Next Article in Journal
Hyperspectral Imaging for Sustainable Waste Recycling
Previous Article in Journal
Does Coal Consumption Control Policy Synergistically Control Emissions and Energy Intensity?
 
 
Search for Articles:
Title / Keyword
Author / Affiliation / Email
Journal
Article Type
 
 
Advanced Search
 
Section
Special Issue
Volume
Issue
Number
Page
 
Journals
Sustainability
Volume 15
Issue 10
10.3390/su15107751
sustainability-logo
Submit to this Journal Review for this Journal Propose a Special Issue
► Article Menu

Article Menu

share Share announcement Help format_quote Cite question_answer Discuss in SciProfiles thumb_up
...
Endorse textsms
...
Comment
first_page
settings
Order Article Reprints
Font Type:
Arial Georgia Verdana
Font Size:
Aa Aa Aa
Line Spacing:
Column Width:
Background:
Article

Annual Effect of the VRF Control Algorithm in Response to the TOU Rate Plan

by
Je-Hyeon Lee
1 and
Young-hak Song
2,*
1
Department of Refrigeration and Air-conditioning, Pukyong National University, Yongso-ro 45, Nam-gu, Busan 48513, Republic of Korea
2
Department of Architectural Engineering, ERI, Gyeongsang National University, Jinju-daero 501, Jinju City 52828, Republic of Korea
*
Author to whom correspondence should be addressed.
Sustainability 2023, 15(10), 7751; https://doi.org/10.3390/su15107751
Submission received: 24 February 2023 / Revised: 21 April 2023 / Accepted: 25 April 2023 / Published: 9 May 2023
Browse Figures
Versions Notes

Abstract

:
Many countries adopt a time-of-use (TOU) rate system, in which electricity rates vary by season and time of day, to reduce power usage during peak power consumption hours. South Korea offers a TOU rate plan that depends on the electricity usage of a building and its contracted power; in this plan, the electricity rate reaches up to 300% depending on the time of day. Hence, electrically powered variable refrigerant flow (VRF) systems are increasingly being installed in small- and medium-sized buildings requiring individual cooling and heating operations. This study aims to develop a new control algorithm to reduce electricity consumption and electricity rates for cooling and heating by VRF systems in university buildings adopting the TOU rate plan and apply it to actual buildings to verify the reduction effect. The proposed control algorithm primarily consists of a module that controls the refrigerant evaporation temperature (cooling) and high pressure (heating) according to the indoor heat load and a module that controls the indoor set temperature based on the hourly electricity rate. The developed algorithm was installed in the controller of a VRF system installed in an actual university building and the annual effect was verified using the method proposed by the International Performance Measurement and Verification Protocol. As a result, power consumption was reduced by 17.8% for heating and 4.0% for cooling due to the application of the control algorithm, and the electricity rates reduced by 19.2% and 7.3%, respectively.

