decarbonsiation awareness permeates – as it should – every electricity transaction and decision. This fundamental distortion between political optics and system reality lies at the heart of often subconscious distrust of politics, but that is for another story.
So, bearing the big numbers in mind, is it the case that if something is recommended in the ISP, does it have to be built?
A summary of the position is on the AER web page and I quote it in part, although it is undoubtedly only the first and not the last word on the ISP’s legal status. In looking at this reference it’s worth bearing in mind that these words can be read in the context of whether the ESB is or should be retained.
“The Australian Energy Regulator (AER) is developing guidelines to make the integrated system plan (ISP) actionable. These include:
- – A new cost benefit analysis guideline
- – A new forecasting best practice guideline
- – Updates to existing regulatory investment test for transmission (RIT–T) instrument and application guidelines.
These guidelines are part of a broader reform led by the ESB, which has resulted in changes to the National Electricity Rules to convert the ISP into action (effective 1 July 2020). The intent of the new rules is to streamline the transmission planning process while retaining rigorous cost benefit analysis.
Under the new rules, the ISP will provide a coordinated whole-of-system plan for the efficient development of the power system that meets power system needs in the long term interests of consumers. The ISP will ‘action’ key projects by triggering RIT–T applications. The RIT–T is a cost benefit analysis that transmission businesses must perform and consult on before making major investments in their networks.
Under the new rules, the ISP will be subject to additional governance arrangements through binding cost benefit analysis guidelines and forecasting best practice guidelines. There will also be changes to the RIT–T instrument and associated application guidelines to be consistent with the new planning process.”
It’s not clear to ITK that the AER’s rules are going to achieve much improvement at all for the simple reason that there will still be no alignment between the RIT-T and the ISP. But let’s start with the process.
Although the ISP does away with some baggage of the RIT-T – mainly around identifying the need – it does nothing to alter the narrow cost benefit analysis [CBA] that represents the basic problem with the RIT-T.
What is going to happen every time is that AEMO and the ISP’s cost modelling will be wrong. Planning bodies never get costs right and everyone knows it.
It’s not AEMO’s core function to be on top of the private sector budget for every asset in the NEM. So with transmission, typically, the initial cost estimate is too low. An example is the proposed new SA-NSW transmission line, although that was produced under the old guidelines. If the cost is higher than estimated we go into the “feedback loop”.
As recently as July 10 AEMO submitted to the AER in regard to the new rules:
“Our work is highlighting that important elements of the draft guidelines are proving difficult to apply and are creating analytical restraints which undermine efficient and timely scenario analysis and system planning, and in doing so, appear to risk the status of the ISP as an actionable system plan. …..
“The collective effort of stakeholders and AEMO in developing the ISP are ultimately undermined if an investment proposed in the ISP faces difficulties in satisfying RIT-T requirements, even where ISP and RIT-T inputs and assumptions are consistent. This outcome would also be inconsistent with the expressed policy intention of Energy Ministers and Governments to provide for an actionable national system plan that achieves the necessary transmission build “ Audrey Zibelman, CEO AEMO
ITK’s conclusion is that as things stand the RIT-T remains the main legal instrument and the main road block for new transmission. Without more fundamental reform to the RIT-T I expect to see transmission development proceed too slowly.
I think State Governments in NSW, Victoria and South Australia also see that this new process is overly bureaucratic and does not achieve the desired objective, and so they will proceed independently. And a reminder that the objective is to have enough transmission built to enable wind, solar and other resources to be built to replace the coal generators that close over the next ten-fifteen years. It’s as simple as that.
The final ISP is a polished document
First thing is to acknowledge what a substantial piece of work the ISP is. Secondly, for those of us that spent time with the draft report, it’s equally worth acknowledging how much improved the final version is.
Those improvements are in the detail of how the modelling is done, the scenario weights, the delineation of firming power by duration and in various other areas. Each improvement is small but in total – like any quality product – they add up.
Of interest to me were the scenario weights. Those weights are subjective, obviously, as there is no history to use as a weighting guide. The point being, I suppose, that the slow change scenario is regarded as unlikely. Yet that frankly is what Federal politics aims to influence.
