Every supply chain interruption gets a cause code. Supplier failure. Port congestion. Weather. Force majeure. In 2023 one of the most-watched interruptions on earth got filed under "drought," and that code was accurate and useless at the same time. The Panama Canal did not run short of ships, labor, or capital. It ran short of rain falling on one forested watershed, and that watershed is the part almost no one who depended on the canal had ever paid for.
That is the pattern this post is about. When a supply chain stops, the root cause is often a living system that is not in your vendor master, not on your risk register, and not in anyone's budget until it fails. You pay for it anyway. You just pay afterward, at the worst possible price.
what a supply chain interruption is
A supply chain interruption is any event that stops or slows the movement of inputs, goods, or services so that a business cannot produce, ship, or sell on schedule. Procurement classifies interruptions by cause: supplier failure, logistics disruption, geopolitical events, cyber incidents, and natural hazards. Insurers underwrite the financial consequence as business interruption and, when the loss starts at a supplier's site, as contingent business interruption. Both generally require physical damage to trigger. A canal that is open but slow is, under most policies, an uninsured delay.
The category "natural hazard" is where nature-driven interruptions land, and the label carries a hidden assumption. Weather is treated as an act of God: unowned, unpriced, uninsurable at the source. Sometimes that is true. Often the "hazard" is a living system that was doing its job for decades, was never funded, and stopped.
A supply chain interruption is often a stock or flow failing on a named landscape, filed as weather because nobody had a line item for the landscape.
the canal ran out of rain
Here is what the interruption looked like from the deck of a ship. In normal conditions the Panama Canal handles about 36 transits a day. In July 2023 the Panama Canal Authority (ACP) cut that to 32 and capped vessel draft at 44 feet, down from 50. By November it was 24 booking slots a day. Ships waited at anchor for days. Carriers added surcharges, lightened loads to meet the draft limit, and some rerouted entirely. The canal carries roughly 5% of global maritime trade, so the queue reached grain, LNG, containers, and finished goods on routes far from Panama.
Here is what it looked like from upstream. Every transit draws roughly 52 million gallons of fresh water from Gatún Lake to lift a ship over the isthmus. The ACP already recovers what it can: water-saving basins on the Neopanamax locks reclaim about 60% of each lockage, and cross-filling on the Panamax locks reuses water between chambers. The efficiency lever has largely been pulled. What remains is upstream. Gatún is not a well. It is rain, caught and held by the forested canal watershed around it, the Chagres River and its neighbors, and the same lake supplies drinking water for about half of Panama's population. In October 2023 rainfall over the watershed was the lowest recorded since 1950, 41% below normal. The lake fell to levels never recorded for that time of year. The locks had water to spend; the forest did not have rain to give them.
In our accounting that is a stock (tropical forest and the soils under it) and a flow (water abundance) failing at the same time on a landscape you could draw a line around. The interruption was hydrological long before it was logistical.
A fair objection: no forest could have made 2023 rain. True. What the forest decides is how much of the rain that does fall reaches Gatún instead of running off, how slowly it drains through the dry season, and how much reservoir capacity is lost to sediment from cleared slopes. Rainfall is weather. Infiltration, dry-season baseflow, and sediment are condition. Condition is what you can fund.
who was already paying the source
This is where the story is better than the usual "nobody pays for nature" complaint, and it is worth being exact.
The ACP does pay the source, in two layers. Since 2009 its Programa de Incentivos Económicos Ambientales (PIEA) has worked with roughly 1,950 farms in the watershed on reforestation, agroforestry, and better pasture. Since 2014 a second modality has paid landowners to keep mature forest standing: $130 per hectare per year, a rate the ACP describes as the highest in the region. In 2025 that forest-protection payment went to 594 families for about 7,600 hectares, a total of B/.788,318 (the balboa trades at par with the US dollar), with the first western-basin payments due in 2026. That is payment for ecosystem services, run by the one entity that could see the dependency: the operator of the locks.