1. Introduction

To reduce greenhouse gas (GHG) emissions, nations around the world have voluntarily submitted their GHG reduction targets since 2016, and all parties have agreed to submit their long-term greenhouse gas emission development strategies (LEDS) and nationally determined contributions (NDCs) to keep the global average temperature increase to below 2 °C by 2020 under the Paris Agreement, with the aim of achieving 1.5 °C [1]. South Korea aims to reduce its carbon emissions by 40% in 2030 compared to 2018, and to achieve this, the building sector has set a target of reducing its carbon emissions from 52.1 million tons of CO2eq in 2018 to 35.0 million tons of CO2eq in 2030, which is a 32.8% reduction [2]. South Korea is proposing electrification in the building sector as the direction toward carbon neutrality, so it is necessary to take measures in terms of demand management, such as energy efficiency and load management, to counteract future increases in electricity demand.
The HVAC system is an essential element in commercial and residential buildings and industrial processes that accounts for more than 50% of the total energy used in buildings. Thus, the importance of improvements in energy efficiency in the HVAC system has continuously increased in an effort to save building energy [3]. A variable refrigerant flow (VRF) system is an air conditioning system that uses electricity as an energy source and is attracting attention as a highly efficient device compared to existing central air conditioning systems. The VRF system consists of a single outdoor unit and multiple indoor units. The main components of the outdoor unit are the compressor and condenser, and those of the indoor unit are the evaporator and expansion valve. The outdoor unit and each indoor unit are connected by refrigerant piping, and more recently, eco-friendly refrigerants, such as R32, have been supplied directly to the indoor units to handle the indoor heat load from an evaporator [4]. The VRF system varies the refrigerant flow rate using variable speed compressor(s) in the outdoor unit and the electronic expansion valves (EEVs) located in each indoor unit. The system meets the space cooling or heating load requirements by maintaining the zone air temperature at the set point. The ability to control the refrigerant mass flow rate according to the cooling and/or heating load enables the use of many indoor units with differing capacities in conjunction with one single outdoor unit. This unlocks the possibility of having individualized comfort control, simultaneous heating and cooling in different zones, and heat recovery from one zone to another. It may also lead to more efficient operations during part-load conditions [5].
Meanwhile, the time-of-use (TOU) electricity rate plan was introduced for demand management to induce changes in electricity consumption patterns by making electricity more expensive during peak hours. Korea’s electricity tariff system is divided into residential, educational, industrial, agricultural, street lighting, and general use contracts according to purpose and currently provides TOU plans with different electricity rates by season (summer, winter, spring, and fall) and time of day (off-peak, mid-peak, and on-peak) for residential, educational, industrial, and general use in certain areas. In Korea’s TOU rate plan, demand and energy charges are determined by the contract demand, the voltage provided, and the option selected. For example, the TOU rate plan for educational buildings with a contracted power of 1000 kW or more is shown in Table 1, and the breakdown for season and time of day is shown in Table 2 [6]. Depending on the voltage amount, the voltage types provided are divided into high-voltage A (3300–66,000 V) and high-voltage B (154,000 V), and users can select Option Ⅰ or Option Ⅱ for each voltage type. High voltage is the voltage supplied to the building by the power company, which determines the capacity of the building’s electrical system. To change between the types of high voltage, it is necessary to change the capacity of the electricity facilities, so temporary changes based on season or period do not occur. Option I has relatively lower demand and higher energy charges, while Option II has higher demand charges and lower energy charges. Since the electricity bill is determined by combining the demand and energy charges, a user should select their option considering energy consumption and usage patterns. The TOU rate plan was introduced for electricity demand management, but in practice, buildings are not using electricity in a way that considers this. Reducing demand during peak hours and shifting demand to off-peak hours is important for demand management effectiveness and increasing user benefits.
Continuous research on the VRF system has been conducted to date. While the control strategy was rarely reported on account of confidentiality protection in the early stage of technology development, recently, many research findings have been reported on relevant subjects, such as comfort control, energy efficiency control, improved humidity control, and defrost operation control [7]. Kim et al. [8] proposed a comfort algorithm that used temperature, humidity, and airflow as control variables. The algorithm operated based on the comfort evaluation index and percentage dissatisfied (PD) value presented in the ASHRAE standard. When the indoor temperature and humidity conditions were outside the comfort zone, air conditioning control was performed by inversely calculating the temperature, humidity, and airflow values to maintain the prescribed comfort conditions. Moon et al. [9] proposed a comfort control algorithm that first calculated the initial temperature and humidity conditions that could reach the thermal comfort range during cooling operations through simulation. Thereafter, the indoor conditions were classified into three categories: cooling, cooling and humidification, and humidification. The HVAC system was operated accordingly to achieve optimal comfort conditions for each category based on the corresponding indoor temperature and humidity. For energy efficiency control, a method has been proposed for controlling the evaporating temperature in cooling operations and condensation temperature in heating operations. Generally, the refrigerant mass flow rate discharged from the compressor was controlled to a constant evaporating temperature or condensation temperature [10]. However, through theoretical analysis, Zhao et al. [11] suggested that this control method causes significant energy consumption due to the expanded pressure ratio of the compressor under partial load conditions and proposed a variable control of evaporating temperature and condensation temperature to improve the energy efficiency. Yun et al. [12,13] developed an algorithm that determines the indoor heat load based on the operating state of the indoor unit and changes the refrigerant evaporation temperature set value during cooling operations and high-pressure set value during heating operations according to the result. By examining the effects through simulation, energy consumption was confirmed to be reduced by 14% for cooling and 22% for heating. Furthermore, Yun et al. applied this refrigerant evaporating temperature control algorithm to actual office buildings and educational buildings and verified the energy-saving effect of the algorithm and its effect on room temperature through a month-long field test [14].
Several studies have been conducted on HVAC system control responses to the TOU rate plan. Aftab et al. [15] conducted a study on air conditioning based on residents’ activities and living patterns to reduce energy and operation costs in HVAC systems, where the TOU charge scheme was applied. The occupancy density was identified using video processing and machine learning, with pre-cooling and pre-heating operations used to reduce energy consumption by approximately 30% during the time slot with lower energy tariffs. This was verified via video processing and machine learning techniques. Additionally, Kang et al. [16] also studied energy cost reduction after predicting power generation amount and load patterns due to renewable energy and hourly electricity tariffs using a building energy management system (BEMS) and taking the prediction results into account for the building operation schedule. The actual measurement results using the optimal operation scheduling based on the long-term data obtained through the BEMS show that the effect of energy reduction was achieved by about 20%. In addition, Avci et al. [17] conducted a study on changing indoor setup temperatures by 0.25 °C increments according to variation in electricity rates and setting the temperature to 25 °C at off-peak rate times and 26 °C at peak rate times, rather than setting a single specific value for the indoor temperature during summer. The simulation analysis results show that energy usage and costs were reduced by 23.6 and 24%, respectively. Previous research has verified that controlling the refrigerant evaporating temperature and high pressure of a VRF system based on load can make the system operate more efficiently. In addition, for the TOU rate plan, precooling/preheating and room temperature control can reduce demand during peak hours and shift demand to off-peak hours, resulting in significant electricity bill savings. However, most previous studies were based on historical operation data or simulations due to the practical problems of studying building air conditioning systems. Based on the results of previous studies, this study aimed to develop a control algorithm to effectively respond to the TOU rate plan, by controlling the indoor set temperature based on the hourly electricity rate and the refrigerant evaporating temperature and high pressure according to the indoor heat load of the VRF systems installed in university buildings. This control algorithm was applied to an actual building to quantitatively analyze energy consumption and operation cost reduction effects based on one year of heating and cooling operations data.