I always like to keep the big numbers of electricity as a reference to decide on the materiality of benefits and costs. Some of those numbers are annual revenue to NEM wide retailers of about $41 billion and regulated asset bases [RAB] of wires and poles companies of about $85 billion (about $67 billion of distribution and maybe $20 billionn of transmission, maybe less). The wires and poles sector is likely more valuable than the generation sector, even allowing for rooftop generation assets.
In that context the transmission investments that would result to 2030 if the ISP is all there was to it would be $10 billion or less, with an annual cost to consumers of $1 billionn. In the following table note that a CPA is a “Contingent Project Application” – it’s basically the final hoop to jump through getting paid by consumers for building new transmission. A PADR is a preliminary step. A reminder that the few transmission projects that have ever passed the RIT-T test in Australia have generally taken about 7 years to move from first sponsorship to being operational.
I leave readers to make up their own mind about value for money. For an average household the cost is about $30 per year, once it’s all built by 2029. For a comparison the NBN cost is $51 billion, the 12 submarines when delivered in the 2050s will be over $100 billion in capital cost etc. The cost of every hour of a NEM wide blackout based on market price cap of $14,500/MWh and say 35GW of maximum demand is nearly $500 million.
The ISP does not have a strategy as such, it has a least cost and least regrets modelling approach based around a series of scenarios.
However, what the modelling produces is akin to a strategy in that it wants to build out new investment as close to load as possible and with as little new transmission as can be managed.
As such resources with high capacity factors, such as solar and wind in North Queeensland and wind in Tasmania tend to be de-emphasized in favour of marginally worse resources close to load.
One side effect, no doubt allowed for in the modelling, is that the portfolio benefits of North Queensland wind and solar are lost until after 2030. Those benefits come from Queensland wind having low or negative correlation with wind from other regions and Queensland solar tending to perform better in winter than solar from other regions.
A couple of new figures from the ISP are on this point. The text advises caution with these graphs.
What I see from them is a relatively flat, delivered-to-market solar cost curve, but a steeper wind delivered-to-market cost curve. And for wind far North Queensland and Tasmania have the advantage, even when transmission is taken into account. Pre -transmission the charts purport to show Darling Downs solar at less than $30/MWh and some wind at just over $30/MWh. Much other analysis shows wind costs for good wind projects being lower than solar costs for good projects.
The ISP states that 43% solar and 57% wind is the mix that minimises the need dispatchable generation. On the face, of it that’s a surprising conclusion as many other studies have ended up with a lower share of solar relative to wind.
Dispatchable capacity will need to reflect seasonality
Finally, for this note there is the question of how much dispatchable generation and when it will be needed. The ISP states that 6-19 GW of new dispatchable capacity is needed by 2040. That is a massive range. There is maybe more that $25 billion difference in capital cost between the two figures.
Also, I’m far from convinced that the ISP authors has their costs right on dispatchable generation just yet.
But putting that to one side, what is done is separate out short, medium and long term storage. In recent times its become clear to ITK, and probably we are the last to realise it, that the seasonal nature of VRE (variable renewable energy, wind and solar) production and demand means that no amount of short and medium term storage (under 10 hours) will likely be enough in summer without a massive over build of wind and solar, and equally there will tend to be a persistent excess in the December half.
No individual excess of demand over production might exceed, say, 100 hours in a typical year and most will be far shorter but if those short runs all come one after another then the short and medium term storage is never fully recharged.
The ISP talks about 24 and 48 hour pumped hydro but other than the Snowy 2.0 pumped hydro scheme that is not being talked about seriously by any proponent in the public domain.
So, for the time being that remains a placeholder as decarbonisation proceeds. It can be done easily enough with existing gas and the remaining coal. Conceptually it could be done by hydrogen which despite its many disadvantages does have a very, very low storage cost. Hydrogen at around $1/kg gets you to $6/gigajoule.
David Leitch is a regular contributor to Renew Economy. He is principal at ITK, specialising in analysis of electricity, gas and decarbonisation drawn from 33 years experience in stockbroking research & analysis for UBS, JPMorgan and predecessor firms.