MFL vs PML for Indian Warehouse Risks: Stack Height, Sprinkler Credits, and Cluster Loss Data

Warehouse property underwriting in India has moved from a tariff-driven, occupancy-classified exercise to a risk-engineered, scenario-based discipline. The de-tariffing of fire insurance in 2007 and the subsequent maturation of risk-based pricing have shifted the methodological centre of warehouse underwriting toward two key scenarios: the Maximum Foreseeable Loss (MFL) and the Probable Maximum Loss (PML). These two figures, while related, address different underwriting questions and produce different programme implications. Conflating them produces underwriting errors that show up as either inadequate capacity or excessive pricing.

The Indian warehouse stock has expanded dramatically since 2018. The Bhiwandi cluster around Mumbai now holds warehouse capacity exceeding 80 million square feet, with much of this constructed as Grade A and Grade B logistics parks. The Manesar-Bawal corridor in Haryana, the Hosur and Sriperumbudur belt in Tamil Nadu, the Whitefield and Nelamangala corridor around Bengaluru, and the Bhiwadi-Neemrana belt in Rajasthan have all developed substantial warehouse clusters serving FMCG distribution, third-party logistics, e-commerce fulfilment, and finished goods storage for adjacent manufacturing.

The loss experience across these clusters has produced a structured evidence base for warehouse underwriting. The 2022 Bhiwandi fires, the 2024 incidents in Hosur and Manesar warehouses, and the recurring monsoon-related stock damage incidents in coastal warehouse clusters have given the Indian market several years of incident data on which to calibrate PML and MFL estimates. The same loss base has also informed risk improvement standards, sprinkler credit methodologies, and the underwriting expectations around fire-cell separation.

This guide is written for Indian commercial property underwriters, risk engineers, and brokers structuring warehouse placements with sum insured ranging from INR 50 crore to INR 5,000 crore and above. It addresses the methodological distinction between MFL and PML, the specific factors that move both figures (stack height, fire-cell separation, sprinkler systems, ignition source control, and operational housekeeping), the credit structure for risk improvements, and the practical use of cluster loss data to validate the scenarios.

The two figures address different scenario assumptions. Maximum Foreseeable Loss (MFL) represents the worst-credible loss scenario that could occur at the insured location, assuming the failure of one or more risk-control features. The MFL is the figure that anchors the total programme limit because it represents the upper bound of credible loss exposure. For a single warehouse facility of 3 lakh square feet holding finished goods inventory worth INR 600 crore, the MFL might be the full INR 600 crore if the scenario assumes failure of sprinklers, failure of fire-cell separation, and unimpeded fire spread across the entire facility.

Probable Maximum Loss (PML) represents the most severe loss that is reasonably probable at the location, with the active risk-control features functioning as designed. The PML accounts for sprinklers operating, fire-cell separation containing the spread to one cell, fire brigade response within reasonable time, and the operational characteristics that constrain ignition and propagation. The PML for the same warehouse might be INR 180 crore, representing fire damage to one fire cell with full inventory loss in that cell plus smoke and water damage to an adjacent cell.

The distinction matters because the two figures drive different programme decisions. The MFL drives the total sum insured and programme limit. The PML drives the working layer pricing and the layer where most reinsurance attention concentrates. Underwriters who price the programme on PML without considering MFL exposure leave gaps in upper-layer protection. Underwriters who set sum insured to MFL but price every layer at MFL-implied severity over-charge the programme and produce uncompetitive pricing.

The ratio between MFL and PML for warehouse risks varies significantly with the physical configuration. A purpose-built modern logistics park with engineered fire-cell separation, automatic sprinkler systems, and adequate fire-water reservoirs may have a PML at 20 to 30% of MFL. A converted warehouse without engineered separation, without sprinklers, and with limited fire-water provision may have a PML approaching 70 to 90% of MFL, because the risk-control features that would otherwise constrain the spread are largely absent. Indian warehouse stock spans this entire range, and the underwriter's PML-to-MFL ratio must reflect the specific facility's configuration.

The documentation expectation for PML and MFL calculations on Indian warehouse placements has tightened through 2024 and 2025. Risk engineering reports now routinely include scenario assumptions, the sprinkler design density and water supply duration, the fire-cell wall ratings, the inventory layout and stack heights, and the historical loss attribution if any prior events have occurred at comparable facilities. Underwriters who accept PML estimates without scrutiny of the supporting scenarios are taking on more underwriting risk than the headline pricing implies.

