Optimal Control of Air-Side Economizer

Abstract

The economizer system is a method of improving energy efficiency through the operating method, which introduces outdoor air through dampers when the outdoor air temperature or enthalpy is lower than the that of the indoor air. The set values used for economizer control include the mixed air temperature and high and low limits. The set values are presented as fixed values in the relevant standards and are controlled to be fixed during actual operation, which may lead to issues such as indoor discomfort, poor indoor air quality, and energy wastage. Therefore, it is necessary to optimize economizer control by determining appropriate set values considering the indoor and outdoor environments. To this end, this paper reviewed the economizer system, control method, control set values, and prediction models in buildings. As a result, it was concluded that optimal economizer control is possible by utilizing a prediction model.

1. Introduction

Korea has set the goal of reducing greenhouse gas emissions by 37% compared to Business As Usual (BAU) by 2030, and, in the building sector, the country aims to reduce greenhouse gas emissions by 32.8% compared to BAU [1]. Efforts are being made to reduce greenhouse gas emissions from the building sector, and the government is promoting the strengthening of energy standards for new buildings, thus activating the green remodeling of existing buildings and gradually making zero-energy buildings mandatory. In the building sector, HVAC (Heating, Ventilation and Air Conditioning) systems account for the largest portions of energy consumption [2]. Therefore, it is necessary to improve the energy efficiency of air conditioning systems to reduce greenhouse gas emissions from buildings. One energy-efficient measure for HVAC systems is system renovation. The lifetime of HVAC systems, including their mechanical and electrical components, is about 15–20 years. Buildings deteriorate over time, and so does the performance of their HVAC systems. To improve the performance and economic feasibility of these systems, renovation and repair are required, which can lead to energy savings [3,4]. Another energy-efficient measure for HVAC systems is to improve their operation method.

In accordance with the Green Building Construction Support Act, the building design criteria for energy saving operated by the notification of the Ministry of Land, Infrastructure and Transport establishes standards for the efficient management of buildings. The standard includes the energy performance index (EPI), which scores indicators related to energy savings and the performance of buildings. Regarding the EPI, the mechanical equipment sector suggests energy-efficient measures through the renovation of the HVAC system and improvement in operation methods. Among the evaluation items within the EPI, the items related to HVAC system renovation are adopting high-efficiency equipment for HVAC blowers, as well as adopting high-efficiency heat recovery systems, while those related to improving systems’ operating methods involve adopting energy-saving controls for ventilation fans in underground parking lots, adopting energy-saving controls such as variable speed control for fans, and introducing an outdoor air-cooling system such as an economizer. The economizer system discussed in the standards refers to a system that reduces the cooling load by introducing outdoor air such as during the mid-period. At this time, if more than 60% of the total outdoor air flow rate is applied, the points can be obtained [5].

An economizer system introduces outdoor air through dampers when the outdoor air temperature or enthalpy is lower than that of the indoor air. There are dry-bulb temperature and enthalpy control for the economizer control, and the set values for control include the mixed air temperature and the high and low limits for free cooling. Conventionally, the set value of the mixed air temperature is kept constant for economizer control. The Facility Engineering Manual sets an high limit of 18 °C, while the American Society of Heating and Air Conditioning Engineers (ASHRAE) standard 90.1 suggests determining the highlimit according to climatic conditions [6,7]. When controlling economizers, the fixed set values are set without accounting for changes in the indoor and outdoor environments, which may cause problems such as indoor discomfort, poor indoor air quality, and energy wastage. Therefore, set values and control methods that are suitable for indoor and outdoor conditions need to be found.

Previous studies have evaluated the energy impact according to the type of economizer system and set values used, but there is a lack of research on determining the set values considering the indoor and outdoor conditions. In addition, big data and machine learning have been utilized in previous studies to predict indoor and outdoor conditions that are affected by various variables as well as to analyze optimal HVAC control. The purpose of this paper is to summarize the theory behind the economizer system, control method, and set values, as well as prediction models for buildings, to determine the optimal approach to economizer control. Section 2 explains the economizer system. Next, the air-side economizer control and set values are explained in Section 3. Then, in Section 4, a method for optimal economizer control is proposed, and finally, the conclusion is presented in Section 5.

