Data study · free, no signup

A house is 22 systems on clocks, from 10 to 200 years.

Every part of a house runs on its own clock. A smoke detector is spent at 10 years; the masonry holding the place up can stand for two centuries. In between sit 22 separate systems, each ageing at its own rate, and you are asked to track all of them at once. This is a component-by-component map of how long the parts of a European house last: 83 tracked parts across 22 systems, every figure sourced and tagged with the confidence behind it.

The data exists. It was never built for you.

There is no free, homeowner-facing guide to how long the parts of a European house last and what each costs to replace. The European references that exist (NEN 2767, BS 7543, VDI 2067, ISO 15686, the national valuation tables) are paywalled professional standards for surveyors, valuers, and asset managers. The popular consumer charts anyone can find online (NAHB, InterNACHI) are American, built on US products, costs, and climate.

This study is the first half of the map a homeowner needs: how long each part lasts. The second half, what each part costs to replace, follows once those figures clear review. The data underneath exists; what was missing is a homeowner’s version of it.

The clocks you reset are the short ones.

Sorted longest-lived at the top to shortest at the bottom, the long clocks (grey) run from forty years to two centuries; the ones you reset while you live there (terracotta) top out at forty.

years · one representative variant per system · source: appkeep component-lifespans dataset

Longest

Load-bearing masonry, 100–200 years. You never replace it on a schedule; an inspection finds it if it ever moves.

Shortest

A smoke detector, on a fixed 10-year replacement cycle (EN 14604 alarms carry a replace-by date). A safety consumable, replaced by the calendar, not by wear.

Your working band

Boiler 15–20, wet-room membrane 20–30, windows 25–35, interior paint 12–16. These fire while you own the place.

See the full table of 22 systems
SystemRepresentative partLifespan (yr)
Carrying structure Masonry (brick/stone) 100–200
Foundation Strip footing (masonry/concrete) 100–150
Roof Clay tile covering 60–100
Walls (exterior shell) Brick facade 60–100
Chimneys / flues Masonry chimney 50–100
Waste water Cast iron drain pipes 50–100
Electrical system Copper wiring (post-1970 installation) 50–80
Stairs & ramps Timber staircase 40–80
Water supply Copper water pipes 40–70
Gas supply Copper gas pipes 40–60
Security Mortice lock 25–40
Miscellaneous Solar PV panels 25–35
Openings (windows, doors, skylights) PVC window frames 25–35
Lightning protection Standard lightning protection system 20–30
Vertical transport Passenger elevator 20–30
Wet spaces Tile-on-membrane waterproofing system 20–30
Data / telecom Ethernet cabling (Cat 5e/6) 15–25
Drainage PVC gutters and downpipes 15–25
Heating system Gas combi boiler 15–20
Interior finishes Internal wall paint 12–16
Ventilation / AC Extract fan (bathroom/kitchen) 10–15
Fire safety equipment Smoke detector 10–10

Download the full dataset (83 variants, JSON)

Nothing big needs replacing for the first fifteen years. Then three clocks strike inside a decade.

Each part sits at the soonest it typically comes due. For the first fifteen years the only replacements are small consumables: a detector, an extract fan, a coat of paint. Then the boiler, the wet-room membrane, and the windows all reach the end of their lives within a decade of each other.

years of ownership · marker sits at the soonest that part typically comes due · big-ticket replacement · smaller upkeep

The axis runs a little past 60 years, so the load-bearing structure that stands for a century or more sits off to the right, mapped in the ladder above. A marker’s position shows when a clock first strikes, not what the job costs: replacement prices are a separate layer, held until those figures clear review.

See the replacement timeline as a table
PartSoonest (yr)Typical range (yr)Type
Detector 10 1010 Smaller upkeep
Extract fan 10 1015 Smaller upkeep
Paint 12 1216 Smaller upkeep
Boiler 15 1520 Big-ticket replacement
Data cabling 15 1525 Smaller upkeep
Gutters 15 1525 Smaller upkeep
Wet room 20 2030 Big-ticket replacement
Lock 25 2540 Smaller upkeep
Windows 25 2535 Big-ticket replacement
Gas pipes 40 4060 Smaller upkeep
Stairs 40 4080 Smaller upkeep
Water pipes 40 4070 Big-ticket replacement
Chimney 50 50100 Smaller upkeep
Drains 50 50100 Big-ticket replacement
Rewiring 50 5080 Big-ticket replacement
Roof 60 60100 Big-ticket replacement

Europe is asking every homeowner to plan decades ahead.

The revised Energy Performance of Buildings Directive fell due for transposition into national law on 29 May 2026, and member states are standing up building renovation passports: a timetable of measures for each home, with residential energy use set to fall 16% by 2030 (European Commission, Energy Performance of Buildings Directive).

A renovation passport is a maintenance plan by another name, and it only works if someone knows when each part of the house reaches the end of its life. That is the number this study supplies. The one place it stays silent is the heat pump: the technology is still young enough that its service life has not settled, so the honest range is wide and this study will not pin it.

Insurers already price one of these clocks.

In Sweden, an insurer’s payout turns on the age of a wet room. Swedish wet-room trade practice (the Byggkeramikrådet and GVK regime) puts the technical life of a tiled bathroom’s waterproofing membrane at 20 to 30 years, with renewal recommended around year 30. Several Swedish insurers treat the waterproofing as written off once it passes 30 years: damage to the structure underneath stays covered, but the membrane’s share of a water-damage claim falls on the owner. The dataset here carries the same membrane at 20 to 30 years, tagged to Swedish and Finnish sources, because a Nordic wet room is a regulated component with its own clock.

How the numbers are built, and what is held back.

Every figure is a sourced range, never a single spurious number. Lifespans are mid-range estimates for European residential stock, synthesised from freely-usable authoritative sources: the German government BBSR component table and the national valuation tables, the Dutch RgdBOEI inspection standard, Finnish Rakennustieto service-life cards and Motiva, Swedish Byggkeramikrådet and Boverket material, with US NAHB and InterNACHI figures used only to cross-check, never as a European number. The paywalled professional standards (NEN 2767, BS 7543, VDI 2067, ISO 15686) are named as references consulted; none of their tables is copied.

Each figure carries a confidence level, from high (three or more sources agree within a narrow band) down to unverified (fewer than two independent residential sources). Physics travels across borders, so a copper pipe corrodes the same everywhere; climate, regulation, and market do not, so every figure is tagged with the countries whose evidence backs it. That tag flags a range resting on US evidence before it is trusted for a Finnish or Dutch house.

Of 83 tracked parts, 77 are verified and ready. Six are held back on purpose, and naming them is the point: the four types of heat pump and the EV charger are too new for their service life to have settled, and prefab bathroom pods have no numeric source we could stand behind. Naming the six gaps is what makes the other 77 figures worth trusting.

Your house keeps its own clocks.

A chart of the average house is a start. Your own house runs on its own clocks, set by what you have and when it went in. The repair-or-replace calculator answers whether one ageing part is in a repair year or a replacement year, and the demo plan projects every clock in a house forward. Both run on the dataset above, and both are free to open.

The guides, by email

Get each new guide the day it lands.

One short email when a new guide goes up. Tell us where your house is and you get your country’s guides as they are written; leave it blank and you get the general ones.

One email per guide. Every one has a one-click way to stop.