Step 3 / 20

Drought and clay, or the loss that takes ten years

8 min of reading

Every peril in the regime breaks something all at once, except one. Clay soil shrink-swell cracks a building over several years, through a slow and reversible movement: the clay loses its water in dry periods and retracts, takes it back and swells, and the masonry sitting on it works until it tears. There is no instant of loss, no water line to photograph, no witness. Yet it is the peril that today accounts for the most disputed share of the regime, and the one whose processing takes longest.

The first difficulty is that the order, for its part, carves up time. It recognizes a quarter, or a run of quarters, on soil moisture indices compared with their historical series. An insured whose cracks opened progressively between 2018 and 2023 therefore faces several possible orders, some recognizing its commune and some not, and it must attach its disorder to a precise period. That attachment is not a filing formality: it is the substantive question, and it is what the expert examination settles.

The notion governing everything is called the determining cause, and it deserves careful statement because it is not intuitive. The question is not whether drought played a part, it almost always plays one; the question is whether it is the determining cause of the disorder, that is, the one without which it would not have occurred. A properly founded building in clay soil cracks under an exceptional drought: the determining cause is the drought. A building founded too shallow in the same soil cracks under an ordinary drought: the determining cause is the design defect, and the regime does not answer for it.

The examination therefore bears on material elements that must be known in order to read a report. The nature of the soil, established by a geotechnical study with borings. The depth of the foundations, measured by inspection pits, and its comparison with what the rules applicable at the date of construction required. The presence of nearby vegetation, a close tree pumping water from the soil far better than a drought does. The existence of earlier disorders, which photographs, sale surveys and meeting minutes often document better than the occupants' memory.

Out of that analysis comes a fork that changes everything for the client, and that is better named early. If the disorder comes from a design or workmanship defect and the building is under ten years old, the route is not the natural catastrophe regime but the builder's decennial liability, with a structural damage policy that prefinances the works without waiting for liability to be allocated. Many files processed for three years as drought should have gone that way, and the time-barring of the route not taken is the true cost of the wrong turn.

One must then know what the indemnity covers, because expectations and reality diverge widely here. The regime answers for direct physical damage, which includes the necessary structural repair, and underpinning a cracked building regularly costs a substantial fraction of its value. It answers neither for the residual loss in market value of the repaired property, nor for loss of amenity, nor for the discomfort of living several years inside a file. And a repair that does not treat the cause lets the movement resume at the first dry summer.

The timetable, finally, must be announced to the client at the first meeting, failing which the relationship degrades over a fact nobody controls. Between the appearance of the cracks, the commune's request for recognition, the order, the examination with borings, any counter-examination and the execution of the works, three to five years is ordinary. The only thing that usefully shortens that delay is the early building of a dated chronology: timestamped photographs of the cracks with a measuring witness, successive readings, and preservation of the construction documents. A file that begins three years after the first cracks begins with no proof of their date.

The worked case

A 900 square meter commercial building, built in 2011 on clay soil, shows stepped cracking on two elevations. The manager photographed them for the first time in September 2019, then let matters run. They worsen markedly in the summer of 2022: a door no longer closes, a through crack appears. He notifies in March 2023. The commune is recognized for the third quarter of 2022 and for the third quarter of 2018; it is not for 2019 or 2020. The examination establishes foundations at 0.60 meters deep, where the 2011 soil study recommended 1.20, and notes an oak planted four meters from the most cracked elevation. How is this file handled?

The analysis

Three questions arise in this order, and the second decides everything. The first is attachment to an order: the disorders observed in 2019 cannot be attached to the 2018 recognition without a chronology dating them, and the September 2019 photographs are here the file's only dated document; the summer 2022 worsening, by contrast, attaches cleanly to the third quarter of 2022, which is recognized. Only one event can therefore be relied on, and it is the 2022 one. The second question is the determining cause, and the report leaves little room: foundations at 0.60 meters where the 2011 soil study recommended 1.20 constitute a workmanship defect, in soil whose nature was known and documented before construction. A properly founded building would probably not have cracked under the 2022 drought, which is the very definition of the criterion: the drought played a part, it is not the cause without which nothing would have happened. The regime is therefore most unlikely to respond, and the insured must hear it early rather than after two years of counter-examination. The third question is the alternative route, and it is the only good news: a building accepted in 2011 falls outside decennial liability, but the exact date of acceptance must be checked, along with the existence of a structural damage policy and any earlier notifications that interrupted time-barring, since a file opened in 2019 on the right basis would have been in time. The oak at four meters, finally, is not a decorative detail: it further weakens attachment to the drought by supplying a local and permanent cause of soil drying, and it will have to be dealt with whichever route is taken, failing which the underpinning will crack in its turn.

What to remember
  • 01A drought order recognizes a quarter, while the disorder takes years: attaching the damage to a precise period is the substantive question, not a formality.
  • 02The determining cause is not the one that played a part, it is the one without which the disorder would not have occurred.
  • 03Foundations too shallow in known clay soil, and the regime stops responding: the route becomes the builder's, if time limits still allow.
  • 04The regime answers for physical damage, not for residual loss in market value nor for loss of amenity.
  • 05Three to five years is an ordinary timetable: only a dated chronology from the first cracks usefully shortens the file.
The notions in this module