The 21-Mile Problem: Why Your CFO Is About to Revisit Every Fuel-Burning Asset on Your Site

If you run a facility in the Philippines that burns LPG, diesel or fuel oil for process heat, hot water or steam, the numbers in your operating cost model are about to be wrong. Not slightly wrong. Materially wrong.

This piece is a warning, not a sales pitch. It is written for the engineering managers, facility managers, plant managers and CFOs who carry the boiler decision on their shoulders — people who, through no fault of their own, are running 2022-vintage assumptions into a 2026 fuel market that no longer behaves the way those assumptions expect. We’d rather you saw it coming than discover it in a board pack three months from now.

If you doubt the scale of what’s happening, you don’t need to take our word for it. The Financial Times editorial board wrote yesterday (18 May 2026) under the headline “The world must brace for an oil supply crunch”:

For the past few weeks, the energy market has been in the eye of the storm. The loss of about 14.4mn barrels a day of crude oil output from the Gulf countries in April due to the Strait of Hormuz closure has been partly offset by draining stockpiles and other temporary reliefs … But a crunch point is approaching. The International Energy Agency warned last week that oil inventories are being depleted at a record pace. Further scarcities in poor countries and price jumps in rich ones loom within weeks. Governments, companies and consumers need to be ready.

The FT goes on to note that the IEA estimates the world has been consuming roughly 6mn b/d more crude than is being produced; that the 400mn-barrel coordinated strategic reserve release announced in March is approaching its limit; that JPMorgan now expects OECD commercial inventories to approach “operational stress levels” by early June; that Brent at around $109 is already 60%+ above pre-war levels even with US-Iran deal hopes priced in; and that the tightest markets are emerging not in crude but in refined products — diesel and jet fuel. Eighty countries have already put emergency measures in place. The FT’s conclusion: “more countries are going to have to learn to live within their more limited energy means.”

And the news is moving in the wrong direction. Today (19 May 2026), President Trump posted on Truth Social that he had suspended a planned US military operation against Iran that was scheduled for tomorrow, at the request of the leaders of Qatar, Saudi Arabia and the United Arab Emirates, who believe a deal is close. Trump said US forces have been instructed to be ready to go forward with a “full, large scale assault of Iran, on a moment’s notice” if an acceptable deal is not reached. Brent slipped 2.15% on the announcement to $109.69, but is still up roughly 50% since the late-February US–Israeli strikes. The ceasefire is fragile, the strait remains largely closed by Iran, and a deal is not the same thing as oil tankers moving again. For any Philippine facility planning its Q3 fuel budget on the assumption that this resolves cleanly, that assumption is now a coin toss.

At the same time, the regional escalation is creeping outward. On Sunday (17 May), a drone struck an electrical generator near the Barakah nuclear power plant in Abu Dhabi — the Arab world’s first nuclear plant. There was a fire but no radiation release. The UAE has not formally apportioned blame, but a senior Emirati official has suggested Iran or one of its regional proxies. This is the first time an Arab nuclear site has been struck, and it is being read across the Gulf as a symbolic warning to both the UAE and the US. The geography matters: Barakah sits roughly 50 km from the southern Hormuz approach. The signal is that the conflict is no longer just about hydrocarbons and shipping lanes; it is about whether civilian energy infrastructure in friendly Gulf states is now considered a target.

And the “Russian backup” is no backup either

One of the levers Manila has reached for is buying Russian crude under a US sanctions waiver. Treasury Secretary Scott Bessent announced this week a further 30-day extension of that waiver — despite promising last month that there would be no second extension — saying the move would “ensure oil reaches the most energy-vulnerable countries” as developing nations bear the brunt of rising oil prices. Read carefully, that is the US Treasury publicly admitting two things at once: that the global supply situation is bad enough to walk back a stated policy commitment, and that countries like the Philippines are in the bucket Washington is now openly worried about. The waiver is a lifeline, but it’s a 30-day-at-a-time lifeline, and the next extension is not guaranteed.

