I’m after training that tightens integrity decisions and risk mitigation, not just slides on standards. After a March MFL run on a 24-in wet gas trunkline, our dig-to-hit ratio was 0.6 and reassessment interval modeling felt shaky — has any API 1160/B31.8S-focused course or POD/POI analytics workshop meaningfully improved your maintenance prioritization and repair criteria?
Kiefner’s ILI Data Analysis plus an API 1160 session that “recalibrates POD/POI with your own dig data” moved the needle for us — by binning POD by wall thickness/coating and re‑ranking with Modified B31G + interaction rules, our next 24‑in wet‑gas MFL went from about 0.6 to about 1.3 dig-to-hit, like finally tuning a fuzzy radio. If you don’t have enough truth digs, try PRCI’s ILI‑U lab to build ROC curves and feed them into your B31.8S reassessment model: https://prci.org/research/ILI. Do you have >30 confirmed digs to fit ROC by joint type?
Building on @kwhite11: ROSEN’s ILI Analytics masterclass paired with an API 1163 validation lab — where we built threat-specific “unity plots” and Bayesian POD/POI updates from our digs — took our wet‑gas MFL dig‑to‑hit from about 0.7 to 1.3 (felt like swapping rose‑colored glasses for readers). Concrete step: segment POD by wall thickness/coating/girth‑weld proximity, correct sizing bias, then re‑run your 1160/B31.8S risk with updated failure frequencies. What tool vendor and tolerance class was your March run?
We got real lift after a hands‑on API 1163 validation lab where we stratified by seam type and wall, corrected a consistent −1.5 mm oversize in the March MFL, and locked a 20‑feature‑per‑stratum validation plan; hit rate jumped and repair picks got cleaner. > correct sizing bias, then re‑run your 1160/B31.8S risk with updated failure frequencies. What tool vendor and tolerance class was — agree, and the trick was pushing the vendor for raw call‑vs‑depth scatter and error distributions so we could update our model, not just accept catalog tolerances. @amartin203 did you see the same when you split by coating/vintage?
In an API 1163 validation lab like @kwhite11 noted, we built a simple “hit matrix” for our 24‑in line by wall/seam and applied a systematic sizing‑bias correction to the March MFL, which nudged our dig‑to‑hit from about 0.6 to about 0.85 and made the reassessment interval less jumpy. Caveat: it only sticks if you’ve got enough verified digs per bucket; otherwise we pooled adjacent buckets for stability.
, I’ve been there — a 0.6 hit rate after a March MFL on a 24‑in wet gas line stung for us too… The one course that tightened calls was C‑FER’s Pipeline Defect Assessment; the practical bit was re‑basing repair criteria with PCORRC using our UT wall distribution and segment‑specific consequences instead of a flat % metal loss. Small caveat: if your liquids loading swings, bake in CO2 rate uncertainty or you’ll overextend your reassessment — do you have any coupon or probe data to anchor growth, @kwhite11?