Now look at who carried the cost of the interruption. Shippers, retailers, commodity traders, and their insurers. The people whose containers sat at anchor had never funded a hectare of the Chagres. They paid for the watershed in demurrage, surcharges, and lost sales, after it failed, and they had no instrument for paying it any other way.
The ACP's next move is instructive too. In 2025 its board approved a reservoir on the Río Indio, with a tunnel to Gatún, scheduled to deliver water around 2031 or 2032; the ACP and trade press put the cost at $1.6 billion, Mongabay at $1.5 billion. How many people it moves is itself contested: the ACP counts roughly 500 families, about 2,000 people; the communities' own 2024 census counts 2,543 people across more than 600 families.
The communities are not only displacees. They are counterparties with a position. Their coordinating body has put forward an alternative, expanding the existing Bayano reservoir with no relocation, which an independent review by Panama's Society of Engineers and Architects found viable; the ACP says Bayano has long been studied and that Río Indio offers more technical advantages. They have argued the consultation fell short of the Escazú Agreement, taken a challenge to the Supreme Court (dismissed on procedural grounds, resubmitted, awaiting admission), and asked the Inter-American Commission on Human Rights for precautionary measures. The ACP, for its part, has committed to titled replacement land, compensation, and a livelihood plan. A legitimate engineering answer and a legitimate objection to it now sit in the same valley.
Hold two numbers side by side. Forest protection across the eastern basin in 2025: $788,318. The new reservoir: $1.6 billion, roughly two thousand years of that payout, online in 2031. The reservoir is not the wrong answer; it is sized for 50 years of demand, including drinking water for more than half the country, and no forest stores what a lake stores. But the ratio shows which lever a dependent reaches for when the shortage arrives, and which one is still funded by a single payor at $130 a hectare. It is also a bill: large, late, and landing partly on people upstream who did not cause the shortage and have not agreed to pay it.
Meanwhile the dry seasons keep coming. On August 20, 2026, the authority announced it would cap daily transits again, at 34 from September 4 and 32 from September 15, as El Niño strengthened. The figures it disclosed are this post in one line: rainfall from May through August was 34% below the historical average, and inflows to the lakes were 44% below normal. The watershed gave up proportionally more than the sky withheld. Some of that gap is physics no forest changes; dry soils drink before they drain. Some of it is condition: how much of the basin is forest, pasture, or bare slope decides how that drink becomes dry-season baseflow instead of runoff and sediment. Rainfall is weather. The distance between 34 and 44 is where condition lives.
what procurement already does right
Respect the playbook before you extend it. When the canal slowed, good procurement teams did exactly what they are paid to do, and most of it worked.
| what procurement does | what it protects | what it does about the watershed |
|---|---|---|
| Dual sourcing and regional suppliers | Continuity when one lane closes | Nothing; the second lane may share the same weather |
| Safety stock and buffer inventory | Weeks of production | Nothing; it buys time, not rain |
| Rerouting and expedited freight | Delivery dates | Nothing; it prices the failure, at a premium |
| Force majeure clauses and business interruption cover | The balance sheet after a covered loss | Nothing; and because BI and CBI usually need physical damage to trigger, a drought-slowed canal mostly did not pay at all |
| TNFD and CSRD dependency disclosure | Board visibility, investor questions | Names the watershed; funds none of it |
None of these are mistakes. They are the right tools for the question "how do we keep shipping when a lane fails?" They do not answer "how do we make the lane fail less often?", because the vendor that decides that is a forest, and forests are not in the vendor master.
