Australia's merchant solar farms just earned 2.6 times more per megawatt than a month earlier. The cause was not more sunshine. It was a grid that finally learned to soak up the surplus.
Every engineer knows the strange arithmetic of a commodity: the more you make, the less each unit is worth. Solar has lived inside that arithmetic for a decade. Fill the midday hours with panels and the midday price falls through the floor, sometimes below zero. The farm that is working hardest earns the least.
This week the Australian grid did something that arithmetic says should be hard. Modo Energy reports that merchant utility solar on the National Electricity Market earned A$107.7k per megawatt-year in October, up 2.6 times from September and the highest since December 2024. The sun did not get stronger. Midday demand rose, rooftop output dipped, and a fast-growing fleet of batteries kept the surplus from drowning the price.
Our lens for this issue: the valuable resource on a modern grid is not the megawatt-hour, it is the hour you can still sell one in. Read what follows with one question in your pocket: in your own business, which of your abundant things is quietly being priced by the scarce thing next to it?
Australia's solar farms had their best month in almost two years. The surprise is what it says about how value is made on a grid.
Imagine a bakery that sells bread only between noon and one. Open a second oven and you do not double your sales; you flood the hour and the price of a loaf collapses. That is the merchant solar farm's problem on Australia's main grid, where rooftop panels and competing wind already crowd the middle of the day. Merchant revenues, Renew Economy notes, have been "increasingly pummelled" by negative daytime prices and curtailment. So when Modo Energy published its October numbers on 2 November, the headline looked like a misprint: A$107.7k per MW per year, 2.6 times September.
Merchant revenue means selling straight into the spot market with no long-term contract. It is the purest measure of what a solar megawatt is worth to the grid at the moment it is produced. Most Australian solar plants hold power purchase agreements, so this number is a thermometer rather than a paycheque, but it is the thermometer investors read.
Spring is the hard season. Mild weather means little cooling demand, rooftop panels hit their peak, and the middle of the day carries more supply than the grid can use. That is when prices go negative and large plants are told to switch off, a practice called curtailment. In South Australia this October, Modo counts 58.6% of available solar output curtailed.
Three things moved at once. Queensland midday demand rose by 1.3 GW as air conditioners started. Rooftop generation fell 9%. And batteries, which charge in the cheap hours and discharge at night, absorbed midday surplus that would once have pushed prices under water. Solar's average captured price jumped from A$17 to A$47 per MWh.
The next page lays out the numbers. The one to hold on to now is not the revenue. It is the direction of the cause: a smaller midday surplus made every remaining solar hour worth more.
Modo Energy puts October merchant utility solar revenue on the National Electricity Market at A$107.7k per MW per year, "up 2.6 times from September and the highest since December 2024." Every state improved. New South Wales led at A$151k per MW per year with a 50.5% capture rate.
Strip out negative dispatch and the figure rises only 3%, to A$110k. In a spring month usually defined by negative prices, they cost solar almost nothing. Modo cautions that most plants hold contracts that encourage negative bidding, so this is a merchant-only view.
Revenue for merchant projects there jumped 166%, the largest rise on the grid. Midday demand rose 1.3 GW while rooftop solar fell 9%, lifting Queensland's capture rate by 21.2 percentage points. Think of a see-saw: shave the surplus on one end and the price on the other end rises.
Rystad Energy's David Dixon says negative-price hours and curtailment are now declining on a trailing twelve-month basis as big batteries arrive. He expects it to continue as the NEM utility battery fleet grows from 26 GWh today to almost 60 GWh by end 2028, counting only projects financed or under construction.
Gas generation on the NEM totalled just 426 GWh, the lowest for any October since the market began, per Rystad. Gas still supplied 8% of New South Wales midday demand, up from 2% in September, so the sun has not yet won the middle of the day everywhere.
Battery revenues went the other way in October as charging costs rose. Modo's Frances Hallen: "In a grid where both are in surplus, solar and storage partially hedge each other." The estimated grid-scale battery fleet earned A$37.62 million on the NEM in October, per Energy-Storage.News.
South Australia's solar revenue rose only 23%, to A$28k per MW per year, about a fifth of New South Wales. Midday demand there fell 29% and it was the only state where capture rate declined. The glut is a local weather system, not a national one.
If scarcity of the midday hour is the product, three things follow. First, the asset that earns most is the one that shifts time. A battery that moves a megawatt-hour from noon to dusk converts an abundant hour into a scarce one. It is a time machine with a depreciation schedule.
Second, solar and storage stop being competitors for the same investor's dollar. Hallen's hedge point matters: when charging costs climb, battery returns dip as solar returns rise. A portfolio that holds both smooths its own cycle.
Third, the shape of demand becomes a design input. Queensland's revenue gain started with air conditioners, not with technology. Any large load that can move to midday, such as cooling, water pumping, EV charging or compute that can wait, is a buyer for the cheapest hour.
The practical move for a technology leader: ask your energy team which of your loads are time-flexible, and what that flexibility is worth per hour. Treat the answer as an asset you already own. This is our inference, not a finding from the sources.