2. Overview of TOU Response Control Algorithm

Figure 1 shows a conceptual diagram of the proposed algorithm. The control algorithm developed to respond to the TOU rate plan consists of two modules: The first module is an operation that increases the efficiency of the VRF system by controlling the refrigerant evaporating temperature (cooling) and high pressure (heating) according to the indoor load. The second module is an operation that controls the indoor temperature according to the hourly electricity price to reduce the electricity bill through peak shifting of the load.
The first module is responsible for what is referred to as “high-efficiency operation” and was applied to all sections except the pre-cooling and pre-heating sections (t1–t2). Excluding the pre-cooling and pre-heating sections (t1–t2), the refrigerant evaporating temperature and high pressure were maintained at their original values because it was necessary to be able to quickly lower or increase the room temperature in those sections. Figure 2 shows a flow chart for high-efficiency operation. In VRF systems using refrigerant as a medium, the refrigerant is a mixture of liquid and gas phases; thus, its flow rate cannot be measured accurately. Hence, the system performance cannot be accurately measured or calculated in practice. Therefore, a previous study [14] proposed a method for predicting the indoor heat load by calculating the required capacity (Q_req) of each indoor unit through the change in indoor temperature and difference between the indoor temperature and the set temperature. For the indoor unit in operation, the difference (dT1) between the current (Tindoor_t) and previous (Tindoor_t−1) indoor temperatures is calculated to identify the changing trend of the indoor temperature, and the difference (dT2) between the indoor temperature (Tindoor_t) and set temperature (Tset) is calculated. Based on dT1 and dT2, the cooling capacity of each indoor unit is calculated to obtain the indoor heat load factor (QPLR) by dividing the sum of the cooling capacities of all indoor units by the rated capacity of the outdoor units. The refrigerant evaporating temperature was selected depending on the indoor heat load factor: 8 °C for a load factor of 0.8 or higher, 11 °C for a load factor of 0.6–0.7, and 14 °C for a load factor of 0.5 or less. The high pressure value was selected as 30 kg/cm2 for a load factor of 0.8 or more, 28 kg/cm2 for a load factor of 0.6–0.7, and 25 kg/cm2 for a load factor of 0.5 or less. Indoor heat load factors of 0.7–0.8 and 0.5–0.6 were set as hysteresis sections to prevent repeated increases and decreases in refrigerant evaporating temperature and high pressure.
Pre-cooling or pre-heating is an operation that lowers or raises the indoor set temperature within a comfortable range, lower or higher than that set by the user, during the time period (t1–t2) prior to peak electricity rate periods. The comfortable temperature range can be calculated using various comfort indicators, but in this study, the user-set temperature (Tset) ± 1 °C was conservatively used. The t1–t2 time period was set to a range of 10–60 min to prevent an increase in energy consumption resulting from excessive pre-cooling and pre-heating. The time required for the room temperature to reach Tset ± 1 °C, i.e., the pre-cooling or pre-heating temperature, was recorded and applied to the same time period the next day. For example, if electricity rates are high at 13:00 in cooling operations, and it took 30 min to reach the pre-cooling temperature of Tset − 1 °C the previous day, on the next day, pre-cooling starts at 12:30. If the pre-cooling temperature is reached 20 min later (at 12:50), pre-cooling starts at 12:40 the next day; however, if the room temperature fails to reach Tset − 1 °C within 30 min, pre-cooling starts at 12:25, with an increase of 5 min the next day. A study to determine the appropriate time using a predictive model in terms of the pre-cooling time will be needed in the future. Indoor heat storage exhibits a time lag phenomenon, which has the effect of dispersing peak load by storing heat in structures such as indoor spaces, building walls, and floor slabs. Thus, through pre-cooling, electricity rates can be reduced in buildings adopting the TOU rate plan [18,19].
Flexible operation is performed in the time zone (t2–t3) when pre-cooling or pre-heating is completed and electricity rates are high. Flexible operation gradually raises the room temperature within the comfortable temperature range. In this study, the comfortable temperature range was set at Tset + 1 °C, and the room temperature of Tset − 1 °C was increased to Tset + 1 °C due to pre-cooling during the cooling operation. However, at this time, if the indoor set temperature was changed suddenly from Tset − 1 °C to Tset + 1 °C, the room temperature would be 2 °C lower than the set temperature. Consequently, all indoor units would switch to thermo-off, and the compressor would stop. As a result, the room temperature would rise rapidly, causing the occupant to feel uncomfortable. Thus, to prevent this problem and operate the VRF system more efficiently, the compressor current was limited to 50% when the load ratio was 50% or more, based on the calculated indoor heat load. Using this constraint, a VRF system equipped with an inverter could be operated at part load with high efficiency to save energy [20] while suppressing occupants’ discomfort by gradually increasing the room temperature without handling all the load. The HVAC system sets the indoor temperature to the highest temperature within the comfortable range and maintains it until the end of the expensive charge period (t3–t4). When the rate returns to a lower rate, the room temperature is changed to the temperature set by the user.

