---

The March diesel spike was a headline. The one building into winter is a line item, and the thread under the post was less interested in the price than in whose contract absorbs it.

Each signal below traces the practitioner debate on the week's most-discussed posts and who the post reached. Audience figures come from LinkedIn's viewer demographics for each post; industry comparisons are against earlier posts on the same topic and are stated as upper bounds. Who reads a post is attention, not demand.

Before the war with Iran, wholesale diesel in Europe, the US Gulf Coast and Singapore sat near $90 a barrel. In March, shipping through the Strait of Hormuz seized up and Asian diesel shot toward $290, then fell almost as fast. By July, prices were back near $120. The climb now is slower, with all three regions moving together and some markets at record highs.

The IEA's September report explains why. Persian Gulf diesel exports fell to just over a quarter of pre-war levels in August, and Ukrainian strikes have nearly halted Russia's fuel exports. Together, about 1.6 million barrels a day of diesel left the market. Global oil inventories are down 507 million barrels since February, 95 million in August alone, and winter heating demand is arriving into depleted stocks. Almost nothing can quickly replace diesel: it moves freight, runs farm equipment, and powers mines, construction sites and backup generators.

The post asked where readers were seeing it hit their costs first. The practitioners who answered mostly pointed to contract terms. A procurement head at a global consumer-goods company: "A spike gets absorbed. A shortage gets passed on, one surcharge at a time, months later. Now is the time to find which contracts pass fuel straight through and which hold a fixed price. Those terms decide who carries the cost." His lag sits in fixed-price supply contracts. Freight reprices faster: a logistics founder noted that "carriers reprice fuel surcharges every week. The brands with the least exposure are the ones whose inventory already sits close to the customer."

The people furthest down the chain said the surcharge is not reaching them. A dispatch assistant at a trucking company said fuel now takes about three-quarters of what the company brings in, and that under contract it cannot simply ask for more per run. A driver trainer said owner-operators are losing money because brokers and shippers are not raising the rates they pay. Those are two voices, not a survey. If they hold, the first cost of this spike sits with small carriers and owner-operators on brokered or fixed rates, while shippers absorb it through weekly surcharges.

The sharpest challenge came from a metallurgical engineer, who argued that "comparing today’s market directly with the March spike—or saying that roughly 1.6 million barrels per day have permanently “left the market”—can be misleading. Oil and refined-product markets adjust." The post didn't say permanently, but his test is the right one: "The real question is not whether diesel prices are high. It’s whether the shortage is persistent enough to keep inventories falling despite increased production and imports." The 507 million barrel figure covers all oil, so it can't settle a diesel argument on its own. The numbers that can are distillate stocks. The week's heating-oil post put US East Coast distillate stocks 32% below their five-year average in September. If distillate stocks rebuild as refiners and exporters respond, he is right. If they keep drawing into heating season, the shortage reading holds.

A large part of the thread rejected the shortage framing outright, blaming oil-company profits or price manipulation. Both things can be true at once: in a product shortage, refining margins widen. For anyone passing costs on, that skepticism may matter as much as the supply data, because customers who doubt the shortage may push back on a surcharge.

Alongside Oil and Gas, the largest viewer industry at 19%, the industries that buy diesel showed up. Transportation, Logistics, Supply Chain and Storage and Truck Transportation each made up 5% of viewers, at least 2.2x their share on earlier oil-related posts. Readers also forwarded it: 14 sends, about 1.9 per 1,000 impressions against a 90-day average of 0.26, which is consistent with the post traveling to colleagues who manage the cost. (Composition: Transportation, Logistics, Supply Chain and Storage 5% vs up to 2.21%; Truck Transportation 5% vs up to 2.21%; Oil and Gas 19%; sends 1.9 per 1k impressions vs 0.26 avg; CXO/VP 15% vs 17.52%.)

> Do US East Coast distillate stocks close the gap to their five-year average before January, or keep drawing through heating season?

Read the original post →

In January, forecasters expected fossil CO2 emissions to rise in 2026. Then the conflict with Iran choked oil and gas flows through the Strait of Hormuz, and prices went vertical. A Washington Post analysis of IEA data now projects 2026 emissions below 2025, a swing of about 472 million tonnes of CO2 from January's forecast.

