For open cycle gas turbines, the apparent story is that orders are backed up for years and years. But:
A) not sure how things stand for the reciprocating engine supply chain
B) US DC developers seem to be able to procure turbines, though maybe it’s vaporware all the way down
- $160-200+/MWh looks OK I guess if speed to market is your main concern, but it should be possible to do quite a lot better with a mix of grid and behind the meter bulk RE, potentially preserving onsite backup too (lots of capital to have sitting around, but most of the gas gen cost is gas itself)
- emissions would be about 1.4 mtpa (including NSW gas extraction scope 3)
- fuel use and emissions would be *way lower* if the gas engines were just backup for when grid power was not available (assuming NSW is able to keep retiring coal and adding clean, given pace of grid demand growth)
- this kind of extra gas demand would be a big complicating factor for gas security policy, as well as for the hopes of the steel sector to swap out coming coal for natgas
- assuming 40% capacity factor this plant would take about 21 PJ gas annually, 4% of current Eastern domestic gas demand, or 17% of NSW, or 7-11% of gas potentially reserved for domestic use under new Cth reservation under development, or 29% of potential annual output of the Narrabri CSG project
- reciprocating engines plus assumed $12/GJ gas price plus guessed 40% capacity factor (reflecting guess on underlying utilisation of DC for inference oriented AI) means something like $160/MWh electricity cost (if no carbon value; $200/MWh at $50/t CO2 price)
www.abc.net.au/news/2026-03...
No idea if this NSW data centre project with associated 700MW reciprocating gas generation will actually happen at full envisaged scale. But some quick thoughts:
- a little surprised that local AI rushers are turning to gas as in US, given local gas adequacy fears
Yeah I am leaning towards “this will have modest near term energy impacts”; people bracing for another post-Ukraine-invasion shock are probably not going to see one. But I won’t be confident in my hunch! Gotta see what actually happens.
And similar to the last couple of years of less-disruptive-but-still-big events in and around oil producers, which produced transitory blips. Will oil markets again settle down rapidly after modest spikes? Or is this chain of events going to prove more sustainedly disruptive? 🤷♂️🤷♂️🤷♂️
It is; but also I’ve been repeatedly surprised in recent years as various serious ructions in and around oil producing or transport regions failed to produce more than blips in a price moderation trend. The current conflict could prove similar (whatever its other effects)
This episode also features:
Bunfight at the IEA Corral - will cowboy tactics prevail?
An irrelevant, inaccessible yet sadly resonant viewing recommendation from me!
A startling twist on BoCA!
Here’s the report, “The Case For Pricing Pollution: reducing emissions, strengthening the economy, and delivering a fair share for Australians”, by Reuben Finighan and Ingrid Burford at The Superpower Institute: www.superpowerinstitute.com.au/work/the-cas...
OUT NOW: a new paper offers two simple tricks to fix the Budget, the economy and the transition: taxation of carbon and gas resource rents. Should Treasury salivate, or should PMO break out in a cold sweat? LMSU assesses! www.letmesumup.net/superpower-i...
(QLD currently in Scyllaville)
- batteries are tough to get into and tough to leave alone. Diverse, fast moving, vastly important to energy, industry, appliances, defence. But also very crowded, subject to everyone else’s strategic policies, and very jumpy on match of supply growth and demand growth
Whither a testing facility for an embryonic Australia battery industry? www.afr.com/policy/energ...
Some thoughts:
- strategic industry policy is hard to steer between the Scylla of “random abandonment of strategy” and the Charybdis of “fail to update strategy in light of experience”
The US provides 14% of the IEA budget, but the budget was only €78m in 2024. ~€11m/yr does not seem like much to find if the alternative is advice constrained against the interests of other members.
Official cooperation on data would be a bigger problem. But it may need to be solved!
U.S. Tells International Energy Agency to Drop Its Focus on Climate Change www.nytimes.com/2026/02/19/c...
Not great for the IEA’s ability to do the job the vast bulk of its members want done.
The biggest increase in energy costs to most households in NSW and QLD over the past twenty years was the distribution network investment surge and utilisation decline in the early 2010s. All under public ownership at the time! Ownership is no inherent protection for the public.
The Fin story hypes up the “transition at risk” angle but the project is still pretty clearly worthwhile (and indeed the added cost of most everything else increases the project benefits along with costs). But “who pays what share” remains up for argument, and certainly has an impact!
www.afr.com/companies/en...
Transmission build costs are way higher than we thought they would be 5 years (and one global inflationary surge and series of interest rate hikes) ago. PEC was the one big 🇦🇺 T project in flight while we discovered this. Who should wear the difference?
Maybe for clarity “more painful” rather than harder - if anything the decline in the returns on investment cements the inevitability of rebalancing at some point. But it won’t be much fun for people & govts when it eventually happens!
(Like it won’t disappear, it’s a genuinely amazing set of industries. But it surely looks different with less of an infinite supply of who-cares-about-returns capital. Slowdown in supply growth: definitely. Maybe slower innovation? Or even more Darwinian cost cutting? Even more export orientation?)
Hugely significant question: the cleantech flood is wonderful for climate but partly a strategic choice, but also a product of unsustainable production-ism. When savings/consumption rebalancing happens, what happens to Chinese cleantech?
The arguments for rebalancing the Chinese economy are stronger than ever but the pain and pressures involved remain unappetising for policymakers, hence fiddly pseudo-supports for consumption that are really production subsidies or shuffling, and drives against involution without altering its causes
We kept the episode simple this time (too much packed in to this book, and summer!) but regular programming will resume in a fortnight. Don’t forget to submit your paper suggestions, book reviews or nerdy comments to www.podpage.com/let-me-sum-u... or www.podpage.com/let-me-sum-u...
The book is “Existential Politics: Why Global Climate Institutions Are Failing And How To Fix Them” by @greenprofgreen.bsky.social Jessica F Green. press.princeton.edu/books/hardco...
OUT NOW: LMSU is back for 2026, kicking off with Summer Book Club. What could be more summery, or summary, than arguing climate policy goes wrong by focusing on managing tonnes rather than transitioning assets?
Get ready to tonne police, or shake ya assets, here: www.letmesumup.net/should-we-do...
This would seem to be another facet of the economic choices/pathologies that @michaelpettis.bsky.social and others chronicle - sticking with growth via investment in production & infrastructure, even as returns diminish, rather than raising consumption (esp of services) has big energy implications
the puppies are just a cover for the true plan: giant space hamsters (miniature)
Top-of-the-dome three things you’d like to try?
