APAC Power Desk

The Atlantic Council flagged on Friday (2026-06-05) that research by its analyst Chhangani on China's Cross-border Interbank Payment System had been cited in Asia Times, one of two such pickups logged that day. The think tank's own pages were throwing technical errors when the citations landed, but the subject is what energy desks should track: the plumbing that lets China settle commodity trade outside the dollar. That matters because the scale is no longer marginal. CIPS, the yuan clearing network Beijing built as an alternative to dollar rails, processed about 920bn yuan ($134bn) a day in March, up from a daily average of 680bn last year, according to figures cited by The Economist. For a system long dismissed as a vanity project, that is a steep climb in a single year. The trade-finance numbers tell the same story. The yuan's share of trade finance has more than doubled, from 2% to 4.5%, since February 2022, the month Russia invaded Ukraine and the West moved to cut Moscow out of dollar clearing. Transaction volumes grew 75% in 2021 and CIPS was processing over $50bn a day by 2022. For energy, the question is who clears the cargo. If the settlement layer for commodity trade keeps shifting onto Chinese rails, the dollar's grip on pricing loosens at the margin, even while benchmarks stay quoted in dollars. Each cargo that settles in yuan is a test of whether the rails hold under volume. The broader internationalisation push is gaining ground. The yuan's share of China's overall international transactions, spanning goods, services and assets, rose above 56% in March after plateauing for much of 2025, The Economist reported. Cross-border portfolio flows reached $712bn in March, 40% above last year's monthly average. Still, the official line is more guarded than the headline numbers. One Chinese analyst said flatly that "the PBOC is aware of the limits of the e-CNY as a tool for RMB internationalisation," a rare admission that the digital-currency leg has a ceiling. Instant-settlement technology also makes transfer errors more likely. Context helps. Even at $134bn a day, CIPS remains small next to the Western rails it is measured against. Britain's CHAPS system processed £8.3 trillion across 4.4 million payments in April 2026, an average £416bn a day, Bank of England data show. CHAPS settles the equivalent of annual UK GDP every nine working days. The energy angle is less about the flows China carries now than the option it is building. The doubling of trade-finance share since 2022 shows sanctioned and sanction-wary counterparties will use the rails when the alternative is exclusion. The pickup of Chhangani's CIPS work across Asia Times and other outlets points to growing official attention on a system that until recently drew little. For commodity desks, the number to watch is the yuan's trade-finance share, which moved from 2% to 4.5% in four years. If it doubles again, the question stops being whether China can settle energy trade outside the dollar and becomes how much already does. The daily CIPS throughput, 920bn yuan and climbing, is the cleaner real-time gauge of how fast that shift is running.

Pakistan's headline inflation accelerated to 11.7% in May, its fastest in two years, after jet fuel prices jumped 94% and diesel 70% from a year earlier, the Pakistan Bureau of Statistics reported on Monday (2026-06-01). The reading was up from 10.9% in April. That matters because the increase is being driven almost entirely by imported energy, not domestic demand. Core inflation, which strips out food and fuel, rose 9% year on year and 8% month on month in urban areas, the bureau said, leaving the gap between the headline and core figures as a fairly direct measure of the energy shock. Motor gasoline rose 62%. The biggest annual increases were registered in sorghum, wheat, jet fuel, diesel and gasoline. The shock traces back to the Middle East. Crude and refined product prices climbed sharply through the first quarter of 2026, particularly after military action in the region on February 28 (2026-02-28) and the subsequent de facto closure of the Strait of Hormuz, the US Energy Information Administration said. For an importer like Pakistan, the product side of that move bites harder than the crude. Refined fuels have tightened faster than barrels. Across Asia, big domestic refiners were ordered to suspend exports of diesel and petrol, and many plants were cutting runs by 10% or more, the data firm Kpler reckoned, draining the regional pool that import-dependent buyers draw on. Supply has thinned from another direction too. Russia's average refinery runs fell to 4.69 million barrels a day in April (2026-04), the lowest in more