3. Methods

To verify the energy and operating cost savings of the developed control algorithm in response to the TOU rate plan, a field test on a VRF system installed in a real building was conducted. In this section, the target building, VRF system, field test, and effect verification method are described.

3.1. Target Building and VRF System

The target building for performance evaluation on the algorithm was a two-story building whose total floor area was 3005 m2. The building consisted of lecture rooms, laboratories, staff rooms, and research rooms for graduate students at the university, located in Tongyeong-si, Gyeongsangnam-do, Korea. The contracted power was 4200 kW, and the following charging scheme was applied due to the TOU pricing scheme: high-voltage A for educational purposes and Option II (refer to Table 1 and Table 2). The VRF system consisted of outdoor units with capacities of 87, 93, and 97 kW and 39 indoor units of 1- and 4-way cassette types. A control device equipped with the algorithm and a power meter measuring outdoor unit power consumption were installed to verify the effect of the algorithm. In this study, the power consumption of the indoor units was not considered.

3.2. Field Test and Effect Verification Method

Table 3 presents operational conditions according to the algorithm applied. To verify the effect of the algorithm, driving with the applied algorithm was alternated daily with driving without the application to minimize the effect of differences in boundary conditions, such as outside air conditions and indoor heating load, on energy consumption. The demonstration was conducted on weekdays and weekends from 08:00 to 21:00, when most heating and cooling operations occurred in the target building, and the occupants freely controlled the on/off settings of the VRF system. In heating operations, when the algorithm was on, the indoor temperature was set at 23 °C while pre-heating and flexible operation temperatures were set at 24 and 22 °C, respectively. Based on the load, the high pressure control was set at 25, 28, and 30 kg/cm2. In contrast, when the algorithm was off, the operation temperature was set at 23 °C, with the high-pressure refrigerant controls set at 30 kg/cm2. The cooling operation followed a similar control method in the winter period, with the indoor reference temperature set to 25 °C, and the pre-cooling and flexible operation temperatures were set to 24 °C and 26 °C, respectively, when the algorithm was on. Furthermore, the refrigerant temperature was adjusted to 8, 11, or 14 °C depending on the indoor heat load. In contrast, when the algorithm was off, the indoor set temperature was set to 25 °C, with the refrigerant temperature control fixed at 8 °C.
Two methods were adopted to apply the developed control algorithm to an actual university building and verify its effectiveness: a method of simply comparing operation data before and after algorithm application and a method proposed by the International Performance Measurement and Verification Protocol (IPMVP). The IPMVP, an international protocol that objectively proposes measurement and verification methods and standards for energy performance [21], is classified into Options A, B, C, and D, according to the characteristics and circumstances of the applied target. The scope of the present study encompassed an overall system, including an indoor thermal environment, indoor unit, and outdoor unit with the controller in the target building. Thus, it followed the standards in Option C.

4. Field Test Results

A demonstration was conducted to verify the effect of the algorithm on heating and cooling operations. The heating demonstration period was a total of 73 days from 18 December 2014 to 28 February 2015, with 37 and 36 days of operations with and without the algorithm, respectively. The cooling demonstration period was 101 days from 22 June to 30 September 2015, with 51 and 30 days of operations with and without the algorithm, respectively. The operation data obtained during the demonstration period were analyzed to verify the energy and operation cost reduction effects of the algorithm.