Carbon Brief's breakdown is the tell: oil demand down about 2.4%, gas down 0.6%, coal up 1.2%. When gas got too expensive, buyers who could switch to coal did. The post called it demand destruction: "Economic pain wearing a climate costume." It expected most of it to come back, with many forecasts showing demand at pre-war levels in 2027. Its argument was about the remainder: when a shock lasts long enough, people make purchases that outlive it, such as a hybrid or EV, a heat pump, or a utility signing for solar and wind instead of fuel it can't price.

The thread pulled that argument in both directions. A climate-finance executive director pushed back on the costume line ("A climate costume? No. This misses the mark.") and argued the remainder is bigger than the post allowed: "The short-term demand destruction that is visible masks a wholesale change in long-term energy demand decisions." His example: "major new LNG import infrastructure investments in the Philippines, Vietnam, and South Korea have now been cancelled. That's 16GW gas-fired power generation capacity demand destruction - permanent loss of 11 million tonnes per annum (MTPA) of expected global LNG consumption." In a second comment, he credited policy for the alternatives that make any of it stick: "none of these climate solutions would be commercially available, deployed at scale, at cheaper economies of scale to step in to meet future energy needs without sustained climate policies reaching back two decades."

From the other side, a senior executive at an energy company argued that "demand loss doesn’t push for using other energy sources but translates into lower output, which equates to GDP loss for everybody who manufactures something." A business developer for oil, gas and construction teams added that coal's gain was not only about price: "Some buyers switch to coal because it's familiar and infrastructure is already in place, not just because it's cheaper. Habit and existing contracts play a big role too."

Both critiques land on the post's own question. Cancelled import terminals are the kind of decision it described, one that outlasts the price spike, and whether they stay cancelled or come back as deferrals is what will decide the answer. A geoscience adviser supplied the historical check: the 2009 financial crisis and COVID produced similar drops with unclear lasting effects, "But the potential substitutions available today are better than they were back then, so perhaps this episode will create longer-lasting effects." A retired fellow framed it as a lesson for importers: "there are NO safe foreign sources, no source safe from disruption."

Oil and Gas was the largest viewer industry at 14%, followed by Renewable Energy Equipment Manufacturing at 9%, consistent with the two industries with the most riding on whether the demand comes back. London Area viewers made up 5% against 3.13% on earlier energy-transition posts. Readers saved it (6 saves) but did not forward it (0 sends). (Composition: Oil and Gas 14%; Renewable Energy Equipment Manufacturing 9%; London Area, United Kingdom 5% vs 3.13%; CXO/VP 16% vs 17.52%; 27 comments.)

> Do the cancelled LNG import projects in the Philippines, Vietnam and South Korea stay cancelled once prices normalize, or return as deferrals?

Read the original post →

Oracle sent a force majeure notice to a Blue Owl unit over possible delays in securing power for Project Jupiter, a 1,400-acre AI campus in New Mexico. Oracle says the project remains on schedule, and a Reuters source says Oracle cannot cancel the lease. About $18 billion in bank loans sit behind that campus, so one power delay becomes a payment question for a tenant, a landlord and a lender group. In late September, the 10-year Treasury yield also hit about 5.2%, the highest since 2007, which raises the return every project in the chain has to clear.

Chipmakers, builders, utilities, lenders, cloud tenants and index investors are priced against one shared schedule, and estimates put AI-linked companies at roughly 40% to 45% of the S&P 500, depending on definition. The post framed it as a deal-structure question: who holds the power risk, and who holds the rate risk? It named four things to watch: force majeure or delay notices, wider spreads on data center debt, lenders asking for more equity, and tenants renegotiating leases.

Behind the paywall: the three public filings a commenter says flagged Jupiter's delay months before Oracle's notice, what the post's unusual audience suggests, where solar practitioners say the money moves next, and Field Notes.

A founder who tracks data center siting answered the post's question with a specific trail. By his account: "Jupiter's on-site power runs on fuel cells fed by the Green Chile gas pipeline. New Mexico's Commissioner of Public Lands denied the pipeline a right-of-way across about 0.6 mile of state trust land in March, and again on July 14. The in-service date filed with FERC then moved from Aug. 15, 2026 to Feb. 1, 2027. The state air permit for the 2.45 GW fuel cell plant is still pending after a pause by the state Supreme Court." If that holds, the delay is about gas supply for on-site generation more than a grid connection. His conclusion: "A force majeure notice is the late signal. The right-of-way denials and permit dockets are the early ones, and they're public." It adds a fifth item to the post's watch list, and it is the one that comes first.

