HPCBS > Integrated Commissioning and Diagnostics > Monitoring and Commissioning of Existing Buildings > High-Information-Content Electrical Load Monitoring
Pier Logo High Performance Commercial Building Systems Berkeley Lab Logo
What's New | Publications | Partners | Glossary | Site Map | Links | Contact Us
Corner   Corner
Program Elements Life Cycle Tools Lighting, Envelope and Daylighting Low Energy Cooling Rule Integrated Commissioning and Diagnostics Indoor Environmental Quality
Corner   Corner
Corner   Corner

High-Information-Content Electrical Load Monitoring

Goal Statement:
To develop and deploy high-speed electrical load monitoring that can provide component-specific information from a centralized location [motor-control center; heating, ventilation and air conditioning (HVAC) service entrance; or whole building], thereby substantially reducing the cost of obtaining this information.

The method being developed is a non-intrusive load monitoring (NILM) system intended to simplify the monitoring of electrical loads or appliances on a building electrical system or subsystem by using a single point measurement instead of measurements for each load of interest.

Conventional vs. Non-Intrusive Monitoring Systems
 
Conventional vs. Non-Intrusive Monitoring Systems

Results:
NILM tests have been conducted at eight sites: the Energy Resource Station on the Iowa Energy Center campus, a Kentucky Fried Chicken fast-food restaurant in Massachusetts, the 160 Sansome building in San Francisco, two schools in the West Contra Costa CA Unified School District, and three Los Angeles (LA) County CA buildings (Internal Services Division, Children's Court, and Communications Building).

Additional activities include:

  • Enabled high-speed communication lines and internet data access for each NILM location (except LA County buildings where manual data download is required).
     
  • Developed algorithms to track variable-speed drive (VSD) loads (fans and pumps). This was a major step, making it possible to track time-varying loads from harmonics generated by VSDs and no other equipment.
     
  • Classified approximately 85 percent of all events at the fast-food restaurant.
     
  • Learned that successful estimation of component-level energy consumption depends strongly on minimizing misclassifications, leading us to explore a rule-based approach to assigning classifications.
     
  • Upgraded the NILM voltage-measurement hardware with a new circuit board that is designed to reduce electrical noise. This circuit board will be available as part of Massachusetts Institute of Technology's (MIT's) effort to market NILM.
     
  • Upgraded NILM software to enhance the speed of computations and allow concurrent result display and data transfer.
NILM Software Architecture Block Diagram
 
NILM Software Architecture Block Diagram

Downloads and Links:

Contact: David Claridge, Texas A&M University (TAMU), (979) 845-1280 or Leslie Norford, Massachusetts Institute of Technology (MIT), (617) 253-8797

Corner   Corner
Last updated
May 27, 2003
Return to top | HPCBS Home