2. Economizer Systems

Overview of Economizer System

An economizer is a system that reduces the amount of cooling energy needed by introducing outdoor air when it is sufficient for cooling, and economizers can be classified into passive and active systems. For instance, a passive system is used to prevent incident heat from being implemented in the design stage of a building so that cooling loads are not generated [8,9]. In contrast, an active system is any mechanical system that reduces the cooling energy or the operating time of a cooling system when cooling a data center or any other facility when the outdoor temperature is sufficiently low. Economizers in active systems can be divided into water-side economizers and air-side economizers. A water-side economizer utilizes cold air to cool the circulating cooling water of an external cooling tower. Figure 1 presents a diagram of an economizer.

Air-side economizers introduce cold air from the outside into the system and use it for cooling, and they are used in HVAC (Heating, Ventilating and Air Conditioning) systems, heat exchangers, etc. Figure 1b shows a conceptual diagram of an economizer. The heat recovery ventilation system reduces cooling and heating costs by recycling the heat energy of the internal air while introducing outdoor air, and it filters pollutants using a filter fitted within the system. If it includes a bypass function, fresh outdoor air is circulated during the intermediate period without being passed through the heat exchange element. In this way, it is possible to reduce energy consumption and extend the lifetime of the heat exchange element [10]. The economizer control employed in an HVAC system controls the outdoor air, ventilation, and exhaust dampers of the system to maintain the set value of the mixed air temperature when the outdoor air temperature or enthalpy is lower than that of the indoor air [11].

Various studies have been conducted on economizers. Kim et al. developed a control method to extend the free-cooling section for a water-side economizer and reviewed the appropriate control conditions through an energy consumption analysis [12]. Jin et al. reviewed the applicability of a cooling water supply and return temperature in relation to energy savings and water consumption in water-side economizers used in data centers [13]. Hwang et al. analyzed the impact of various design and operation variables on the energy consumption of cooling systems to determine how power consumption can be reduced by introducing an economizer system into a data center. The authors also conducted a theoretical analysis and examined field applications to derive an operating method that minimizes power consumption [14]. Kim et al. analyzed an economizer control method suitable for domestic climate characteristics to save energy used for cooling by employing an economizer in a building with a gas engine VRF heat pump system [15]. Kim et al. attempted to derive energy-saving measures through the efficient operation of outdoor air cooling in an energy recovery ventilator (ERV). They utilized EnergyPlus (v.9.3.0) software and examined the energy-saving effect created by the bypass outdoor air-cooling control of each type of ERV, the ratio of the outdoor air cooling operation time, the heat exchange airflow rate, the energy consumption of the fan power, and the heat exchange efficiency [16]. Jiang et al. analyzed the main factors affecting the applicability of air-side economizers [17]. This type of economizer system is not only considered an energy-efficient measure in domestic regulations, but it is also used in various systems as an energy-efficient measure for domestic mechanical systems. This paper reviewed theories and papers related to control methods, set values, and optimal control to achieve energy savings in air-side economizer control applied to HVAC systems.

3. Air-Side Economizer Control

3.1. Types of Economizer Control

The Handbook of Air Conditioning and Refrigeration Engineering defines outdoor air cooling as introducing outdoor air when it is lower in temperature than the indoor air, i.e., when it is cold enough to be used for cooling. According to the ASHRAE Standard 90.1, an economizer is a system that cools air by introducing outdoor air through the adjustment of its dampers to reduce mechanical cooling when the outdoor air temperature is low. An economizer, as defined by the ASHRAE, represents the control methods for outdoor air cooling presented in the Handbook of Air Conditioning and Refrigeration Engineering. Common economizer control methods include dry-bulb temperature and enthalpy control. The dry-bulb temperature control method introduces outdoor air by comparing the dry-bulb temperature of the outdoor and return air, and enthalpy control is based on the enthalpy value [18]. In other words, the dry-bulb temperature control method adjusts the opening rates of the outdoor and return air dampers when the outdoor air temperature is lower than that of the indoor air, using it for cooling. Otherwise, the operation of the outdoor air damper is kept to a minimum to introduce the least possible amount of outdoor air for ventilation. According to the law of conservation of mass, the mixing of outdoor air and return air can be expressed as Equation (1). The outdoor air intake rate can be expressed as the ratio of the mixed air flow rate to the outdoor-air flow rate, as seen in Equation (2). Through Equations (1) and (2), the mixed air temperature can be derived as shown in Equation (3).

