Water infrastructure
Most of what a water company owns is buried, most of the cost is in the pipes, and most of it was laid by people who are no longer alive.
Water infrastructure is unusual among utilities in that almost all of it is invisible, most of it is very old, and its condition is largely unknown until it fails.
What the assets are
| Asset | Typical design life | Note |
|---|---|---|
| Impounding reservoirs and dams | 100+ years | Very long-lived; the constraint is siltation and safety inspection |
| Trunk mains and aqueducts | 80–120 years | Failure is rare and consequential |
| Distribution mains | 50–100 years | The bulk of the network by length. Cast iron laid a century ago is still in service |
| Service pipes | 40–70 years | Often the customer’s responsibility, and a large share of leakage |
| Treatment works | 25–40 years for plant | Civil structures last far longer than the equipment inside them |
| Pumping stations | 20–40 years | Electrical and mechanical plant on shorter cycles |
| Sewers | 80–150 years | Victorian brick sewers remain in service across British cities |
The renewal arithmetic
- A network of 100,000 km of distribution main
- An average asset life of 100 years
- Steady-state replacement, so the network neither ages nor rejuvenates
- Required annual replacement
100,000 km ÷ 100 years = 1,000 km/year - As a percentage of the network
1,000 ÷ 100,000 = 1% per year - At a typical replacement rate of 0.2%
100,000 × 0.002 = 200 km/year - Implied replacement cycle at that rate
100,000 ÷ 200 = 500 years
Sustaining a 100-year asset life requires replacing about 1% of the network annually. Typical actual replacement rates across many countries have been a fraction of that, which means networks are ageing rather than being maintained in steady state.
Assets do not all fail at their nominal design life — many cast iron mains have far exceeded theirs. The arithmetic shows the direction of travel rather than a prediction of imminent failure.
Why it is hard
- Cost is dominated by excavation and road reinstatement, not by the pipe. Replacing a main under a busy street can cost several times the same work in a field
- Condition is largely unknown. Many utilities do not know precisely what material a given main is, or when it was laid
- Failures are localised and repairable, so deferring renewal has no immediate visible consequence — until the failure rate rises
- The benefit of renewal accrues over decades while the cost falls in one price review period, which is an awkward fit for five-year regulatory cycles
- Sewers are worse: they are inspected less, fail less visibly, and their failure mode is collapse rather than leakage
What is changing
Condition assessment is improving — acoustic and satellite leak detection, in-pipe inspection, and pressure and flow analytics all narrow where to dig. Trenchless techniques such as pipe bursting and lining reduce the excavation cost that dominates. Neither changes the underlying arithmetic: a network with a hundred-year life needs about one percent replaced every year.
Sources
- Water Services Regulation Authority — Water company performance and price review data. Open — attribution required · Open Government Licence v3.0
- American Water Works Association — AWWA standards, water audits and utility benchmarking. Public but restricted · Publisher terms — public access, reuse not clearly granted
- Water UK — Water UK industry statistics and company reporting. Public but restricted · Publisher terms — public access, reuse not clearly granted