Modelica-based system modeling for studying control-related faults in chiller plants and boiler plants serving large office buildings

https://doi.org/10.1016/j.jobe.2021.102654Get rights and content

Highlights

  • Presented high-fidelity models for chiller/boiler plants under faulty conditions.

  • Conducted a comprehensive fault impact evaluation of the 13 control-related faults.

  • Identified the critical faults based on their impacts on energy and service quality.

  • Revealed non-linear relationships between the fault impacts and the severity level.

Abstract

Dynamic modelling of the faulty operation of chiller plants and boiler plants can help identify their impacts and support the development of fault detection methods. However, adequate models are seldom reported in the literature. In this study, we aim to develop high-fidelity models for approximating the dynamic behaviors of chiller plants and boiler plants under control-related faults. Specifically, we first designed a typical configuration of the chiller plants and the boiler plants; we then modeled both the physical systems and controllers of these typical plants with Modelica. When developing the Modelica models, we created a hierarchical model structure while modules in each layer can be redeclared and parameterized at upper layers. This model structure facilitates the implementation of fault scenarios through intuitive model modifiers. At last, we applied the proposed models in a comprehensive fault impact evaluation of the thirteen control-related faults of chiller and boiler plants. In this evaluation, the proposed models are coupled with the EnergyPlusTM thermal load model to study the impact of various faulty scenarios. Based on the fault impact evaluation results, we identified the faults that have the most significant impacts on the operation of the chiller and boiler plants, respectively. We also found that the relationship between the impacts of the studied faults and the severity level of the faults can be highly non-linear. This study contributes to the literature by providing the first dynamic models of chiller plants and boiler plants which can be used to study control-related faults on a large-scale.

Introduction

Chiller plants and boiler plants account for about 35% of the primary energy used by commercial building cooling and 21% of the primary energy used by commercial building heating, respectively, in the U.S [1]. The energy efficiency of chiller plants and boiler plants can be significantly affected by operation faults [2,3]. Cheung and Braun [4] found the electricity consumption of chiller plants under faulty conditions can be increased by up to 14.8%. García, Álvarez, etc. pointed out that [5] poor air inlet settings can lead to 20% higher fuel consumption by boiler plants. Owing to their significant impacts on the energy efficiency of chiller plants and boiler plants, substantial efforts have been devoted to better understand those faults and thereby eliminate them when operating chiller plants and boiler plants. Those efforts can be divided into two groups: fault impact analysis and automated fault detection & diagnostics (AFDD). Fault impact analysis aims to quantify the impacts of various faults during the building operation. It helps building operators to identify the critical faults and guide researchers to identify the critical research directions [4,6]. AFDD attempts to develop methods to detect operational faults and then isolate the causes of the detected faults. AFDD has been an active area of research in HVAC systems. There have been a considerable number of AFDD methods for chiller plants and boiler plants proposed over the past two decades [[7], [8], [9]].

For both the fault impact analysis and AFDD, modeling faulty chiller plants and faulty boiler plants is indispensable. It is so far the most common way to quantify the impacts of faults [6]. In general, fault models of chiller plants and boiler plants can be categorized into three groups. In the first group, those fault models can be obtained by modifying the parameter values of fault-free models. For example, Basarkar, Pang, etc. [10] modified the rated capacity of a fault-free chiller model and the rated efficiency of a fault-free boiler model to model a refrigerant leak fault and a fouled water tube fault, respectively. Those fault-free models were developed in EnergyPlusTM [11]. In the second group, the models of faulty building systems are obtained by introducing new parameters to fault-free models while those parameters define the degree or extent of the studied faults. For example, Cheung and Braun [4] added six parameters to the chiller model in EnergyPlusTM to calculate how the chiller power is influenced by faults such as overcharging, excess oil, non-condensable in the refrigerant, and condenser fouling. In the third group, the models of faulty building systems are developed by treating the fault(s) explicitly based on underlying physics. In this case, major increases in the modeling detail are usually required if fault-free models are leveraged. For example, Shohet, Kandil, etc. [12] developed a physics-based model of a non-condensing boiler to investigate faults that occur within boilers. When developing this model, they added a significant number of new equations to the existing fault-free boiler models. For instance, a governing equation for the combustion process was added for abnormal combustion conditions caused by faults such as excess air. Cheung and Braun [13] developed gray box models for components in chillers, such as compressors, condensers, and expansion valves. They calibrated the coefficients of those models with data sampled from real chillers under normal and faulty conditions.

