NIST logo
*
Bookmark and Share

Fault Detection and Diagnostics for Commercial Heating, Ventilating, and Air-Conditioning Systems Project

Summary:

NIST is working to measure and improve the operational performance of commercial buildings by leveraging previously untapped capabilities within modern automation and control systems.  This requires developing a measurement science that enables automatic detection and diagnosis of equipment faults, sensor failures, and control errors in the heating, ventilating, and air-conditioning (HVAC) systems of buildings.  The resulting fault detection and diagnosis (FDD) software (“FDD tools”) will utilize existing sensors and controller hardware, and will employ artificial intelligence, deductive modeling, and statistical methods to automatically detect and diagnose deviations between actual and optimal HVAC system performance.

Description:

Objective: To improve the operating efficiency of commercial heating, ventilating, and air-conditioning (HVAC) systems by 10% to 30% through development and demonstration of the enabling measurement science for detecting faults and control errors in commercial HVAC equipment and systems, and transferring the measurement science to the private sector by 2014.

What is the new technical idea? The new idea is to leverage the untapped capabilities of modern building automation and control systems by developing embedded FDD tools that monitor the performance of subsystems and automatically detect faults. Advances in building automation technologies provide the prospect of very data–rich performance surveillance in buildings that can be applied on system-wide scales that are necessary to optimize overall system performance and achieve the goal of zero energy consumption. The key to realizing efficiency improvements is combining new measurement technology and performance metrics with analysis techniques that can be implemented in building automation and control products. The resulting systems would have a distributed, embedded intelligence that can detect and respond to faults and operational errors and inefficiencies. Microcontroller technology has advanced in a way that simplifies the implementation of the proposed embedded intelligence. Success is likely because of increasing industry demand for the technology as a way to reduce energy consumption and associated environmental impacts[1].

What is the research plan? NIST will develop and demonstrate a comprehensive set of tools and algorithms for detecting faults and control errors in a wide range of commercial HVAC equipment and systems following a sequence of steps that has proven successful in the past:

  • Develop a rapid prototyping platform for initial development of FDD algorithms that includes a Matlab® interface to HVACSIM+ (a sophisticated system simulation package developed at NIST)
  • Obtain building automation system data from real buildings and laboratory studies of normal and faulty operation
  • Test rapid prototypes using collected data and simulations
  • Develop an FDD expert system framework  (“FDD Expert Assistant” or FDD-EA) combining equipment-level FDD algorithms with new learning algorithms that adapt to operator feedback, improving FDD reliability and acceptance by minimizing false alarms and helping operators cope with cascades of true alarms
  • Conduct field tests in cooperation with existing and anticipated industry CRADA partners

In past years the project has focused on the early steps of this plan and resulted in several equipment specific FDD tools. The current focus has shifted to refinement and testing of a prototype FDD-EA as a means to make those tools more autonomous and more adaptive. Autonomy means tools that are less dependent on a human expert intervening in order to infer critical FDD parameter values from the available data. Adaptation means the tool improves its reliability by “learning” from its mistakes, through interaction and feedback from building operators, a capability that enhances FDD acceptance.

NIST will leverage existing CRADAs with industry partners to speed commercialization of FDD research results. NIST will also leverage academic ties to sponsor components of the research through contracts and grant solicitations. The outcome will be a suite of proven algorithms for FDD that can be embedded in hardware that is practical for commercial adaptation.

 


[1] Turner, C. and M. Frankel.  2008. "Energy Performance of LEED for New Construction Buildings.  U.S. Green Building Council.

Major Accomplishments:

Research Outcome:

  • Veronica, D. A. 2013. Automatically Detecting Faulty Regulation in HVAC Controls. HVAC&R Research, Vol. 19, No. 4, 412-422.

Potential Research Impacts:

  • Veronica, D. A., Detecting Cooling Coil Fouling Automatically – Part 1: A novel Concept. HVAC&R Research, Vol. 16, No. 4, 2010 (FY 2010)
  • Veronica, D. A.,  Detecting Cooling Coil Fouling Automatically – Part 2: Results Using a Multilayer Perceptron. HVAC&R Research, Vol. 16, No. 5, 2010 (FY 2010)

Realized Research Impacts:

  • Schein, J., Bushby S.T. "Fault Detection and Diagnostics for AHUs and VAV Boxes," ASHRAE Journal, Vol 47 No. 7, July 2005 (5 citations)
  • Schein, J., and Bushby, S.T. "A Hierarchical Rule-Based Fault Detection and Diagnostic Method for HVAC Systems," HVAC&R Research, Vol. 12, No. 1, 2006 (16 citations)
  • Schein, J., Bushby, S.T., Castro, N.S. “A Rule-Based Fault Detection Method for Air Handling Units,” Energy & Buildings, Vol. 38, Issue 12, December 2006, pp. 1485-1492, 2006 (44 citations)

Improved Energy Efficiency of Operations:

  • Cooling and heating mode fault-applied performance data for heat pumps provided to industry and academia. This well characterized data provides a basis for manufacturers to develop new fault detection products (FY 2010).
  • Rapid prototyping tool developed for investigating FDD algorithms. Rapid prototyping and testing of candidate algorithms will enable researchers to identify the most promising fault detection approaches (FY 2011).
  • Variations of the Fault Detection and Diagnostics – Expert Assistant (FDD-EA) tool for single-duct air handling units, dual-duct air handling units, variable-air volume boxes, and terminal units incorporated into the California Energy Commission Universal Translator tool (FY 2013).

Realized Technology Transfer Impact:

  • Industry adoption of VPACC and APAR fault detection algorithms for performance monitoring of HVAC products, based on NIST measurement science work, leading to better performing and more energy efficient HVAC systems (FY 2006).
Vent pipes

Start Date:

October 1, 2011

Lead Organizational Unit:

el

Facilities/Tools Used:

Virtual Cybernetic Building Testbed

Staff:

Project Leader:  Dr. Daniel A. Veronica

Contact

General Information:
Dr. Daniel A. Veronica, Project Leader
301-975-5874 Telephone

100 Bureau Drive, M/S 8631
Gaithersburg, MD 20899-8631