than 16 years, OilX estimated, as Ukrainian drone strikes knocked out processing capacity. Moscow was weighing limits on diesel and jet fuel exports, Interfax reported on 2026-05-27, a step that would pull more middle distillate out of an already short market. Asia sits at the center of the exposure. The Gulf supplies between 40% and 80% of the seaborne crude bought by China, India, Japan and South Korea, and in 2025 Asia absorbed roughly 87% of the crude and 86% of the LNG that passed through Hormuz, according to the Economist. Japan draws about 95% of its oil from the Middle East, around 70% of it routed through the strait. The subsidy bill is the usual pressure valve, and it is straining. In Thailand, diesel subsidies were costing the government more than 1bn baht (£22m) a day, and petrol stations had seen panic buying with some posting out-of-stock signs. Pakistan has far less fiscal room to absorb the same hit, which is why the cost is showing up directly in the consumer price index rather than buried in the budget. The pull on fuel is global. US exports of crude and petroleum products hit a record 14.2 million barrels a day in the week of 2026-05-11, 33% above the same week of 2025, the EIA reported, while total US stocks including the strategic reserve fell about 24.1 million barrels, one of the five largest weekly declines on record. Barrels leaving the US tighten the same product market Pakistan imports from. What to watch now is duration. The Economist cited bank estimates that if Hormuz disruptions persist for five more weeks, euro-area inflation alone could rise by nearly a percentage point over the following year; for thinner-buffered importers the pass-through is faster and sharper. Pakistan's May print captured a market that was still tightening, not one that had peaked. If Russian export curbs land and Asian refiners stay defensive on runs, June's distillate prices set the next leg, and the country has little cushion left to keep it out of the inflation numbers.

Lightsource bp began construction on its Lower Wonga hybrid project near Gympie in Queensland on Friday (2026-06-05), combining an approximately 380MWdc solar farm with a 281 MW / 843 MWh battery, the company said. The plant is being built to help supply one of Australia's biggest single energy consumers. That matters because the offtaker is industry, not the grid. The project is intended to power a smelter, the kind of round-the-clock load that historically anchored coal and gas baseload contracts. Pairing utility-scale solar with hours of storage and pointing it at a single industrial buyer is a different proposition from selling spilled megawatts into a pool. The head of Lightsource bp in Australia called it a "new phase" for the global power sector. The phrasing is promotional, but the structure underneath it is real: co-locating generation and storage so the combination can firm a profile rather than dump intermittent output. The demand pulling these projects forward arrived in concrete form on Thursday (2026-06-04). The US-based AI specialist Iren said it would build an 800 MW data centre near the Bundey sub-station, about 75 kilometres north-east of Adelaide, with media reports putting the cost near $10 billion. Iren cited South Australia's target of reaching 100 per cent net renewable energy by 2027. South Australia is where the firming build-out is most visible. Neoen has committed to two batteries totalling 400 MW and 1,600 MWh at Goyder, alongside a 412 MW wind farm already built and a planned 400 MW second stage; the wider Goyder complex could host 1,200 MW of wind, 600 MW of solar and more than 1,600 MW of storage. Genaspi Energy won a Capacity Investment Scheme contract for a 300 MW solar farm and a 300 MW, 1,200 MWh battery near the same sub-station, which is also the entry point for Project EnergyConnect's 800 MW link to New South Wales. Capital is closing on these at scale. Edify Energy reached financial close in May on its Smoky Creek and Guthrie's Gap stations, billed as Australia's largest solar-battery hybrids, delivering 720MWp of solar and 2,400MWh of storage. The Lower Wonga groundbreaking is one project in a pipeline that is no longer experimental. The driver is load growth that power systems are scrambling to meet. Power demand from data centres could reach 9 to 17 per cent of US electricity supply by 2030, or up to 790 terawatt-hours, US battery firms told Reuters, even as grid connection queues and a China-dependent supply chain slow how fast they can build. The same squeeze is appearing in Australia, where industrial and AI loads are bidding for firmed renewable supply ahead of the transmission needed to serve them. Across Southeast Asia the