4.1. Heating Demonstration Results

First, Table 4 shows the results of a simple comparative analysis before and after applying the algorithm during the heating demonstration period. All data are expressed as daily mean values. The operating time of each indoor unit was calculated based on the daily operation time of 39 indoor units, while the operation time of each outdoor unit was calculated based on the daily operation time of the three outdoor units. As explained in Section 2, the part load was calculated using the indoor and set temperatures. The coefficient of performance (COP) was calculated by dividing the part load by the measured outdoor unit power consumption. The results indicate that the outside air temperatures during the periods of applying and not applying the algorithm were very similar, at 3.9 and 3.8 °C, respectively. The room temperature was at 24.3 and 25.1 °C for the algorithm on and off conditions, respectively; the algorithm-controlled room temperature was approximately 0.8 °C lower due to the flexible operation section. In addition, the indoor set temperature without algorithm application and the standard indoor set temperature for algorithm application were both 23 °C; thus, the indoor temperature was about 2 °C higher than the set temperature. Considering the sensor that measures room temperature is located inside the cassette-type indoor unit installed on the ceiling and is affected by the hot air discharged from the indoor unit, the air temperature measured is higher than that at the occupant’s location. Therefore, in the VRF system, the room temperature is controlled by adjusting a set difference between the air temperature near the ceiling and where the occupants are located. The field test site in this test was set to have 2 °C of adjustment, i.e., the sensor’s set temperature was 2 °C higher than the actual set temperature. This phenomenon occurs when the warm air in a indoor unit rises by convection, which is observed only during heating operations. During cooling operations, the ceiling air temperature and the air temperature at the occupant’s location are almost the same. The indoor unit operating time was 139.5 h for algorithm application and 147.6 h for non-application, with 8 h of difference, and the outdoor unit operation time was similar, with 24.2 h and 24.1 h, respectively. The part load calculated from the operating condition of the indoor unit was 21.2% for the operation with the algorithm, resulting in a relatively low indoor temperature and a short indoor unit operation time; this was 3.8% higher than that of the operation without the algorithm at 17.4%. With a low load of about 20%, the operation was conducted at a high pressure of 27.2 kg/cm2, which was made possible by the high efficiency of the algorithm. As a result, the COP at the time of application of the algorithm was significantly improved, reaching 2.69, compared to 1.93 without application. As a result, despite the large part load, the daily average power consumption decreased by about 12.7% from 324.6 to 283.5 kWh, with the average daily electricity rate decreasing by approximately 13.9% from 32,943 to 28,368 KRW.
In the simple comparison result for the entire heating period, an increase in the COP and a reduction in power consumption and electricity rates were confirmed as a result of operation with the algorithm. However, this operation differed in boundary conditions, such as outdoor temperature, VRF system operation time, and indoor heat load. As a result, verification of this effect using regression analysis, as proposed by IPMVP, was introduced. A regression analysis was conducted based on the operating data without algorithm application, with a regression model derived to predict the electricity consumption and electricity rates according to each boundary condition. The power consumption and electricity rates calculated by inputting boundary conditions when applying the algorithm to the regression model were compared with the actual measured values. This method enables the verification of the effect under the same boundary conditions. The form of the regression model used in this study is shown in Equations (1) and (2) [14,21], with the model’s coefficients derived through regression analysis, shown in Table 5. The variables used in the regression models of heating power consumption and operating costs were selected through correlation analysis. They were outdoor temperature, cumulative operation time of the outdoor unit per day, and indoor load factor, which were highly correlated [14].
Powerh = a∙Tout + b∙Toper + c∙Qindoor + d
Costh = a∙Tout + b∙Toper + c∙Qindoor + d
Powerh denotes the power consumption of the outdoor heating unit (kWh); Costh denotes the heating electricity rates (KRW); Tout denotes the outdoor temperature (°C); Toper denotes the cumulative operation time of the outdoor unit per day (h); Qindoor denotes the load factor (%); and a, b, c, and d denote regression model coefficients (unitless). The coefficient of determination of the regression model was 0.95 for the power consumption prediction model and 0.96 for the electricity rates prediction model, both exceeding the 0.75 required by the IPMVP and showing sufficient accuracy to be used for effect verification. In addition, the p value of the F statistic was also confirmed to be statistically significant, as both models exhibited values less than 0.05.
Figure 3 shows the results of verifying the heating effect on electricity consumption and electricity rates using regression analysis. The bar graph is the actual measured power consumption and electricity rate, while the dotted line graph is the power consumption and electricity rates calculated through the regression model for operation without the algorithm (baseline). In the period when the algorithm was not applied, the actual measured power consumption and the value calculated through the model were almost consistent. During the algorithm application period, the actual measured power consumption was less than the value calculated by the model. The difference between the measured value and the value calculated from the model indicates a saving effect due to the application of the algorithm at the same outdoor temperature, outdoor unit operation time, and indoor heat load. The same pattern is also seen in the electricity rates. The comparison showed that power consumption and electricity rates were reduced by 17.8% and 19.2% for the heating and cooling models, respectively, due to application of the algorithm.

4.2. Cooling Demonstration Results

Table 6 shows the results of a simple comparative analysis before and after applying the algorithm during the cooling demonstration period. The outdoor air temperature was 25.5 °C with the algorithm, which was approximately 0.6 °C higher than 24.9 °C without application. The room temperature was 26.0 °C with the algorithm and 25.7 °C without, showing a difference of approximately 0.3 °C. As the VRF system controls the room temperature within ±1 °C of the set value, it seems that the room temperature without the algorithm is controlled at 0.7 °C higher than the set temperature of 25 °C. For this reason, even though the algorithm was applied, the difference in room temperature was not as large as expected. The operation time of the indoor unit was 227.1 h with the algorithm and 221.4 h without, and it was 5.7 h longer when applying the algorithm. The operation times of the outdoor unit with and without the algorithm were similar at 28.1 h and 27.6 h, respectively. The part load was 21.9% with the algorithm and 22.5% without. Even during cooling operations, low-load operations with a part load of 20%, similar to that of the heating operations, continued to occur. Due to the low load, the refrigerant evaporating temperature rose significantly to 11.8 °C when the algorithm was applied compared to the previous 8 °C, whereas the COP values were 3.35 and 3.30, respectively, showing no effect on efficiency. When the algorithm was applied, the average daily electricity consumption decreased by approximately 4.1% from 188.9 to 181.2 kWh, and the average daily electricity rates decreased by approximately 6.9% from 21,209 to 19,754 KRW.
A regression model analysis was conducted to verify the effect of the algorithm under the same boundary conditions. Unlike in heating operations, for the cooling operations, the correlation of power consumption and electricity rates with outdoor unit operation time was not high; thus, a regression model was derived with only outdoor temperature and indoor heat load as variables. The regression model equations for cooling operations are shown in Equations (3) and (4) [14,21], and the coefficients of the derived regression model are shown in Table 7. The variables used in the regression model of cooling power consumption and operating costs were selected through correlation analysis. They were the outdoor temperature and indoor load factor, which were highly correlated [14].
Powerc = a∙Tout + b∙Qindoor + c
Costc = a∙Tout + b∙Qindoor + c
Powerc denotes the power consumption of the outdoor cooling unit (kWh); Costc denotes the cooling electricity rates (KRW); Tout denotes the outdoor temperature (°C); Qindoor denotes the load factor (%); and a, b, and c denote the regression model coefficients (unitless). The coefficient of determination of the regression model was 0.93 for the power consumption prediction model and 0.88 for the electricity rates prediction model, both of which satisfied the IPMVP criteria. Furthermore, the p value of the F statistic was also confirmed to be statistically significant, as both models exhibited values less than 0.05. Figure 4 shows the cooling effect verification results using regression analysis. The outdoor temperature and indoor heat load of the operation with the algorithm were entered into the regression model, derived through the operation data without the algorithm, to calculate the power consumption and electricity rate, as well as their effectiveness, by comparing them with the actual measured values. As a result, the operation with the algorithm reduced power consumption by 4.0% and electricity rates by 7.3% compared to the operation without the algorithm at the same outdoor temperature and indoor heat load.