Among the physical variables affecting warehouse PML, stack height is the most consequential. Stack height refers to the vertical extent of stored inventory, typically measured from the floor to the top of the highest pallet or rack. Modern Indian logistics parks routinely operate at stack heights of 9 to 12 metres for ambient storage, with selective vertical storage in mezzanine and double-deep configurations extending higher. Cold storage and pharmaceutical warehouses may operate at lower heights for product-handling reasons, while quick-commerce fulfilment centres in metro geographies trend toward higher density.

Stack height affects PML through several mechanisms. The first is fire propagation dynamics. A fire in stored combustibles at low stack height (under 3 metres) propagates primarily horizontally with limited vertical thermal column development. A fire at high stack height (above 6 metres) generates a thermal column that pulls air, accelerates combustion, and produces faster horizontal spread along the ceiling. The flame heights and heat release rates from high-stack storage fires can be several multiples of low-stack equivalents.

The second is sprinkler effectiveness. Conventional ceiling-only sprinkler systems designed to densities such as 0.25 gpm/sqft over 1,500 sqft (typical for Class III commodity at moderate stack heights) become inadequate at high stack heights. For storage above approximately 3.7 metres for many commodity classes, in-rack sprinklers or higher-density ceiling sprinklers (specifically designed under NFPA 13 or FM Global guidance for high-piled storage) are required to constrain a fire to the fire cell. Indian warehouse stock often operates with sprinkler systems designed for moderate stack heights while actually storing inventory at heights that exceed the design envelope.

The third is fire-cell wall effectiveness. Fire-cell walls rated for 2-hour or 4-hour fire resistance are designed to contain fire within the cell for the rated period. The effectiveness depends on the height to which the wall is built, the integrity of penetrations, and the absence of paths for fire to bypass through the roof or floor. Walls that stop short of the roof line, or that have unprotected openings for conveyor systems and dock doors, are compromised regardless of their nominal rating.

The underwriter's stack height assessment should include the actual operating stack heights (not the design stack heights), the commodity class being stored, the sprinkler design density and the design stack height for which the sprinklers are rated, and any deviation between sprinkler design and actual operating practice. Where the deviation is material, the PML must reflect the inadequacy: a sprinkler system designed for 3 metre stack height storing inventory at 9 metre actual height should be modelled as essentially non-functional for fire control purposes, with PML moving toward MFL.

The 2022 and 2024 Bhiwandi warehouse fires both involved actual stack heights significantly exceeding the design parameters of the installed sprinkler systems. The post-event analysis showed that sprinklers operated but were unable to penetrate to the seat of the fire because of the height of the stored goods and the operational characteristics of the commodities involved (textile finished goods with high heat release and significant smoke generation). The underwriting lesson is that stack height verification, not stack height assumption, is the discipline that supports credible PML estimates.

Sprinkler credits have historically been the most discussed underwriting credit in Indian fire and special perils pricing. Tariff-era practice provided a defined percentage credit for the presence of an approved sprinkler system. De-tariffed pricing requires underwriters to assess sprinkler performance, not just sprinkler presence, and to grant credits commensurate with the actual fire control benefit.

The sprinkler credit framework for Indian warehouse risks should consider five elements. The first is design adequacy. The sprinkler system must be designed to a recognised standard (NFPA 13, FM Global, or the relevant Indian standards) for the commodity class and stack height actually being stored. A sprinkler design certificate from the installation date is necessary but not sufficient; the underwriter or risk engineer must verify that current operations remain within the design envelope.

The second is water supply adequacy. Sprinklers without adequate water supply do not constrain fires effectively. The water supply assessment includes the reservoir capacity (typically requiring 2 to 4 hours of design demand), the pumping capacity (with redundancy via diesel and electric pumps), the pipe network sizing, and the pressure availability at the sprinkler heads. Indian warehouses with sprinkler systems but inadequate fire-water reservoirs, often a feature of facilities upgraded incrementally rather than designed from inception, should not receive full sprinkler credit.

The third is maintenance and testing. Sprinkler systems require periodic testing of flow, pressure, and head condition. The risk engineering report should document the testing regime, the frequency, the responsible third party, and the last test date. Indian warehouses where sprinkler testing has lapsed, or where the testing has been internal only without third-party validation, present an unverified protection level. The 2024 Hosur warehouse incident involved a sprinkler system that had not been tested for over 18 months, with several heads found to be impaired on post-event inspection.

The fourth is operational compatibility. Warehouse operations that obstruct sprinkler coverage compromise effectiveness. Examples include stacking goods above sprinkler heads, installation of mezzanine floors without corresponding sprinkler protection, accumulation of dust on sprinkler heads in dusty commodity environments, and operational temperatures that risk premature head activation or false trip. The underwriter's site walkthrough should look for these operational issues, not rely solely on the design drawings.

