Skip to Main Content
CDM Smith Pride Logo
A woman working at a desk with dual monitors displaying charts and diagrams in an office environment.A woman working at a desk with dual monitors displaying charts and diagrams in an office environment.

Find lead faster: shrinking LCRI replacement lists with data-driven tactics

Katie Deheer, data science expert at Trinnex, explains how data-driven strategies help utilities reduce uncertainty and plan more realistic, defensible replacements.

Practices

UtilitiesProgram Management

Markets

Local, County, and State GovernmentMunicipal UtilitiesState Government

Priorities

Lead in Drinking Water

The Lead and Copper Rule Improvements (LCRI) require utilities to submit and maintain a baseline service line inventory, update it annually, and replace all lead and galvanized requiring replacement (GRR) service lines within mandated timeframes, often while inventories still contain significant unknowns.

"Unknown service lines directly drive your annual replacement targets, even when many of them ultimately turn out to be non-lead," says Katie Deheer, data science expert at Trinnex. "Predictive modeling helps utilities reduce uncertainty so they can meet LCRI obligations more efficiently and realistically." 

Middlesex, NJ prediction map

With milestone deadlines approaching, utilities are using data-driven tactics such as predictive modeling to reduce uncertainty, set more realistic replacement targets and prioritize verification work.

"Unknowns increase replacement targets. Predictive modeling helps utilities reduce uncertainty before those targets are set."

Katie Deheer, AI Consultant and Strategist

The reality: unknowns inflate your targets

LCRI replacement planning is based on counts of known lead service lines, galvanized requiring replacement (GRR) and unknown service lines. When inventories are heavy on unknowns, utilities can find themselves planning for replacement volumes that realistically aren't feasible, outpacing available funding, workforce capacity or construction seasons.

"That's where utilities can get stuck," Deheer says. "If you don't reduce unknowns ahead of your baseline inventory and replacement planning, your annual targets can become unmanageable."

The following four tactics show how utilities can use predictive modeling to reduce unknowns and improve replacement planning.

Blue circle with a white number 1 in the center.

Chip away at unknowns before baseline targets are set

Consider using predictive modeling to support the categorization of unknown service lines based on relative lead probability, informed by verification results and available contextual data (such as construction era, parcel information, and infrastructure records). Where the analysis indicates sufficiently low likelihood, some lines may be candidates for reclassification as non-lead.

In some systems, this approach has been observed to help reduce the number of unknown service lines prior to key deadlines, which may help moderate replacement targets that are otherwise influenced by unknown classifications.
 

Blue circle with white number 2 inside

Build a verification pool that actually represents your system

Design field investigations to produce a statistically representative verification dataset, instead of relying on convenience field investigations or redundant locations.

Evaluate existing field data for bias and fill gaps intentionally so the model’s performance and assumptions are defensible when they are reviewed by regulators or stakeholders.

Blue circle icon with the number 3 in white.

Prioritize resources based on likelihood of lead

Prioritize investigations and replacements in locations with the highest probability of lead or GRR, instead of excavating unknowns at random.

Improving the lead “hit rate” (for example, from ~10% to ~70–90% in high-probability areas) helps accelerate replacements.

Blue circle icon with white number 4 inside.

Treat your inventory like a living data set

Update the inventory and model regularly as new field verification data is collected, so that prioritization and target setting reflect current conditions.

Using an integrated workflow for inventory management, field data collection, modeling and reporting (for example, a single platform that connects these steps) helps keep periodic updates consistent and audit-ready.

A note on accuracy and defensibility

A chart displaying the average cost and accuracy of various service line material verification methods.

“No identification method is 100% accurate,” Deheer notes. “What matters most is transparency, performance metrics, and responsible application.” Utilities can support responsible use of predictive methods by setting conservative classification thresholds, evaluating model performance over time, and maintaining clear, defensible documentation of the data, assumptions, and decision logic informing classifications.

 

What’s the bottom line?

LCRI deadlines won't wait for perfect data, and neither should utilities. A data-driven approach cuts through uncertainty, directs field work to the highest-risk lines, and produces an inventory that's both current and defensible.

“The key is starting now and continuing to refine,” says Deheer. “That's what makes compliance achievable and cost-effective.”

Curious to learn more?

Reach out to an expert to find out how we delivered this project.

Katie Deheer

Katie Deheer

Product Management

At Trinnex, Katie leads the development of innovative digital solutions that leverage data to support agility and smart decision making for infrastructure leaders.

More resources on this topic

View All Insights
Construction workers with excavators on a street.
Insight

Is your utility ready for the LCRI? Lessons from Illinois' early experience

Illinois utilities have been operating under stricter lead regulations since 2022, and Chicago's former water commissioner, Andrea Cheng, PhD, PE, shares lessons on navigating new sampling requirements, boosting resident engagement, and coordinating complex inventory efforts.
Worker checks data on smartphone during underground utility inspection.
Insight

Piloting no-dig lead pipe detection technologies: lessons learned from the field

A recent pilot study where no-dig technologies could not accurately find lead showed how important it is to test technologies on specific systems before deploying them large scale.
Person writing on a whiteboard with terms related to communications strategies.
Insight

Ten tips for connecting with the public

Aligning communities around infrastructure improvements sets a course for a brighter future. Here are 10 tips from communications experts to enhance community engagement on your next project.
Glowing digital envelopes on a dark background, symbolizing email communication.

Sign up for Lead and Copper Rule updates and more

Sign up for our newsletter to stay in the know on regulations, recent projects and the latest innovations to tackle lead in drinking water.

Sign Up