numbers are larger and the gaps wider. Power demand from data centres, EVs and green industrial parks is forecast to grow by more than 100 TWh over the next three to four years, requiring upwards of $200 billion, according to a report flagged by ESG News. Singapore's conditional awards to import up to 3.4GW of firmed solar from Indonesia could lift regional installed solar capacity by more than 70 per cent, though those projects still face hurdles before they are bankable. The constraint is execution, not ambition. The same report points to an estimated $18 billion annual shortfall in grid investment by 2035, and subsea interconnector economics remain punishing, with European precedents showing development costs above US$60 million and cable booking deposits of 10 to 20 per cent of value paid years in advance. For commodity desks the read-through runs to coal. Each gigawatt of firmed renewables contracted to an Australian industrial load is a gigawatt that does not buy thermal coal or import LNG, and the packet's signal chain flags Australian renewables build as bearish for Newcastle coal and, downstream, JKM. JKM (Asian LNG) sat at $18.76 on Friday (2026-06-05), with the coal ETF proxy down 6.07 per cent on the day, though neither move can be pinned to a single project. The question Lower Wonga does not answer is whether firmed solar can hold a price low enough to keep a smelter running. South Australia's power spot was $109.19 on Friday (2026-06-05), a reminder that firming is expensive when the wind drops. Watch whether the next wave of industrial offtakers signs at fixed prices or demands grid backstops the system cannot yet guarantee.

CATL began operations at the world's largest energy storage testing and validation platform on May 28 (2026-05-28), the Chinese battery maker said, opening a 10-hectare campus in Xiamen backed by roughly RMB3 billion in investment. The Xiamen Energy Storage Validation Research Institute is pitched as open, shared infrastructure that any player in the global storage sector can use. That matters because the reliability problem it targets is already visible in the field. According to CATL, nearly one in five large-scale energy storage power stations worldwide are underperforming, and nearly half of all storage systems face grid-connection delays of more than two months. Those are not marketing numbers. They describe a fleet that is being built faster than it can be validated, and they come from the company with the most to gain from convincing buyers the problem is solvable. The centerpiece is a Thermal Safety and Combustion Laboratory, which CATL calls the world's first large indoor combustion facility, fitted with a 20MW calorimeter to measure heat released during thermal events. Safety is the gating issue. CATL frames sodium-ion and other high-density chemistries as the next storage battleground, possessing very high theoretical energy density but needing substantial work on safety before they scale. The timing tracks demand that is running ahead of the grid. The IEA projects AI and data centres alone could account for as much as 4% of global electricity use by 2030, accelerating the case for grid modernisation and new capacity. Storage sits directly in that path, smoothing intermittent supply and buffering load that traditional generation struggles to follow. China's advantage here is partly about cost. Chinese data centres can secure power for around three cents per kilowatt-hour, according to official figures, roughly half the rate many American operators pay. Cheap power lowers the bar for deploying storage at scale, and it gives Chinese manufacturers a domestic proving ground that foreign rivals cannot easily match. The investment numbers around the sector are large but uneven. Global renewable investment is projected to reach $2.2 trillion this year, more than double the sum going into fossil fuels, and over 40% of the IEA's $3.3 trillion estimate for total energy spending. Solar leads that flow, and storage is the necessary complement when grid bottlenecks loom. Southeast Asia shows where the strain concentrates. Power demand from data centres, EVs and green industrial parks across the region is forecast to grow by more than 100 TWh over the next three to four years, requiring investment above $200 billion, with more than half flowing into data centres as operators chase faster grid access. Storage and grid capacity, not generation, are the chokepoints. The cancellation rate underlines the gap between announcement and delivery. Only around 60% of the $540 billion in announced green investments across power and EV supply chains is considered likely to proceed under current conditions, the 2026 Southeast