$$m_{mix}{T}_{mix}=m_{oa}{T}_{oa}+m_{ra}{T}_{ra}$$

(1)

$$\mathsf{\alpha }={\displaystyle \frac{m_{oa}}{m_{mix}}}$$

(2)

$$T_{mix}=\mathsf{\alpha }{T}_{oa}+(1-\mathsf{\alpha })T_{ra}$$

(3)

The amount of outdoor air introduced can be expressed by Equation (4) based on the law of conservation of mass and the law of conservation of energy, while the change in the opening ratio of the outdoor air damper according to the outdoor air temperature can be expressed as shown in Figure 2.

$$\alpha ={\displaystyle \frac{T_{ra}-T_{mix}}{T_{ra}-T_{oa}}}$$

(4)

As for the enthalpy control method, latent heat is considered, which was not the case for dry-bulb temperature control, and the temperature control values for the dry-bulb temperature are replaced with enthalpy. In other words, it is a control that introduces outdoor air by considering the enthalpy of the outdoor and indoor air at the same time and enables the introduction of outdoor air when its enthalpy is lower than that of the indoor air. The relationship for enthalpy control can be expressed as Equation (5), and the amount of outdoor air intake can be expressed as Equation (6).

Table 1. Outdoor air intake ratio according to control type.

Control Type	Condition for Economizer Control	Outdoor Air Damper’s Opening Ratio
Dry-bulb temperature control	$T_{mix}<T_{oa}<T_{High}$	Fully open
	$T_{Low}<T_{oa}<T_{mix}$	Open according to the rate of the outdoor air intake ratio based on the outdoor air temperautre
	$T_{oa}<T_{Low},T_{High}>T_{oa}$	Minimally open
Enthalpy control	$h_{mix}<h_{oa}<h_{High}$	Fully open
	$h_{Low}<h_{oa}<h_{mix}$	Open according to the rate of the outdoor air intake ratio based on the outdoor air temperautre
	$h_{oa}<h_{Low},h_{High}>h_{oa}$	Minimally open

illustrates the relationship between outdoor air temperature and outdoor air intake when controlling the economizer. When the outdoor air temperature is greater than the set value for the mixed air temperature and less than the high limit, the outdoor air damper is fully opened, and when the outdoor air temperature is between the low limit and the mixed air temperature, the outdoor air damper’s opening ratio changes according to the outdoor air temperature. Further, when the outdoor air temperature is less than the low limit or greater than the high limit, the damper opens by a minimal amount; in such cases, the minimum opening ratio is considered the ventilation amount.

$$m_{mix}{h}_{mix}=m_{oa}{h}_{oa}+m_{ra}{h}_{ra}$$

(5)

$$\alpha ={\displaystyle \frac{h_{ra}-h_{mix}}{h_{ra}-h_{oa}}}$$

(6)

The economizer system is an energy-saving system, and enthalpy control leads to more energy savings than dry-bulb temperature control. Son et al. evaluated the existing economizer control performance through simulations and found that using enthalpy control during the cooling period is an effective energy-saving method [19]. Kim et al. used TRNSYS, a dynamic energy simulation program, to evaluate the performance of economizer control methods; they analyzed the existing dry-bulb temperature control and enthalpy control in terms of indoor thermal comfort and energy consumption. As a result, both control methods satisfied the indoor thermal comfort and IAQ criteria. And the results of the HVAC energy consumption analysis confirmed that enthalpy control saves more energy compared to dry-bulb temperature control (Figure 3) [20]. Thus, energy savings are possible when dry-bulb temperature control is applied compared to when an economizer is not used, and additional energy savings are possible if enthalpy control is operated.

