American Business Conferences 2021 virtual Pipeline Leak Detection Congress

Crestwood Midstream, "small leaks still challenging". Williams on stress corrosion cracking. Siemens’ Pipeline 4.0. ProFlex on offshore leak detection. Leak detection in Gulf of Mexico. Black Bear Transmission’s best practices for monitoring. Marathon Pipeline on leak detection regulations. Flyscan’s hyperspectral airborne survey. Case histories from Kairos Aerospace. HiFi Engineering’s fiber optic monitoring. Intelliview’s AI/smart camera.

Speaking at the American Business Conferences 2021 virtual Pipeline Leak Detection Congress, Reagan Nguyen (Crestwood Midstream) presented a ‘pro-active approach’ to pipeline leak detection. While it might seem obvious, the best leak prevention is having no leaks at all! The approach includes ‘call before you dig’ advocacy, keeping landowners on board and maintaining up to date data on your assets. Threat assessments, leveraging the API 1160 methodology for hazardous liquids and ASME B31.8S for natural gas, will lead to appropriate monitoring and mitigation strategies.
Today’s leak detection systems are challenged by small leaks below the sensitivity of the system. Identification and isolation of a leak can take time. Advanced systems can perform well but they are expensive to install, test and maintain. They do not ‘prevent’, they help mitigate.

Amy Shank (Williams) presented on stress corrosion cracking, a ‘time-dependent’ threat and a hard-to-model issue. SCC is generally found withing 20 miles of compressor stations where heat and stress is highest. 2015 was a wake-up call for industry, and PHMSA reporting has shown a significant subsequent increase in crack tool mileage. Shank gave a shout-out to Rosen’s EMAT detector, capable of finding smaller cracks than hydro tests.

Nico Jansen van Rensburg (Siemens Energy) asked ‘what does digital technology bring to the table?’ Siemens Pipeline 4.0 promises an ‘integrated approach to optimizing midstream assets’. Pipeline 4.0 embraces a slug of digital goodies, from the digital twin, AI, wireless, IoT and cyber security. So why aren’t all pipelines @4.0? Digital transformation faces barriers from the ‘feature shock’ of maintaining complex technology, reliability issues, and the new skills required. The solution is not just technology but requires a paradigm shift in the way operators, technology providers and service personnel interact.One P4.0 example is Siemens’ IoT-based ‘ spontaneous leak detection as-a-service’, a solution that embeds leak detection technology from ProFlex.

In a follow-up presentation, ProFlex Techologies’ Scott Bauer drilled down into the partnership that combines PFT’s Pipe-Safe solution with Siemens’ IoT. The system leverages remote pressure monitoring and artificial intelligence to locate leaks with a 20-50 foot accuracy. The system uses a ‘modified negative pressure wave’ approach. A burst creates a back-propagating pressure wave that can be caught with acoustic monitoring at monitor stations. Edge-processed data is sent to the SE cloud for further analysis to pinpoint the rupture. SMS/emails are sent to key personnel with the location. The SLDS* system is on show at Siemens’ closed loop demonstrator in Houston.
* Spontaneous leak detection system.

Stuart Mitchell (ProFlex) presented a case study of an offshore leak detection system deployed on EnVen Energy’s ‘Lobster’ Gulf of Mexico platform. The Siemens/ProFlex negative pressure monitor described above was installed on the platform’s subsea pipelines with the SLDS computer on the platform. Testing involved short duration simulated leaks over a two week period. A millisecond sampling rate produced some 4.5 gigabytes of data. An ML model was trained on the data and deployed on four risers in 2021, integrated with the platform’s control system.

Ronda Louderman (Black Bear Transmission) outlined best practices in effective pipeline monitoring and material validation. Over time, regulation has evolved from a ‘best efforts’ stance to a more prescriptive approach requiring more technical expertise, contractor qualification and attention to field personnel’s attention and attitude. Inspectors want to see ‘safety as a priority, transparency and honesty’. This may include owning-up to past issues, along with mitigation plans in place.
Operators need to develop different plans for new-builds, legacy assets and acquisitions. Louderman outlined a methodology spanning assessment of what needs to be captured and a plan of action. The latter needs to prioritize high consequence areas, but also lines with a history of leaks, maintenance issues and third party strikes. Lines with missing documentation and acquisitions without due diligence need special attention. Senior management needs to be kept in the loop on audits and new regulations. Louderman suggested inviting management along to conferences or at least providing a post-event write-up. She concluded with some pointers to resources including regulation status, the PHMSA community toolbox and the Federal Register.

