Twenty Bidders for the Rs 7,280 Crore Magnet Scheme: Underwriting a Factory India Has Never Insured

Energetica India Magazine reported on 14 August 2026 that the Ministry of Heavy Industries had received 20 bids under the Rs 7,280 crore rare earth permanent magnet (REPM) manufacturing scheme. Saur Energy followed on 18 August 2026 with the shape of the contest: 20 bidders competing for five beneficiary slots, which it read as a signal of how sharply the domestic critical-minerals market has changed.

Five of those bidders will end up building sintered neodymium-iron-boron (NdFeB) magnet plants in India. Each will need a property and engineering programme, first for the construction phase and then for operations. And each will walk into a market that has never rated this occupancy, because until now India did not have one.

That matters more than it sounds. Indian property underwriting still runs largely on occupancy analogy: an underwriter reaches for the nearest known class, applies its rate and its warranties, and moves on. There is no near class here. A magnet plant is not a foundry, not an electronics assembly line, not a general engineering works. It is a powder-metallurgy operation running reactive rare-earth metal in an inert atmosphere, with a hydrogen inventory, vacuum furnaces above 1,000 degrees, and a feedstock chain that runs through Chinese export licensing.

The broker who arrives with the process mapped, the hazard zoned, and the wordings checklist drafted will set the terms. The broker who treats it as a metal-fabrication risk will place a policy that responds to the wrong events.

Sintered NdFeB magnets are made by powder metallurgy, not casting or machining. The sequence drives every exposure worth pricing.

1. Strip casting and hydrogen decrepitation. Alloy is cast into thin flakes, then exposed to hydrogen gas, which the alloy absorbs and which cracks it apart along grain boundaries. This is the step that puts a flammable gas inventory inside a metals plant: hydrogen storage, distribution, a vacuum-and-purge cycle, and the leak-detection discipline that goes with it.
2. Jet milling to fine powder. The decrepitated material is milled to particle sizes in the low single-digit micron range under inert gas. At that fineness, rare-earth metal powder is both pyrophoric, meaning it can ignite on air contact without an external ignition source, and a dust explosion hazard in suspension.
3. Aligning and pressing. Powder is pressed in a magnetic field to orient the grains, then isostatically compacted. Powder transfer between milling and pressing is where inert containment is most likely to be breached in practice.
4. Vacuum sintering and heat treatment. Green compacts are sintered in vacuum furnaces above 1,000 degrees Celsius, then annealed. These furnaces are the highest-value single assets on site and typically the longest to replace.
5. Machining, grinding and coating. Sintered blocks are cut and ground to tolerance, generating wet grinding swarf of fine rare-earth metal, one of the most under-managed ignition sources in the whole plant. Coating adds a surface-treatment chemical load.
6. Magnetising and testing. Final magnetisation creates strong field hazards around equipment and personnel, and handling risk on assembled magnet stacks.

Read that list as an underwriting brief. Steps 1 to 3 are a flammable-gas and combustible-metal-dust risk. Step 4 is a machinery breakdown and long-lead-time risk. Step 5 is an ignition-source and housekeeping risk. Nothing in it is covered adequately by an assumption borrowed from a general engineering rate.

Underwriters who have handled pharma or food-processing accounts will recognise the deflagration chain: a primary ignition inside a mill or conveyor lifts settled layers, and the secondary explosion across the building is what destroys the plant. That mental model transfers here, and the discipline that goes with it, dust explosion process safety of the kind Indian pharma sites have had to build, is the right starting vocabulary.

What does not transfer is the extinguishing assumption. Rare-earth metal powder is a combustible metal, and water on a burning metal fire can accelerate it rather than control it by liberating hydrogen. A site protected on the standard sprinkler-everywhere logic can be worse off in the powder zones than one that is deliberately not sprinklered there and is instead served by Class D extinguishing media, inert containment and isolation.

Three consequences for the placement.

• Fixed protection has to be zoned, not uniform. Ask what medium protects the milling, powder-transfer and grinding zones specifically, and whether the answer is Class D media, inert-gas suppression or dry sand and shovel discipline. Uniform sprinkler coverage stated on a proposal form is a warning sign here, not a rating credit.
• Grinding swarf handling is a named question. Wet grinding sludge from sintered magnet machining is fine, reactive, and prone to self-heating as it dries. How it is collected, kept wetted, stored, and disposed of is a housekeeping control an underwriter should write into the risk report.
• Hydrogen and inert gas need their own hazard study. Hydrogen decrepitation puts flammable gas next to reactive metal powder. The site should be able to show a hazard study covering the gas system, the purge cycle, leak detection, and the confined-space and asphyxiation exposure created by inert atmospheres.

The five scheme winners will be in construction before they are in production, so the first programme is an erection all risks placement with delay in start-up (DSU) attached, not a fire policy.

The DSU indemnity period is where most of these placements will go wrong. The default habit is to size it against the contract programme: take the contractual completion date, add a margin, and call it done. That understates the exposure on a plant whose critical path is imported vacuum sintering furnaces, jet mills and magnetising equipment. If a furnace is damaged in transit or during erection, the recovery clock is not the contractor's schedule. It is the manufacturer's order book plus fabrication plus shipping plus re-erection plus re-commissioning.

Size the DSU indemnity period on the replacement lead time of the longest-lead imported item, evidenced by a written quotation from the equipment supplier, and put that evidence in the submission file. An underwriter shown a supplier lead-time letter will accept a longer indemnity period far more readily than one asked to take 24 months on trust.

