I've been finding the reporting, opinion and debate online to be all-over-the-place.
It can be confusing, vague and not very informative. Journalists often just repeat what politicians have said and not offer much of their own insight or analysis. I've seen plenty of numbers, graphs and statistics that are presented without context or sources.
So, I decided to go read the AEMO, CSIRO and Frontier Economics reports myself—and it turns out, a lot of it is pretty accessible—you can just go read it.
For context: I'm a newbie when it comes to energy policy. I work as a computer programmer and I'm a bit of a numbers guy that likes to understand how things work.
In this post I'll try to summarise my understanding and walk you through it.
Let's start with some important concepts.
AEMO is a semi-government organisation that runs the wholesale electricity market called the National Electricity Market (NEM).
AEMO performs an array of gas and electricity market, operational, development and planning functions. It manages the National Electricity Market (NEM)[1]
The NEM facilitates the exchange of electricity between generators and retailers. Retailers resell the electricity to businesses and households.[2]
In short: AEMO's main job is to ensure the market runs smoothly between buyers and sellers to keep the lights on for Australian homes and businesses.
There are different kinds of participants in the market, including e.g.
There are more, but that should give you an idea.
Participants register with AEMO and pay a fee to participate in the market. My understanding is that fees are in proportion to involvement in the market.
To give you a sense, in AEMO's 2024 annual report, they had around $486M in total revenue, which I believe mostly comes from the fees collected from market participants.
AEMO's main expenses in 2024 were mostly labour (~1500 employees), IT systems for running the NEM, followed by spending on consultants and contractors.
AEMO also does some planning and forecasting work and publishes a report every two years called the Integrated System Plan (ISP).
The most recent one was 2024, the next one for 2026 is in development.
The ISP is a plan for investment in the NEM to ensure a reliable and secure power system through Australia’s transition to a net zero economy.[3]
The Integrated System Plan (ISP) is a whole-of-system plan that provides an integrated roadmap for the efficient development of the National Electricity Market (NEM) over the next 20 years and beyond.[4]
Its primary objective is to optimise value to end consumers by designing the lowest cost, secure and reliable energy system capable of meeting any emissions trajectory determined by policy makers at an acceptable level of risk.[4:1]
In short, the ISP aims to set a roadmap into the future to:
AEMO's ISP is trying to make some predictions about the future—which is not an easy thing to do!
AEMO's ISP report identifies the three most likely scenarios about the demand of electricity in Australia through to 2050:
- Step Change, which fulfils Australia’s emission reduction commitments in a growing economy.
- Progressive Change, which reflects slower economic growth and energy investment with economic and international factors placing industrial demands at greater risk and slower decarbonisation action beyond current commitments, and
- Green Energy Exports, which sees very strong industrial decarbonisation and low-emission energy exports. [3:1]
In plain language:
In the AEMO ISP 2024 report[3:2], they estimate the likelihood of each scenario:
After extensive consultation, AEMO assigned likelihoods of
- 43% for Step Change
- 42% for the similar Progressive Change
- and 15% for Green Energy Exports
So, step-change and progressive demand scenarios are pretty close. AEMO reckons step-change is ever-so-slightly more likely.
So for each scenario of future demand, we now want to know:
What is the best way to meet that demand with investment in generation, storage, and transmission infrastructure?
AEMO calls it the "optimal development path":
The optimal development path is the lowest-cost, resilient, pragmatic path to the NEM’s energy future of net zero by 2050, while also meeting government policies.[3:3]
The optimal path aims to:
My understanding is that AEMO uses a software simulation model[5] that crunches the numbers to find this optimal development path.
They run a bunch of simulations that take a bunch of inputs with their best guesses based on historical evidence, assumptions and forecasts about costs, timelines and more… the simulation plays out what would happen with different mixes of generation, transmission and storage technologies and infrastructure.
At the end, they take the simulation results and pick the one that meets all the requirements for the lowest cost as the "optimal development path" for that scenario.
My understanding is the biggest cost components are:
One of the major inputs to AEMO's ISP comes from CSIRO, which is a report called GenCost that looks at the capital costs of electricity generation + storage.
GenCost is a collaboration between CSIRO and AEMO to deliver an annual process of updating the costs of electricity generation, energy storage and hydrogen production technologies with a strong emphasis on stakeholder engagement.[6]
GenCost is an important input into AEMO's Integrated Systems Plan (ISP).
Updated every two years, the ISP outlines the lowest-cost investment to ensure Australians can access reliable, secure and affordable electricity, while meeting our national emission reduction targets.[7]
CSIRO's GenCost provides one of the biggest and most important inputs to AEMO's ISP report, which covers capital costs of power generation + storage.
AEMO takes GenCost as an input for the ISP and tries to estimate the total cost of the system, including other costs:
In addition, AEMO adds up all the other costs and works out the "optimal path" to comply with their other constraints to meet the demand in each of the three most likely scenarios of future demand: step-change, progressive and green-export.
Until recently, CSIRO and AEMO haven't included nuclear in the GenCost and ISP reports.
CSIRO's most recent 2023-24 GenCost report includes nuclear capital costs for the first time.
So, hopefully it's clearer about what AEMO and CSIRO's roles are—and what the purpose of the different reports are.
The Liberal–National Coalition recently released analysis by Frontier Economics that supports introducing nuclear power in Australia.
Before we get into their analysis… who are Frontier Economics? Until last week, I'd never heard of them.
We are the independent economic consultants for complex commercial, litigation, regulatory and public policy decision-making. Our results for clients are well regarded because they are underpinned by our economists' 30+ years of sector expertise, technical experience, and robust economic modelling and analysis.[8]
I had a few initial questions:
Frontier Economics say they have done these reports at no charge to the coalition, which was surprising to me.