Asia Green Economy Report found. In Vietnam, Thailand and Indonesia, between 50% and 60% of renewable projects were cancelled over the past five years on regulatory uncertainty, permitting and limited grid capacity. That is the context CATL is selling into. A shared validation campus is useful only if the underperformance it documents can be fixed cheaply enough to keep deployment economics intact. The report behind the regional numbers warns electricity demand growth will outpace infrastructure, with annual grid investment shortfalls estimated at $18 billion by 2035. There is a competitive read too. CATL launched the centre alongside mass production of sodium-ion batteries, a chemistry it wants positioned as the next standard. Owning the test bench that certifies safety and performance is a way to set the terms on which rivals' cells get judged. The unresolved question is whether validation infrastructure moves the underperformance figure at all. Nearly one in five stations underperforming and half facing connection delays are problems of installation, grid integration and operation as much as cell chemistry. A 20MW calorimeter in Xiamen measures combustion risk well. It does not clear an interconnection queue. Watch two things. First, whether sodium-ion moves from mass-production announcements to bankable safety data that buyers accept. Second, whether the grid-connection delays CATL flags start to compress, because without that, even validated batteries sit idle. The cost of capital is cheap in China and dear elsewhere; that asymmetry, more than any lab, will decide where the next gigawatt-hours land.

The economics of building new wind farms in Australia are "getting worse, not better," the chief executive of one of the country's largest energy utilities said in remarks reported on Friday (2026-06-05), and the federal Capacity Investment Scheme is not changing the math. Rising construction costs, connection and transmission delays and a fickle off-take market are increasingly blocking projects from reaching a final investment decision. That matters because the Capacity Investment Scheme was meant to be the backstop that got projects built, the subsidy that underwrote revenue so developers would commit capital. If even supported wind cannot clear the investment-decision hurdle, the build-out of new generation slows, and the warning is coming from the head of a major utility rather than a marginal player. The same message came out of Tasmania. At the 8th Annual Tasmanian Energy Development conference in Devonport on Wednesday (2026-06-03), developers of island wind projects offered an almost identical account, with one presentation opening with a warning to the audience before the speaker reached any numbers. Acen put a figure on it. Pollington, speaking for the developer, said it had spent about $25 million on its Robbins Island and Jim's Plains projects, which have sat in the development pipeline for close to a decade. Robbins Island shows where the money goes. Acen has to build a long, winding transmission line to connect the project to Tasmania's grid, and that cost keeps eating into a business case the company has been trying to make work for nearly ten years. The complaints landed at back-to-back industry gatherings. At the Clean Energy Council's Australian Wind Industry Forum in Melbourne on Tuesday (2026-06-02), the supply side was, as one attendee put it, largely talking to itself, a room of developers comparing the same problems. Two warnings, from a major utility boss and from a mid-scale developer at the other end of the country, point the same way. The cost stack that decides a wind project is moving against developers faster than revenue support can offset it. None of this is new in kind. What has shifted is the framing. Developers are no longer treating transmission and connection delays as teething problems but as the factor that decides whether a project is built at all. The off-take market is the other squeeze. Wind developers need long-term contracts to finance construction, and they say those deals are getting harder to win even where a subsidy scheme sits behind them. A project can clear planning, secure support, and still stall because no buyer will sign for the output at a price that covers the build. The test now is whether any of these long-stalled projects reaches a final investment decision in the months ahead. Robbins Island, a decade and roughly $25 million into development, is the one to watch. If Acen cannot make the numbers work with the transmission line attached, the message from the conference circuit stops being a complaint and becomes a forecast about how much new Australian wind actually gets built.

