P2-7. Enabling Tools > Outcomes

Technical Outcomes:

• Phase 1: Typical HVAC system and component faults were embedded in the VCBT simulation engine and the FDD Test Shell was interfaced to the VCBT. Adding the virtual faults allowed testing of FDD tools. Development of the communication tools between the VCBT and FDD Test Shell was successful. This allowed BACnet compatible building control systems to provide data to the Test Shell, which in turn will decrease development time and reliability testing of FDD tools.

• Phase 2: Data acquisition capability was added to the FDD Test Shell and off-line testing with data from real buildings was completed, but on-line testing was not completed. The BACnet Data Source (BDS) was developed to extend the FDD Test Shell capabilities by adding a component that can exchange messages with BACnet controllers as a way to collect the data needed by the FDD tool. The data can be made available in real time or stored in a file for later retrieval and analysis. Off-line testing was completed with data sets from over a half dozen buildings. On-line testing was not completed due to difficulty in obtaining a demonstration partner able and willing to allow real time access.

• Phase 3: The performance of the WBD/OAE diagnostician in the blind tests was analyzed for low, normal, and high sensitivity. Using the OAE at its low sensitivity setting, 8 of the 15 faults were correctly identified, no false positives were obtained, and 7 faults were not detected (false negatives). For the normal sensitivity setting, 11 of the 15 faults were correctly detected, no false positives were reported, and 4 faults were not detected. With the high sensitivity setting, 10 of the 15 faults were correctly identified, 2 false positives were reported, and 3 faults were not detected (false negatives). In each of the three sensitivity test sets, two of the false negatives were due to the conditions masking the faults so that even though the fault existed in the model, conditions (e.g., temperatures) were not appropriate for system performance to be affected by the presence of the faults.

• Phase 3: In three of the four blind trials on the WBD/WBE diagnostician, it correctly identified anomalous energy use caused by a fault. The single false negative (undetected problem) was due to the small magnitude of the fault (this fault had an energy impact below the WBE reporting threshold of $1/day).

Market Outcomes:

The VCBT and FDD Test Shell, as enhanced during this project, will cut development time and improve the robustness of FDD methods deployed in building control systems.

Back to Previous Page