Environmental Energy Technologies Division News

Environmental Energy Technologies Division News
  • EETD News Home
  • Back Issues
  • Subscribe to EETD News
  • Print

Diagnostics for Building Commissioning and Operations

Many buildings fail to perform as well as expected because of problems that arise at various stages of the life-cycle, from design planning to operation. These problems originate in different parts of the building, including the envelope, the HVAC system and the lighting system. Consequences include increased energy costs, occupant discomfort, poor productivity, health problems and higher maintenance costs. Examples include leaking ducts, stuck dampers and disabled or poorly tuned control loops. Such faults can be detected either by active testing, for example as part of building commissioning, or by passive monitoring during day-to-day operation.

Fault detection involves a determination that the observed performance of a system or component differs significantly from the expected performance. A human operator can make this determination by inspecting and analyzing measurements from a building, for example temperatures and flow rates in an air-conditioning system, or the process can be automated using computer- based methods. Most computer-based fault detection methods use a model of correct or intended operation to predict the expected behavior, as shown in Figure 1. If these models can be configured using design information or manufacturers' data, they can be used during commissioning to verify that the design intent has been achieved and that this level of performance is maintained during operation. Once a fault has been detected, the next step is to diagnose its cause by combining information from different sensors and operating conditions. The final step is to estimate the cost and benefits of fixing the fault in order to estimate the urgency of effecting a repair.

Computer managment of data generated during the life-cycle of a commercial building.

Figure 1. A model-based fault detection scheme

Information Monitoring and Diagnostic System

LBNL has been leading a multi-institutional research project in the diagnostics area for several years. The project involves partnering with an innovative building operator to evaluate a prototype Information Monitoring and Diagnostic System (IMDS) installed in a commercial office building. The IMDS archives measurements from high-quality sensors every minute. It includes a powerful data visualization tool, which can be used on-site or accessed via the Web. The IMDS has been used to identify and correct a series of control problems. It has also allowed the operators to make more effective use of the building control system, freeing up time to take care of other tenant needs. They believe they have significantly increased building comfort, potentially improving tenant health and productivity. The reduction in the time required to operate the building is worth about $20,000/year, which could pay for the IMDS in about five years. A control system retrofit based on findings from the IMDS is expected to reduce energy use by 20% over the next year, worth over $30,000/year. The project has also included the evaluation of simple chiller models for fault detection, concluding that such models can be used as reference models to monitor operation and detect faults. The ability of the IMDS to measure cooling load and chiller power to 1% accuracy with a one-minute sampling interval permits the detection of faults that would otherwise go undetected. A virtual tour of the IMDS can be found at Diagnostics for Building Commissioning and Operations web site.

Screenshot of simulation models of the building subsystems

Figure 2. Screen shot showing an implementation for a dual-duct air-handler system.

On-Line Model-Based Performance Analysis

As part of a recent analysis of the Philip Burton Federal Building in San Francisco, EETD researchers have developed simulation models of the building subsystems such as the air-handling units. A comparison of simulated performance and actual performance has proved useful for performance validation and operational analysis. As an extension to this work, we have developed an online implementation of the simulation models that can be used for performance analysis of air-handling units in real time.

The simulation models are based on first principles and can be configured from normally available design information. Researchers have incorporated the models into a software tool that can be used to improve the control of air-handling systems and validate performance at the same time. The software tool is in the form of an add-on for conventional feedback control loops, and has been developed using the BACnet protocol for communication with the building control system. Use of the BACnet standard potentially allows the tool to be deployed on a variety of different systems without having to reengineer the communication interfaces for each new application. Field trials of the tool are now being carried out in the Phillip Burton Federal Building.

— Philip Haves, Mary Ann Piette, and Tim Salsbury

For more information, contact:

  • Philip Haves
  • (510) 486-6512; fax (510) 486-4089

This research is supported by the California Institute for Energy Efficiency, California Energy Commission, and the U.S. Department of Energy, Office of Building Technologies, State and Community Programs, and the USDOE, Federal Energy Management Program.

↑ home | ← previous article | next article →