ib vogt reached financial close on its 99 MWp Project Luca in Ajuy, Iloilo on Friday (2026-06-05), signing an Omnibus Loan and Security Agreement that secures a $75m senior debt facility from Rizal Commercial Banking Corporation, with RCBC Capital Corporation as lead arranger. That matters because the loan was underwritten by a domestic Philippine bank, not a development finance institution or an offshore lender, and it closed under the country's Green Energy Auction framework. Financial close on a GEA-tendered project is the moment a tariff award turns into steel and panels. Local lender appetite at this scale tells you the auction terms are clearing a bankability test that has stalled plenty of Southeast Asian renewables. Project Luca pairs 99 MWp of solar PV with a 4 MW / 16 MWh battery, ib vogt's first solar-plus-storage hybrid in the Visayas. Once running, the plant is expected to generate more than 160 GWh a year, enough to power over 85,000 households and cut roughly 70,000 tonnes of CO2 annually, the developer said. Look closely at the battery. Four megawatts of storage against 99 MWp of solar is a token, not a buffer, a unit sized for grid-code firming and ramp control rather than for shifting bulk solar into the evening peak. The hybrid label is doing heavier lifting than the kit, the norm for this first wave of Philippine solar-storage where batteries are bolted on to satisfy connection rules rather than to arbitrage the day. The timing was not unique. The same day (2026-06-05), Levanta Renewables reached financial close and issued notice to proceed on its 166-MWp Barotac solar-and-storage project, also in Iloilo, targeting commercial operations by mid-2027 under the GEA-4 round. Two closes in one province on one day point to a pipeline that is finally converting. The demand case behind all this is not subtle. Power consumption from data centres, EVs and green industrial parks across Southeast Asia is forecast to grow by more than 100 TWh over the next three to four years, requiring upwards of $200bn in investment, according to a recent sector report. Solar megawatts are the cheap, fast part of meeting that. Getting the electrons to load is not. The binding constraint is the grid. The same report flags an estimated $18bn annual shortfall in grid investment across the region by 2035. A 99 MWp plant in Ajuy is only as useful as the line that evacuates it, and connection queues, not capital, are increasingly what set the build-out pace. Cross-border ambition sharpens the contrast. Singapore's conditional awards to import up to 3.4 GW of firmed solar from Indonesia could lift the region's installed solar capacity by more than 70%, yet those projects remain stuck on subsea-cable economics, with development costs that can exceed $60m and booking deposits running 10-20% of cable value. Domestic project finance is closing now; the interconnector dream is still negotiating deposits. ib vogt said Project Luca strengthens a Philippine pipeline it puts at more than 1,000 MW. One close does not make a pipeline. But a domestic bank writing $75m of senior debt against a GEA tariff is the kind of precedent that makes the next ticket easier to underwrite. What to watch is whether GEA-4 keeps producing tariffs lenders will fund, and whether battery sizing on the next round of hybrids grows beyond compliance minimums. The grid connection timeline for Luca and Barotac will say more about Philippine renewables in 2027 than either financing announcement.

