1、Hye-Sun Jin and Jin-Kyung Kang is a master degree student at Department of Architectural Engineering, Ewha Womans University, Seoul, South Korea. Bo-Hye Choi and Sung-Im Kim is a Ph.D student at Department of Architectural Engineering, Ewha Womans University, Seoul, South Korea. Jae-Han Lim and Seun
2、g-Yeong Song is a professor at Department of Architectural Engineering, Ewha Womans University, Seoul, South Korea. Conditional Demand Analysis for Estimating the Electric Energy Consumption of Household Facilities in Apartment Buildings Hye-Sun Jin Bo-Hye Choi Sung-Im Kim ASHRAE Student Member ASHR
3、AE Student Member ASHRAE Student Member Jin-Kyung Kang Jae-Han Lim, PhD Seung-Yeong Song, PhD ASHRAE Student Member ASHRAE Associate Member ASHRAE Associate Member ABSTRACT The purpose of this paper is to derive interaction variables for Conditional Demand Analysis (CDA), which can estimate the elec
4、tric energy consumption of a specific household facility. There have been considerable attempts to reduce household energy consumption because households, especially apartments, are considered as a substantial consumer of electric energy. Although it will be helpful to determine the energy consumpti
5、on at the end-use level, there is only information about the total energy consumption of primary energy, such as electricity, natural gas, oil, and district heating. Therefore, it is necessary to disaggregate the total energy demand into components that are attributable to specific end uses, such as
6、 heating, cooling, ventilation, domestic hot water, lighting, cooking, appliances, etc. With this information, the use of electric energy, especially by household facilities, can be easily controlled by occupants because it reflects the behavior of occupants directly. Using this information, we can
7、easily determine how much electric energy has been used based on the occupants behavioral patterns and, make decisions to reduce energy costs and choose energy efficient facilities. CDA is a methodology that is based on statistical regression of end-use facilties. Using this method, we can appropria
8、tely estimate the distribution of the total electric energy among end-uses. Furthermore, it can be directly related to the guidelines for residents to voluntarily participate in preventing excess electric energy consumption. INTRODUCTION Recently, interest in reducing energy consumption and greenhou
9、se gas emissions has increased in all economic sectors. The national energy consumption of buildings comprises 16-50 % of the toal energy consumption and is equivalent to 30 % of the worldwide average (Lukas G. 2009). The energy consumption in apartment buildings in Korea is increasing steadily by a
10、n average of 2.8 % per year because of increased residential area, large-sized household appliances and so on (Korea Energy Statistics Information System, KESIS). Because of these factors, apartment buildings are perceived as a national energy consumption source, and numerous attempts have been made
11、 to reduce energy consumption in apartment buildings. Accordingly, energy-related policy, which has been focused solely on estimates of future energy consumption during the last decade, has been changed to manage and limit the demand for energy through various measurements of energy demand by integr
12、ating building information from the Building Administration Information System (BAIS) and energy consumption data, including electricity, gas, and district heating provided by energy suppliers. However, because the building energy consumption statistics data are limited mostly to electricity, gas, o
13、r other primary energy types, it is very difficult for actual users of the building, i.e., building residents and its owner, to understand details about building energy use patterns and to make them voluntarily participate in building energy savings. To let them know the exact energy consumption sta
14、tus and to provide guidelines for energy savings, there is a need to grasp a detailed overview of end-use energy consumption, including heating, cooling, domestic hot water, ventilation, lighting, and cooking based on the apartments monthly and annual total energy consumption information. Direct and
15、 accurate measurement of the end use of energy in apartment building is ideal; however, because considerable costs and time can be incurred, the need for estimating end-use energy consumption through the end-use energy consumption model will be pointed out. To that end, an analysis model of end-use
16、energy consumption will be derived. Therefore, in this study, after consideration of end-use energy consumption models, we define end-use energy consumption models for electricity, gas, and district heating systems in apartment buildings in Korea. In addition, the major interaction variables of cond
17、itional demand analysis are determined, and an estimate of the end-use energy consumption in apartment buildings is made using regression analysis based on the classification of end-use energy. DEFINITION OF A HOUSEHOLD END-USE ENERGY CONSUMPTION MODEL An end-use energy consumption model is a set of
18、 equations designed to disaggregate a households total annual fuel consumption into end uses, such as heating, cooling, ventilation, domestic hot water, and so on. Using this model, information about the distribution of the total energy consumption based on the end-use of the energy can be obtained.