On paper, Russian crude diversifies away from the Gulf. In practice, Reuters reported yesterday (Moscow, 18 May 2026) that Ukraine has been systematically targeting Russian energy infrastructure as peace talks have stalled. The list inside the last six weeks alone is striking:

• NORSI (Lukoil) — Russia’s fourth-largest refinery and its second-largest gasoline producer, ~320,000 bpd — suspended operations on 5 April following a drone strike.
• Tuapse — an export-focused refinery on the Black Sea, ~240,000 bpd — halted on 16 April; a second drone strike caused a major fire on 28 April.
• Syzran (Rosneft, ~170,000 bpd) and Novokuibyshevsk (Rosneft) — both halted from 18 April after drone-damage to processing equipment.
• Perm — ~250,000 bpd, halted on 7 May after a drone-caused fire.
• Astrakhan gas processing plant — fire on 13 May (12 bcm/yr gas, 3 mtpa stable condensate, plus gasoline/diesel/LPG).
• Moscow refinery (Kapotnya, ~220,000 bpd) — struck on Sunday (17 May); 12 people wounded.

That is roughly one million barrels per day of Russian refining capacity offline or impaired in six weeks, much of it on the diesel and gasoline side — precisely the products the FT identifies as the tightest global markets. The world’s third-largest crude exporter is having its refining fleet taken apart drone-strike by drone-strike. “We’ll just buy Russian” is not the off-ramp it sounded like in March.

And neither Venezuela nor the US shale basin can fix the diesel deficit

If you’re asking “why don’t we just buy from somewhere else,” the answer sits at the refinery gate. Petron’s Limay refinery — the only one in the country, ~180,000 bpd, ~30% of national demand — was configured for Middle East medium-sour crude (~32° API, 1.5–2% sulfur). The product slate the country needs (diesel-heavy) is locked to that crude slate. Swap the crude and the slate changes:

• Venezuelan heavy crude (Merey / Boscán, ~16° API, ~3% sulfur) physically can’t be processed at Limay. It needs a coker or upgrader to crack the heaviest fractions, and Limay doesn’t have one. Retrofitting one is a 3–5 year, >$1 bn capex project. Even US refiners can only run Venezuelan heavy at a subset of Gulf and East Coast plants that have coking units.
• US WTI / shale (Eagle Ford, Bakken, Permian, ~40–45° API, <0.5% sulfur) is run-able, but produces a gasoline-heavy, naphtha-heavy, diesel-light slate. The Philippines burns far more diesel than gasoline (trucks, jeepneys, gensets, cold chain, plus power-gen on the islands). Putting US shale through Limay would leave you with surplus gasoline to export at a discount and a diesel shortage you’d still have to import.
• Russian ESPO is lighter and sweeter than what Limay was designed for, so the same slate distortion applies, milder. Petron has confirmed buying ~2.48 mn bbl of Russian crude already, after losing access to ~4 mn bbl of Middle East supply.

The operational implication for your site: even if Manila secures alternative crude tomorrow, the country cannot refine its way out of a diesel shortage. The 70% of refined product the Philippines imports finished comes from Asian refineries facing the same problem at larger scale. Switching crude source does not produce more diesel at the pump. The only thing that does, at the facility level, is using less of it. That is the lever sitting on your site.

And LPG is structurally worse than diesel

If your site runs on LPG — hot water, process heat, kitchens, autoclaves, boilers in the 30–120 °C range — the picture is sharper still. Of every fuel the Philippines depends on, LPG is the worst-positioned, for four reasons that compound:

1. Higher import concentration than crude. 91% of Philippine LPG is sourced from the Middle East — a tighter origin profile than the 98% Gulf-crude figure once you account for inter-Asian refinery balancing on the crude side. There is no domestic LPG production at scale. The product comes from Qatar, Saudi Arabia, the UAE and Iran — all upstream of the same chokepoint.
2. The LPG carrier fleet is specialised and small. LPG ships are pressurised or refrigerated vessels, not interchangeable with crude or clean-product tankers. The global VLGC fleet cannot be redirected from other trades on short notice, and newbuild lead times are years, not months. When the supply pinches, you cannot “just charter more ships.”
3. The national buffer is thin and shrinking. National LPG supply was reported at 33–36 days at peak; PNOC’s emergency 22,000 MT shipment in May is roughly five days of national consumption. The April 13 removal of LPG and kerosene excise taxes is a price cushion, not a supply solution. Industrial LPG buyers are reporting allocation cuts before pump price moves — supply is being rationed before it is being repriced.
4. There is no commodity-grade substitute at the burner. An LPG-fired industrial boiler cannot trivially switch to diesel. Burner heads, gas trains, regulators, pressure-vessel certification and combustion-tuning records are not interchangeable. A “fuel switch” on an LPG asset is a six-to-twelve-month engineering project at best, and the substitute (diesel) is itself supply-constrained and price-volatile. Sites that planned to ride out the crisis on LPG-to-diesel optionality are discovering the optionality is not there.

From a CFO’s point of view, this is the worst of all worlds: the most concentrated supply origin, the smallest dedicated carrier fleet, the thinnest national buffer, and the highest cost of substitution at the asset. It is also the category where the heat-pump retrofit math is most compelling. An LPG-fired hot-water duty at 75–90 °C is exactly the duty an industrial R290 or CO₂ heat pump replaces with a COP of 3.5–4.5 and zero combustion. The Capex is bounded, the lead time is months not years, and the operating cost lands in a band that is decoupled from any future LPG shipment that does or does not arrive. If you are running anything on LPG — hospital, hotel, food manufacturer, dairy, brewery, cold chain ancillary — this is the fuel to address first.

The Philippines, which imports 98% of its crude from the Middle East and depends on Asian refineries for most of its refined product, now finds that the supposed alternative supply is being systematically degraded too. Below is what that means for your site, why the fuel-cost picture is moving, and what we’d suggest doing in the next ninety days.

What changed: the fuel input prices

The Strait of Hormuz has been effectively closed to commercial shipping for over seventy consecutive days, the longest sustained closure in its modern history. The Philippines imports ninety-eight per cent of its crude from the Middle East. Of the refined product we don’t import as crude, ninety-seven per cent of liquid petroleum and ninety-one per cent of LPG comes from Asian refineries that themselves source from the Gulf.

The result, as of mid-May 2026:

• Diesel peaked at ₱130+ per litre in some regions in late March; currently ₱75–90 wholesale, with industrial delivered prices materially higher.
• LPG is the country’s tightest fuel, with national supply down to 33–36 days at peak. PNOC’s emergency 22,000 MT shipment for May was a band-aid, not a solution. Industrial LPG users are seeing sustained price increases and growing supply uncertainty.
• Fuel oil has tracked Brent crude’s move from $90 to $105–115 per barrel sustained, with $154 per barrel projected if the closure extends another twelve weeks.
• Electricity from Meralco bottomed-out in May 2026 at ₱14.33/kWh — but only because ERC accelerated a refund, suspended the GEA-All, and forced suppliers to absorb the line rental cap. These are one-shot levers. Generation charges themselves rose materially. The May 2026 reprieve is not repeatable.

If your facility is running a 2022-vintage spreadsheet that assumes diesel at ₱54, LPG at pre-war contract pricing, and Meralco at ₱11.50/kWh, throw it away.

Why your payback math is now wrong

Heat pumps replace combustion-based heat with electricity at a coefficient of performance of typically 3.0 to 4.5. That is, one kilowatt-hour of electrical input produces three to four-and-a-half kilowatt-hours of useful heat output.

The payback case has always come down to one ratio: the cost of a unit of useful heat from your boiler versus the cost of a unit of useful heat from a heat pump. Three variables drive that ratio: the fuel price, the boiler efficiency, and the electricity price.

In March 2022, with diesel at ₱54/L and Meralco at ₱11/kWh, a typical industrial heat pump retrofit had a simple payback of 24–30 months. Attractive but not urgent.

The cost of the boiler nobody talks about

There is a second-order cost that engineering managers track but rarely surface to the board: supply continuity risk.

If your plant cannot produce because the LPG delivery did not arrive, the cost of that lost production day is usually an order of magnitude larger than the fuel cost itself. Food manufacturers know this. Cold chain operators know this. Hospitals running steam-sterilisation autoclaves know this. Hotels with 200 occupied rooms and no hot water know this.

