Logo image
Development of a boiler monitoring system for predictive maintenance of boiler tube failures
Journal article   Open access

Development of a boiler monitoring system for predictive maintenance of boiler tube failures

Garikai Tawanda Marangwanda and Innocent Musonda
JOURNAL OF ENERGY RESOURCES TECHNOLOGY PART A-SUSTAINABLE AND RENEWABLE ENERGY, Vol.2(7), pp.1-36
01/07/2026
Handle:
https://hdl.handle.net/10210/520213

Abstract

Energy & Fuels Science & Technology Technology
Unplanned outages in coal-fired power stations are significantly driven by premature boiler tube failures, contributing to over 30% of forced shutdowns across certain utility fleets. This article presents a case study on the design, development, and pilot implementation of a predictive boiler monitoring system targeting supercritical boiler reliability. The boiler monitoring system showcases a transformative path toward predictive maintenance and operational efficiency in legacy power infrastructure. The system integrates a 3D interactive model, digital shadow, and predictive analytics to monitor and forecast boiler tube degradation. Over 240,000 inspection data points across multiple boiler circuits were digitized and analyzed with the goal of determining the remaining wall thickness and time to minimum allowable limit for the boiler tube walls. A predictive model based on the power law regression was trained successfully using historical data measured during three outages. The monitoring system was able to forecast boiler tube wall thickness with success. At 1 year, the deviation was 2.5%, which increased to 8.1% for 9 years, a value that is acceptable based on ASME Boiler and Pressure Vessel standards. As such, by knowing the expected boiler tube wall thickness, maintenance planning is enhanced through reduction of expected outage time of 4 weeks per year to address repairs. Another benefit is the automated generation of quality assurance, quality control, and cutting instruction reports.
pdf
Research (68)1.17 MBDownloadView
Open Access CC BY V4.0
url
https://doi.org/10.1115/1.4071543View
Published (Version of record) Open

Metrics

1 Record Views

Details

Logo image