1 · Soak. At noon, surplus solar would crash the price. Batteries charge instead, taking supply off the market.
2 · Lift. With less surplus, fewer hours go negative and fewer plants are curtailed. The average price solar captures rises.
3 · Shift. The stored energy is sold at dusk when demand peaks, displacing gas.
The consensus reading of this week is cheerful: batteries are fixing solar's cannibalisation problem. Our Move 37 is an uncomfortable one. The jump is evidence that solar is still a weather-dependent trade, and the real news is how small the cost of negative prices became.
Look at the base. September was weak, so a 2.6× rise starts from a low floor; October's A$107.7k is "the highest since December 2024", which means it only returned to earlier territory. And the cause Modo gives is partly meteorological: "higher midday cooling demand and lower wind and solar generation tightened supply." A grid that rewards solar when there is less of it is not a solved glut. It is a seesaw that tilts back when the weather turns.
One month is not a trend. Modo's own data attribute 12.6% of New South Wales solar revenue for the month to a single price event on 12 October. Strip out one afternoon and the story is softer.
The glut has not gone, it has moved. South Australia curtailed 58.6% of available solar output and its capture rate fell. Its revenue is A$28k per MW per year, about a fifth of New South Wales.
Batteries are a self-limiting cure. Battery revenue fell as charging costs rose. If the fleet grows from 26 to almost 60 GWh by 2028 as Rystad expects, spreads that pay for storage today get competed away. Nobody in these sources models that endpoint.
Contracts hide the truth. Modo notes most solar assets keep PPAs or certificate contracts. A merchant spike pays the minority that stayed exposed.
NASA's Neil Gehrels Swift Observatory re-entered the atmosphere at 7:39 a.m. EST on Wednesday 4 November over North Sumatra, almost 22 years after launch, per the US Space Force. On 20 October, drag had pulled it down to 275 km (171 miles) and into safe mode. NASA expected most of it to burn up, though "some components may have survived." No debris findings were reported.
Hannover Medical School, with Renji Hospital in Shanghai, reported early results in seven patients with alpha-1 antitrypsin deficiency, using base editing aimed at the SERPINA1 gene. One patient no longer needed weekly replacement infusions at the time of reporting. Very early data. These details come from a secondary summary; we could not open the paper's abstract, so treat them as provisional.
Fraunhofer's ZenBio platform combines cell-free protein synthesis with AI-guided testing. The goal is to evaluate up to 1,500 enzyme variants in parallel. The first phase of the automated workflow is reported complete; the full 1,500-variant system is still a target, not a result.
Three stories this week, one pattern. A grid's solar farms earned more when the middle of the day became less crowded. A telescope's last act was to come down where physics predicted. A biotech platform promised 1,500 experiments at once and delivered the first phase of the plumbing. In each, the exciting object is not the one that creates value. The bottleneck beside it is.
Bottlenecks are unglamorous. They are the battery next to the panel, the workflow around the model, the hour that is still sold out. They are also where pricing power lives, and they move. Today's scarce hour becomes tomorrow's glut once everyone builds toward it. The leader's job is to ask not where abundance is, but where the next constraint will stand, and to be there with something useful before the crowd.
1. Modo Energy (F. Hallen), "NEM merchant solar revenues rose 159% to $108k/MW/year, in October 2026," 2 Nov 2026 — https://modoenergy.com/research/en/australia-nem-solar-revenues-oct-2026
2. Renew Economy (S. Vorrath), "Solar bonanza: Merchant revenues soar to highest level in two years as gas generation hits all-time low," 3 Nov 2026 (quotes Rystad Energy's D. Dixon) — https://reneweconomy.com.au/solar-bonanza-merchant-revenues-soar-to-highest-level-in-two-years-as-gas-generation-hits-all-time-low
3. Energy-Storage.News, headline: Australia's grid-scale BESS fleet earned est. AU$37.62M in the NEM in October 2026, 2 Nov 2026 (headline only; article not opened) — https://www.energy-storage.news/
4. NASA, "NASA's Swift Observatory Re-Entered Atmosphere," updated 4 Nov 2026 — https://science.nasa.gov/blogs/swift/2026/11/02/nasas-swift-observatory-to-re-enter-atmosphere/
5. Next Horizon, "5 Science News Stories — November 3, 2026" (secondary) — https://www.nexthorizon.space/2026/11/5-science-news-stories-november-3-2026.html
6. Journal of Hepatology, base editing for alpha-1 antitrypsin deficiency (abstract not retrievable; not independently verified) — https://doi.org/10.1016/j.jhep.2026.09.028
7. Phys.org, "Automated platform optimizes enzyme production" (Fraunhofer ZenBio; via item 5, not opened) — https://phys.org/news/2026-11-automated-platform-optimizes-enzyme-production.html
Note: the feature's "so what" section and the Contrarian verdict are Inflection's own inference, not findings of the sources.
Which of your abundant things is being quietly priced by the scarce thing standing next to it?
Researched, written & designed with Claude. Typeset in Poppins & Lora on the Anthropic palette.