19、 This energy consumption information can be used to predict and address future energy demands from the national perspective and to improve building energy efficiency from the standpoint of residents. The end-use energy of apartment buildings from the Building Energy Efficiency Rating Systems or the
20、Building Energy Consumption System Regulates of Korea is now divided into 4 categories, i.e., heating, domestic hot water, ventilation, and lighting, whereas, it is classified into 7 categories, i.e., heating, cooling, domestic hot water, ventilation, lighting, cooking, and appliances, based on ISO1
21、2655 (energy performance of buildings presentation of measured energy use of buildings). As such, this study categorizes the end-use energy consumption in accordance with the domestic legal basis in the first phase, and additionally the categories that form a large portion of the other end uses in a
22、ctual buildings are classified according to ISO12655. This study designated end-use energy consumption as the energy consumption targeted for billing and determined end-use energy uses as heating, cooling, ventilation, lighting, cooking, and appliances. The general structure of the end-use energy co
23、nsumption model is shown as Figure 1. Figure 1 General structure of End-Use Energy Model. After considering the general structure of end-use energy consumption, we propose end-use energy consumption models that are adequate for domestic circumstances in apartment buildings of Korea. Based on the def
24、inition and classification of the previously mentioned end-use energy model, every detail of the end-use energy consumption is defined. In general, the heating systems can be divided into individual heating and district heating systems, which are fit for the embedded radiant floor heating system of
25、apartment buildings in Korea, and analyze the end-use energy consumption of each system (Figure 2). The heating energy source from individual heating system is either completely liquefied natural gas (LNG) or electricity. The energy used to produce hot water by the district heating systems is the to
26、tal energy consumption during heating periods including the piping losses, electric energy for hot water circulating pumps and heat exchangers efficiency, instead of the heating load of housing unit. That is, the energy required to produce and supply heat by the District Heating Corp. is not taken i
27、nto account. The cooling energy is the electric energy used to provide air conditioning to the building, in which has been equipped with packaged type air conditioners (PAC) in general. The energy of the domestic hot water supply is the energy required to produce and transport hot water from the hot
28、 water supply installations in the building. The ventilation energy is the electric energy measured through the ventilation unit in the individual apartment units (in this case, as in the others, the ventilation fans for parking space in basements are included in the energy for appliances). The ligh
29、ting energy is the energy associated with electric lighting devices. The cooking energy is the energy used for cooking within the individual apartment units. The appliance energy is the energy consumed by home appliances, etc. In the case of individually heated apartments, the gas consumption is mea
30、sured through gas meter, and electricity is measured through a Watt-hour meter, a panel board and a household unit, in order. In the case of apartments with district heating, the details are equal to individual heating for the end-use energy; in terms of energy flow, district heating is measure with
31、 a calorimeter (flowmeter) for heating and hot water supply and finally used for heating and hot water energy. (Refer to Figure 2) Figure 2 (a) Household End-Use Energy Consumption Model in Individual Heating System and (b) Household End-Use Energy Consumption Model in District. Method for Estimatin
32、g Household End-use Energy Consumption As can be seen via the relevant model, the amount of total energy consumption in an apartment building can be expressed as the sum of its end-use energy consumption. In other words, by considering interaction variables related to the characteristics of weather,
33、 construction, occupancy and use, from the total energy consumption value for electricity, gas, and district heating, each end-use energy consumption value of heating, cooling, ventilation, domestic hot water, lighting, cooking, and appliances can be derived. Calculated in this manner, the equation
34、can be given as Equation (1): (1) Estimation of End-use Energy Consumption. The techniques used to model residential end-use energy consumption can be grouped into two categories: “top-down” and “bottom-up” as shown as Fig 3 (Lukas G. 2009) Figure 3 Estimation method of End-Use Energy Consumption In
