Elsevier

Automation in Construction

Volume 22, March 2012, Pages 203-211
Automation in Construction

An online fault diagnosis tool of VAV terminals for building management and control systems

https://doi.org/10.1016/j.autcon.2011.06.018Get rights and content

Abstract

Proper design, reliable control and health of variable–air–volume (VAV) terminals are essential for reliable operation and energy efficiency of VAV air-conditioning systems. This paper presents an online fault diagnosis tool for pressure-independent VAV terminals and its implementation. The fault diagnosis tool is developed using a robust fault detection and diagnosis strategy for VAV terminals. Cumulative sum (CUSUM) control chart with estimated process parameters is utilized to detect faults in VAV terminals. A rule-based fault classifier is designed and utilized to find the sources of faults in VAV terminals. The fault diagnosis tool relies only upon the sensor data and control signals that are commonly available in building management control systems (BMCS). The fault diagnosis tool can be installed in BMCS to online read the operating data of all VAV terminals in a building through local area network and diagnose faults of VAV terminals in the building. The fault diagnosis tool has been validated in a large-scale office building located in Hong Kong and can effectively find all faults occurring in VAV terminals.

Highlights

►An online fault diagnosis tool is developed for VAV terminals to improve energy and control performances. ►CUSUM control charts with estimated process parameters to detect faults in VAV terminals. ►A rule-based fault classifier with expert rules and fault isolation algorithms to diagnose faults in VAV terminals. ►A technical approach to implement online fault detection and diagnosis for VAV terminals in large-scale office buildings.

Introduction

It has been a focus issue to reduce carbon-emission and slow down global climate warming, which will be a long-term and arduous task for human being. In order to solve this issue, human being has begun to focus on energy conservation and the development of green energy. VAV system is a popular type of heating, ventilation and air conditioning (HVAC) system in office and commercial buildings for energy saving, heat comfort and most adapting to space with variable load conditions. However, VAV systems tend to have more faults due to the complexity of VAV systems and their control systems. If a fault fails to be detected, diagnosed and removed in time, it will result in the increase of energy consumed by the system, the deterioration of space heat comfort and the damage of components. Fault detection and diagnosis (FDD) has been approved to be an essential and efficient supporting tool in fixing faults timely and reducing the impacts of them in building HVAC system applications. Therefore, it is very significant to develop suitable detection and diagnosis methods that can be used in the VAV systems of buildings.

Research on fault detection and diagnosis techniques for HVAC systems has become an active and intensive area since the late 1980s and early 1990s [1]. Cui and Wang [2] presented an online adaptive strategy for the fault detection and diagnosis of centrifugal chiller systems. The benchmarks of the performance indexes were provided by simplified reference models. Six physical performance indexes and a set of rules were used to detect and diagnose faults in centrifugal chiller systems. An online adaptive scheme was developed to estimate and update the thresholds for detecting abnormal performance indexes. Yoshida et al. [3] developed dynamic FDD model for VAV air handling units using recursive autoregressive exogenous algorithm. One methodology, based upon frequency response of the model was evolved for automatic fault detection and diagnosis. A short and simple method was also used to detect the faults of VAV units operating in the same zone by comparing their behavior. Seem et al. [4] looked into VAV terminal on-line control. Two indices were calculated from building management system (BMS) driven data for VAV box online monitoring and fault detection. One index provided an indication of how well the controller maintains the process output while the other provided the indication of mechanical components (actuators, dampers) operating status. Xiao and Wang [5] presented an online fault diagnosis tool to be used to assist building automation systems for online sensor heath monitoring and fault diagnosis of air handling units (AHU). The tool employed a robust sensor fault detection and diagnosis strategy based on the Principal Component Analysis (PCA) method. Wang and Qin [6] proposed a fault detection and validation strategy for the flow sensor of VAV terminals using PCA methods. The faulty sensor is reconstructed after it is isolated by the strategy, and the FDD strategy repeats using the recovered measurements until no further fault can be detected. Thus, the sensitivity and robustness of the FDD strategy is enhanced significantly. Many other methods have also been developed to detect and diagnose the faults in HVAC systems concerning various faults of sensors, equipment and control faults [7]. However, very few efforts were made on the application and implementation of online fault detection and diagnosis for the VAV terminals [8].

It can be concluded that research on fault detection and diagnosis for VAV terminals has been very insufficient. There are a number of causes of the problem. The first cause is that sensor information is insufficient to detect and diagnose faults in VAV terminals. VAV terminals are instrumented with the minimum number of sensors sufficient to implement basic local-loop and supervisory control strategies. The second cause is that the data that is collected overwhelms building operators. There is little effort to consolidate the information into a clear and coherent picture of VAV terminal status. The third cause is that building operators may overlook symptoms of a failure because they may not fully understand the control strategies implemented [9].