The work presented in both reports is funded and directed solely by Frontier Economics. Consultation with various government and private sector parties was sought, to ensure we modelled the inclusion of nuclear power in the NEM most accurately.[9]
The cynical side of me expected their analysis and reporting to be biased and partisan.
Their co-founder and managing director Danny Price has this to say:
There is a large amount of ill-informed and misleading cost comparisons being shared about nuclear power in Australia, and Australians deserve better analysis and commentary to make the right decisions for our energy future.
— Danny Price, Managing Director, Frontier Economics
[9:1]
.. and to be totally honest, I completely agree with this sentiment.
I want to be informed, but everything I read online is trying to persuade me.
So, who is Danny Price anyway?
He is a leading expert in energy economics, and has 30+ years of experience advising governments, regulators and private firms on energy market design, energy reform and implementation, regulation, power trading and hedging, and asset sales and acquisitions.
He has advised many of the world’s largest energy utilities in countries including Australia, the US, Singapore, Kenya, India, New Zealand, China, Finland and Jamaica.[10]
He's done work for both the federal Liberal-National Coalition as well as South Australian Labor, as well as internationally. Been in the business 30+ years.
Frontier Economics released two reports:
Report 1 – Developing a base case to assess the relative costs of nuclear power in the NEM
Report 2 - Economic analysis of including nuclear power in the NEM
Report 1 replicates AEMO's ISP model to set up a base case as a comparison point.
It looks at AEMO's two most likely future demand scenarios: step change and progressive change.
Report 2 takes this model with the two demand scenarios, but makes one change:
What if nuclear power was introduced starting in 2036?
Frontier Economics make some additional assumptions about the capital, operating and decommissioning costs of nuclear power plants.
The results of their analysis is very interesting—it turns out there are tradeoffs!
Their analysis found:
First question I had was, what's the explanation? Why is it so different?
On the face of it, the capital costs of nuclear capacity can appear expensive compared to a solar or wind plant (which is about, respectively, $1,800/kW and $2,500/kW) and this simple cost comparison is often used to condemn consideration of nuclear power.
However, wind and solar only produce electricity about a third of the time on average and in the Australian NEM these sources of intermittent electricity generation tend to produce electricity at the same time and tend not to be operational at similar times. [11]
That makes sense. Solar arrays all come online during the day. They're all off at night.
This means that, broadly, the NEM needs about three times as much wind and solar capacity to produce the same electricity as a generator that can operate on command and, more or less, continuously – that is dispatchable generators.
In addition to the large amount of renewable capacity that has to be built and paid for to produce enough electricity, capacity also has to be developed to store electricity at times when there is surplus production from renewables to allow this to be discharged when there is a shortfall of renewable generation (which occurs roughly about half the time). [11:1]
In short:
If you just do a very rough comparison of capital costs:
It looks like 3 x $2k/kw should still be cheaper than $10k/kw.
Frontier Economics explains that the cost difference comes from all the additional storage and transmission costs that needs to be developed to handle the 3x generation capacity.
I think that's enough for now. If you made it this far—thank you very much for reading.
In short:
If you want to learn more, I'd highly recommend reading these reports yourself. It sounds intimidating, but my own experience was that they're well-structured and have a fairly straightforward writing style.
I hope this summary is helpful for other people like me that are newbies to the topic. Feedback is very welcome.
Power plants have a maximum capacity, but most don't run at 100% of their full capacity for a variety of reasons.
Solar fluctuates every day, wind can come and go, power plants might need to be scaled down or turned off for maintenance.
Solar in particular is highly variable.. it varies from zero at night, up to 100% during the day.
Coal and nuclear are at the other end in terms of variability: they can run continuously at nearly any level of capacity (above some minimum level).
Over the long term, a helpful simplification is to work out the percentage of actual electricity generated compared to the theoretical maximum.
This concept is called the capacity factor.
Typical capacity factors of different technologies:
Of course, this is still a major simplification that's helpful for making comparisons.
Sometimes the comparison doesn't tell the whole story. e.g
A solar array with a 30% long term capacity factor is quite different to a nuclear plant that's running continuously at 30%—even though they both appear to have the same capacity factor.
Can someone please tell me what this concept is called?
Is the concept related to reliability or variability? Or is it more that some generators are "dispatchable" where they can be dialed up and down to whatever the operator wants?
I kept getting confused between Megawatts (MW) and Megawatt-hours (MWh), this is how I understand it:
I'd love it if someone could clarify for me why capital costs are quoted in different units.
I believe that Frontier Economics is talking about peak capacity, so let's say you built a reactor with a 1000MW capacity.. $10k/KW is a $10B cost, is that right?
For CSIRO GenCost, I need to re-read the report I think.. it's not clear to me how $/MWh works.
One thing I've seen confused is the difference between cost vs price.
AEMO, CSIRO and Frontier Economics are estimating how much it costs to build and upgrade system infrastructure, which is different to estimating the market price of electricity.
Politicians will often talk about the price of electricity but these reports are talking about costs.
It's a simplification where the assumption is that lower costs will result in lower electricity prices for consumers.
There's a great website called OpenElectricity that makes information about the NEM visible and accessible to the public.
It's interesting to see how the demand varies as well as the proportion of coal vs renewable generation changing through the day—and over longer term.
Some things are unsurprising: emissions intensity lines up with fossil fuel usage.
Some are surprising: negative electricity prices!
Draft: 18th Dec 2024
Published: 20th Dec 2024