The fifth is integration with fire-cell separation and detection. Sprinklers work best when integrated with fire-cell separation that limits the area requiring sprinkler activation and with detection systems that provide early warning. Stand-alone sprinklers without complementary detection and separation have lower effectiveness than integrated systems. The credit structure should reflect the integration, with full sprinkler credit typically conditional on the presence of complementary controls.

The quantitative credit range for Indian warehouse risks varies by insurer but typically falls between 15% and 40% of base rate for fully compliant, well-maintained, integrated sprinkler systems. Sprinkler systems with deficiencies in any of the five elements above receive proportionally smaller credits, and systems with material deficiencies may receive no credit despite being physically present. The underwriter's credit decision should be documented with reference to the specific findings, not granted as a blanket reduction.

Fire-cell separation is the architectural mechanism for limiting fire spread within a warehouse facility. The intent is to contain a fire to one defined cell, allowing sprinklers and fire brigade response to control the loss before it propagates to adjacent cells. Effective fire-cell separation is the single biggest difference between a warehouse PML at 20% of MFL and a PML approaching 80% of MFL.

The design standards for fire-cell separation in modern Indian warehouses typically draw on National Building Code 2016 requirements supplemented by FM Global or NFPA guidance for specific commodity classes. The standards specify maximum cell areas (typically 3,000 to 9,000 sqm depending on commodity hazard class and protection level), required wall ratings (2-hour, 3-hour, or 4-hour fire resistance), opening protections (fire doors with appropriate ratings and automatic closers), and continuity requirements (the wall must extend from floor to roof, with appropriate fire-stopping at penetrations).

The operational reality at many Indian warehouses deviates from design intent. Common deviations include: walls that stop short of the roof level, allowing fire to spread above the wall through the void space; fire doors propped open or wedged for operational convenience, defeating the door's containment function; conveyor openings between cells that are unprotected by fire dampers or curtains; service penetrations (electrical conduits, HVAC ducts, plumbing) that have not been fire-stopped after installation; and accumulation of stored goods against fire walls, transferring heat through the wall and overwhelming the rating.

The underwriter's assessment of fire-cell separation must therefore involve physical inspection, not document review. The risk engineer should walk the perimeter of each fire cell, verify wall continuity to the roof, check fire door closures and seals, inspect penetration fire-stopping, and document the operational practices around stored goods placement. Deviations should be photographed and noted in the report, with PML implications calculated specifically for the observed deficiencies rather than the design intent.

For warehouse facilities where fire-cell separation is structurally inadequate or operationally compromised, the PML calculation must reflect uncontrolled spread potential. The underwriter has three options: price the risk at the higher PML, require remediation as a subjectivity, or decline the placement. Indian commercial practice has generally favoured pricing with subjectivities, but the 2024 and 2025 loss experience has produced increasing willingness among underwriters to require pre-binding remediation for material fire-cell deficiencies.

The 2022 Bhiwandi cluster fire, which spread across multiple ostensibly separate warehouse units, was a defining event for Indian fire-cell underwriting. Post-event inspection found that the fire-cell walls between units, while nominally 4-hour rated, did not extend to the underside of the metal roof deck, allowing the fire to spread through the void. The lesson, internalised across the underwriting community, is that fire-cell separation is a function of execution and operation rather than design specification.

The three principal Indian warehouse clusters have produced distinct loss patterns that inform underwriting assumptions for risks in each location. Understanding the cluster-specific patterns supports more accurate PML and MFL calibration than generic warehouse parameters.

The Bhiwandi cluster around Mumbai is the largest single warehouse concentration in India, with capacity exceeding 80 million square feet predominantly serving FMCG distribution, textile finished goods, and e-commerce fulfilment for the Mumbai metropolitan region. The loss pattern shows three recurring features. The first is high textile finished goods exposure, with several major fires through 2022, 2024, and 2025 involving textile inventory in densely stacked configurations. The second is the monsoon flood exposure, with the 2024 monsoon producing significant water damage claims at ground-floor and basement storage areas. The third is the operational density issue, with many Bhiwandi warehouses operating beyond their design capacity in terms of stack height and aisle width, particularly during peak inventory periods around festivals.

The underwriting calibration for Bhiwandi placements should reflect these patterns. Textile-dominant inventory warrants conservative sprinkler credit assumptions because of the high heat release and smoke generation characteristics of textile fires. Ground-floor and basement exposure should be modelled for monsoon water damage, with sublimits and franchise structures aligned to the cluster experience. The operational density should be verified through actual site inspection at peak periods, not just at design capacity.

