Weather-driven demurrage: anticipate in 10 days, plan in 45

Weather related demurrage can be anticipated 10 days ahead, and planned more efficiently with 45.

The vessel is still days away from arrival, but the factor that will decide whether it berths directly or waits in line already exists: the weather condition that closes the port. The problem is that most operations only see this once the vessel is already in the line up, the berth is locked, and the demurrage invoice has already started running. By then, the decision that mattered (rerouting cargo, notifying the client, adjusting the schedule) has already passed.

It does not have to be this way. The weather condition that closes a port has an identifiable pattern before it happens, and that pattern serves two different horizons. With 10 days of lead time, you can anticipate the real probability of the port closing, in time to act. With 45 days, you can see the trend and plan capacity and resources more efficiently. Neither is a crystal ball, but both are worth more than finding out about the problem at the dock.

In this article

• The real pain: demurrage, waiting time, and the lack of transparency about when the port resumes operations
• What causes the pain: the weather hazard that delays berthing and unberthing, and why it leaves a trace
• How to avoid it before the impact: visibility, anticipation rules, and communication across the chain
• Proof from those who already anticipate
• Why ask the AI Climate Agent
• Frequently asked questions

Demurrage is the effect. Weather delayed berthing is the cause.

Demurrage does not arise only from bureaucracy. Much of it comes from a berthing or unberthing decision delayed by a weather condition (swell, gale, heavy rain, cyclone) that nobody saw coming far enough in advance. The hazard is the weather condition. The risk it brings is the vessel standing idle, the berth occupied longer than planned, and uncertainty about when the operation will return to normal.

This cost does not stay with the carrier alone. It spreads across the links that depend on that vessel: the freight forwarder who needs to reorganize pickup, the importer or exporter watching the delivery window close, the terminal losing berth productivity, the trucking company rescheduling the fleet that was going to collect the cargo. The more links in the chain, the more people making decisions without knowing when the port will resume operations.

The scale of the problem is well documented. Globally, about US$ 18 billion a year is lost to waiting and poor planning at ports, and a vessel spends an average of 8% of its voyage time idle at anchorage. In Brazil, accumulated demurrage at ports reached US$ 2.3 billion in 2024, 15% above 2023. A single anchored Panamax costs between US$ 15,000 and US$ 25,000 per day.

The weather that closes a port leaves a fingerprint, and it appears in two different horizons

Public weather forecasts cover an area of about 25 km and speak to the region, not the berth. They have not learned from that specific port's history, do not account for depth and coastline, and so cannot distinguish an event that closes the bar from one that merely makes maneuvering difficult.

By cross referencing hyperlocal historical data (meteorological and oceanographic) with that port's real maneuvering data, it becomes possible to recognize the pattern each weather condition leaves on port operations: the fingerprint of how that specific port reacts to swell, wind, or rain. That pattern serves two different horizons, and the two should not be confused.

With 45 days of lead time, you can identify the trend: whether demurrage risk at that port is increasing or decreasing over the coming weeks. This is not a point forecast. It is a directional reading, and its value lies in planning capacity, scheduling, and resources more efficiently, not in promising the port will close on a specific day.

With 10 days of lead time, you can identify the concrete probability of impact happening within that specific window. This is the horizon that allows you to actually anticipate demurrage: deciding on berthing, rerouting cargo, or notifying the client in time to act, not just to plan.

The difference between planning the season with the 45 day trend and deciding the next maneuver with the 10 day probability is the difference between two complementary tools. Both prevent surprises.

Avoid before the impact, don't react after it

Anticipating demurrage has three fronts, and all three depend on seeing the weather before it happens, not on reacting well afterward:

Operational planning with visibility. Knowing the trend for the next 45 days, whether demurrage risk at that port is increasing or decreasing, shifts the question from "what to do if the port closes" to "what capacity and resources to prepare for the coming weeks."

Anticipation with well defined rules and actions. With the 10 day probability and the identification of each berth's closure patterns, you can set clear triggers, which risk level activates which action and who decides, turning the forecast into an automatic decision before demurrage starts accruing.

Assertive communication across the chain's links. The freight forwarder, the importer, the trucking company, and the terminal react better when they receive the same information, at the same time, with the same time window. Whoever communicates late, replans late.