Three other EAR-phase points worth settling before binding:

• Marine cargo and transit on the imported furnace and mill packages, with the delay-in-start-up trigger consistent between the marine cargo section and the DSU section. A gap between the two is a common way a transit-caused delay ends up uninsured for its financial consequence.
• Testing and commissioning cover with a duration that reflects a first-of-kind ramp, not a nominal four weeks. Commissioning a vacuum sintering line and a hydrogen system is where early-life damage is most likely.
• The handover boundary between EAR and the operational property programme, so there is no window in which the plant is neither an erection risk nor an operating one.

Once operating, the plant carries two overlapping covers on the same assets: fire and special perils on the material damage side, and machinery breakdown on the equipment side. Vacuum sintering furnaces sit exactly on the seam, and that is where claims disputes are made.

The scenario is ordinary enough. A heating element fails, or a vacuum seal loses integrity, or a control system misfires, and the furnace is damaged. Was that a breakdown, or was it a fire? The material damage insurer will argue the proximate cause was internal machinery failure and point at the machinery policy. The machinery insurer will argue there was a fire and point back. The insured sits in the middle with a stopped production line and two adjusters.

Fix it at placement, not at claim. Three drafting moves.

1. Write the boundary into the wordings explicitly. State which policy responds to furnace damage arising from internal failure with consequential fire, and which responds to fire from an external cause. If both sections sit with the same insurer, a single combined section with one deductible is cleaner than two policies with a shared argument.
2. Align the deductibles and the indemnity periods. A machinery breakdown section with a 12-month indemnity attached behind a material damage section with 24 months creates a shortfall on exactly the loss the plant is most likely to suffer.
3. Cover refractory, elements and consumable parts deliberately. Machinery policies routinely exclude parts of limited working life. On a vacuum furnace, that exclusion can swallow the loss. Get the carve-out negotiated and priced rather than discovered.

The reason this scheme exists is that the magnet supply chain runs through China, and Chinese export licensing has already demonstrated it can throttle rare-earth and gallium flows. That creates an exposure that no standard business interruption wording addresses, and the interaction is worse than either problem alone.

The standard mechanism first. Contingent business interruption in most Indian programmes responds only when a named or unnamed supplier suffers physical damage of a type insured under the policy. A licensing regime, an export quota, or an administrative refusal causes no physical damage anywhere, so the trigger never fires. This is the same structural gap that export-control-driven supply disruption exposed for Indian EV and electronics firms, and it applies with full force to a magnet plant dependent on imported rare-earth feedstock.

Now the compounding effect, which is the part most brokers miss. Suppose the plant suffers an insured fire in the pressing hall and loses three months of production. Reinstatement finishes on schedule, but the feedstock the plant had queued has been reallocated, and the next licensed shipment is four months out. The physical damage loss is three months. The actual outage is seven. Most BI wordings will pay for the period during which the damage affected the business, and an insurer can reasonably argue that the extra four months flowed from a supply constraint, not the damage.

What to do about it at placement:

• Name the suppliers. A contingent BI extension that names specific upstream feedstock and alloy suppliers, with agreed sub-limits, is worth far more than an unnamed extension with a physical-damage-only trigger.
• Test the wording against a non-damage trigger. Ask the insurer directly, in writing, whether a supplier's inability to deliver for regulatory or licensing reasons is within cover. Get the answer on file whichever way it goes.
• Address the extension-of-outage point specifically. Negotiate wording that does not cut off the indemnity the moment the damaged asset is technically reinstated, where the business genuinely could not restart for reasons connected to the loss.
• Treat inventory policy as risk control. Buffer stock of feedstock is an operational answer to an insurance gap, and an underwriter should be told about it, because it shortens the tail on exactly the scenario that worries them.

The awkward part of a genuinely new occupancy is that neither side can point at a domestic loss record. That is not a reason to price on instinct. It is a reason to price on engineering evidence, and the broker should decide in advance what evidence counts.

The file that makes this account rateable contains, at minimum:

• A process hazard study covering the full sequence from hydrogen decrepitation through magnetising, done by someone who has assessed combustible metals and pyrophoric powder before, not a generic dust survey.
• A hydrogen system safety case: storage quantities and location, leak detection, ventilation, purge procedures, and separation from the powder-handling zones.
• Zoned fixed protection documentation naming the medium for each zone, with a written justification for any area deliberately left unsprinklered because water is the wrong medium there.
• A swarf and powder housekeeping standard covering collection, wetting, storage and disposal of grinding sludge, with an audit frequency.
• Compartmentation drawings. Fire separation between milling, pressing, sintering, machining and finished-goods storage is the single largest lever on the PML. A monolithic shed pushes the PML toward whole-site.
• Supplier lead-time letters for the furnaces, mills and magnetisers, which underpin both the DSU and the operational BI indemnity periods.
• A written PML memo setting out the worst credible single event and why, rather than a bare percentage for the insurer to argue with.

Expect the placement to need treaty and facultative reinsurance support given the values and the novelty, and expect the lead underwriter to want a narrative they can defend internally. Give them one. On a class with no precedent, the underwriting file the broker builds becomes the de facto benchmark for the next four plants.

That is the strategic point for brokers watching this scheme. Five plants will be built. The rating basis, warranties and wordings settled on the first one will be copied onto the rest, because the market will have nothing else to copy. Whoever does the engineering work now writes the standard.