Physics Plus AI: A New Era for Pipeline Leak Detection

Replace-in-kind fixes are supposed to be the easy path back to normal operation. But once a change crosses a technical or regulatory threshold — like a decades-old foundation rating — the schedule, not the engineering, becomes the real constraint. Traci Purdum, Chemical Processing's EIC, reads Andrew Sloley's latest Plant InSites column.

In October’s Distilled News, executives at the CIEX R&D Innovation Summit explain how AmSty, Hexion and 3M are using AI and operational excellence to drive innovation. Also: BASF’s possible takeover of Evonik, plant closures from Goodyear and Westlake, and Sumitomo Chemical’s production-scale bio-based LCP.

How AI is reshaping pipeline leak detection? Chemical Processing Editor-in-Chief Traci Purdum talks with KROHNE's Maximilian Ihring and Daniel Vogt to answer that question. The following conversation, edited for clarity, examines how conventional systems rely on static hydraulic models that lose accuracy as conditions change. KROHNE's PipePatrol NEO layers supervised machine learning on proven physics-based models, improving sensitivity and reducing costly false alarms while keeping physical laws as guardrails. Also discussed: challenges unique to chemical pipelines, field results showing major performance gains and the importance of data quality. The bottom line — human operators must keep authority over critical decisions as AI evolves. This episode is sponsored by KROHNE

In this episode of Chemical Processing's Distilled podcast, Editor-in-Chief Traci Purdum talks with water treatment expert Brad Buecker about the hidden risks of remote cooling systems scattered around large plants. Buecker explains how scale formation and microbiological fouling quietly degrade heat transfer, why cooling towers concentrate impurities and complicate chemistry, and why monthly vendor visits often aren't enough. He stresses online instrumentation, reliable biocide feed and trained staff who act on data. Buecker also shares cautionary tales, including a boiler kept online to earn $30,000 that required more than $2 million in repairs, and his family's connection to Legionnaires' disease.

In a chemical facility, choosing the person who feels safest is understandable. These are high-risk environments where poor decisions can affect production, people and plant. Nobody wants to experiment with leadership when the consequences of getting it wrong can be dire. The problem is that competence is not the same as readiness, and familiarity is not the same as low risk. When succession decisions repeatedly favor the people who already look, sound and operate like current leaders, organizations may believe they are reducing risk. In reality, they may be reproducing it. Listen to this In Case You Missed It episode of Lauren Neal's workforce column. You can read the column here.

Schneider Electric's Thomas Kwan explains why chemical plants should map their loads before buying software, and why "the optimizer decides" isn't an operating philosophy. Jonathan Katz, executive editor of Chemical Processing, recently spoke with Thomas Kwan, an energy transition expert at the Schneider Electric Sustainability Research Institute, about his most recent Energy Saver column for Chemical Processing, titled "The Flexible Chemical Plant: Scheduling Energy Use Without Sacrificing Throughput." Kwan, whose official title at Schneider is global vice president of strategic innovation and industrial ecosystems, wrote about how electrified compressors, separations and utilities can behave more like power brokers whose price refreshes every hour. He noted that plants running through a price spike are paying for steady-state, continuous operation they may not need, and he suggested that, to save energy, they must consider operational flexibility itself an energy asset. Kwan asked readers whether they knew which loads in their operation can move by minutes or hours, what that movement is worth under their energy tariff, and when product commitments, catalyst constraints or safety margins erase the benefit before it reaches the ledger.

In September's Distilled News, a federal appeals court strikes down the EPA's methylene chloride ban, three chemical makers agree to pay $455 million over PFAS, and BASF and Kemira look at new production.

In this episode of the Chemical Processing Distilled Podcast, Brad Buecker of SAMCO Technologies joins editor-in-chief Traci Purdum to dig deeper into reverse osmosis for boiler makeup water. Buecker contrasts RO with sodium softening, explaining how RO removes up to 99% of dissolved ions, cutting corrosion, scaling and blowdown compared to softening alone. He stresses that raw water quality and pretreatment reliability — not snapshot testing — determine RO success and warns against microbiological fouling, chlorine damage and cutting corners on monitoring. Buecker also covers normalization programs, cleaning triggers and design considerations for plants weighing a switch from softening to RO.

An Eastman Chemical veteran-turned-leadership coach explains why technical excellence doesn't translate to leadership skill — and how organizations can close the "broken rung" that pushes women out of the pipeline right after their first promotion.

ICIS's Al Greenwood explains why ethane-fed commodity producers are cashing in while specialty chemical makers struggle to pass on higher costs.  Read the full article.