Chasing leaks across a large distribution network is expensive and exhausting. Non‑revenue water (NRW) often sits in the double digits, the network keeps aging, and environmental dynamics make buried assets harder to read. In this landscape, two tools get discussed together but often confused: L‑band “active leakage” techniques that surface moisture anomalies as points of interest (POIs), and likelihood‑of‑failure (LoF) modeling that scores the risk of failure at the pipe level.
They serve different decisions.
Below, we unpack the main drivers of water loss, what LoF modeling delivers for planners, specifically Rezatec’s Pipeline Risk (PLR), where L‑band signals can help, and how to turn risk into field results.
The Anatomy of Water Loss in Distribution Networks
NRW blends real losses (leaks and bursts) with apparent losses (metering/data). Many utilities experience 20–30%+ losses before treated water reaches customers. Aging mains are part of the story, but so are site conditions: ground motion, soil type and moisture, vegetation vigor, slope/topography, and weather patterns. These factors vary by neighborhood, which is why a static “find and fix” mindset can underperform. A risk‑led approach focuses limited resources where the combination of likelihood and consequence is highest.
What LoF modeling delivers at the pipe level
PLR calculates three things at once: Likelihood of Failure (LoF), Consequence of Failure (CoF), and overall Criticality. Scores appear directly on a digitized map of your network – letting planners scan risk spatially, not just in spreadsheets. Risk is computed at roughly 100 m (~330 ft) pipe segments, which is granular enough to guide crews to the right blocks without implying pinpoint leak detection. Each segment includes a Certainty Index (0–1) to convey confidence and a set of Pipe LoF Influencers that make the model explainable: material, soil characteristics, ground motion, vegetation change, break history, and more.
Those details turn a red segment from “worry here” into “worry here because…,” which is what engineers and operations leaders need to justify action.
Why the inputs matter (and why PLR is not satellite‑only)
Accuracy in risk modeling depends on the breadth and relevance of inputs. PLR integrates 110+ variables across utility network, environmental, and satellite sources. Approved materials cite, among others, C‑band ground deformation monitoring and multispectral vegetation indices (e.g., NDVI). The value is not any one signal; it’s how the system learns your network’s unique failure signature across conditions. That’s why utilities use PLR to prioritize both near‑term investigations and longer‑horizon renewals.
Where L‑band techniques fit—and typical constraints
L‑band methods are designed to sweep large areas and flag moisture anomalies as POIs. As a triage input, POIs can help crews decide where to listen first or where to sanity‑check suspicious zones—especially if you account for recent rainfall and irrigation. Constraints reported in public materials include: POIs are often broad and not pipe‑specific; urban false positives can be common; historical baselines may be limited/costly; and L‑band is not a predictive planning tool. In short: L‑band can add another signal to field triage, but it’s not a substitute for pipe‑level, forward‑looking risk.
From risk to results: a practical workflow
-
Map risk and consequence: Use LoF, CoF, and Criticality to rank your top segments. Criticality helps avoid chasing small leaks while high‑impact mains degrade.
-
Explain the “why”: Review Pipe LoF Influencers alongside the Certainty Index to select the first wave of segments for validation.
-
Validate where it pays: Guide crews to those segments for non‑invasive checks. If you have L‑band POIs, treat them as supplementary—LoF sets the order of operations.
-
Act on findings: Focus acoustic leak detection, condition assessment, or renewal planning on validated, high‑risk segments. Tie actions to Criticality for service reliability.
-
Close the loop: Feed what you learn in the field back into the model so predictions improve with your data.
Answers to common questions
-
Q: Is PLR a leak detection system?
A: No. It predicts failure risk to support planning and helps target leak detection and inspections. -
Q: Does PLR use L‑band?
A: No. Approved materials reference C‑band deformation and multispectral vegetation indices, along with network and environmental data. -
Q: How specific are the outputs?
A: Pipe‑level risk with roughly 100 m segmentation, plus a Certainty Index and a breakdown of key risk drivers.
Water loss is a planning challenge as much as a detection challenge. L‑band POIs can help triage, but pipe‑level LoF, CoF, and Criticality—plus certainty and influencers—provide the roadmap for where to act and why.
See how this looks on your network – request a walkthrough below