Using digital twins and data analytics to make informed operational decisions – a centrifugal compressor case study
Andy Jones A B and Denis Marshment A BA Worley Digital, Perth, Australia.
B Corresponding authors. Email: andy.jones@worley.com; denis.marshment@worley.com
The APPEA Journal 61(2) 544-547 https://doi.org/10.1071/AJ20047
Accepted: 25 March 2021 Published: 2 July 2021
Abstract
An Australian operator with an expansive portfolio of liquefied natural gas, pipeline gas and oil assets wanted to gain greater visibility of how their compressor fleet was operating. To help the operator gain a greater understanding of the performance of their assets, Worley Digital built digital twins of 17 of their compressors. This involved detailed first principle calculations supported by hundreds of thousands of offline simulations. This created the thermodynamic correlations necessary to accurately model the most important operating statistics surrounding each stage of those assets. We then used dimensional analysis and data analytics to benchmark the operation of these machines. This represents a hybrid approach; first principles engineering and simulation coupled with data science and analytics. A suite of tools was then delivered to the operator in their preferred real-time data analytics and visualisation platform, providing them with information about the efficiency of their compressors. They were then able to perform net present value calculations based on different operating and maintenance strategies. This gave them the detailed information they needed in order to be able to make more informed decisions. An example of this might be a decision to bring a scheduled maintenance activity forward based on the knowledge that there are net operational and financial incentives to do so. The resulting tool has proved extremely valuable, quantifying some business-critical metrics that have led to several financially impactful operational decisions.
Keywords: digital twin, digitalisation, data science, transformation, centrifugal, compressor, optimisation, degradation, monitoring, data analytics, operational intelligence, online real-time monitoring, coal seam gas, gas processing facility, compression train, dimensionless analysis, polytropic head, polytropic efficiency, head losses, capacity maximisation, fouling, liquid carryover, fuel emissions calculations.
Andy Jones is currently the Digital Operations Lead and Digital Twin Subject Matter Expert (SME) for Worley Digital and is a process engineer and data scientist with 27 years of experience in the oil and gas, mining, water, chemical and nuclear sectors, in varying locations worldwide. He is an industry expert in the specification and development of operational digital twins. Most recently, he has been combining the greatest strengths of traditional first principles engineering and simulation-based applications with best of the data science approach including machine learning and artificial intelligence. He is also pioneering augmented reality (AR) and virtual reality (VR) customer solutions. He has a long history of utilising best available technology to enhance profitability and reduce downtime. He graduated from the University of Nottingham, United Kingdom in 1994 with an BSc (Hons) in Chemical Engineering and possesses a graduate certificate in data analytics from Deakin University. |
Denis Marshment leads the Worley Data Science Customer Solutions practice for Worley. He brings a wealth of experience in the application of digital technologies to the energy and resources sector including digital twins, predictive analytics, asset performance management, field mobility, forecasting and optimisation. Denis is a highly accomplished consultant who has led large-scale implementations of EAM integrated Asset Performance Management and Mobility solutions including the management of large multidisciplinary project teams tasked with technical implementation, integration and content development Denis also has over 20 years of hands-on experience in building and implementing maintenance and reliability programmes. |
