Independent Technical Evaluation and Modernization of a Transmission Integrity Management Risk Model: An Anonymized TIMP Risk-Assessment Case Study

Abstract

This study evaluates the technical adequacy, internal consistency, current applicability, and modernization needs of a transmission integrity management risk framework using five anonymized artifacts: Transmission Integrity Management Program (TIMP), TIMP risk-assessment methodology, TIMP risk-model workbook, a transmission class-analysis procedure, and a structure-code dictionary. The assessment combined structured document review, formula-level reverse engineering, independent recalculation of all available route and segment metrics, HCA/MCA/class-location reconciliation, operational GIS-method assimilation, descriptive diagnostics, rank-correlation analysis, and comparison with current federal and Texas requirements and peer-reviewed pipeline-risk literature. The program documents establish several sound governance principles, including conservative treatment of missing data, periodic review, subject-matter-expert review, management of change, quality assurance, and use of risk ranking to prioritize integrity activities. The implemented workbook, however, is a dimensionless relative-risk index rather than a probabilistic failure or consequence model. It contains 594 dynamic risk segments across 11 routes and 11.7998 miles of coated-steel legacy pipeline, materially narrower than the 2024 TIMP portfolio description of approximately 103.66 miles across 30 systems containing steel and high-density polyethylene. The consequence-area review found that the program defines HCA and MCA identification processes, reports 18 HCAs totaling approximately 3.59 miles, and includes 28 MCA and 20 HCA segment entries in its portfolio table; however, another program passage reports 16 HCA segments, and the supplied materials do not reconcile these counting bases. The operational GIS procedure provides a practical workflow based on separate class-location and PIR/PIC geometries, structure inventory, a cluster boundary adjustment, HCA/MCA flags, and map outputs. Its integration requires controls because 660 feet is incorrectly labeled as 220 meters, the cluster provision is described too broadly, several structure codes combine distinct class and identified-site tests, multi-family dwelling counts are not explicit, MCA roadway evidence is not operationalized, and blank values cannot distinguish No, Unknown, and Not evaluated. The workbook itself contains no MCA status, PIR/PIC, identified-site, qualifying-roadway, or HCA-method field and treats HCA as a mutually exclusive ordinal population category alongside Classes 1-3. HCA-labeled workbook segments represent 8.63% of modeled mileage but 15.20% of total risk, with a length-weighted risk density of 21.90 compared with 12.36 for Class 3; the difference is driven primarily by consequence and pipe factors rather than higher mean threat. Routes 6, 10, and 8 account for 56.9% of the workbook total-risk index. The workbook route “risk per mile” sums segment risk-density indices and is therefore strongly segmentation dependent (Pearson r = 0.925); a length-weighted aggregation changes rankings materially (Spearman rank correlation = 0.418). Multiple threat categories are effectively static, the leak-history window ends in June 2016 with no nonzero leak inputs, several within-category weights do not sum to 1.00, and an additive longitudinal factor appears in implemented PipeFAC but not in the methodology. The legacy model remains useful as a transparent triage tool for the limited asset subset it represents, and the operational GIS procedure can be retained as the evidence-production layer after controlled revision.