The Manesar-Bawal corridor in Haryana has developed substantial warehouse capacity serving the automotive industry, with finished vehicle storage, component warehouses for tier-1 and tier-2 suppliers, and emerging EV-related storage for batteries and components. The loss pattern shows two recurring features. The first is the EV battery storage risk, with several incidents involving lithium-ion battery storage facilities through 2024 and 2025 that have produced complex fire situations with extended firefighting requirements and significant smoke and water damage to adjacent inventory. The second is the metal components fire risk, where stored steel and aluminium components produce limited direct fire exposure but generate heavy structural damage and contamination from any adjacent fire.

The underwriting calibration for Manesar placements should specifically address battery storage configurations, including the separation of battery storage from other inventory, the use of specialised fire suppression where battery storage is significant, and the contingent business interruption exposure from automotive supply chain disruption. The metal components fire profile suggests close attention to structural protection and adjacency separation rather than commodity-class fire suppression alone.

The Hosur and Sriperumbudur belt in Tamil Nadu serves electronics, automotive, and emerging semiconductor supply chains. The loss pattern shows two recurring features. The first is the high-value electronics inventory exposure, with finished goods values per square foot among the highest in the Indian warehouse stock and contingent business interruption exposure for the downstream manufacturing customers. The second is the monsoon and cyclone exposure from the eastern coast weather systems, with the 2023 Cyclone Michaung event producing notable Tamil Nadu warehouse claims.

The underwriting calibration for Hosur and Sriperumbudur should address the high inventory value per unit area, with attention to commodity-specific PML calculations rather than generic warehouse parameters. The catastrophe modelling should include cyclone exposure with appropriate return period analysis, and the BI exposure should reflect the downstream supply chain impact for electronics manufacturers operating just-in-time inventory practices.

For each cluster, the underwriter should maintain a current loss database, ideally drawing on the Insurance Information Bureau (IIB) aggregate data supplemented by the insurer's own portfolio experience. Cluster-specific loss ratios, severity distributions, and incident pattern data provide the empirical anchor for PML and MFL estimates that would otherwise rely on theoretical models. To explore how a cluster-level intelligence layer can support broker advisory and underwriting quality assurance, Request Access to Sarvada's platform.

Where the underwriting assessment identifies risk-control deficiencies but the cedant insurer is willing to write the risk, the placement typically includes subjectivities or warranties requiring specific risk improvement actions. The structure of subjectivities and the discipline around their compliance materially affects whether the placement performs as priced or produces underwriting losses.

Subjectivities on warehouse placements typically fall into three categories. Compliance subjectivities require the insured to demonstrate ongoing compliance with codes and standards (factory licence, fire NOC, environmental clearances, BIS standards for electrical installations). Improvement subjectivities require specific remediation of identified deficiencies within defined timelines (installation of additional sprinkler heads, upgrade of fire-water reservoirs, fire-stopping of identified penetrations, training of operations staff on fire-cell discipline). Operational subjectivities require adherence to defined operating practices (maximum stack height limits, prohibition on smoking in storage areas, housekeeping standards, fire drill frequency).

The effectiveness of subjectivities depends on enforcement. Subjectivities that are written into the policy but never inspected for compliance provide nominal protection only. Insurers with active risk engineering follow-up programmes verify compliance at defined intervals, typically with a mid-term inspection at six months and a pre-renewal inspection at ten months. Compliance failures trigger policy review with options including premium adjustment, coverage modification, or non-renewal.

For brokers advising warehouse clients on placements with subjectivities, the strategic discipline is to manage the subjectivities as a project. Each subjectivity is logged with an owner, timeline, completion criteria, and evidence requirement. The risk improvement plan is reviewed quarterly with the client's operations team and reported to the insurer at defined milestones. Brokers who run this discipline well sustain favourable underwriting positions for their clients across multi-year renewal cycles; brokers who allow subjectivities to lapse without remediation find their clients facing premium increases or coverage restrictions at renewal.

The credit structure for completed risk improvements is the natural counterpart to subjectivities. Where a client has invested in sprinkler upgrades, fire-cell remediation, or operational training, the renewal underwriting should recognise the improvement with corresponding credit adjustments. The risk engineering report at renewal should explicitly compare the previous year's findings with the current state, documenting the improvements completed and the residual concerns. This documented trajectory provides the basis for credit conversations with the underwriting team.

The broader implication for warehouse underwriting in India is that the most effective programmes evolve across multi-year cycles, with annual loss experience, risk improvement progress, and PML refinement informing each renewal. Underwriters who treat each renewal as a standalone exercise lose the institutional benefit of the longer-term relationship. Brokers who structure their warehouse practice around multi-year risk management partnerships, rather than transactional placement, produce better outcomes for clients on both pricing and capacity dimensions.