Definition
LDCC measures the fraction of a distribution system's total pipe network where AI-based leak detection has been demonstrated to work reliably, as opposed to the fraction where detection coverage is assumed but unverified.
The Measurement Gap
Measurement Gap and Decision Supported
The Invisible Failure Mode: A water utility deploys an AMI-based leak detection system across its 1,200-mile distribution network. The vendor reports 85% network coverage based on meter density and algorithmic reach. The utility's water loss reduction plan assumes this coverage is real and projects a 30% reduction in real losses over 3 years. After 18 months, actual leak detection rates in different zones vary dramatically: the system performs well in dense residential areas with closely spaced meters but has functionally zero detection capability in industrial zones with sparse metering and large-diameter transmission mains. The 85% coverage claim was accurate by meter count but misleading by pipe-mile: the system's demonstrated coverage is closer to 52% of the network by length, and the undetected zones contain the oldest, most leak-prone infrastructure.
What Existing Metrics Miss:
Vendor-reported coverage metrics typically count sensors or meters, not demonstrated detection performance by network segment. IWA water balance methodology quantifies total real losses but does not distinguish between network segments with and without functional detection capability. Without LDCC, utilities cannot identify the "dead zones" where leak detection investment has not translated to actual detection performance.
The Decision LDCC Enables:
Where in the distribution network does the leak detection system actually work, and where are the detection dead zones that require supplemental monitoring, different technology, or targeted infrastructure renewal? LDCC converts a binary coverage claim into a spatial map of detection confidence.
Cross-References:
LDCC connects downstream to non-revenue water reduction program effectiveness. Dead zones with low LDCC concentrate undetected real losses. LDCC also connects to Pipe Condition Grade Validation Rate (PCGVR, GMI-WT-023) because detection dead zones often overlap with areas where pipe condition data is sparse. Customer Leak Alert Positive Predictive Value (CLAPV, GMI-WT-024) measures detection quality for the subset of leaks detected at customer meters, while LDCC measures the spatial extent of detection capability across all technologies. Demand Forecast Hydraulic Impact Score (DFHIS, GMI-WT-025) is affected by undetected leaks in LDCC dead zones that create unmodeled demand.
Formula
Plain language: Map the distribution system into detection environment segments (combinations of pipe material, diameter range, pressure, and ambient conditions). For each segment, assess whether the AI has demonstrated detection capability (confirmed finds in that segment type, or validated negative results). The coverage is the fraction of total pipe length in segments with demonstrated capability.
Notation:
LDCC = Sum(Length_demonstrated_s) / Total_pipe_length
Where:
- Length_demonstrated_s = pipe length in segment type s where the AI has at least one confirmed detection or validated negative (confirmed via field investigation)
- Total_pipe_length = total system pipe length
- s = detection environment segment (e.g., "6-inch DI pipe, 60 PSI, suburban residential")
Units: Percentage. 100% = every segment type has demonstrated detection. The gap (1 - LDCC) represents pipe length where "no detection" may mean "no capability" rather than "no leaks."
Proposed formula, pending validation pilots.