Batteries set the wholesale price in Australia's main grid nearly a third of the time in the first quarter of 2026, and average wholesale costs fell 12% as they did it, pv magazine Australia reported on Thursday (2026-06-04). That matters because the bullish case for cheaper Australian power now rests on one technology doing a job it was not built to do. Batteries are setting prices because they increasingly run as the marginal unit, charging when solar floods the grid at midday and discharging into the evening peak. The arbitrage works over hours. It does not work over days. Regulators have already banked the savings. The Australian Energy Regulator's final Default Market Offer confirms household power costs falling up to 10.7%, with bigger cuts for businesses, as cheap battery and renewable capacity coincides with soft coal and gas markets. Retailers are being told to pass it all through. But the same report that credits batteries for the drop carries its warning in the headline: long-duration storage will define the market. Look at what the largest buyers are doing. BloombergNEF expects solar to become the biggest single source of power within a decade, yet still sees coal and gas supplying 51% of the incremental generation feeding data centres by 2050, purely because those plants run around the clock. That is the firming gap four-hour batteries cannot close. Google's answer was to put $1 billion of 100-hour Form Energy batteries into a recent data centre project, a bet that short-duration storage is not enough. Then there is demand, which the price story treats as static. Australia's biggest grid hit a record for power demand in the fourth quarter of 2025, even as renewables supplied more than half of national generation for the first time. A 12% fall driven by a midday solar surplus assumes that surplus keeps arriving and that load does not chase it higher. Electrification and data centres argue the other way. So the market is watching the right number and drawing the wrong conclusion. It reads a 12% wholesale decline as the front edge of a permanent reset. The more cautious read is that batteries have flattened the easy part of the curve, the predictable daily swing between solar glut and evening peak, and left the hard part untouched. If that read is right, the error shows up in the hours batteries cannot cover. A run of low-wind, overcast days drains a four-hour fleet by early evening and hands pricing back to gas and coal peakers, the same plants the cheap-power narrative assumes are leaving. The cross-market tell sits in coal. A bearish Australian power signal is already feeding bullish pressure into Newcastle thermal coal and, through it, Asian LNG, which is not the path you would expect if firming were a solved problem. JKM spot sat near $18.76 on Friday (2026-06-05). None of this argues that the decline is fake. It argues that it is conditional, and that the condition is weather rather than technology. The grid that produced January's demand record did so while renewables ran above 50%, which tells you the system can already swing hard in both directions within a single season. What would confirm the contrarian view is straightforward to watch. If the next quarter of data shows the 12% wholesale fall holding only across sunny, windy weeks and thinning out during still, cloudy stretches, the drop is a daylight phenomenon, not a durable one. Watch the evening-peak settlements rather than the quarterly average, and watch how much long-duration capacity actually reaches the grid instead of the press release. Google's 100-hour purchase is the template. The open question is whether anyone in the National Electricity Market is matching it at the scale the falling-price story quietly assumes.

Akaysha Energy says the Waratah Super Battery, built on the site of the demolished Munmorah coal plant in New South Wales, is now running at 700 MW and 1,680 MWh, its full rated storage capacity, after engineers returned one of two damaged transformers to service. That matters because Waratah is the largest battery in Australia by capacity, and its job is to act as a "shock absorber" for the NSW grid, absorbing faults on the main transmission lines so that more power can flow on existing infrastructure. The jump to 700 MW, or 82 per cent of the project's 850 MW rating, lifts output from previous operating levels of around 400 MW. The capacity came back online after repairs to one of the three transformers damaged in what Akaysha called a "catastrophic" incident last October (2025-10), when the project was still in the final stages of commissioning. Full storage is restored. Full contract delivery is not. Of the 700 MW now available, Akaysha says 350 MW will continue to serve the System Integrity Protection Scheme, the contracted shock-absorber role, while the other 350 MW will be deployed into the National Electricity Market for arbitrage and frequency control services. That split tells you something about the economics. The merchant half earns whatever volatile spot and frequency markets pay; the contracted half is locked. The delays have already cost money. The Australian Energy Regulator says late delivery of the full SIPS contract, originally due in May last year (2025-05), has cut payments to Waratah and its "paired generators" by more than $90 million. For a project meant to underwrite reliability as coal retires, that is a pointed reminder that even flagship storage assets carry execution risk. The transformer failure was not a market event. It was hardware, on a single piece of equipment, on a site that until recently burned coal. Akaysha is not standing still. The company recently reached full output at its 415 MW, 1,660 MWh Orana battery in the central west of NSW, is building the Elaine battery in Victoria, and has just won a firming tender in South Australia for its Brinkworth project. The pipeline is real, and it is concentrated in the states doing the most to replace thermal generation. The wider context