Pipeline Integrity Management and Data Science Blog

Risk Modeling Combined with Machine Learning Supercharges Integrity Analysis

Many pipeline operators have separate departments for managing integrity data and risk analysis. This detached approach makes it challenging to form a complete picture of pipeline integrity.

At OneBridge, we’re dedicated to elevating integrity management, which is why we’ve partnered with C-FER Technologies to embed sophisticated risk models into our data management software.

By combining machine learning and risk analysis into a single platform, operators advance towards a truly comprehensive integrity management program.

C-FER: industry-leading, quantitative risk models

A pioneer in the development of pipeline risk modeling, C-FER Technologies helps the oil and gas industry “improve safety, operational efficiency and environmental performance.” Their innovative risk models have been validated by companies worldwide and cited by PHMSA to support pipeline risk management decisions.

According to ASME standard B31.8S, there are nine threats to pipeline integrity; C-FER’s models address each threat using one of the following two methods to estimate probability of failure (PoF).

Historical-based – This technique uses industry failure rates along with attribute-based adjustment factors to determine the PoF of a particular threat. A historical-based approach applies to: construction defects, equipment failures, incorrect operation & maintenance and seismic hazards.

Structural Reliability – This technique uses engineering assessment models which incorporate pipeline condition data to derive a PoF. Threats evaluated using this approach include: manufacturing defects, equipment impact, external corrosion, internal corrosion and stress corrosion cracking.

C-FER’s analysis is quantitative rather than qualitative, as it uses data to estimate uncertainty rather than subjective input (such as SME interpretation).

However, quantitative analysis is only as good as the information that feeds into it. A platform that confirms data quality, such as OneBridge’s Cognitive Integrity Management (CIM), is ideal for supporting C-FER risk models.

CIM: machine learning and data science

Cognitive Integrity Management makes it possible to validate, normalize and align a wide range of datasets, thereby laying the groundwork for through pipeline risk analysis.

Our cloud-based platform leverages machine learning to ingest data using algorithms that constantly improve and “learn”. The more data the program encounters, the faster and more efficient it becomes. Automated checks ensure that the integrity data ingested into your system is correct and up-to-date.

Cloud computing allows CIM to manage large quantities of information. The software handles data uptake that could traditionally take an operator weeks, or even months, in just a few hours. This automated process frees up integrity engineers to spend more time making decisions based on the results of analysis and risk modeling.

CIM combines datasets from different departments, vendors, sources and formats into a single platform to provide exceptional data visibility and accessibility. The result is a verified pool of integrity data—a critical ingredient for quantitative risk models.

Combining data management and analysis for superior pipeline integrity

Breaking through data silos is an important step towards creating a streamlined integrity management process. At OneBridge, we’ve solved the problem of silos and we’ve gone even further, incorporating C-FER risk modeling into our data management software.

CIM’s advanced machine learning algorithms thoroughly verify data before feeding it to C-FER’s industry-leading risk analysis. This approach delivers the most accurate integrity analysis and PoF possible.

Within CIM is a convenient Microsoft Power Business Intelligence dashboard, which displays:

  • Real-time analytical results
  • Risk for each pipeline segment, on a rolling-mile basis
  • PoF across all nine pipeline threats

Invaluable to integrity engineers is CIM’s ability to roll up and drill down into information related to specific PoF results, as well as zero in on particular pipeline segments. This data-informed approach strengthens an operator’s confidence when choosing dig locations or prioritizing in-line inspections.

As an enterprise-wide software solution, Cognitive Integrity Management puts advanced machine learning and industry-validated risk modeling together to give your entire team a comprehensive view of pipeline threats.

Setting a new industry standard for pipeline integrity

“We believe our collaborative solution, combining CIM data management capability and machine learning with C‑FER’s risk algorithms, will deliver the most advanced integrity and quantitative risk management solution available for O&G pipeline operators today.” ~Mathew Bussière, Manager of Pipeline Integrity & Operations, C-FER

When there’s separation between the steps in an integrity process it leaves room for errors, slows the results of risk analysis and creates gaps of knowledge. OneBridge’s CIM solves these problems using machine learning to validate integrity data and connect it directly to C-FER’s advanced risk models.

Real-time access to trusted PoF results advances your pipeline integrity team to the forefront of the industry, protects the public and the environment and ensures the safety of valuable pipeline infrastructure.

Contact OneBridge to find out how you can advance your integrity program to the leading edge using CIM combined with C-FER risk modeling.