Cyclone Sadeem (May 2026) Commercial Claims: Wind Versus Flood Apportionment for Bengal Industrial Losses

Cyclone Sadeem intensified into a Very Severe Cyclonic Storm and made landfall near Bakkhali on the night of 9-10 May 2026. Kolkata recorded gusts around 155 km/h and roughly 251 mm of rainfall, the heaviest near the city in about three decades, with four-metre waves reported off Digha and Puri before the system weakened inland through 10 May. For brokers running Bengal and Sundarbans-area corporate accounts, the headline damage figure is not the problem. The problem is that almost every meaningful loss arrived through two doors at once.

A typical affected site, a textile unit in Howrah, a warehouse on the Kolkata-Diamond Harbour stretch, a food-processing plant near Canning, took roof and cladding damage from wind in the first hours, then standing water from drainage failure and river overflow over the next two days. Under a Standard Fire and Special Perils (SFSP) policy with the STFI add-on, both storm and flood are insured perils. So the buyer assumes the whole loss is simply payable. It is not that clean.

Storm, Tempest, Flood and Inundation are grouped under one STFI label, but they are distinct perils with distinct treatment in most wordings. Storm and cyclone damage (wind, flying debris, pressure) often sits under the base material-damage cover or a wind sub-head. Flood and inundation frequently carry a separate, higher excess and sometimes a separate sub-limit, because insurers price standing-water exposure differently from wind. When a single event triggers both, the surveyor has to apportion the damage between causes, and that apportionment decides which deductible bites, whether a sub-limit caps the recovery, and whether a specific add-on responds at all. That is the whole ball game for a Bengal account this month.

Indian fire-policy claims turn on proximate cause: the dominant, effective cause in the chain of events, not necessarily the first or the last. Where only one peril operates, this is academic. After Sadeem, two perils operate on the same asset, and that is where money is won or lost.

The practical test surveyors apply is separability. Where wind damage and water damage hit different parts of the same property, or the same part in a clear sequence, they get treated as distinct losses with distinct causes. Roof sheets stripped by 155 km/h gusts are a storm loss. Finished-goods stock soaked by 30 cm of standing water on the shop floor two days later is a flood loss. Each then attracts its own excess and its own sub-limit, even though one cyclone caused both.

The harder cases are genuinely concurrent or indivisible damage. A switchgear room flooded after wind first tore open the wall presents a chain: wind let the water in. A broker can argue storm as the proximate cause of the whole sequence, which usually carries the lower flood-versus-storm excess and avoids a flood sub-limit. The insurer will often argue the water did the actual damage, pushing it to the flood treatment. Neither side is obviously right, which is exactly why the wording, the survey photographs and the timeline matter so much.

The takeaway for the desk: do not accept a blanket flood characterisation for the entire site loss. Force the analysis down to the component and the sequence, because that granularity is where the recoverable number actually lives.

Before arguing causation, read the schedule and the STFI clause as actually issued, not the marketing summary. Three places decide the apportionment economics.

The excess structure. Many de-tariffed property programmes now carry differentiated excesses: a base excess for fire and allied perils, and a higher AOG (act of God) or specifically flood excess, often expressed as a percentage of claim amount subject to a minimum and maximum in rupees. A loss characterised as storm may attract the lower excess; the same loss characterised as flood may attract a five- or ten-lakh-rupee-plus deductible. On a one-crore loss, the characterisation can swing the net recovery by several lakh.

Sub-limits and first-loss limits. Some wordings cap flood or inundation at a sub-limit well below the full sum insured, particularly for accounts in known flood-prone districts. Storm may sit at full value. If 60 percent of a Bengal warehouse loss is pushed into a flood sub-limit that is already exhausted by the stock damage, the balance is uninsured even though the policy looked adequate on paper.

Add-on triggers. Debris removal, architect and surveyor fees, and escalation add-ons usually respond to any insured peril, but some are drafted to respond only to specified perils. Earthquake and STFI exclusions inside an otherwise wide cover are common. Confirm the add-on actually names the operative peril.

One more mechanism cuts across all three. Average, the under-insurance clause, applies independently of the wind-versus-flood split. If the sum insured was set on book value rather than reinstatement value, the average clause will scale down whatever the apportioned figure turns out to be. The pro-rata reduction bites before any deductible, so a 40 percent under-insurance on a Bengal plant erodes the recovery on both the storm head and the flood head alike. Check sum-insured adequacy on the same pass, because there is little point winning the causation argument if average then halves the result.

The discipline here is to map the full loss against each of these three mechanisms for both a storm characterisation and a flood characterisation, then quantify the gap. That gap is the size of the argument worth having.

For manufacturing and processing accounts, the property loss is often the smaller number. The Business Interruption (BI) loss, lost gross profit while the plant is down, is where Sadeem will hurt most, and the wind-versus-flood split flows straight through into it.

BI cover responds only if there is an admitted material-damage claim from an insured peril at the insured premises. So the MD characterisation is load-bearing twice. If part of the physical damage is denied or pushed below a flood sub-limit, the BI flowing from that portion of damage is at risk too, because the material-damage proviso may not be satisfied for that head. A broker who settles the property side carelessly can quietly forfeit a much larger BI entitlement.

