LFP Cathode Active Material Plant Risk Profile in India 2026: Calcination Fire, Iron Phosphate Dust and Process-Safety Underwriting

Most of the Indian battery insurance conversation has lived at two ends of the chain: cell and pack assembly under the PLI scheme, and the fire-prone tail of recycling and second-life storage. The middle, where the active material itself is made, has stayed off the underwriting map because there was no domestic production to rate. That changes in 2026.

Himadri Speciality Chemical announced India's first commercial plant for Lithium Iron Phosphate (LFP) cathode active material (CAM) in Odisha. Phase 1 targets 40,000 MTPA at an estimated Rs 1,125 crore, with the company aiming to be operational in roughly 27 to 36 months and to scale to 200,000 MTPA over five to six years, enough to feed close to 100 GWh of cell capacity. Press reporting around the wider project envelope has cited figures up to about Rs 4,800 crore as the build-out progresses.

For a broker, the headline is not the capex. It is that CAM production is a high-temperature inorganic chemical process, not an assembly line. It carries calcination kilns running at 700 to 900 degrees Celsius, large volumes of fine metal-oxide and iron-phosphate powder, inert-atmosphere furnaces purged with nitrogen, and milling and handling steps that generate combustible dust. None of this has an established Indian loss history, which means property and liability underwriters will be rating it partly from first principles and partly by analogy to specialty chemicals and pharma process plants.

That gap is the opportunity. The broker who walks into the placement with a structured view of the hazard, a defensible PML basis, and a wordings checklist will set the terms of the conversation. The broker who treats it as a generic factory fire policy will leave the client under-covered and the account mispriced. This profile lays out the genuine exposures and where the standard market wordings will fall short.

Strip away the battery branding and a CAM plant is a precursor-to-powder chemical operation. The dominant industrial route for LFP is high-temperature solid-state synthesis: blend lithium, iron and phosphate precursors, then calcine and sinter the mixture in a controlled-atmosphere kiln. Understanding the process gives you the loss scenarios.

The risk concentrates in a few zones:

• Calcination and sintering kilns. Roller-hearth, pusher or rotary kilns hold material at 700 to 900 degrees Celsius, frequently under an inert nitrogen blanket or a mild reducing atmosphere. These are continuous, hard to shut down cleanly, and represent both the highest fixed asset value and the longest replacement lead time on site.
• Powder handling and milling. Precursor blending, jet or ball milling, and pneumatic conveying generate fine particulate. Iron-phosphate and carbon-coated powders in suspension can present a combustible-dust hazard, with the classic deflagration and secondary-explosion chain familiar from pharma and food process plants.
• Drying and solvent steps. Spray drying and any wet-process or coating chemistry bring solvent and thermal-fluid loads. Solvents such as NMP are more central to downstream electrode coating than to LFP CAM itself, but utilities, heat-transfer oil and binder chemistry still add ignitable inventory.
• Inert-gas and gas-handling systems. Nitrogen purging creates an asphyxiation and confined-space hazard for personnel, and any hydrogen-bearing reducing atmosphere adds a flammable-gas exposure.

The loss menu therefore runs wider than a single fire peril. You are underwriting kiln fire and thermal runaway in process equipment, dust deflagration in milling and conveying, machinery breakdown on high-value rotating and heating equipment, and contingent third-party exposure if off-spec CAM degrades the cells of a downstream buyer. Each maps to a different section of the programme, and each needs to be named, not assumed.

The Indian Standard Fire and Special Perils Policy (SFSP) responds to fire, explosion and a defined list of perils, but it was not built to think in dust-deflagration terms. A CAM plant forces that thinking, and it is where a lazy placement quietly fails the client.

Combustible metal-oxide and iron-phosphate dust behaves like other process dusts: a primary ignition in a mill, dryer or conveyor can disturb settled layers and trigger a far larger secondary explosion across the building. The severity driver is housekeeping and ductwork design, not the headline kiln. Underwriters who have seen pharma and food-processing dust losses will recognise the pattern, and brokers should borrow that vocabulary directly.

The practical underwriting test is whether the site has carried out a Dust Hazard Analysis, classified its dust by explosibility (the Kst and Pmax values that drive vent sizing), and engineered protection accordingly: explosion venting, isolation valves, suppression on enclosed equipment, and disciplined housekeeping to stop layer accumulation. A plant that can show this should earn materially better terms than one that cannot.

For the broker, three placement actions follow. First, confirm the explosion peril is clearly within the material damage cover and not narrowed by an exclusion or a restrictive warranty. Second, position dust protection as a rating credit, getting the client's capex on venting and suppression reflected in the premium rather than ignored. Third, set the deductible structure to recognise that dust losses tend to be either small and frequent or catastrophic, with little in between, which argues for a sensible base deductible on attritional events and clear PML discipline on the tail.

The hardest number in this placement is the Probable Maximum Loss. With no domestic CAM operating history, the insurer cannot lean on portfolio experience, so the PML narrative the broker supplies will largely decide the capacity and price.