The Department of Energy reported diesel demand disruption of 20–40 per cent in March. Four hundred and twenty-five fuel stations closed nationwide. Industrial LPG buyers are reporting allocation cuts. None of this shows up in your Capex justification spreadsheet, but all of it shows up in your operating risk register.

A heat pump powered by Meralco-grid electricity has continuity risk too — brownouts in Luzon and Visayas have already begun. A heat pump powered by rooftop solar plus battery does not. That is the only architecture that genuinely removes Hormuz exposure from your facility.

The numbers we are actually seeing on real sites

We work with a range of customers across the Philippines. We are not at liberty to name them, but the patterns are consistent:

• An aviation maintenance and repair facility in Manila running heat pump hot water for hangar and washing operations: verified energy savings against the previous LPG-fired boiler system, with operational cost reductions that survive even under our Scenario 1 fuel-price assumptions.
• A food manufacturer in Bulacan with process water heating requirements at 75–85 °C: payback under the original LPG cost basis was 30 months. Under May 2026 cost basis, the same configuration is approaching 12 months. Paired with the rooftop solar installation we are now scoping, sub-12-month payback is realistic.
• Cold chain operators in Pangasinan and Bulacan looking at iCOOL and AquaHERO configurations to replace ammonia and HFC-based systems where service support has become unreliable, with R290 platforms offering both performance and supply chain advantages.

Across the board, the value proposition that we struggled to land in 2024 now lands in the first meeting.

The four mistakes to avoid

If you are a facility manager who is now under pressure from your CFO to “do something about fuel costs,” there are four common mistakes worth flagging.

What we recommend doing in the next ninety days

This applies to any facility burning LPG, diesel or fuel oil for process heat — not just heavy industry. Karnot’s heat pump range covers 6 kW to 350 kW per unit, with cascade and multi-unit configurations beyond that, so the technology fits everywhere from a boutique hotel hot-water duty to a mid-sized food processing plant. The question is no longer whether a heat pump fits your site — it is which configuration, and how fast you can move:

1. Pull the last twelve months of fuel invoices. Calculate the cost per kWh of useful heat output at the actual boiler efficiency on your meters (not the nameplate).
2. Build a sensitivity table for fuel costs at +25%, +50%, +100% over current. Run the same model for Meralco at +5%, +15%, +25% generation charge increases. These are not stress tests. They are realistic scenarios for the next eighteen months.
3. Request a heat pump feasibility study based on hourly load data, not annual totals. The OEM partners we work with can deliver this in two weeks.
4. Assess your rooftop area for solar. Industrial sites usually have far more capacity than the facility manager initially assumes. Most light-industrial roofs in Luzon can host 200–400 kWp.
5. Look at the Energy-as-a-Service option. For facilities that don’t want the Capex on their balance sheet, EaaS structures pass the equipment cost to a third party and share the savings. We can model this either way.

A note on the technology

Karnot Energy Solutions Inc. is a Philippine cleantech company headquartered at the Low Carbon Innovation Centre in Mapandan, Pangasinan, with a BOI-SIPP application in progress. We commercialise CO₂ and R290 (natural refrigerant) heat pump and cooling platforms under the iHEAT, iCOOL, iCRYO, AquaHERO and iSAVE product lines. Our OEM partners have deployed over ten thousand units across Europe, including in markets that have already lived through severe gas supply disruption (Germany, the Netherlands, the Nordics) and built the operational track record to prove the technology in continuous industrial duty.

CO₂ and R290 are natural refrigerants with global warming potentials of 1 and 3 respectively, compared to several thousand for legacy HFC refrigerants. This is becoming material as F-gas phase-down regulations tighten worldwide.

We are not a foreign vendor passing through. We are a Philippine company, with manufacturing partnerships in Europe and Asia, solving a Philippine engineering problem with technology that has been proven elsewhere. We are happy to discuss specific applications, share customer references where permitted, and provide feasibility studies for facilities that want a defensible engineering case rather than a sales pitch.