Despite the encouraging results from fault modeling for chiller plants and boiler plants, existing fault models have two major drawbacks, limiting their potentials for supporting general fault-related studies. First, they tend to ignore fast building dynamics and adopt ideal control. Most of the existing fault models are implemented by modifying or adding parameters to fault-free models. In the literature, EnergyPlusTM is used frequently as the fault-free model for implementing fault models. However, the extent to which faults can be approximated is subject to the basic assumptions of EnergyPlusTM. Specifically, EnergyPlusTM assumes that fast dynamics are negligible [14]. Therefore, the fault models implemented in EnergyPlusTM may not capture the fast building dynamics over short-term periods. However, those fast building dynamics can play an important role in determining the impacts of operation faults, especially the control-related ones.

Second, the usage of existing fault models can be labor-intensive, especially when it comes to large-scale fault-related studies. In those studies, it is common to consider multiple faults occurring at different operating conditions. The combination of faults and operation conditions can easily generate a large number of simulation cases. This number can be further increased substantially if fault occurrence probability is considered [15]. On the other hand, a simulation-based validation of AFDD methods may require a co-simulation set up to communicate the simulation models with the testing methods in real-time. However, configuring the existing fault models for this purpose can be troublesome [16].

Some studies aim to mitigate those issues in large-scale studies. For example, Li and O'Neil developed a software framework for facilitating fault impact analysis [6]. This framework can automatically generate EnergyPlusTM input files based on predefined fault conditions. Wang and Karami [17] proposed a virtual testbed to evaluate the developed AFDD methods with simulation data. However, most of those studies rely on ad hoc software implementation and very few of those studies consider those issues when developing fault models, limiting their abilities for supporting large-scale studies.

In this paper, we present high-fidelity models for approximating the behaviors of chiller plants and boiler plants under faulty conditions. Compared to existing ones, the proposed models have two advantages: first, they better characterize the dynamic patterns in the system operation. In those models, control architecture and control logic are faithfully implemented. Thus, they can be used to study control-related faults, such as incorrect staging control due to sensor bias and high steady-state errors owing to mistuned feedback control. In this study, we used Modelica [18], which is an equation-based object-oriented modeling language, to establish the system model. Modelica is very suitable for modeling multidomain systems [19,20] that contain not only the physical system but also the control system. Second, they are readily extensible, supporting large-scale investigations to explore different faulty conditions/scenarios. Those models are established in a hierarchical structure while modules in each layer can be redeclared and parameterized at upper layers. fault scenario can be described through intuitive model modifiers. In sum, the proposed models, for the first time, provide a solution to study control-related faults in chiller plants and boiler plants on a large-scale. We applied the proposed models in a comprehensive fault impact evaluation on thirteen control-related faults. In this evaluation, the proposed models are coupled with the EnergyPlusTM thermal load model to represent various faulty scenarios.

The rest of this paper is organized as follows. In Section 2, a detailed description of the studied chiller plant and the studied boiler plant is provided. After that, the studied control-related faults are discussed in Section 3. Then, system models of the studied plants are elaborated in Section 4. We elaborate on how we implement models in Modelica, validate the models, and extend the models for large-scale fault-related studies. After that, a comprehensive fault impact analysis is conducted in Section 5. Conclusions can be found in Section 6.

Section snippets

Studied system

The studied system provides chilled water and hot water to a prototypical large office building in the U.S. This office building consists of twelve floors while each floor is served by one air handling unit (AHU). Each AHU has one cooling coil where chilled water cools the air leaving the AHU. There are five thermal zones on each floor, and each zone is served by one variable air volume (VAV) terminal. In each VAV terminal, hot water heats the air entering the thermal zone. More detailed

Operational faults

We consider three types of operational faults that are commonly observed in chiller plants and boiler plants. Those faults are sensor bias, leaking valve, and untuned Proportional-Integral (PI) control. The following sections elaborate on how we describe those faults quantitatively.

