Utilities must prioritize large inventories of linear assets to develop defensible capital plans. The task is complicated by the volume and variability of data that must be considered. This joint investigation by New Castle County, DE (NCC) and Jacobs evaluates how different risk scoring formulas influence pipe asset prioritization in rehabilitation programs. The study compares the traditional risk = LOF × COF methodology, which multiplies likelihood of failure (LOF) and consequence of failure (COF) into a single value, with alternative additive approaches.
Our objectives were to evaluate how each scoring method affects rehab prioritization, 10‑year rehabilitation schedules and budgets, and identify the risk prioritization methodology best aligning with NCC’s operational goals and risk tolerance.
This study reflects ongoing collaboration between NCC and Jacobs to create risk scores and manage their sewer and stormwater assets. Analysis incorporates NCC’s GIS asset inventory, work order history, and condition assessments. Additionally, remaining useful life (RUL) estimates and cost estimates were generated, and a predictive AI model was used to flag pipes with high potential defect acceleration using an acceleration factor (AF). NCC is reviewing modeled results and comparing them with field observations, maintenance trends, and planned rehabilitation.
Three prioritization techniques were compared using identical datasets and business rules and differing prioritization techniques.
- Traditional: Single risk value calculated using LOF × COF
- Alternative 1: Refined priority value calculated as COF + 2×LOF
- Alternative 2: Refined priority value calculated as COF + 2×LOF + AF
When LOF or COF scores were missing, substitutes were made using percent‑life‑used or a peer comparison of pipes within the same basin and age range, for the same material and diameter. For comparison purposes, a 10‑year program was constructed under equal annual budget assumptions, filling each year with the highest‑priority assets.
Findings:Assessing the OptionsWhen assessing the possible prioritization outcomes, the scoring methods produced meaningfully different prioritization spreads. Traditional LOF × COF provided an irregular range of possible outcomes while the additive options created a more normalized range of outcomes. When assessing the prioritization outcomes, we focused on two hypothetical example pipes.
Comparing Outcomes Using NCC DataThe three prioritization methodologies have been applied across NCC’s full pipe inventory, generating priority scores for every asset. To validate these modeled results, NCC selected a representative subset of pipes to apply manual, engineering‑judgment‑based prioritization, drawing on institutional knowledge. This manually ranked set will serve as a ground‑truth benchmark. The study then compares the modeled priorities to both NCC’s hand‑assigned rankings and NCC’s planned rehabilitation projects to evaluate alignment, highlight discrepancies, and identify opportunities to refine the risk scoring approach.
SignificanceChoice of risk formula has direct implications on rehab prioritization, construction workload, budget stability, and long‑term system performance. For NCC, adopting COF + 2×LOF+AF offers tangible benefits: it highlights condition‑driven assets that require timely intervention while still assessing criticality, targets pipe defect degradation, and supports proactive rehabilitation. More broadly, this study provides a replicable, transparent method for utilities to reevaluate their risk modeling and integrate rehabilitation-centric prioritization into capital planning and budgeting.
