The power of prediction

Transferring hydrocarbons safely, efficiently, with no unexpected failures or shutdowns, and with minimal impact on the environment remains the overarching goal of pipeline projects worldwide. To achieve these aims, operators have at their disposal a range of digital and automation solutions such as pipeline modelling, augmented operations, and compressor and pump station simulations – all of them enabled by rapid, secure, high-availability communications and dedicated fibre optic backbones – that offer visualisation along the entire pipeline network, often from a single location.¹

From a technology perspective, artificial intelligence (AI), machine learning (ML) and data analytics are increasingly becoming embedded in pipeline operations, gathering vast amounts of real-time data on cathodic protection and leak detection systems¹ using wireless sensors, and making that information available quickly and securely to personnel via a distributed control system (DCS).

Asset performance management (APM) is a key element of this digital revolution. Monitoring the health and efficiency of critical pipeline assets with ‘APM 4.0’ enables operators to identify and address process upsets before they incur damage. Using quantitative risk analysis and the current state of machine health solutions, customers can leverage the data available to them to optimise operations, reduce energy consumption and emissions – and maintain competitive advantage.

In this article we will explore how APM has evolved from traditional run to failure or time-based modelling to today’s data-driven solutions, how ABB’s portfolio of ‘vendor agnostic’ APM tools can help pipeline customers monitor and enhance the physical integrity of their systems and take a closer look at APM 4.0 in action at a compressor station in Poland.

The brave new world of data driven APM

First, a brief history of APM illustrating how it has evolved over the few decades, driven by technology. Taking a bottom-up approach, most enterprises began by simply fixing equipment when it broke, which has obvious drawbacks in terms of unscheduled downtime and maintenance costs.

Then, using an automotive analogy, industry began to think ‘rather than running out of gas on the side of the road, we’ll fill up with gas each week,’ which led to a time or calendar-based perspective on APM. The drawbacks of using a schedule to assess a potential failure is that it treats every piece of machinery as being of the same importance, rather than prioritising critical assets and also uses time as the only ageing parameter.

Manufacturers began to deduce that ‘maybe breakdowns are dependent on how many kilometers or miles I drive,’ which constituted a significant leap forward into usage-based maintenance, whereby failure modes and effects analysis (FMEA) and reliability-centred maintenance (RCM) were also used to prioritise critical assets. However, this too had its shortcomings, in that it failed to leverage real-time data, instead basing maintenance on how and when an asset has broken down in the past. In real-world operations, research has shown that…

To access the full version of this article and get a free trial subscription to World Pipelines, sign up here!

Read the article online at: https://www.worldpipelines.com/special-reports/15012024/the-power-of-prediction/