Build the PML around physical separation and the worst credible single event, not around the whole-site replacement value. The questions that move it:

1. Is the kiln hall fire-separated from milling, drying and warehousing? Compartmentation with rated walls and controlled openings is the single biggest lever on a fire PML. A monolithic shed under one roof pushes the PML toward total loss.
2. How is finished CAM and precursor stored? Large bagged or silo inventory of fine powder is both a fuel load and a high-value sum insured. Warehouse fire separation and stock turnover policy matter as much as the process zone.
3. What is the kiln replacement lead time? Imported high-temperature kilns and controlled-atmosphere systems can carry long procurement and commissioning windows, which drives the business-interruption tail far more than the material damage figure.
4. What detection and fixed protection covers the high-value zones? Early aspirating detection, fixed suppression and a credible emergency response shorten the loss and support a lower PML.

The machinery angle deserves its own line. Calcination and milling equipment sits squarely in machinery breakdown territory, and a kiln refractory failure or drive breakdown can stop production without any fire at all. A CAM programme that buys fire cover but skimps on machinery breakdown leaves a real gap, because the most likely production-halting event in early operation may be an equipment failure during ramp-up, not a blaze.

Write the PML up as a short engineering memo, not a single percentage. An insurer presented with a reasoned worst-case scenario will hold price discipline far better than one handed a bare number to argue with.

On a project of this scale, the material damage loss is rarely the largest number. The business interruption loss usually is, and CAM has features that make the BI tail unusually long.

Three of them stand out. Custom kilns and controlled-atmosphere lines carry long replacement lead times, so an indemnity period of 12 months is almost certainly too short. For a first-of-kind Indian asset relying partly on imported equipment, 18 to 24 months is a more honest starting point, and the broker should make the client confront that rather than default to a year.

Second, this is a single-site, first-mover asset feeding a concentrated set of cell-maker customers. A long outage does not just lose margin, it can lose offtake contracts and hand share to imports, which is a consequential loss dimension that a vanilla gross-profit BI calculation will not capture. Worth discussing supplier and customer extensions explicitly.

Third, the supply chain runs both ways. The plant depends on precursor and lithium feedstock, and its customers depend on it. That argues for genuine attention to contingent business interruption, in both directions, rather than treating it as boilerplate.

For the placement, that means three numbers the broker must get right: the indemnity period (lean long), the BI sum insured built on projected rather than current throughput as capacity scales, and a waiting period that does not strand the client on the high-frequency early-life events. Get those wrong and the policy will technically respond while leaving the client badly short on the loss that actually threatens the business.

Property dominates the early conversation, but the liability tail on a CAM plant is real and under-discussed. CAM is an intermediate product sold business-to-business into cell manufacturing, and that shapes the exposure.

The central scenario is off-spec material. If a batch of CAM carries the wrong particle morphology, moisture, or trace contamination, it can degrade the cells built from it, surfacing as capacity fade, swelling or thermal events far downstream and long after delivery. That is a classic product liability and pure financial loss chain, and the broker should map where the client's liability ends and the cell-maker's begins. Indian product liability cover, and the recall economics behind it, deserve specific attention here, because a defective intermediate that has already been built into thousands of cells generates recall and rework costs out of all proportion to the value of the powder itself.

Three liability strands to scope:

• Product liability and financial loss to customers from off-spec or contaminated CAM, including the cost of cell-maker rework and downstream recall, which standard third-party bodily-injury and property-damage wordings may not fully reach.
• Public and environmental liability around a chemical site handling fine powders, process effluent and inert and reducing gases, with the pollution and gradual-deterioration carve-backs read carefully.
• Employee exposure from confined-space and nitrogen-asphyxiation hazards, high-temperature work and dust inhalation, which ties the property risk back to workers' compensation and process-safety discipline.

The practical move is to read the product liability wording against the actual sale chain. Many off-the-shelf policies exclude or sub-limit pure financial loss and recall, which is exactly the exposure a CAM seller faces. If the client is contracting to supply gigawatt-scale cell makers, the indemnity limits and the recall extension should be sized to that contract, not to the modest invoice value of a CAM batch.

This is a placement to lead from the front, because the market does not yet have a settled template for CAM and the broker who supplies the structure usually keeps the account. A workable approach.

Start with engineering, not capacity. Commission or insist on a process-safety and fire-engineering survey before going to market, covering dust hazard analysis, kiln and inert-gas safety, compartmentation and detection. A clean survey is the single most valuable document in the file and turns a speculative submission into a rateable one.

Then build the programme in layers that match the hazard:

• Material damage and business interruption as the core, on a fire and special perils plus engineering basis, with the explosion peril and combustible-dust scenario explicitly confirmed in cover and a long indemnity period.
• Machinery breakdown and erection cover for the kilns and process lines, recognising that ramp-up breakdown may precede any fire loss, and that during construction the build itself needs erection-all-risks cover.
• Product liability with recall and financial-loss extensions sized to the cell-maker offtake contracts, not the powder invoice value.
• Public, environmental and employee covers consistent with a hazardous chemical site.

On the market itself, expect to involve GIC Re and treaty capacity given the values and the novelty, and expect the lead underwriter to want a story they can defend internally. Give them the underwriting narrative: process description, hazard zoning, protection investment, PML memo and BI basis. Tariff-era instincts do not fit a first-of-kind risk, so the broker who shows up with engineering and a structured PML will get both better price and broader cover than one who treats it as another factory.

The wider point: India is about to build a CAM cluster from zero. The brokers who learn this risk now, while there is one announced plant, will own the segment when there are ten. Treat this profile as the first draft of that playbook and refine it against each real survey you see.