That gap is not procurement's fault. There has been no payable. You cannot cut a purchase order to a watershed.
the bill you already pay
Every row in that table costs money. Some of it is prepaid (safety stock, a disclosure workstream); most of it arrives after the living system has already failed; none of it reaches the landscape.
| pay-after bill | fund-the-source alternative |
|---|---|
| Canal surcharges, demurrage, reroute fuel | A share of the watershed's condition, funded before the dry season |
| Expedited air and truck freight to cover a late vessel | The same money pointed at the forest that decides how late the vessel is |
| Business interruption premium, plus the delay loss no policy pays because nothing was physically damaged | Fewer failures on the landscape that no policy is actually written against |
| Inventory write-downs and missed revenue | A named place kept in the condition that keeps the SKU moving |
| A $1.6 billion reservoir sized for 50 years of canal and drinking-water demand, online in 2031 | Not instead of the reservoir: the watershed above both lakes, funded now, so the new lake fills too |
you cannot offset the watershed
Two reflexes show up as soon as a company sees its nature dependency, and both deserve a plain answer.
The first is disclosure. TNFD's LEAP approach and CSRD's biodiversity standard (ESRS E4) will help you locate the dependency, name it, and report it. That is genuine progress; you cannot fund what you cannot find. But a disclosed watershed is not a wetter watershed. The report is the map. It is not the payment.
The second is the credit. Once the dependency has a name, someone will propose buying a water, carbon, or biodiversity credit to "offset" it. A credit retired in another basin does nothing for the basin your SKU stands on. Offsetting the watershed is not funding the watershed. If you want the ledger-versus-factory argument in full, it lives in hold the factory, not the ledger; this post only needs the conclusion.
If you want to see how far the dependency graph runs beyond the parcel next door, read the dependency your map can't see and your place depends on a place. If your interruption risk is terroir, the specific character a place gives a signature product, the supply chain risk hiding in your signature product is the luxury-goods version of this argument. The canal is the commodity version: the same logic with grain and LNG instead of jasmine.
the sku has a place behind it
Here is the turn. Behind every SKU that stopped moving there is a place: a watershed, a floodplain, a forage year, a canopy. That place exists whether or not anyone writes a check. The Chagres forest was catching rain before there was a canal to need it. Ensurance is how a dependent funds that place's present condition. It is not what the place is.
In practice that means three things a treasurer can hold. First, the object of payment is a named living system, not a unit of avoided something. Second, the payment funds present condition (the forest standing, the soil holding water this season), which is the thing your interruption risk actually tracks. Third, the instrument is a certificate: a funding position tied to one named place, held by whoever depends on it, with proceeds routed to that place's stewardship. That is the whole gloss. If you want the mechanics, specific ensurance is where certificates live.
A fourth thing, because the Río Indio makes it unavoidable: who decides. Funding a watershed's condition does not buy a say over the land upstream. Title, consent, and terms come first; the payment follows the terms the steward agreed to, it does not write them. A dependent's check is a payment for present condition, not an easement by another name and not a lever over someone else's land use. If what a payor actually wants is control, that is a different transaction and should be named as one.
A treasurer will also raise the free-rider objection, and it is correct: one shipper funding the Chagres is a rounding error shared with thousands of vessels, and the ACP is already the natural payor there. Panama is the illustration, not the invoice. Your check belongs in the basin where you are the concentrated dependent: the watershed above your plant, the floodplain around your distribution center, the growing or grazing landscape behind your top input. That is where one payor's funding moves condition, and where the co-payors, if any, are a short list you could name.
Two honest notes on our stage. The instruments are live and the volumes are small. There is no Panama Canal watershed certificate to point you to today; the canal is here because it is the clearest public record of a supply chain interruption that was really a watershed failing, and because one payor was already doing half the right thing. And a first engagement with us is not a demo. It is one conversation about one named place your operation depends on and one bill you already pay for it.
For insurers, the same logic reads two ways: fewer physical-damage triggers on the landscape a policy is written against, and a way to help insureds fund the non-damage delay risk that standard BI and CBI leave uncovered. For investors, it reads as the difference between a company that reroutes around its nature dependency every dry season and one that has started funding it.
what to do with this
Pick one SKU that has stopped moving in the last three years. Ask where the interruption actually began, not which port, but which landscape. Then ask who, if anyone, is paying for that landscape's condition today.
If the answer is "nobody who depends on it," you have found your nature line item. It is already in the budget. It is just filed under freight.