Jeffrey and John [10] used the CUSUM control chart to detect faults in VAV terminals and tested it using emulation, laboratory, and offline field data sets. The fault detection method was offline tested using simulated and artificial faults. However, some critical problems were not involved in their study, such as how to determine the parameters of CUSUM control chart in the application of VAV terminals, how to isolate the faults occurred in VAV terminals and how to online implement it in building air-conditioning systems. In this study, a detailed discussion is given for determining the parameters of CUSUM control chart, such as control limits, the process error mean, and the process error standard deviation, to facilitate its practical implementations in VAV air-conditioning systems. A rule-based fault classifier is developed and utilized to diagnose the faults of VAV terminals. The fault classifier uses a rule set of twenty expert rules including several rules given in the reference [11] to isolate fifteen fault sources in VAV terminals. A technical approach is presented to implement online fault detection and diagnosis for all VAV terminals in a large-scale office building and online test the fault detection and diagnosis strategy through all varies and amount of real faults occurred in VAV terminals.

This paper presents an online fault diagnosis tool for VAV terminals in buildings, which is achieved by implementing a robust fault detection and diagnosis strategy. In the strategy, CUSUM control charts with estimated process parameters are utilized to detect faults in VAV terminals. A rule-based fault classifier consisting of a set of expert rules and fault isolation algorithms is designed to isolate fault sources. The tool relies only upon the sensor data and control signals that are available in common energy management and control systems of buildings and is implemented in buildings through local network communication technique. The remainder of this paper is organized as follows. Section 2 is the overview of methodology for the fault diagnosis tool. Section 3 is the description of the fault diagnosis tool. Section 3.1 is the verification of the fault diagnosis tool. Finally, a conclusion is drawn.

Section snippets

CUSUM control chart

The CUSUM control chart is a popular signal-based FDD method [12], which has been widely used in industrial process control to monitor production quality. The CUSUM algorithm was first proposed by Page [13]. Useful information of the CUSUM control chart can be found in [14], [15], [16]. The basic concept behind CUSUM chart is to accumulate the error between a process variable and its expected value. When the CUSUM values fall outside the control limits, the control process is considered to be

Typical VAV terminal instrumentation

The fault diagnosis tool is developed to detect and diagnose faults in pressure-independent VAV terminals. VAV terminals are common and key components in a VAV air-conditioning system. A typical VAV air-conditioning system consists of an air handling unit with a variable speed supply fan and cooling coils; supply and return air ducts; outdoor, return and exhaust air dampers; and VAV terminals in individual zones. Fresh air is transported to AHU by fresh fan. Return air is back to AHU through

Verifications and discussions

The fault diagnosis tool has been online implemented and validated on a office building located in Hong Kong. The BMCS installed in the building gathers sensor measurements and control signals using 5-minute sampling interval. In this study, the in-control mean and standard deviation of the temperature error (μ0 and σ0) between the measured zone temperature and its setpoint are estimated from historical, fault-free operating data. The slack parameter k = 3, the upper control limit hS = 240, and the

Conclusions

An online fault diagnosis tool is presented in this paper to find and identify the faults in VAV terminals of buildings. The fault diagnosis tool includes 5 functions: operating data acquisition, fault detection, fault diagnosis, automatic generation of operating charts and automatic fault report. The fault diagnosis tool uses CUSUM algorithms to accumulate the normalized temperature error between the measured zone temperature and its setpoint, and detect faults in VAV terminals. It utilizes a

Acknowledgments

The research work presented in this paper is financially supported by Building Intelligent Control Research Fund (JRP 0901) granted by Swire Properties Ltd. The authors appreciate the technical assistance from Honeywell Limited (Hong Kong) during the conduct of research presented in this paper.

References (28)

  • J.M. House et al.

    An Overview of Building Diagnostics, Diagnostics for Commercial Buildings: from Research to Practice

    (1999)
  • S. Jeffrey et al.

    Application of control charts for detecting faults in Variable–Air–Volume terminals

    ASHRAE Transactions

    (2003)
  • H. Wand et al.

    Data driven fault diagnosis and fault tolerant control: some advances and possible new directions

    Acta Automatica Sinica

    (2009)
  • E.S. Page

    Continuous inspection schemes

    Biometrika

    (1954)
  • Cited by (0)

    View full text