If storage is minor, then the transition will cost the moon.

I've been mulling over the 2024-25 GenCost report and AEMO's ISP over the last week to try to decipher in my own mind exactly what their authors were attempting to achieve with both documents.   Both documents have been completely politicized  since they were released and it seems obvious to me that the government has directed all and sundry to pretend they are something they are not.  So here is a quick overview of my thoughts and analysis.

GenCost

GenCost is an attempt to model comparative pricing of different energy sources if you were:

1. Building new green-field generation plant.
2. Doing it in a green future environment. 

You can see this by means of their discussion around future coal technologies and really only pricing in ultra-supercritical coal or carbon-capture technologies. Overall the document is bound by a great number of assumptions mostly that the continuation towards net-zero will continue and therefore a number of existing technologies will be seen as unviable.

But to be clear, it is NOT the comparison of the cheapest forms of electricity production and makes no statement about whether any of these options will be cheaper than the current system.

The other issue with GenCost is that parts of it make no sense.

This is figure 6.2 in the report which as far as I can tell it suggesting that the storage component of VRE costs is some small ratio of the cost of generation. This is simply not the case, more on that later.

Integrated Systems Plan

I think this is the most confused document that I have ever read, and it certainly has caused massive polticial issues because the government pretended it is something it is not and then seems to have forced everyone involved to do the same. 

As far as I can tell the document is actually a plan of all the major things that need to happen in order for the electricity network to transition to meet the government's net-zero targets.  Aside from the fact that it does not contain ALL the things that are required, the major issue with it is that it is totally undeliverable and completely under-costed.

I have significant issues with the modelling. My experience with economists within the energy sector is they have limited real world experience, they will model efficiency and "optimal path" based on what their spreadsheets tell them rather than physical reality and their models suffer from extreme bias due to this.

This is likely to lead to estimations of effort and risk/cost analysis being way off the mark and therefore real world "deliverability" being completely understated. As I have said before, the more you have to do the more likely for things to go wrong.

As per my concerns with the GenCost report, as I have spoken about earlier, is deliverability and cost. Let's continue with the with the same problem from above, VRE storage, to demostrate my concerns.

The ISP states:

“In total, the NEM is forecast to need 36 GW/522 GWh of storage capacity in 2034-35, rising to 56 GW/660 GWh of storage capacity in 2049-50. ”

and has this chart - Storage installed capacity and energy storage capacity, NEM (2024-25 to 2049-50, Step Change) 

Snowy Hydro 2.0  (which I will discuss another day) will provide 320GWh of storage capability to the NEM and "hopefully" be available in 2029. It is  completely over budget at $13 billion (which doesn't include Humelink now at $5 billion) so the projects true costs are more like $20 billion, but let's go with $13B to be kind.  On that number this works out to be around $40.625m/GWh or $40,625/MWh or $40.63/kWh. 

So there is around 200GWh of other storage to fill the gap over the next 10 years, at $40.625m/GWh that equates to just another $8.125 billion, but that number is not realisable.

Snowy 2.0 is not a baseline for comparative storage projects in Australia and extrapolating out its $/MWh does not provide a real world estimate of what the additional storage will cost because you need the same scale to get the same costs.  Something Queensland is finding out 

Work on the $14.2b Borumba Pumped Hydro project south-west of Noosa was due to begin next year and deliver electricity by 2030.

The 2,000-megawatt facility was a signature part of the former state government's renewable energy plan to power two million homes.

But a Queensland Hydro report, commissioned before the election has revealed the project will now cost $18.4b and will not be ready until 2033 at the earliest.

Borumba is a step-down from Snowy Hydro 2.0 , being  2GW/48GWh . At $18.4b it will cost $383m/GWh or $384,000/MWh or $384/kWh.  That's nearly 10 times the price of Snowy 2.0

And it gets worse the smaller you go.

Genex kidston pumped hydro is a 2GWh pumped hydro scheme, it reportedly cost $777M, however that didn't included $150M in free money the Queensland government gave them as it also required a 190km HV line they couldn't afford. It took 5 years+ to build and is considered cheap due to the existing suitable geography because it was an old mine site. All in (including debt ) its over a $1bn in required return and that is ignoring O&M. That’s $500m/GWh or $500,000/MWh or $500/kWh. 

If you keep going smaller, into the realm of chemical battery storage the prices don't get any  better and the standard costs of the lower end of large scale storage in the real world looks to be around $500/kWh. 

Once you get to consumer batteries it's $1500/kWh or more.

So the additional 200GWh is not just another $8.1bn it is going to be more like another $80-$100bn of required investment, and this doesn't include any additional system services or other network augmentations that will be required.  

If storage is only a minor part of the VRE all-in costs, as per the GenCost report, then we are all in a lot of trouble.