  • 1)

    Sensor bias

Sensor bias refers to differences between measured and actual values of observed variables. A sensor bias can be described byu=uˆ+ewhere u and uˆ are the measured value of an observed variable from sensors

System models

In this section, we elaborate on how we develop high-fidelity models of the chiller plant and the boiler plant with Modelica. We also discuss how those high-fidelity models can support large-scale fault-related studies.

Simulation settings

We conduct a fault impact analysis on thirteen faults, as shown in Table 4. The sensor bias faults affect the input of Controller #1, #2, #3, #5 in the chiller plant and that of all the controllers in the boiler plant, the leaking value fault impacts the actuation of the output signals of Controller #4 in the chiller plant, and the untuned PI control faults influence output signals of Controller #2 and #3 in the chiller plant and Controller #2 in the boiler plant. Note the simulation is

Conclusions

In this study, we present a set of high-fidelity models for a typical chiller plant and a typical boiler plant. Those models faithfully represent the control architecture of the studied plant and thereby can be used for studying the control-related faults on a large scale. The usage of proposed models is demonstrated via a comprehensive evaluation of the impacts of common control related faults on the energy performance of the plants. In this evaluation, sensitivity analysis is performed to

Author statement

Sen Huang: Conceptualization, Methodology, Software, Formal analysis, Writing - original draft; Wangda Zuo: Methodology, Formal analysis, Writing - Reviewing and Editing; Draguna Vrabie: Methodology, Reviewing, and Editing; Rong Xu: Methodology, Software.

Conflicts of interest

The authors whose names are listed immediately below certify that they have NO affiliations with or involvement in any organization or entity with any financial interest (such as honoraria; educational grants; participation in speakers’ bureaus; membership, employment, consultancies, stock ownership, or other equity interest; and expert testimony or patent-licensing arrangements), or non-financial interest (such as personal or professional relationships, affiliations, knowledge or beliefs) in

Declaration of competing interest

The authors declare that they have no known competing financial interests or personal relationships that could have appeared to influence the work reported in this paper.

References (31)

  • S. Huang et al.

    Amelioration of the cooling load based chiller sequencing control

    Appl. Energy

    (2016)
  • S. Huang et al.

    Improved cooling tower control of legacy chiller plants by optimizing the condenser water set point

    Build. Environ.

    (2017)
  • J. Åkesson et al.

    Modeling and optimization with Optimica and JModelica.org—languages and tools for solving large-scale dynamic optimization problems

    Comput. Chem. Eng.

    (2010)
  • D. Westphalen et al.

    Energy Consumption Characteristics of Commercial Building HVAC Systems Volume I: Chillers, Refrigerant Compressors,and Heating Systems

    (2001)
  • S. Katipamula et al.

    Review article: methods for fault detection, diagnostics, and prognostics for building systems—a review, Part I

    HVAC R Res.

    (2005)
  • Cited by (15)

    • Assessment of HVAC system operational fault impacts and multiple faults interactions under climate change

      2022, Energy
      Citation Excerpt :

      Lu et al. [24] analysed the effects of 359 fault cases on various impact indicators, including energy usage, thermal comfort levels, operational cost, ventilation metrics, etc., under the cold climate. Huang et al. [25] evaluated the annual energy variations of chiller and boiler plants and thermal unmet hours caused by 13 control-associated faults. In addition, some of the literature evaluated the fault impacts at the component level.

    View all citing articles on Scopus

    This research is partially supported by the of the U.S. Department of Energy, Energy Efficiency and Renewable Energy, Office of Building Technologies, Emerging Technologies Program, under Contract no. DE-AC05-76RL01830. This research is also partially supported by the National Science Foundation under Awards No. IIS-1802017. This work also emerged from the IBPSA Project 1, an internationally collaborative project conducted under the umbrella of the International Building Performance Simulation Association (IBPSA). Project 1 aims to develop and demonstrate a BIM/GIS and Modelica Framework for building and community energy system design and operation. The authors acknowledge the valuable comments from Dr. Hayden Reeve, Senior Technical Advisor at the Pacific Northwest National Laboratory.

    View full text