Further, ASHRAE Standard 90.1 classifies the controls into the fixed dry-bulb control, differential dry-bulb control, fixed enthalpy control, and differential enthalpy control methods depending on whether information about the ventilation air is considered. The outdoor air conditions for each control method are shown in Figure 4 [21] and

Table 2. Conditions for economizer control.

Control Type	Condition for Economizer Control
Fixed dry-bulb control	$T_{oa}<T_{set}$
Differential dry-bulb control	$T_{oa}<T_{ra}$
Fixed enthalpy with fixed dry-bulb temperature control	$h_{oa}<h_{set}$
Differential enthalpy with fixed dry-bulb temperature control	$h_{oa}<h_{ra}$

. The fixed dry-bulb control method is operated when the outdoor air temperature is lower than the set temperature value for the indoor air, while the differential dry-bulb method is operated by comparing the indoor air temperature with the outdoor air temperature. The fixed enthalpy control method implies that the economizer is operated when the outdoor air enthalpy is lower than the set value for the indoor air, while the differential enthalpy control method implies that the economizer’s operation is determined based on the outdoor air enthalpy and the measured enthalpy of the indoor air [22]. Among the control methods suggested by the ASHRAE, the enthalpy controls reduce energy consumption more than the dry-bulb temperature controls, and, among the enthalpy controls, the differential enthalpy control results in the most energy savings [23].

Nassif et al. proposed a new control strategy for an economizer that utilizes a different damper strategy (split-signal damper control) to provide better performance than the existing control strategy. This control controls only the outdoor air damper and opens the other two dampers to 100%, which reduces the static pressure of the dampers and minimizes energy consumption in the fan. When this strategy was evaluated with the existing damper control strategies, it was confirmed that energy was saved in the supply and return fans [24]. Typical economizer control is classified into dry-bulb temperature and enthalpy control, and the enthalpy control is a more effective energy-saving method than the dry-bulb temperature control. Additionally, related research is being conducted with the goal of saving more energy compared to existing control methods.

3.2. Economizer Set Value

The economizer control set values are the mixed air temperature and the high and low limits ($T_{high}$,$T_{low}$) that determine whether the economizer is operated. In the existing control method, when controlling the economizer, the set value is kept constant. The mixed air temperature is set to 13 °C in the case of the dry-bulb temperature control. The high limit in the economizer control method is based on climate conditions according to ASHRAE standard 90.1, and the details are as shown in

Table 3. High limit for economizer control—ASHRAE Standard 90.1.

Category		High Limit
Control Type	Climate
Fixed dry-bulb temperature control	0B, 1B, 2B, 3B, 3C, 4B, 4C, 5B, 5C, 6B, 7, 8	24 °C
	5a, 6a	21 °C
	0A, 1A, 2A, 3A, 4A	18 °C
Differential dry-bulb tempeature control	0B, 1B, 2B, 3B, 3C, 4B, 4C, 5A, 5B, 5C, 6A, 6B, 7, 8	$T_{Return}$
Fixed enthalpy with fixed dry-bulb temperature control	All	65 kJ/kg or 24 °C
Differential enthalpy with fixed dry-bulb temperature control	All	$h_{Return}$ or 24 °C

. For the fixed dry-bulb temperature control, the high limit is suggested to be set to 18 °C, 21 °C, and 24 °C depending on climatic conditions, whereas, for the differential dry-bulb temperature control, the high limit is suggested to be the same as the return air temperature. Furthermore, the fixed enthalpy in the fixed dry-bulb temperature control is set as 65 kJ/kg or 24 °C regardless of the climate. In contrast, the differential enthalpy in the fixed dry-bulb temperature control is set as the return air enthalpy, which is 24 °C. In addition, in the Facility Engineering Manual, the set value for the high limit is suggested to be 18 °C [6].