5. Conclusions

This study developed an algorithm to reduce power consumption and electricity rates under a TOU rate plan by changing the control settings of a university building’s VRF system and verified its effect through a long-term field test. The controlled set values were refrigerant evaporating temperature (cooling), high pressure (heating), compressor current limit, and indoor set temperature. By controlling these values according to the electricity rate by time of day and indoor heat load, an algorithm consisting of high-efficiency operation, pre-cooling, pre-heating, flexibility operation, and indoor set temperature shift operation was developed. The developed algorithm was loaded into the control device of the VRF system and applied to a VRF system installed in a university building. For this experiment, heating operations were conducted for 73 days, from 18 December 2014 to 28 February 2015, and cooling operations were conducted for 101 days, from 22 June to 30 September 2015. The algorithm on and off operations were conducted one day at a time in an alternating manner. The algorithm effect was analyzed using a simple comparative analysis and regression analysis proposed by IPMVP. The purpose of the regression analysis was to calibrate for the difference in boundary conditions such as outdoor temperature, indoor load, and outdoor unit operation time between the algorithm on and algorithm off states. In this study, the difference in boundary conditions was not large because the algorithm on and off operations were conducted one day at a time in an alternating manner. As a result, the results of regression analysis and simple comparative analysis were almost similar. Applying the algorithm resulted in a 17.8% reduction in power consumption and a 19.2% reduction in power bills for heating and a 4.0% reduction in power consumption and a 7.3% reduction in power bills for cooling.
During the cooling and heating demonstration periods, the indoor heat load was continuously maintained at 20%, with the high pressure held at 27.2 kg/cm2 by lowering the pressure, while the refrigerant evaporating temperature was operated by raising the evaporating temperature to 11.8 °C. However, the COP improvement effect for high-efficiency operation showed a significant difference between heating and cooling, with a 39.5% improvement in heating and a 1.5% improvement in cooling. In the target VRF system, the cooling was controlled by increasing the evaporating pressure. This was performed by controlling the evaporating temperature without directly controlling the pressure for system stability, while the heating was controlled by directly controlling the pressure. A difference in compressor frequency was observed even when the cooling and heating systems were operated at a similar partial load range. This difference can be attributed to variations in the control method and compressor frequency range; thus, a detailed refrigeration cycle data analysis will be required for an accurate cause analysis. In this study, the comfort range was conservatively set to the user-set temperature ± 1 °C. The reduction effect was insignificant because the difference in room temperature before and after applying the algorithm was insignificant, especially in cooling operations. Subsequent studies will require the utilization of comfort indicators such as predicted mean vote (PMV) and adaptive comfort to improve evaluation of the algorithm.

Author Contributions

Conceptualization, J.-H.L. and Y.-h.S.; Methodology, J.-H.L. and Y.-h.S.; Investigation, Y.-h.S.; Data curation, J.-H.L.; Writing—original draft, J.-H.L.; Funding acquisition, Y.-h.S. All authors have read and agreed to the published version of the manuscript.

Funding

This research was supported by a grant of the project for Infectious Disease Medical Safety, funded by the Ministry of Health & Welfare, Republic of Korea (HG22C0044).

Institutional Review Board Statement

Not applicable.

Informed Consent Statement

Not applicable.

Data Availability Statement

Data was obtained from SAMSUNG Electronics and are available from the authors with the permission of SAMSUNG Electronics.

Conflicts of Interest

The authors declare no conflict of interest.

