Water damage in an empty house, or how to multiply a bill by ten
A 2 mm leak in an occupied house means a mop. In an empty one it means a six-month project. Anatomy of a claim that should almost never happen.
A weeping flexible connector passes roughly one litre an hour. Twenty-four litres a day. Almost nothing: in an occupied bathroom you hear it, you see the mark, you tighten the fitting and that is the end of it.
Multiply by forty-two days — the average interval between visits to a house looked after "by the neighbour" — and you have a thousand litres under a floor screed. At that point it is no longer a leak. It is a project.
I. What actually happens over six weeks
Water does not sit still. It travels, downwards and sideways, following whatever routes the building offers.
First days: the screed absorbs. Nothing is visible, including to anyone visiting. The tiles feel dry.
Second week: water reaches the insulation beneath the screed. Waterlogged polystyrene loses most of its thermal performance and never dries out in a closed house. On underfloor heating, efficiency starts to fall with nobody the wiser.
Third week: the partitions. Plasterboard wicks up twenty to sixty centimetres depending on the lining. Skirtings swell. Floating engineered flooring lifts and cups.
After a month: the smell. That is often the first human signal — the neighbour who comes in for the post and finds it "a bit musty". By then mould is established in the linings, and the treatment is no longer drying but stripping out.
II. The six starting points, by frequency
Across the houses we take into management, the origin of the claim breaks down fairly consistently.
- Flexible connectors — the braided stainless hoses under basins and behind WCs. Real service life: five to seven years. Almost nobody replaces them before failure.
- Washing machine and dishwasher — supply hose under standing pressure, door seal, a solenoid valve stuck open.
- Hot water cylinder — tank corrosion, a scaled safety group that no longer reseats. A holed cylinder releases 200 litres at once, then keeps refilling.
- Roof and terrace waterproofing — cracked upstands, outlets blocked with pine needles. Typical of the flat roofs on contemporary villas at Pyla.
- Buried irrigation — a pipe split by frost waters the subsoil for a whole spring, and the excess consumption only shows on the annual bill.
- Swimming pool — a leak at the main drain or in the plant room, emptying the pool and saturating the ground around the foundations.
III. Insurance: what the policy actually says
This is where owners discover, at the worst possible moment, a clause they never read.
Most household policies contain, for second homes, an unoccupancy clause. It typically provides that beyond a continuous period of absence — often thirty, sixty or ninety days depending on the policy — certain cover is reduced, made conditional on precautions, or suspended.
The precautions required look almost identical from one insurer to the next: water shut off at the main, systems drained during frost, heating maintained on frost protection, and periodic visits capable of proof.
The important word is capable of proof. In a dispute, the insurer will not ask whether someone was calling in: it will ask you to demonstrate it. A dated visit report with photographs is worth infinitely more than a statement written after the event. That is one reason we time-stamp and archive every visit — not out of administrative habit, but because that document becomes evidence the day there is a claim.
The subject is worth going into: see Insurance and unoccupancy, the clause nobody reads.
IV. Detect early, or not at all
There is no single solution. There is a chain, and it is worth what its weakest link is worth.
Level zero: the valve closed. A house whose mains supply is shut off between stays cannot flood. It is free, it is absolute, and it is incompatible with a water heater left in service or with occasional letting.
First stage: contact sensors. A pad on the floor under each water point, at five or ten euros each. It detects water once it reaches the floor — so late, but well before six weeks.
Second stage: the motorised valve. An actuator on the main supply, driven by the sensors, which shuts the water off automatically. It is the only device that turns an alert into an action when nobody is there.
Third stage: flow analysis. A smart meter that learns the house's consumption profile and flags the anomaly — a continuous overnight flow, for instance — before water ever reaches the floor.
We set out the real limits of each of these, including what they cannot do, in Detecting a leak from six hundred kilometres away.
V. The human factor is still decisive
An alert is only useful if somebody can respond to it.
I labour the point because it is the blind spot of almost every device sold to absent owners. The sensor sends a notification. The owner is in Brussels, in a meeting, and sees it two hours later. He calls a plumber he does not know, who cannot get in, and who will not come.
A complete chain means: detection, notification, somebody with the key, the mandate and the number of the plumber who knows the house, and a written record of what was done. The first three links are sold in shops. The fourth cannot be bought; it has to be organised.
Frequently asked questions
How much does water damage cost in a second home?
It depends entirely on how long it goes undetected. A leak caught the same day costs the price of a fitting and an hour of a plumber's time. The same leak found six weeks later means lifting floors, drying out, replacing insulation, renewing linings and redecorating: from a few hundred euros to tens of thousands.
Should the water be turned off every time you leave a second home?
Yes, as soon as the absence exceeds a few days, unless an automatic shut-off device is fitted. It is the only measure that removes the risk rather than reducing it, and most insurance policies mention it explicitly among expected precautions.
Does insurance cover water damage occurring during a long absence?
In principle yes, but settlement may be reduced or refused if the policy's unoccupancy clause was not observed: water shut off, frost protection, periodic visits. The particular conditions, which set the maximum absence periods, must be read, and proof of visits kept.
Is a connected leak detector enough to protect an empty house?
It detects; it does not act. Without a motorised valve to close the supply, and without someone able to be on site within the hour, the alert simply dates the incident more precisely. A complete arrangement combines detection, automatic shut-off and human response.
How do I prove to my insurer that the house was being supervised?
Through dated, illustrated visit reports kept over several years. A handwritten log can suffice; a time-stamped digital record is stronger. What matters is that the evidence existed before the claim rather than being reconstructed afterwards.
Updated May 27, 2026