is a storage market straining to keep up with demand. BloombergNEF told its New York summit in April that utilities want far more energy storage than developers can deliver, with high battery pack prices, shipping bottlenecks and other supply-chain constraints damping near-term deployment. That squeeze is sharpest in the United States, where Reuters reported battery firms are seeing surging interest from power-hungry AI data centres but face long grid-connection queues and a supply chain heavily dependent on China. Waratah sidesteps the queue problem by being already built and connected. Its lesson is the other one. Getting steel and copper to site, and keeping it running, is where these projects stumble. Investors have noticed the demand even where the engineering is hard. Fluence Energy shares closed at $24.16 on 8 May 2026, up 98.2 per cent in a single week, after the company disclosed master supply agreements with two hyperscalers and a record $5.6 billion backlog. The stock is still down roughly 39 per cent year to date, and the company carries negative stockholders' equity of $265.88 million against cash of just $36.59 million. Demand is not the constraint. Balance sheets and hardware are. For traders watching the NEM, the signal from Waratah is incremental supply of fast frequency response and arbitrage depth in NSW, with another 350 MW of merchant capacity now chasing spot and frequency spreads. That tightens the bid in volatile intervals and, at the margin, caps the upside on price spikes the battery is built to clip. The open question is the remaining capacity. Waratah is at 700 of 850 MW, and the second damaged transformer is still out. Akaysha has not said when the final 150 MW and full SIPS delivery arrive. Until they do, the AER's $90 million-plus payment cut keeps accruing against a battery that, on paper, is finally storing everything it was built to hold.

Lightsource bp began construction of the Lower Wonga hybrid project near Gympie in Queensland on Friday (2026-06-05), pairing 380 megawatts of solar with a 281 MW and 843 megawatt-hour battery in what the developer calls one of the largest solar-battery hybrids in the country. The head of Lightsource bp in Australia framed the start as a "new phase" for the global power sector. That matters because the project is not chasing the wholesale market. Rio Tinto, which is spending $7.2 billion shifting its Gladstone aluminium operations from coal to green power, has contracted to take 40 per cent of Lower Wonga's output. The interesting shift here is the buyer: a heavy industrial load locking in firmed renewable supply directly, rather than a utility balancing a portfolio. Gladstone is among Australia's largest single energy consumers, and Rio Tinto is spreading its bets across several developments at once. Alongside Lower Wonga, it has signed offtake deals with the neighbouring Smoky Creek and Guthrie's Gap solar-battery hybrids, plus two projects yet to break ground: the 1.4 GW Bungaban wind project, which includes battery storage, and a separate 1.1 GW development. The portfolio is being assembled to replace coal-fired baseload at scale. Smoky Creek and Guthrie's Gap are the anchor of that supply. Edify Energy reached financial close on the two Queensland projects in late May (2026-05-20 and 2026-05-21), a combined 600 MW (720 MWdc) of solar and 2,400 MWh of battery storage, which Edify describes as the largest solar and battery hybrids currently under construction in the country. Lightsource bp is building elsewhere too. Its Goulburn River project near Merriwa in New South Wales pairs a 49 MW, 562 MWh long-duration DC-coupled battery with a 585 MWdc solar farm, one of the firm's first hybrids of this design. The DC-coupling matters for cost and for how much solar energy can be captured rather than clipped. The logic behind all of this is straightforward. Solar provides the lowest-cost scalable electricity, and battery storage shifts that energy into periods of higher demand, which strengthens flexibility and reliability across the grid. Bolted onto a smelter that runs around the clock, the storage is what makes the offtake credible. But credibility on paper is not the same as delivery. BloombergNEF warned in May (2026-05-19) that even as developers rush to bring battery projects online and demand stays strong, high battery pack prices, global shipping bottlenecks and other supply-chain constraints are dampening near-term deployments. That is the gap between announced megawatt-hours and energised ones. The risk is concrete, not abstract. BNEF pointed to Vistra restructuring its 350 MW/1,400 MWh Moss Landing Phase III project with PG&E because of battery supply uncertainty, a reminder that financial close and even construction starts do not guarantee on-schedule storage. For a buyer like Rio Tinto stitching together five projects to retire coal, slippage on any one shifts the whole transition timetable. There is a coal read-through, though it should be held loosely. Every gigawatt-hour that firmed solar feeds into Gladstone is a gigawatt-hour that coal-fired generation does not, which over time trims thermal coal demand into one of Australia's industrial heartlands. The signal direction points bearish for Newcastle coal, but the volumes are years out and the consensus in the data here is thin. What to watch is conversion. Two of Rio Tinto's contracted projects, the Bungaban wind development and the 1.1 GW project, have offtake but no construction start; whether those move from signed deals to ground-breaking will show how much of this pipeline is real versus optioned. The other thing to track is the batteries themselves. Lower Wonga, Smoky Creek and Guthrie's Gap together represent more than 3,200 MWh of storage entering construction in a single Queensland cluster, into the same supply chain BNEF flagged as stretched. Energised capacity, not announced capacity, is the number that will tell you whether the Gladstone transition is on schedule.

