Microcracks and the 12-Month Wall: Underwriting Ground-Mount Solar Hail Damage and the BI Indemnity Gap in 2026

India crossed 272 GW of non-fossil capacity by the end of January 2026, with solar the single largest contributor, and the Ministry of New and Renewable Energy put total non-fossil capacity at roughly 283 GW by 31 March 2026. FY 2025-26 added record solar volumes, much of it ground-mount in Rajasthan, Gujarat, Madhya Pradesh and Karnataka. That growth has dragged a weather peril most Indian property underwriters used to treat as a rounding error into the centre of the file: hail.

Severe convective storms, the meteorological family that produces large hail, drove an estimated US$60 billion of insured losses globally in 2025, and the solar account has felt it directly. Individual hailstorm events overseas have each caused more than US$50 million of module damage, and the reinsurance market has responded by tightening solar terms, raising deductibles and pushing waiting periods out.

India is not insulated. Unseasonal hail over the central and western solar belt now arrives in pre-monsoon convective bursts that did not feature in the historical loss record those sites were priced against. For a broker, the practical problem is that a standard Standard Fire and Special Perils (SFSP) or operational package policy was never engineered for a tracker-mounted glass surface tilted into the sky across hundreds of acres.

The exposure is also concentrated in a way that should worry an accumulation underwriter. India's ground-mount build is clustered in a handful of high-irradiance states, which means a single convective event can hit multiple insured parks in the same portfolio on the same afternoon. A peril that was once spread thinly across a property book now lands as a correlated solar loss.

Hail is no longer a peril you note and move past on a solar risk. It is the peril most likely to produce a disputed claim, because the damage it causes is frequently invisible at the point of survey and the loss it triggers outlasts the indemnity period you bound.

The reason hail breaks solar wordings is physical, not legal. A hailstone striking a module rarely shatters the front glass in the way a cricket ball would. More often it transmits a shock that fractures the silicon cells beneath an intact glass surface. These micro-cracks are invisible to the naked eye and to a routine post-event walkthrough. The panel still generates, the glass looks fine, and a fast claim survey can pass it as undamaged.

That is the trap. Micro-cracks propagate under thermal cycling and mechanical stress, gradually isolating cell areas, creating hot spots, accelerating degradation and, over months, dragging the array's output below its warranted curve. The yield loss is real and largely irreversible, but it shows up after the survey, after the claim, and often after the indemnity period has closed.

Detecting it requires electroluminescence (EL) imaging, where a current is driven through the module in darkness and a camera captures the cells that no longer light up. EL is the only reliable way to separate a genuinely sound panel from one that is quietly failing. It is slower and costlier than a visual inspection, which is exactly why a vanilla policy never contemplates it.

Forensic practice now sorts hail damage into three buckets: cosmetic (appearance only), functional (minor measurable loss) and critical (significant degradation or a safety hazard such as hot-spot fire risk). That categorisation drives whether a claim is paid and at what quantum. If your material damage wording does not name EL testing as the agreed method of establishing damage, the insurer and the insured will argue about whether a module that passes a flash test but fails EL is even "damaged" at all.

On a conventional solar placement, two heads of cover are supposed to catch a hail loss. The material damage section pays to repair or replace physically damaged property. The business interruption section pays the revenue lost while that damage is being made good. Micro-cracking slips between them.

First, the material damage trigger usually turns on "physical loss or damage". An insurer can plausibly argue that a module which still produces, still passes a power test and shows no visible breakage has suffered no covered damage, even though EL shows cells fracturing. If the wording does not define damage to include latent cell cracking confirmed by EL, you have handed the insurer a defence.

Second, even where damage is conceded, BI responds only to loss flowing from that damage during the indemnity period. Silent degradation produces a slow yield decline that may not breach the policy's revenue threshold until long after cover has expired. The loss is genuine; the consequential loss is simply out of time.

There is also a machinery dimension. Inverters, transformers and tracker motors can be hit by the same storm, and whether that sits under property perils or machinery-breakdown cover changes the deductible and the waiting period that applies. A broker placing solar without aligning the SFSP, machinery breakdown and BI sections to a single agreed damage definition is leaving three different adjusters to fight over one storm.

Even a clean claim runs into a structural problem: time. The standard BI indemnity period in Indian operational solar placements is twelve months. That number was inherited from general property practice, where a damaged factory can be rebuilt and re-equipped inside a year. A solar plant cannot.

Replacement modules are not held in bulk on site. Specialised crystalline and tracker-matched panels, inverters and transformers carry lead times that frequently run to a year or more, lengthened further by the global supply position and by India's own module sourcing and approved-list requirements. Re-engineering a damaged string to accept a different module, and finding the scarce specialist labour to do it, adds months again.

