How Irrigation Flow Monitoring Detects Leaks and Failures

Establish expected flow for each operating state

The controller should know which valves, pumps, and master valves are commanded when a reading is evaluated. A rotor zone, drip zone, and two-zone concurrent program can have very different normal flow. Expected ranges may begin with design estimates, but field measurements under stable conditions provide a better baseline.

Ranges need room for normal pressure and seasonal variation without becoming so broad that broken equipment is missed. Changes should be reviewable and attributable; silently relearning a broken line as normal defeats the purpose of monitoring.

• High flow can indicate a broken pipe, missing head, stuck valve, or unexpected concurrent demand.
• Low flow can indicate a closed supply, failed valve, clogged filter, pump problem, or restricted line.
• Flow with all outputs off can indicate seepage, a valve that did not close, or another untracked water use.
• No reading can indicate a meter, wiring, power, or communication fault rather than zero water use.

Account for timing and meter behavior

Flow is rarely stable the instant a valve opens. Lines fill, air moves, pressure changes, pumps start, and meters need enough movement to report accurately. Define startup delay, evaluation window, persistence, and recovery criteria so brief transients do not create repeated false alarms.

Meter size, pulse rate, minimum detectable flow, installation orientation, straight-pipe requirements, water quality, and wiring distance affect results. The controller cannot compensate for a meter that is incorrectly selected or installed.

Choose a proportional response

Not every exception should immediately close water to an entire property. A modest deviation may create an advisory for inspection; a sustained severe over-flow may stop the active station; flow with every station off may justify closing a master valve. The response should consider confidence, consequence, time of day, occupied areas, freeze protection, and whether shutdown creates another risk.

Maximum runtimes and total water limits provide an independent layer of protection. They remain valuable if a flow sensor fails or an exception does not fit the learned pattern.

Give technicians a usable incident record

An alert should identify the property, meter, active outputs, expected range, observed values, duration, automatic action, and current state. Preserve readings before and after the event so staff can see whether the exception was sudden, gradual, intermittent, or tied to a particular program.

After repair, record the cause and resolution. Over time, confirmed incidents improve thresholds, maintenance planning, and the business case for flow monitoring without hiding uncertainty behind a single red alarm indicator.