The pipe did not fail last Tuesday. It started failing months ago.
- Adelitics Team
- May 19
- 4 min read
Updated: May 20
Most bursts are not surprises. They are the final event in a deterioration sequence that began long before water reached the surface.
That sequence generates detectable signals: and utilities that know how to listen and interpret them can avoid most pipe breaks.
That is the core argument behind leak before break: the gap between first signal and catastrophic failure is not an inevitability. It is a decision window. And for utilities still operating primarily in reactive mode, that window is where the real cost is paid.
The burst you saw was not the failure
When a large water main ruptures, the visible event (the flooding, the emergency crews, the road closure) is the end of a process, not the beginning of one. By the time water reaches the surface, the pipe has typically been deteriorating for months or years.
Cast iron mains corrode from the outside in. Cement-lined pipes develop hairline fractures under cyclic pressure. Ground movement stresses joints that have held for a century. These processes do not announce themselves. But they do leave traces.
As a pipe wall weakens and develops micro-cracks, pressurised water begins to escape through defects at very low flow rates. This generates acoustic emissions: vibration that travels through the pipe wall and surrounding soil. In many cases these signals are present weeks or months before structural failure occurs.
What the deterioration process actually looks like
Pipeline failures typically follow a progression. Corrosion or material fatigue creates localised weakness. Pressure fluctuations (daily cycles of demand, transient events, temperature variation) load and reload that weakness over time. Eventually, a small defect forms. Water escapes. The surrounding soil begins to wet and soften. The defect grows.
In many failure modes, this process takes months. In some, it takes years. Catastrophic failure, the event that produces headlines and emergency response, is the terminus of the sequence, not its beginning.
The implication is direct: if the signals are there, and if monitoring systems are capable of detecting them, the question is no longer whether early warning is possible. It is whether utilities are positioned to act on it.
The window utilities are missing
A leak identified before it becomes a burst is a fundamentally different
operational problem from a burst that has already occurred.
When water is escaping through a small defect under live pressure, a targeted repair can be scheduled. The pipe stays in service. Supply is maintained. There is no emergency excavation, no flooded basement, no reputational event. The intervention is planned, contained and proportionate.
When the defect reaches catastrophic failure, none of that is true. As just one of thousands of examples, in October 2019, a trunk main failure in Finsbury Park sent water through approximately 150 homes, suffered water damage and displaced 90 families into temporary accommodation. Eighty firefighters responded. Supply to the surrounding area was interrupted for hours. The broader cost (emergency repair, insurance, restitution, temporary accommodation, regulatory scrutiny) is rarely captured in a single figure, but it is always larger than the cost of the repair that preceded it.
The window between first signal and catastrophic failure is where the financial case for preventative water network maintenance is most concrete. Every hour inside that window is an hour in which the outcome is still preventable.
What acting in the window looks like
Across the Adelaide CBD, a continuous acoustic monitoring network covering approximately half of the water main network (around 300 permanent acoustic loggers) demonstrated what acting in this window actually produces.
Published research from the programme found that approximately 55% of developing leaks associated with cracked pipe were detected before uncontrolled failure, meaning before unplanned customer interruption or third-party damage. SA Water has reported publicly that the smart network has helped detect around half of all water main leaks and breaks in the Adelaide CBD since 2017, enabling repair crews to act while the city sleeps.
These are not marginal gains. They represent a systematic shift in when utilities learn about deterioration, and therefore in what they can do about it. Earlier detection produces earlier intervention. Earlier intervention produces fewer catastrophic outcomes.
The work undertaken in Adelaide with SA Water ultimately evolved into Adelitics.
Capital investment for the Adelaide CBD programme was approximately AUD $900,000 in sensor hardware and deployment at the time of commissioning — a figure that needs to be read against the cost of even a single major trunk main failure, which can reach tens of millions of dollars when insurance, restitution and emergency response are included.
The arithmetic is not complicated. What requires attention is the organisational commitment to act on the signals rather than wait for the burst.
The organisational condition
Technology does not close the decision window on its own. Sensors generate alerts. Analytics interpret those alerts. But someone has to act on them, and that requires workflows, dedicated capacity and genuine operational trust in what the system is telling you.
When acoustic monitoring is treated as an additional responsibility layered onto existing roles, it competes with higher-priority reactive tasks and loses. The monitoring alert for a slowly developing leak does not carry the same urgency as a burst that is already on the street. Without deliberate organisational design, the window closes by default.
The utilities that extract value from leak-before-break monitoring share a common characteristic: they treat it as an asset management strategy, not a technology deployment. The distinction matters because it determines whether the decision window is used or lost.
Pipe failure is rarely a sudden event. In many cases, the evidence of what is about to happen is already in the acoustic record, embedded in signals that have been building for months. The question is not whether those signals exist. It is whether the utility has the monitoring infrastructure, the analytics capability and the organisational commitment to act on them before the window closes.