Various studies have been conducted on economizer set values. Lee et al. confirmed that the set value for the mixed air temperature that leads to the least energy consumption is not constant depending on the building load when using the dry-bulb temperature control [25]. Wang et al. developed a steady-state energy consumption model to determine the optimal temperature for the supply air that minimized energy costs when operating an economizer control and provided the optimal temperature for the supply air according to the outdoor air and indoor conditions [26]. Lee et al. evaluated the energy-saving effects by changing the set value for the temperature of the air supply when controlling the dry-bulb temperature and enthalpy and confirmed that additional energy savings were possible compared to existing economizer controls when the set value was changed according to the outdoor temperature [27]. Seong et al. identified the optimal conditions in terms of energy consumption based on changes in the high limit’s set value in order to suggest the optimal economizer control and the optimal setting for the efficient introduction of outdoor air [28]. Bakke evaluated the influence of the set value of the low limit on energy consumption during economizer operation [29]. In existing economizer controls, the set value is kept constant, but energy can be saved by setting an optimal value according to the indoor and outdoor conditions.

4. Prediction Model for Optimal Air-Side Economizer Control

As seen in previous studies, the economizer is an energy-efficient system, and the existing control set values are mixed air temperature and the high and low limits, which determine whether the economizer is operated. These set values are generally kept constant and set without proper consideration of the indoor and outdoor conditions. This can result in energy wastage. Therefore, in order to optimize economizer control, it is necessary to reset the set values according to changes in indoor and outdoor conditions. The indoor and outdoor conditions of a building are affected by numerous factors, including ambient weather conditions, the building’s structure and characteristics, and the operation of sub-components, such as lighting and HVAC systems. Given this complexity, it is difficult to accurately predict these conditions [30].

In accordance with the automation of buildings, technologies such as big data and artificial intelligence are combined with building’s equipment technology to predict buildings’ energy consumption, predict the performance of HVAC systems, optimize systems, and use these as tools for real-time control. Prediction models can be classified as (1) model-driven (white-box), (2) data-driven (black-box), and (3) grey-box according to the modeling method employed. The model-driven method needs a high level of expertise and experience with respect to the basic laws of physics and extensive knowledge and requires many input variables, depending on the model, to be predicted. In contrast, the data-driven method uses the empirical correlation of data without any knowledge of the physical process; hence, it does not require a high level of theoretical knowledge about the output value to be predicted, making it easy to develop a prediction model. Most statistical and machine learning techniques follow this method. Among the data-driven models, the machine learning model, which is associated with a field within computer science that uses knowledge gained through learning from past experiences to make decisions, is attracting attention as a tool for developing prediction models for buildings. Finally, the grey-box method uses a combination of the model-driven and the data-driven methods.