References

  1. 2050 Carbon Neutral. Available online: https://www.korea.kr/special/policyCurationView.do?newsId=148881562 (accessed on 24 April 2023).
  2. The 2050 Carbo Neutrality and Green Growth Commission. Available online: https://2050cnc.go.kr (accessed on 24 April 2023).
  3. Chua, K.J.; Chou, S.K.; Yang, W.M.; Yan, J. Achieving Better Energy-Efficient Air Conditioning–A Review of Technologies and Strategies. Appl. Energy 2013, 104, 87–104. [Google Scholar] [CrossRef]
  4. Lee, J.H.; Yoon, H.J.; Im, P.; Song, Y.H. Verification of Energy Reduction Effect through Control Optimization of Supply Air Temperature in VRF-OAP System. Energies 2018, 11, 49. [Google Scholar] [CrossRef]
  5. Big Ladder Software. Engineering Reference—EnergyPlus 8.6. Available online: http://bigladdersoftware.com/epx/docs/8-6/engineering-reference/variable-refrigerant-flow-heat-pumps.html#variablerefrigerant-flow-heat-pump-model-system-curve-based-model (accessed on 24 April 2023).
  6. Korea Electric Power Corporation (KEPCO). Available online: https://cyber.kepco.co.kr/ckepco/front/jsp/CY/E/E/CYEEHP00209.jsp (accessed on 24 April 2023).
  7. Zhang, G.; Xiao, H.; Zhang, P.; Wang, B.; Li, X.; Shi, W.; Cao, Y. Review on recent developments of variable refrigerant flow systems since 2015. Energy Build. 2019, 198, 444–466. [Google Scholar] [CrossRef]
  8. Se-Chang, K.; Jung-Min, L.; Tanabe, T.; Suhgoo, K.; Hwan, O.S.; Jae-Boong, C. A Study on Development of Cooperative Control Logic for VRF System. KIBIM 2011, 1, 119–120. [Google Scholar]
  9. Kim, M.H.J.; Won, J. An Experimental Study on the Performance of a VRF Air Conditioning System with a Thermal Comfort Control Algorithm. J. Korean Inst. Archit. Sustain. Environ. Build. Syst. 2015, 9, 247–252. [Google Scholar]
  10. Chen, W.; Zhou, X.; Deng, S. Development of control method and dynamic model for multi-evaporator air conditioners (MEAC). Energy Convers. Manag. 2005, 46, 451–465. [Google Scholar]
  11. Deyin, Z.; Xu, Z.; Ming, Z. Variable evaporating temperature control strategy for VRV system under part load conditions in cooling mode. Energy Build. 2015, 91, 180–186. [Google Scholar]
  12. Young, Y.G.; Hyeon, L.J.; Jun, K.H. Development and application of the load responsive control of the evaporating temperature in a VRF system for cooling energy savings. Energy Build. 2016, 116, 638–645. [Google Scholar]
  13. Young, Y.G.; Hyeon, L.J.; Jun, K.H. Dynamic target high pressure control of a VRF system for heating energy savings. Appl. Therm. Eng. 2017, 113, 1386–1395. [Google Scholar]
  14. Hyeon, L.J.; Sunggeun, S.; Young-hak, S. Verification of refrigerant evaporating temperature control effect in VRF systems in actual buildings. Energy Build. 2021, 250, 111295. [Google Scholar]
  15. Aftab, M.; Chen, C.; Chau, C.-K.; Rahwan, T. Automatic HVAC control with real-time occupancy recognition and simulation-guided model predictive control in low-cost embedded system. Energy Build. 2017, 154, 141–156. [Google Scholar] [CrossRef]
  16. Kang, S.J.; Park, J.; Oh, K.-Y.; Noh, J.G.; Park, H. Scheduling-based ressreal time energy flow control strategy for building energy management system. Energy Build. 2014, 75, 239–248. [Google Scholar] [CrossRef]
  17. Avci, M.; Erkoc, M.; Rahmani, A.; Asfour, S. Model predictive HVAC load control in buildings using real-time electricity pricing. Energy Build. 2013, 60, 199–209. [Google Scholar] [CrossRef]
  18. Pyeong-Chan, I.; Ki-Min, K. An Evaluation of Control Technique for Building HVAC Systems Using Thermal Mass Capacity of Building Structure. AIKRA JAMS 2007, 9, 247–254. [Google Scholar]
  19. Lee, K.-H.; Yang, S.-K.; Han, S.-H. Reducing Peak Cooling Demand Using Building Precooling and Modified Linear Rise of Indoor Space Temperature. SAREK 2010, 22, 86–96. [Google Scholar]
  20. Kang, W.H.; Lee, J.M.; Yeon, S.H.; Park, M.K.; Kim, C.H.; Lee, J.H.; Moon, J.W.; Lee, K.H. Modeling, Calibration, and sensitivity analysis of direct expansion AHU-Water source VRF system. Energy 2020, 199, 117435. [Google Scholar] [CrossRef]
  21. John, W. International Performance Measurement Verification; Efficiency Valuation Organization: Washington, DC, USA, 2012; Volume 1. [Google Scholar]
Sustainability 15 07751 g001 550
Figure 1. Algorithm concept. (a) Cooling. (b) Heating.
Figure 1. Algorithm concept. (a) Cooling. (b) Heating.
Sustainability 15 07751 g001
Sustainability 15 07751 g002 550
Figure 2. Flow chart of the high-efficiency operation.
Figure 2. Flow chart of the high-efficiency operation.
Sustainability 15 07751 g002
Sustainability 15 07751 g003 550
