Article   |     |   24.03.2021

Big data and asset performance management.

How can big data help improve the reliability and performance of your assets?

Today, companies are collecting petabytes of data daily, in all aspects of operations from engine and drive systems, to pipes and pumps. But how do we effectively utilise this data to improve and enhance operations?

As noted in our ‘data to decision’ whitepaper <LINK NEEDED>, many organisations have a wealth of data but have concerns over quality, authentication and relevance applied to both historical and new data.

What we should realise is that the concept of ‘big data’ may be new, but the methods of analysing data aren’t; Lloyd’s Register has been collecting data and interpreting it for various assets for over 250 years. It’s the systems and tools that help us understand and analyse such vast amounts of data that are changing the industry.

These changes are defined by the four V’s: The velocity – the rate at which we gather data; its variety, including structured and unstructured data from images, audio, sensors to logs; its volume, from kilobyte to yottabyte being generated in volumes, unheard of even five years ago; and of course its veracity.

Condition monitoring

Parallel to the benefits of utilising subsea and topside technologies to enhance production, are the added constraints around maintenance and replacement of parts. Replacing faulty subsea units can result in long periods of downtime and in-turn loss in production. Above the surface things aren’t much different as faulty machinery can lead to reduced production capacity and of course, unplanned downtime. Particularly on unmanned platforms where visual checks can’t be conducted.

As we begin to explore and produce in deeper waters and harsher environments, preventive maintenance and understanding of components and equipment will be intrinsic to operations. The ability to understand analytical trends and interpreting the geological, engineering and production data will be a sure-fire way to success.

Condition monitoring; analysing a particular parameter of a condition in machinery such as vibration or temperature has significantly helped in assessing the state of components and reducing the need to for visual inspections and unplanned maintenance shutdowns.

From condition monitoring to predictive maintenance

The advent of systems and technologies to enhance these asset management methods (such as predictive maintenance), has significantly improved operations, allowing for 24/7 monitoring and the analysis of this data to be enhanced – in essence, future proofing operations.

Today, big data and the application and analysis of data in current and previous operational periods (comparing performance of previous components and operating levels), allows for inherent risks such as sudden component failure to be significantly reduced as component lifecycles are better understood.

How does data analysis enhance production?

Identifying failure events before they fully develop will not only save cost and time in operations. By applying these analyses to the FEED stages in a platform’s lifecycle will allow industry to build for resilience. Combine this with RFID – remote frequency identification and the possibilities are endless as a component’s lifetime is determined, its replacement is already in transit before the fault occurs, minimising turnaround and downtime.

Why stop there?

Combining data samples around parameters such as temperature and vibration, with production information (flow, voltage etc.) we can also begin to see how different production levels have an impact on the assets’ equipment and components.

Such information allows for a truly optimised operation as an asset is run to the most efficient production rate, whilst ensuring that equipment and components lifecycles are at maximum capacity.

The true application of big data is the ability to understand what data sample is needed, how to analyse and interpret, and of course how to apply the outcome. Without factoring each of these functions, big data may just be too much to handle.

icon-chevron-lefticon-chevron-righticon-chevron-upicon-close-circleicon-closeicon-facebookicon-globeicon-insightsicon-instagramicon-linkedinicon-logo-largeicon-logo-originalicon-logo-sgc-largeSGC Engineering LLCA Vysus Group Companyicon-menuicon-newsicon-searchicon-sectorsicon-servicesicon-statsicon-studyicon-twittericon-webinaricon-youtubevysus-leaf-optim