Proof from those who already anticipate

This is not theory. At the Port of Coquimbo, in Chile, anticipating 25 swell events (218 hours) helped avoid port omissions. In Mejillones, the benefit reached US$ 305 thousand per year. And Santos Brasil reduced vessel waiting time from seven to three days by integrating weather intelligence into its operation.

Why ask the AI Climate Agent, and not a generic chatbot

A generic chatbot answers with what it knows from the public model, something like "conditions may worsen between Thursday and Friday." i4sea's AI Climate Agent answers with what it has learned about that specific port: which berth, which window, which probability, and what to do about it. What changes is not the artificial intelligence being used, but the data behind it: proprietary high resolution historical reanalysis for the operation, cross referenced with that port's real maneuvering history.

This also solves the lack of transparency about when the port resumes operations. Instead of waiting for a generic bulletin, anyone in the operation can ask the Agent directly, on WhatsApp or the portal, and receive the expected window, the probability, and the recommendation, with that port's history available for audit.

Try the AI Climate Agent

Will the vessel you're tracking berth or wait? Ask the Agent with the port, the cargo, and the window: it responds with the probability of impact within the 10 day window, the trend for the coming weeks, and the recommendation, with the port's history available for audit.

Frequently asked questions

How far in advance does the Agent identify a weather impact at the port?

Two windows, with different uses. With 45 days, the trend of more or less demurrage risk at that port, useful for planning capacity and resources. With 10 days, the concrete probability of impact happening within that specific window, which is the horizon that allows you to anticipate demurrage and act in time.

Is this the same thing as a weather forecast?

No. Weather forecasts cover a broad area and do not know that port's history. The Agent cross references hyperlocal data with the port's real maneuvering history to recognize the specific pattern of that operation.

Does the Agent replace ChatGPT or another general purpose AI I already use?

For the question "when will weather impact my port, with what probability, and what should I do about it," yes. A general purpose model does not have that port's calibrated history or the operational triggers defined by your team.

Does this help contest demurrage with the carrier?

The decision trail (what was forecast, the probability, the recommendation sent, and when) stays on record, and that trail supports a demurrage dispute or a force majeure claim.

Does this work only for ports that already use i4sea, or for any port?

It works for any port in Brazil, Chile, Peru, and Mexico, not only for existing i4sea clients.

Sources

1. MarineTraffic, global analysis of 200,000 vessels published via Wärtsilä Insights (28/07/2020): ships spend 4% to 9% of voyage time anchored waiting for a slot (basis for the figure used in the text). https://www.wartsila.com/insights/article/routine-losses-ships-spend-up-to-9-percent-time-at-anchorage · Waiting and poor planning at ports cost about US$ 18 billion per year: ShippingWatch. https://shippingwatch.com/Ports/article11550328.ece

2. Valor Econômico, 03/04/2025: port delays generate an extra cost of US$ 2.3 billion for Brazil in 2024, 15% above 2023 (Bain & Company study). https://valor.globo.com/empresas/noticia/2025/04/03/atraso-em-portos-gera-custo-extra-de-us-23-bi-ao-brasil-em-2024.ghtml

3. i4sea, Maritime and Navigation page: an anchored Panamax costs between US$ 15,000 and US$ 25,000 per day, combining berth waiting time and the cost of idle cargo.

4. i4sea, benefits study at the Port of Coquimbo (Chile), 2026: anticipating 25 swell events (218 hours) helped avoid port omissions.

5. i4sea, commercial master deck: a benefit of US$ 305 thousand per year at the Port of Mejillones (Chile).

6. i4sea, commercial master deck: Santos Brasil reduced vessel waiting time from 7 to 3 days with weather intelligence in its operation. Public coverage of the case: G1, 18/02/2026. https://g1.globo.com/sp/santos-regiao/porto-mar/noticia/2026/02/18/inteligencia-climatica-evita-tragedias-e-reduz-filas-no-porto-de-santos-entenda.ghtml

7. i4sea, AI Climate Agent: lead times of 45 days (trend for planning) and 10 days (probability for anticipation), the fingerprint of weather impacts at each port, and coverage of Brazil, Chile, Peru, and Mexico (statements by CEO Mateus Lima, 04 and 06/09/2026).