Figure 3. Heating effect verification results. (a) Power consumption. (b) Electricity rates.
Figure 3. Heating effect verification results. (a) Power consumption. (b) Electricity rates.
Sustainability 15 07751 g003
Sustainability 15 07751 g004a 550Sustainability 15 07751 g004b 550
Figure 4. Cooling effect verification results. (a) Power consumption. (b) Electricity rates.
Figure 4. Cooling effect verification results. (a) Power consumption. (b) Electricity rates.
Sustainability 15 07751 g004aSustainability 15 07751 g004b
Table
Table 1. TOU rate plan for educational buildings.
Table 1. TOU rate plan for educational buildings.
ClassificationDemand Charge
(KRW/kW)		Energy Charge (KRW/kWh)
Time PeriodSummer
(June to August)		Spring, Fall
(March to May, Sepember to October)		Winter
(November to February)
High-voltage AOption Ⅰ6090Off-peak68.568.572.5
Mid-peak113.282.9111.7
On-peak179.1103.4150.4
Option Ⅱ6980Off-peak64.064.068.0
Mid-peak108.778.4107.2
On-peak174.698.9145.9
High-voltage BOption Ⅰ6090Off-peak67.067.070.8
Mid-peak110.581.2108.8
On-peak173.4101.0146.1
Option Ⅱ6980Off-peak62.562.566.3
Mid-peak106.076.7104.3
On-peak168.996.5141.3
Table
Table 2. Season and time period classification.
Table 2. Season and time period classification.
ClassificationSummer, Spring, FallWinter
Off-peak23:00–09:0023:00–09:00
Mid-peak09:00–10:00
12:00–13:00
17:00–23:00		09:00–10:00
12:00–17:00
20:00–22:00
On-peak10:00–12:00
13:00–17:00		10:00–12:00
17:00–20:00
22:00–23:00
Table
Table 3. Operation conditions.
Table 3. Operation conditions.
OperationIndoor Set Temp.High Pressure (Heating) and
Refrigerant Temp. (Cooling)
Algorithm offEven days
(08:00–21:00)		Heating: 23 °C
Cooling: 25 °C		High pressure: 30 kg/cm2
Refrigerant temp.: 8 °C
Algorithm onOdd days
(08:00–21:00)		Heating: 23 ± 1 °C
Cooling: 25 ± 1 °C		High pressure: 25, 28, 30 kg/cm2
Refrigerant temp.: 8, 11, 14 °C
Table
Table 4. Summary of heating results.
Table 4. Summary of heating results.
Algorithm onAlgorithm off∆ *Savings Rate (%) **
Outdoor temperature (°C)3.93.8−0.1-
Indoor temperature (°C)24.325.10.8-
Operation time of indoor unit (hour/day)139.5147.68.15.5
Operation time of outdoor unit (hour/day)24.224.1−0.2−0.7
High pressure (kg/cm2)27.230.02.8-
Part load (%)21.217.4−3.8−21.7
COP (-)2.691.93−0.76−39.5
Electricity (kWh/day)283.5324.6−41.112.7
Operation cost (KRW/day)28,36832,9434574.813.9
* Value of (Algorithm off − Algorithm on). ** (∆/Algorithm off value) × 100 (%).
Table
Table 5. Heating regression model coefficients.
Table 5. Heating regression model coefficients.
abcd
Power consumption−8.9718910.205814.9866326.71493
Electricity rates−922.877481028.77608537.234392388.92745
Table
Table 6. Summary of cooling results.
Table 6. Summary of cooling results.
Algorithm onAlgorithm off∆ *Saving Rate (%) **
Outdoor temperature (°C)25.524.9−0.6-
Indoor temperature (°C)26.025.7−0.3-
Operation time of indoor unit (hour/day)227.1221.4−5.7−2.6
Operation time of outdoor unit (hour/day)28.127.6−0.5−1.9
Refrigerant evaporating temperature (°C)11.88.0−3.9-
Part load (%)21.922.50.62.8
COP (-)3.353.300.051.5
Electricity (kWh/day)181.2188.97.74.1
Operation cost (KRW/day)19,75421,20914556.9
* Value of (Algorithm off − Algorithm on). ** (∆/Algorithm off value) × 100 (%).
Table
Table 7. Cooling regression model coefficients.
Table 7. Cooling regression model coefficients.
abc
Power consumption8.176918.60785−208.54925
Electricity rates1174.760361062.31636−31,961.83441
Disclaimer/Publisher’s Note: The statements, opinions and data contained in all publications are solely those of the individual author(s) and contributor(s) and not of MDPI and/or the editor(s). MDPI and/or the editor(s) disclaim responsibility for any injury to people or property resulting from any ideas, methods, instructions or products referred to in the content.

Share and Cite

MDPI and ACS Style

Lee, J.-H.; Song, Y.-h. Annual Effect of the VRF Control Algorithm in Response to the TOU Rate Plan. Sustainability 2023, 15, 7751. https://doi.org/10.3390/su15107751

AMA Style

Lee J-H, Song Y-h. Annual Effect of the VRF Control Algorithm in Response to the TOU Rate Plan. Sustainability. 2023; 15(10):7751. https://doi.org/10.3390/su15107751

Chicago/Turabian Style

Lee, Je-Hyeon, and Young-hak Song. 2023. "Annual Effect of the VRF Control Algorithm in Response to the TOU Rate Plan" Sustainability 15, no. 10: 7751. https://doi.org/10.3390/su15107751

Note that from the first issue of 2016, this journal uses article numbers instead of page numbers. See further details here.

Article Metrics

Back to TopTop