Stack the real sequence: storm, then EL diagnosis to confirm which modules are actually compromised, then procurement, then shipping and customs, then installation and recommissioning. Twelve months is gone before the replacement panels reach the substation. The insured carries the revenue loss in the gap, even on a paid claim. The problem compounds at scale: a single severe storm over a several-hundred-megawatt park can compromise tens of thousands of modules at once, and there is no scenario in which that volume of matched, approved-list panels arrives, clears customs and gets installed inside a year.

Insurers have noticed the tail. Reported practice has stretched BI waiting periods from around fifteen days out to forty-five and even ninety days, pushing risk back onto the insured at the front of the loss while the indemnity period still expires at the back. That is the worst of both ends: a longer self-insured opening and an unchanged twelve-month ceiling. The economically correct response is not to accept a longer waiting period silently; it is to extend the indemnity period to match the replacement reality.

For ground-mount solar exposed to hail, a twenty-four to thirty-six month indemnity period is the honest number, and it should be negotiated before binding, not discovered during a claim. The cost of that extension is modest against the revenue at stake, and it is the only lever that converts a structurally short policy into one that actually pays the loss to its end. A broker who bolts an extended indemnity period onto the slip at renewal, rather than at first placement, has already let one storm season pass uncovered for the part of the loss that matters most.

This is where a broker earns the fee. Two clauses decide whether a hail loss is paid in full or argued down, and both are negotiable at placement.

Define damage by EL, not by appearance

The material damage section should state that physical damage includes latent cell cracking and electrical discontinuity established by electroluminescence imaging following an insured peril, and that EL is the agreed method of loss assessment for solar modules. Specify who pays for EL testing and at what coverage density (full-array versus sampled) so the cost of proving the loss does not itself become a dispute. Where possible, secure a baseline EL scan at commissioning so post-event imaging has a clean reference and the insurer cannot attribute cracks to manufacturing or transit.

Extend and align the indemnity period

Set the BI indemnity period against realistic module and inverter lead times, not the default twelve months. Twenty-four to thirty-six months is defensible for hail-exposed ground-mount sites. Add an extended supply chain or loss of profits lead-time provision so procurement delays inside the indemnity period are explicitly covered rather than treated as the insured's own delay.

Three more levers matter:

• Reinstatement basis. Confirm replacement is to current technology and efficiency, since the original module may be discontinued, and that betterment is not deducted where like-for-like is genuinely unavailable.
• Degradation carve-back. Replace any blanket performance-degradation exclusion with wording that carves back degradation caused by an insured event.
• Deductible structure. Match the time and monetary deductible to the waiting period so the insured is not exposed twice for the same opening period of the loss.

Better wording costs money, and the broker has to defend the premium to a developer whose IRR model assumes a thin insurance line. The argument is straightforward: the cheap policy does not actually transfer the hail risk, so its low premium is illusory.

Expect insurers to price hail exposure through three levers rather than headline rate alone. The first is a separate, higher hail deductible, often percentage-based on the module sum insured, which keeps attritional cosmetic claims off the policy. The second is a stretched waiting period, which must be negotiated alongside the deductible so the two do not silently double the insured's retained opening loss. The third is the indemnity period itself, where moving from twelve to twenty-four or thirty-six months carries real cost but is the only lever that fixes the structural gap.

Underwriters will also reward physical mitigation. Modules certified to the tougher hail impact classes, the ability to stow trackers to a steep defensive angle when a storm is forecast, and a credible pre-event stow protocol all reduce expected loss and should pull rate down. A broker who presents the stow capability and the IEC hail rating in the submission gives the underwriter a reason to compete.

This is also where India-specific structure helps. A parametric hail or convective-storm trigger sitting on top of the indemnity policy can pay quickly on a measured event, bridging the long EL-and-procurement tail that the indemnity BI only settles at the end.

Most solar hail disputes are lost in the first seventy-two hours after the event, when the wrong evidence is gathered and the wrong assumptions harden. The fix is a pre-agreed claims protocol attached to the policy, not improvised after the loss.

The protocol should require, in sequence: immediate visual and drone survey of the affected blocks; preservation of weather data and any hail-pad or radar evidence to fix the storm as the proximate cause; EL imaging of a statistically valid sample (and full-array EL where sampling shows material cracking); and a power-performance baseline comparison against the commissioning EL reference. Naming the surveyor or loss-adjuster panel with solar EL competence in advance avoids the delay of appointing one mid-loss.

1. Fix the cause first. Without contemporaneous weather and radar evidence, the insurer can later argue the cracks predate the storm or arose from transit or installation.
2. Image before you replace. Once panels are removed, the chain of evidence breaks. EL has to happen in situ, before any reinstatement begins.
3. Quantify the tail honestly. Model the yield decline across the full indemnity period, not just the weeks immediately after the storm, because the degradation curve is the loss.

Document the duty of fair presentation under the principle of utmost good faith at inception: disclose the site's hail history, the module hail rating and the stow capability. A clean disclosure file is what stops an insurer from reopening coverage when a large degradation claim lands two years later. The broker who builds this protocol into the policy schedule, rather than leaving it to the developer's plant manager, converts a likely dispute into a payable claim.