Various studies have been conducted to develop prediction models for building control. Li et al. predicted the cooling load using an SVM (support vector machine) model [31]. Simon et al. predicted the cooling load using a data-based PENN model [32]. Deba et al. developed a cooling and heating load prediction model for a building using an ANN (artificial neural network) model [33]. Chou et al. developed a heating and cooling load prediction model using an ANN, SVR, CART, CHAID, and GLS and derived an optimal model by comparing the developed models [34]. Xu et al. developed a load prediction model for residential buildings using the MLP neural network model and optimized the developed model using various optimization algorithms [35]. Kim et al. developed an indoor CO₂ concentration prediction model. At this time, the ANN, random forest (RF), SVM, and k-nearest neighbor (KNN) algorithms were used, and five-point data for variable air volume (VAV) terminal unit control and monitoring were used. In addition, as a result of the performance evaluation and comparison of the developed model, it was confirmed that the measured value and the predicted value had a relative error within about 5%, and the ANN model had excellent performance among the machine learning models [36]. Kang et al. developed a model to predict the cooling energy consumption of the VFR system using an ANN model. Further, through a performance evaluation of the prediction model, it was confirmed that the accuracy of the prediction model satisfied the evaluation criteria [37]. Ahmad et al. predicted the electric energy consumption of buildings using artificial intelligence techniques such as SVMs and ANNs [38]. Guang et al. developed a ESN-based neural network model to predict the cooling energy [39]. Kim et al. produced an ANN, an SVM, and a Gaussian process model that predicted the energy consumption of an office building [40]. Mocanu et al. derived an optimal model by using various machine learning techniques, such as the neural network model, Boltzmann machine learning, and SVR, to predict energy consumption [41]. Shin et al. developed a performance prediction model for an air-cooled heat pump system using ANN, SVM, RF, and K-NN models [42]. Choi et al. developed an occupancy prediction model based on the recurrent neural network model and, by optimizing the developed model and evaluating its performance, derived the optimal model for each prediction time [43]. Kwon et al. predicted the cooling load of a large building using an ANN model and, based on the cooling load prediction model, confirmed that the energy consumption for cooling and production costs can be reduced by applying the model to the chiller of an actual large building [44]. Sholahudin S. et al. predicted the load using a dynamic neural network model for the operation of an HVAC system and confirmed that it accurately predicted the instantaneous heating load using only a small number of input variables [45]. Kallio et al. developed an indoor CO₂ concentration prediction model using a machine learning model with the goal of improving buildings’ energy efficiency through the prediction of indoor environmental changes [46]. Shin et al. developed a COP prediction model using an ANN and SVM for the operation of a hybrid geothermal system [47]. Kim et al. used an ANN model to optimize the supply air flow rate and temperature of a VAV system [48]. Lee et al. developed an energy prediction model for a building using an ANN to derive the optimal control set values that take into account real-time environmental changes during HVAC control [49]. Likewise, these prediction models are widely used for HVAC systems and building control.

Lee et al. developed a load prediction model using an ANN model and proposed an economizer control method that varies the mixed air temperature set value using the developed model and evaluated it. As a result of the evaluation, it was confirmed that energy was saved compared to the existing control method (Figure 5) [25]. Lee et al. proposed a control method that resets the mixed air temperature by considering indoor and outdoor air conditions when controlling the economizer. For this, a prediction model that predicts indoor CO₂ concentration and energy consumption was developed. Based on the predicted results from the developed model, the set values of mixed air temperature are derived in real time. The developed control was evaluated using a co-simulator of TRNSYS and Matlab, and it was confirmed that energy was reduced compared to the existing dry-bulb temperature control [50]. In this way, it can be concluded that the prediction model can be used for optimal economizer control.

5. Conclusions

In this paper, existing studies on economizer systems, control methods, and set values were reviewed. In addition, the existing prediction models for buildings were reviewed, and the detailed conclusions are as follows:

(1)

An economizer system that reduces energy consumption during cooling by introducing outdoor air was developed, which can further be divided into passive and active systems. Also, active systems are classified into water-side economizers and air-side economizers, and air-side economizers are applied to HVAC and HRV systems. The air-side economizer applied to the HVAC system introduces outdoor air for cooling through the damper to adjust of the outdoor, ventilation, and exhaust damper to maintain the mixed air temperature set value when the outdoor air temperature or enthalpy is lower than the indoor air temperature or enthalpy.

(2)

Economizer control methods include dry-bulb temperature control and enthalpy control, and in terms of energy, enthalpy control can save more energy than dry-bulb temperature control. In addition, the economizer set values include high and low limits and mixed air temperature. In the existing control method, these set values were constant values, so changing indoor and outdoor environments were not considered. Having a fixed set value can cause problems such as discomfort, reduced indoor air quality, and energy waste. Therefore, in order to optimize economizer control, it is necessary to reset the set values considering the indoor and outdoor environments.

(3)

It is difficult to predict changing indoor and outdoor conditions because this is affected by various factors. As buildings are becoming more automated, technologies such as big data and artificial intelligence are being combined with building technology to predict building conditions and HVAC system performance in advance, and they are used as tools for system optimization through real-time control. Therefore, it is expected that real-time optimal control is possible by predicting the factors affecting the economizer control set value.