Jason Dalton (Marathon Pipe Line) reported on the current situation of leak detection requirements. PHMSA’s 49.CFR 195.444 states that operators must deploy a leak detection system on HCA*s. Systems are evaluated in terms of their effectiveness and how quickly a leak can be addressed. Alongside regulations, the public perception is one of near perfect leak detection capability and ‘immediate’ mitigation. Dalton suggested that the API RP 1175 could form the basis of a leak detection program. But such programs can be challenging, especially for smaller operators, as they are resource-intensive. Expectation management is needed, as not all leaks can be detected instantly. Detection timeframes can be as much as days or weeks. In summary, while the midstream has better leak detection than ever before, implementation is not straightforward and industry benchmarking is difficult due to operator hesitancy to share leak detection system performance.
* High consequence areas.

In a follow-up presentation, Dalton drilled-down into the use of pattern recognition to characterize leaks from pressure data. He advocates training ML models to recognize all operational events, not just leaks, from the data. Models should predict what behavior will result from an operation and deviations from expected behaviors should trigger an alarm. Historical data on slack line events (degassing) and pump performance will provide baseline records against which future, possibly anomalous, behavior can be recognized. This kind of pattern recognition does not require ‘buzzwords’ or fancy ML/AI supercomputers*. It all boils down to understanding how your system operates and empowering the people closest to the action.
* Similar real-time pattern recognition software can run on a Raspberry Pi. See for instance ‘BP Lower 48 - failure analysis on the Raspberry Pi’.

Eric Bergeron presented Flyscan’s ‘next generation’ airborne right-of-way inspection and slow leak detection technology. Initial Flyscan development was funded by PHMSA. Equity funding to date totals $11 million, notably from Enbridge. Flyscan’s dual mission is to catch leaks and threats early. The UV laser Raman detector is said to detect a single cup of crude oil mixed with topsoil from a distance of 500 feet. Three RGB cameras feed onboard real-time pattern recognition software that distinguished facilities and threats. A hyperspectral camera provides further leak detection capability.

Steve Deiker reported on some case histories from Kairos Aerospace’s pipeline emissions monitoring program. Kairos performs large-scale repeat methane monitoring over an area of interest. Data is exclusive to the operator. A combination of spectral imagery and optical cameras feeds a data pipeline into the cloud. Most central and western US petroleum basins have been surveyed in 2021. Leveraging the survey results has enabled operators to eliminate 18.3bcf of methane emissions, with ‘a greater GHG impact than Tesla!’ The surveys have shown huge variations in per-mile of pipe emissions between operators. Using Kairos’ technology, operator DCP Midstream received an environmental excellence award from the GPA Midstream Association.

Steven Koles (Hifi Engineering) presented a different approach to monitoring, combining high fidelity fiber optic monitoring with machine learning. Fiber deployed along the length of a pipeline beats conventional point monitors, providing a continuous measure of pressure, temperature and other variables along the whole length of the line. HDS, HiFi’s distributed sensor offering combines specialized fiber, big data acquisition systems and an AI/ML data processing back end. The system can detect leaks and encroachments (landslips) along with normal events such as pigging runs. HiFi is also working on flow rate estimation from the fiber data. In conclusion, fiber can be an important element of ESG strategies that align with API 1175.

Also of note at the event were, Intelliview’s AI/smart camera system as used by Chevron and Enbridge, Atmos International’s leak detection solution and Pipeline Project Services’ drone methane mapping system that feeds data into the Esri GIS cloud.

More from the ABC Congress home page.

Click here to comment on this article

Click here to view this article in context on a desktop

© Oil IT Journal - all rights reserved.