35、 the case of the top-down approach, we can analyze how the total energy consumption can be compared to the national indicators, such as macroeconomic indicators (gross domestic product (GDP), employment rates, and price-indices), climate conditions, housing construction/demolition rates, and estimat
36、es of appliance ownership and the number of units in the residential sector. Using this approach, it is possible to estimate over long periods and to reflect social trends, including macroeconomic and socioeconomic effects; additionally, there is a simple input value. However, this approach provides
37、 rough analysis without the classification of end-use energy and has a high dependence on energy consumption data. Compared to the top-down approach, the bottom-up approach is a methodology that infers the standard expected energy consumption of the residential sector at the local and national level
38、 and includes two different methods: the statistical method and the engineering method. While the statistical method includes macroeconomic and socioeconomic impacts and uses billing data or a simple survey, there is a high correlation between the independent variables, and it also requires a large
39、sample survey to ensure diversity. Although the engineering method for the bottom-up approach is a new technological modeling method, it is not dedicated to economic factors. The calculations are complex, and the approach requires detailed input values. The most significant difference between the st
40、atistical and engineering methods is that, whereas the former reflects the actions of residents, the latter makes assumptions about the use of end-use energy because it does not have any details about the occupancy of a residence. Each of the modeling technologies varies based on the input value, ca
41、lculation technology or simulation technique, and the results are applied differently. The energy consumption in the apartment building system can be resolved by being metered directly or through engineering estimation, but in regard to the engineering method, it is not dedicated to the behavior of
42、the occupants, is based solely on theoretical inquiry and cannot systematically cope with changes in price, income, household size and so on. Additionally, significant costs arise in the direct measurement. Therefore, we attempt to use the statistical method as the methodology for estimating energy
43、consumption in the apartment buldings, as it is not problematic in terms of estimating the impacts of the occupants behaviors and costs. Statistical techniques consist of regression, Conditional Demand Analysis (CDA) and neural networks. First, regression compares the total energy consumption with t
44、he data that are expected to affect energy use. CDA performs regression analysis based on boilers, air conditioning, various electric appliances, etc. Third, a neural network is a simplified mathematical model based on the closely interconnected parallel structure of biological neural networks. Beca
45、use CDA reflects the macroeconomic and socioeconomic impacts, the behaviors of the occupants and actual users of the building, and performs the regression analysis based on the specific energy end uses, we would like to use CDA as an estimation of the end-use energy consumption in the apartment buil
46、ding. Definition of Conditional Demand Analysis (CDA) and Analysis of Major Interaction Variables. CDA is a regression model in which an apartments total end-use energy consumption is expressed as binary variables or count variables. In previous researches, Parti et al. (1987.) recognized that the l
47、imitations for predicting the end-use energy consumption of existing apartment buildings are based solely on theoretical considerations and not occupants behavioral patterns. Thus, the energy consumption does not reflect any changes in systemic characteristics, e.g., price, income and housing sizes.
48、 However, they predict the amount of end-use energy consumption using the econometric method without any theoretical engineering data or direct metering. An empirical analysis has been made based on records of the electricity costs of more than 5,000 apartment buildings, and the information may corr
49、espond to the climate data, ownership of appliances, demographic variables, etc. Covariance analysis has been carried out to predict the end-use energy consumption, and these functions are used for the calculation of the energy usage reflecting fuel prices and income levels, etc. In this study, a basic CDA model can be expressed in the form of Equation (2) as follows: (2) At this time, the value of UEChetis based on the usage pattern of other appliances investigated in the building energy usage pattern survey. As in equation (2) above, the energy consumption by the
