A portfolio water-loss program should connect valid signals, field response, loss estimates, repairs, verification, recurrence, escalation, and capital action.

Define what the portfolio means by a water-loss event

Separate a suspected leak, confirmed leak, pipe break, fixture or valve failure, equipment malfunction, overflow, continuous low flow, irrigation problem, cooling-tower loss, abnormal process use, billing or meter error, unauthorized use, and a legitimate operating change. Define the property and systems in scope, data available, response hours, severity levels, notification paths, authority to isolate water, resident or tenant obligations, insurer or regulatory notice, and the consequences the program is intended to reduce.

Use consequences as well as volume. A small release above electrical equipment, archive space, healthcare use, occupied units, or a critical process may require a faster response than a larger controlled loss to drainage. The program should distinguish water conservation, utility-cost control, property-damage response, environmental health, business continuity, and capital planning. KOW can help organize the technical record; emergency response, plumbing repair, remediation, safety control, insurance, and regulatory decisions remain with the responsible qualified parties.

Map supply, meters, end uses, and normal behavior

For every property, record water sources, utility accounts, master meters, submeters, AMI or interval feeds, domestic and process distribution, isolation valves, pressure zones, irrigation, cooling towers, boilers, kitchens, laundry, pools, laboratories, medical equipment, tenant or resident service, fire protection where relevant, alternate water, sewer or deduct meters, and the systems that receive or depend on the water. Identify owner, operator, utility, tenant, vendor, and emergency contacts for each boundary.

Build expected patterns at the interval the data can support: occupied and unoccupied flow, weekday and weekend, seasonal irrigation, cooling load, cleaning, turnover, construction, filling events, known batch processes, and legitimate baseload. Compare bills, meter data, equipment runtime, weather, occupancy, and operational logs. WaterSense and FEMP describe water balances as a way to reconcile supply with measured and estimated end uses and expose unexplained loss; the balance is a starting point for investigation, not proof that every unexplained gallon is a leak.

Design alerts that lead to an owned response

Select signals by property and consequence: continuous flow during expected low use, sudden rate change, daily or hourly variance from a weather- or occupancy-aware baseline, meter-to-submeter imbalance, equipment makeup or blowdown ratio, tank-level change, surface moisture, pressure loss, valve position, or a utility AMI notification. State the data interval, units, threshold, persistence, delay, suppression, escalation, and responsible recipient. A portfolio-wide percentage threshold can miss a small but damaging event or repeatedly flag a legitimate high-use property.

Document meter range, accuracy, multiplier, clock, communications, data gap behavior, device power, alarm testing, sensor placement, valve and shutoff capability, cybersecurity or network ownership, and what happens when the monitoring service is unavailable. EPA notes that flow monitors, surface sensors, and sensor-activated shutoffs can work together and can integrate with a building system. Automatic shutoff is not appropriate for every service; evaluate life-safety, sanitation, process, resident, equipment, and operating consequences before authorizing a control action.

Validate, contain, and estimate the event

The receiving person should confirm the property, meter, time, current rate, comparison period, known activity, data freshness, and whether another source corroborates the alert. Route the event by severity to the site team and qualified responders. Protect people and sensitive property, locate the affected area, isolate only through authorized procedures, maintain required services, document damaged or wet areas, notify responsible parties, and preserve photographs, meter and trend data, valve actions, equipment states, and the chronology of response.

Estimate the loss using the best supported method: interval consumption above a valid baseline, flow rate multiplied by a documented duration, master-to-submeter imbalance, tank or level change, measured discharge, equipment cycle data, or a bounded range where start time or flow is uncertain. State whether the estimate includes only purchased water or also sewer, energy, treatment, business interruption, material damage, drying, remediation, and restoration. Do not convert a billing spike into a precise leak volume without checking meter reads, estimates, adjustments, occupancy, weather, and operating changes.

Close repair, restoration, and verification separately

The event record should identify the failed component or supported cause, affected area, isolation, temporary work, repair scope, contractor or responsible party, materials, permits or authority involvement where applicable, pressure or operational testing, restored valves and controls, sanitation or water-quality steps, drying and remediation status, damaged-finish restoration, equipment reset, date normal service returned, and open limitations. An invoice marked complete does not prove every downstream consequence is closed.

Verify the result in the same evidence language that identified the event. Review post-repair interval flow, minimum-use periods, meter balance, pressure, equipment makeup, alarms, moisture or condition monitoring, occupant reports, and the next comparable operating cycle. Define the verification period and acceptance owner. Test and maintain monitoring devices and shutoff functions. Where the repair is temporary or the cause remains uncertain, set an expiration, inspection frequency, additional investigation, and explicit condition for permanent work.

Escalate recurrence into a capital and portfolio decision

Roll up events without losing the property record. Useful measures include events by severity and end use, response and isolation time, estimated and verified loss, repair and restoration cost, open days, recurrence by component or location, repeat alarms, false and unverified alerts, properties without interval data, meter and sensor reliability, permanent-repair status, and capital actions initiated. Keep normalized consumption or water-use intensity separate from incident control; a relatively efficient building can still have a material single event.

Set escalation triggers for recurring fixture or valve failures, corroded distribution, pressure problems, freezing, failed isolation, inaccessible piping, unreliable meters, obsolete controls, repeated cooling-tower overflow, chronic irrigation loss, or temporary repairs that outlive their intended period. Connect the event history to PCA and CNA findings, water-management plans, renewal projects, insurance or resilience work, and budget prioritization. The portfolio loop is complete when the alert produces an accepted disposition and the record changes the next operating or capital decision.

Primary sources

Sources and technical references.

Source review current as of September 2, 2026.