The audience was not the usual energy crowd. IT Services and IT Consulting made up 22% of viewers, Software Development 12% and Financial Services 8%, and a large share of viewers worked at the company in the news, which likely explains much of that skew. Readers kept the post: 14 saves on 38 reactions, about 0.37 per reaction against a 90-day average of 0.13, though the thread itself was three comments. The energy-side detail that explains the delay came from the comments. (Composition: IT Services and IT Consulting 22%; Software Development 12%; Financial Services 8%; Greater Seattle Area 5% vs 2.5%; saves 0.37 per reaction vs 0.13 avg; CXO/VP 21% vs 17.52%.)

> Does the Green Chile pipeline's FERC in-service date hold at Feb. 1, 2027, or move again before Jupiter's 2028 opening?

Read the original post →

US solar installs fell 14% in 2025, yet solar was 54% of all new US generating capacity, and solar, wind and storage together were 92%. America added about 43 gigawatts of solar, less than in 2024, with utility-scale down 16%. Wood Mackenzie's base case has installs roughly flat for a decade, with no year topping 43 gigawatts until 2033. The post's read was neither "soaring" nor "stalling": "Solar already won on cost. It's now hitting limits that have little to do with demand: grid connection queues, equipment tariffs and tax-credit timing."

The top state was Texas at 11 gigawatts, more than double California's 4.7. The chief operating officer of a national energy brokerage said that matches what his firm sees: "Developers go where the interconnection path is shortest and the market lets a project get paid without much ceremony, and ERCOT has been that place for years. My read is that the next constraint is less about building and more about who takes the output. Midday solar keeps pushing prices down, so the value shifts to storage and to buyers who can shape load around it." The exposure in that account sits with whoever holds the price on midday output: a merchant generator, or a buyer locked into a fixed-price, as-generated contract. A director at an energy company made the grid version of the point: "Solar is definitely getting cheaper and easier to build, but getting that power where it needs to go is becoming the bigger challenge. If the grid can't keep up, we won't get the full benefit of all that new generation"

The post reached the smallest audience of the four signals, 1,875 impressions, but a concentrated one. Utilities made up 9% of viewers, at least 1.6x their share on earlier solar posts, and Renewable Energy Power Generation 6%, also at least 1.6x, consistent with the people who run interconnection and buy the output. Three of the ten commenters were founders or startup operators. (Composition: Utilities 9% vs up to 5.54%; Renewable Energy Power Generation 6% vs up to 3.61%; Renewable Energy Equipment Manufacturing 13%; CXO/VP 18% vs 17.52%; commenters 3 of 10 founders.)

> Does the US beat 43 gigawatts of solar in a single year before 2033, and does ERCOT keep the lead once queue and congestion costs catch up?

Read the original post →

- San Diego shut off power on purpose in a heat wave: SDG&E put up to 35,000 customers in line for planned shutoffs across more than 30 communities, meant to stop wind-driven sparks from starting wildfires. The post traced the cause to weather: heat calls for more power on the lines, and Santa Ana winds call for fewer live lines. Its broader point was that cost and reliability are moving the wrong way together. Berkeley Lab found California had the biggest inflation-adjusted price rise of any state in the five years through 2024, which the post tied to wildfire costs, so customers fund fire prevention and then live with the shutoffs it requires.

---

The thread running through this week: when the physical system tightens, the cost lands on whoever holds a fixed price or a fixed schedule with no index, pass-through or exit. In diesel, commenters closest to the trucks say it stops at small carriers on brokered or fixed rates. At Oracle's campus, the power-delivery risk sits in a lease and roughly $18 billion of bank debt, and one commenter's account puts the early warning in public permit dockets. In solar, practitioners place it with whoever holds the price on midday output.

If your project or portfolio sits on one of these chains, the question for this quarter is which party your contracts leave holding the cost.

Reply if any of this is playing out at your company, or contradicting what you're seeing on the ground. Every reply goes directly to our analyst desk and feeds our intelligence.