China's carbon emissions rose about 2% in the first quarter of 2026, Carbon Brief reported on Thursday (2026-06-04), even as the country kept adding wind and solar at a furious pace. The reason was awkward. Generation from coal and gas climbed 4% over the same period, after falling through 2025. That matters because it undercuts the assumption beneath most bearish coal and oil forecasts for China, namely that new renewables capacity automatically displaces fossil generation. In early 2026 it did not. The clearest evidence sits in the wind fleet. China's installed wind capacity grew 23% year-on-year, an addition of 120GW, Carbon Brief said. Yet the average capacity factor fell from 27% to 22%, an 18% drop, and wind output showed almost no growth. Run the counterfactual and the gap is stark. Had capacity factors held, the new wind, solar and nuclear should have delivered an extra 160 terawatt hours in the quarter versus a year earlier, more than the 120TWh rise in power demand. Instead clean generation rose just 60TWh. Coal and gas covered the difference. The squeeze tightened in March. Power demand grew only 3.5% and hydropower output rose 9%, conditions that should have eased the call on thermal plants, yet fossil generation still increased 4.2%. Weather explains part of it: weak wind speeds and nuclear refuelling outages. The import side mattered too. The de facto closure of the Strait of Hormuz cut China's crude oil and natural gas imports by around 20% in April (2026-04), the Centre for Research on Energy and Clean Air said, pushing coal power up for a fourth straight month. The crude shortfall hit refiners directly. Plunging imports forced Chinese processors to cut runs sharply in April (2026-04), with the state-owned sector dropping to multiyear lows, Bloomberg reported, after the near-halt of Hormuz shipments choked a vital crude channel. Oil-product demand tells a mixed story. Apparent consumption rebounded in January and February on transport, then slipped slightly, Carbon Brief said. The macro backdrop is violent. The EIA described global oil markets as a period of heightened volatility tied to Hormuz, which carried nearly 20% of world oil supply before military action began in late February (2026-02). Crude oil implied volatility has averaged 78% since the conflict started, with daily Brent crude implied volatility peaking at 106% on 12 March (2026-03-12). One popular claim drew a pushback. Carbon Brief said the data do not support the idea that the Hormuz closure has driven a marked jump in China's coal-chemicals output. The coal story here is about power generation, not feedstock substitution. Underneath the power numbers, heavy industry is shrinking. Cement production fell 7% and crude steel output 5% in the quarter, as real estate investment contracted another 11% following a 17% reduction in 2025. Weaker industrial demand should be bearish for coal. It was not enough to offset the renewables shortfall. April brought no relief. Total power generation rose an estimated 6.6% year-on-year, but weak wind, subdued solar and extended nuclear outages pushed coal to a fourth consecutive monthly gain, with thermal commissioning in the first quarter surging, the Centre for Research on Energy and Clean Air said. The signal to watch is whether wind capacity factors recover as 2026 runs on, or whether China has built a fleet the grid cannot lean on. If the latter, every bearish coal thesis tied to nameplate renewables growth needs revisiting, and the Hormuz import gap keeps the marginal tonne of coal firmly in the stack.