Data-driven Performance Analytics, Health Monitoring and Optimization for Chiller Plants

Technology Overview

Water-cooled chillers often consume between 40-60% of the total energy used in commercial buildings. According to the 2017 Building and Construction Authorities’ (BCA) Building Energy Benchmarking Report, retrofitting and performance improvement measurements on chillers increased plant efficiencies by 42% on average, from 1.1 kW/RT to 0.63 kW/RT.

A Singapore company has developed a technology solution to enabling prescriptive/predictive maintenance as well as continuous optimization of air-conditioning systems.

The system works by tapping on Building Management System (BMS) data to perform analytics to extract insights on chiller plant performance. Prior to deployment, the system scans through the historical data (at least 6 months data is required) and performs correlation analysis to ensure data quality. The end user will be notified about potential faults in the sensor measurements.  

The primary targeted end users are commercial and manufacturing buildings who are operating chillers of 5 years old or greater, and with BMS which is not capable of providing intelligent performance analytics. The system aims to utilize data-driven analytics approach to extend the life of existing chillers by helping the facility engineers to operate the chillers at the best possible efficiency. This helps building owners to prevent or delay the need for immediately retrofitting entire chiller systems.

The technology provider is seeking partnerships with BMS vendors, chiller plant system integrators and maintenance organisations. Potential partner companies from France are welcome to submit their interest to co-innovate under the France-Singapore Joint Innovation Call. Eligibility of participating French companies for this call can be found in this document.

Technology Features & Specifications

The developed technology enables data collected from multiple software systems (e.g. Building Management System (BMS), Power Quality Monitoring System (PQMS)) to undergo intelligent data processing which are visualized through a web-based dashboard. Useful insights are derived from the trends and patterns extracted from the data using intelligent algorithms. The visualizations and insights are beneficial to understand the correlation between existing chiller setpoints, operation, changeover schedules, and energy consumption. Existing chiller plant configurations can be tuned by the facility management team based on these insights to improve chiller performance.

Patterns extracted from historical data can be used to identify the optimal loading pattern for the individual chillers and understand the efficiency of system components (e.g. chillers and water pumps). This information is used to optimize chiller sequencing schedule (i.e. distribute the cooling amongst chillers) and chiller set points. The intelligent algorithms help identify anomalies in the operations in order to identify bottlenecks and unwanted sources for power dissipation. The detected system anomalies are displayed on the dashboard to alert the facility management team, thus minimizing reaction time to take precaution measures.

Some of the examples of recommended actions to the end user are:

  • Optimal load regions for each chiller based on historical data
  • Alarm when chiller is operated outside suggested optimal load region
  • Alarm in the events of threshold breach (operation/health anomalies)
  • Optimal scheduling based on performance characteristics of chillers.

Potential Applications

The technology provides integration with Facility Management Systems that is applicable to the following industries and provides for optimization of all centralized air conditioning facilities.

  • Production Industries
  • Factories
  • Automobiles
  • Construction
  • Office Properties
  • Shopping Malls
  • Educational Institutions

Market Trends and Opportunities

According Singapore BCA Building Energy Bench Marking Report, 532 Commercial Buildings with annual operating cost for chiller plants is about 720 Million SGD. 30% savings can be anticipated with the implementation of such optimization solutions.

Customer Benefits

  • The chiller plant operation set points generally follow design conditions and manually tuned based on domain knowledge.
  • Predictive analytics using historical data of chiller plant operations helps optimize energy consumption. 
  • Data collected from the Facility Management Systems (FMS) can be utilized effectively using state-of-the-art data analytics and artificial intelligence models to improve chiller plant processes and operations. Some of these measures include active control of chiller set points, optimal load balancing, and flow rate control. 

Make an Enquiry