for Predictive Analytics in Electrical Utilities
FEATURED PAPER
By Godwin Ndem and Dr. Khamis AI-Romaimi
Faculty of Business and Technology
School of Computer Science and Engineering
University of Sunderland
Sunderland, UK
Abstract
Electricity suppliers are putting money into condition monitoring and data analytics so that they can foresee equipment failure rather than having to respond to it. Although there is a great deal of technical literature available on sensing, analytics and modelling, there is much less advice concerning how to handle such investments as projects and programmes that require changes to decisions, working practices and organisational habits. This article puts forward a programme and project management framework for implementing predictive analytics initiatives in electricity suppliers. At the project level, the framework consists of five phases (align and prioritise, prepare the data and sensing, develop and validate, integrate and deploy, and operate and scale), each phase being concluded by a stage gate which poses a specific question. At the programme level, a coordinating layer oversees a series of projects, the shared capabilities and the benefits across different types of equipment. Six cross-cutting disciplines apply to both levels. The framework is based on recent reviews of the technical field, a recent study into predictive analytics for substation equipment, ISO 21502:2020 on project management, ISO 21503:2022 on programme management and the PMBOK Guide, Seventh Edition. It is a conceptual framework and has not yet been tested empirically. Its aim is to provide sponsors, programme managers, project managers and engineers with a common structure for planning, governing and maintaining such initiatives.
Keywords: programme management; project management; electrical utilities; stage gates; benefits management; governance; data governance; predictive analytics
- Introduction
Power transformers, circuit breakers, switchgear, cables and protection equipment are amongst the largest, most robust and capital intensive of electrical utility assets. Failure of any of these can result in supply outages, dangerous and costly situations and extensive repairs. Ultimately, power utilities are shifting from a reactive, scheduled servicing methodology to one based on condition and predictor maintenance supported by sensing and analytical technologies.
The technical aspects of this transition have been well documented. Surveys and reviews have listed the methods, the enabling technologies and the hurdles, generally grouped into what is called predictive maintenance (Krupitzer et al., 2020; Zonta et al., 2020; Achouch et al., 2022). The management aspects have been less analysed. Despite the widespread presence of diagnostic technologies, many utilities continue to use manual inspections, portable dissolved gas analyses, vibration analysis, infrared thermography, and static nameplate data to determine the state of equipment, according to Ghosh and Mittal (2026). In a survey of 24 large US electric utilities, only two had previously tried federated learning; the authors attribute low adoption partly to the conventional separation of operating and technical roles in power utilities and partly to the computer skills involved in adapting existing infrastructure to a new approach.
So it’s organisational as well as algorithmic. If a predictive analytics project is to deliver value, it must improve the decisions made from the model’s output. That can only happen if the entire data set, system, person and process ecosystem is aligned. That means the project team needs to be managing a programme. A project might be able to demonstrate its model on one equipment class, but taking that model to the whole of a utility is a programme, with shared data, shared skills and value emerging only when multiple projects come together.
The article provides a management framework for these types of initiatives within electrical utilities and makes three contributions. Firstly, it defines the features that set those initiatives apart from traditional capital projects. Secondly, it defines a five-phase project framework with phase gates and deliverables for each phase. Thirdly, it introduces a programme level as well as six cross-cutting disciplines that link the technical work to business value.
More…
To read entire paper, click here
How to cite this paper: Ndem, G., Al-Romaimi, K. (2026). A Programme and Project Management Framework for Predictive Analytics in Electrical Utilities; PM World Journal, Vol. XV, Issue XI, November. Available online at https://pmworldjournal.com/wp-content/uploads/2026/11/pmwj170-Nov2026-Ndem-Al-Romaimi-PPM-Framework-for-Elect-Utilities-Predictive-Analytics.pdf
About the Authors
Godwin Etia Ndem
Sunderland, United Kingdom
Godwin Etia Ndem is a project management and engineering professional with over six years of experience in project planning, monitoring, execution, quality management, risk management, and asset management. He holds an HND in Electrical/Electronic Engineering from the Institute of Management and Technology (IMT), Enugu, and a PGD in Public Administration from the University of Uyo, Nigeria. He is currently pursuing an MSc in Project Management at the University of Sunderland, UK. He also has training in HSE, Fire Watch, and Construction Safety, with a strong interest in safe, efficient, and quality project delivery. He can be contacted at goddyndemnigeria@gmail.com
Dr Khamis Al-Romaimi
Sunderland, United Kingdom
Dr Khamis Al-Romaimi completed his PhD from University of Sunderland. He received MSc degree in Electrical Utility Engineering from Curtin University of Technology, Australia. He received Bachelor of engineering in Automatic and Control System Engineering from University of Sheffield, and he received his Mechanical and Manufacturing Engineering from University of Manchester-Stockport. His employment experience in electricity and water sectors in operation, maintenance, data analysis, strategic planning, program management, asset planning, economics, system studies, design, projects management, load dispatch centre, risk management, asset management and business continuity development. He is currently lecturer in Engineering. He can be contacted at Khamis.al-romaimi@sunderland.ac.uk