The indemnity period is the second pressure point. Flood damage to a Bengal plant frequently extends the outage well beyond wind damage: silt in machinery, water-logged electricals, mould in stock and contamination clean-up all add weeks. If the maximum indemnity period was set at three months on the assumption of a quick fire-and-rebuild scenario, a flood-driven outage can run past it, leaving the tail uninsured. Sadeem-hit accounts should be re-modelling the realistic restoration timeline now, not at renewal.

Three quantification points worth pressing:

• Increased cost of working. Renting alternative space, air-freighting replacement parts, running overtime. These are usually recoverable and materially shorten the loss if documented from day one.
• Trends and circumstances. Adjust the standard turnover for the post-cyclone market, demand spikes or supply disruption can cut both ways. Build the assumption explicitly rather than letting the loss adjuster impose one.
• Savings. Wages or power not incurred during shutdown are deducted. Keep the deductions honest and narrow, not assumed at a round percentage.

For the full method, brokers should align the BI workpapers with the property apportionment from the outset so the two files tell one consistent story.

The IRDAI-licensed surveyor controls the apportionment in practice, so broker behaviour in the first fortnight after appointment matters more than anything done later. After Sadeem, surveyor bandwidth across Bengal is stretched thin, which cuts both ways: reports are slower, but a well-prepared file gets disproportionate attention.

The goal is to make the storm-versus-flood story easy for the surveyor to accept on the favourable reading, and hard to recharacterise later.

1. Lock the timeline. Pull the IMD bulletin sequence, local gust records and any CCTV or plant-log timestamps showing when wind damage occurred versus when water entered. A documented sequence (wind first, water later) supports treating storm as the proximate cause of consequential water ingress.
2. Photograph by cause, not by area. Wind damage to the roof line, water marks at the recorded flood height, and the path of ingress should be captured as separate evidence sets. Mixed photos invite a blanket flood call.
3. Segregate damaged stock physically. Wind-and-debris-damaged goods kept apart from water-logged goods make component-level apportionment defensible and speed the salvage assessment.
4. Get a preliminary causation memo in writing. A short broker note framing the causation argument, filed with the surveyor early, becomes the reference the report responds to.

Salvage deserves the same care, because it is part of the apportionment rather than a separate afterthought. Water-logged electronics and stock often retain meaningful salvage value, and how that salvage is credited affects the net loss on the flood head specifically. Do not let salvage be netted off carelessly against the storm head, where the deductible is usually lower, when the damage that created it was a flood loss. Tie each salvage recovery back to the cause that produced it, so the credit lands on the correct head and does not quietly inflate the deductible drag on the wrong one.

Where the loss is large or the apportionment is genuinely contested, a second surveyor or a loss-assessor engaged by the insured can be justified. The cost is small against a seven-figure swing in recovery, and the presence of independent expertise changes the tenor of the negotiation.

Property is not the only file Sadeem opened. Bengal's logistics and port-linked accounts carry exposures that sit outside the SFSP policy and need separate, parallel handling.

Goods in transit during the cyclone, on trucks stranded on the Kolkata-Haldia corridor or in coastal feeder movements, fall under marine or transit cover, not the property policy. The peril wording differs: a marine cargo policy on Institute Cargo Clauses responds to its own list of perils, and storm or flood damage to cargo is treated under marine principles, including general average if a vessel was involved. Brokers should not let a transit loss drift onto the fire claim where it does not belong, and vice versa.

Stock-throughput and warehouse accounts present a boundary question: was the stock damaged as static property (fire/STFI) or as goods in the logistics chain (marine/transit)? The answer dictates which policy, which excess and which sub-limit applies, and for multi-location operators the accumulation across both classes matters for the eventual treaty recovery and for next renewal's pricing conversation.

A few practical flags for the desk:

• Open-cover declarations. Confirm in-transit consignments were declared and that the open cover was live on the loss date. Lapsed or undeclared transits are a common and avoidable gap.
• Port and yard exposure. Containers and break-bulk sitting in yards near Haldia or Kolkata may fall in a coverage seam between marine and property. Establish custody and the operative policy before the surveyor does it for you.
• Aggregation across classes. A single corporate group may have property, marine and BI claims from one event. Coordinate them, because insurers will aggregate for their own reinsurance recovery, and a coherent multi-class presentation strengthens the whole position.

Sadeem lands in the middle of a property market that was already repricing eastern-coast cyclone and STFI exposure. The claims experience from this event will feed straight into 2026-27 renewal terms for Bengal accounts, and brokers should get ahead of that rather than be surprised by it.

Expect three movements. First, STFI and AOG excesses on coastal and flood-prone Bengal risks will rise, and percentage-of-claim deductibles with higher rupee minimums will spread further down the account size range. Second, flood sub-limits will be applied or tightened on accounts in districts that took standing water, sometimes regardless of the individual account's own loss. Third, the rating base for storm and flood perils will firm, reflecting both the IIB-informed minimum-rate direction in fire treaties and reinsurers' own cyclone-load reassessment after a second severe Bay of Bengal system in recent years.

The broker's job is to protect good risks from being repriced on portfolio-wide assumptions. That means walking into renewal with risk-improvement evidence: bunding and plinth heights raised, critical electricals relocated above the recorded Sadeem flood line, drainage upgrades, and a documented business-continuity plan. An account that can show concrete flood-mitigation capex has a real argument for keeping its flood sub-limit and excess closer to the prior terms.

The accounts that come through this well will be the ones whose brokers treated the claim and the renewal as a single, eighteen-month exercise rather than two transactions.