Energy consumption of 270 schools in Tianjin, China

Abstract

With the rapid development of education cause, the increasing energy consumption of school buildings is gradually causing widespread concern in recent years in China. This paper presented an analysis of energy consumption of 270 schools located in the city of Tianjin, China. The analysis focused specifically on calculating the space heating energy consumption indexes and non-heating energy consumption indexes of different types of schools, aiming at providing reliable and precise data for the government to elaborate policies and measures. The space heating energy consumption of schools adopting district heating and gas boiler were 92.04 kWh/(m²·a) and 64.25 kWh/(m²·a), respectively. Comparing to the schools without a canteen, the non-heating energy consumption index of schools with a canteen can increase by 8%–37%. Furthermore, clustering of different energy sources, the total primary energy consumption indexes were also presented. Space heating energy consumption accounted for approximately 64%–79% of the total primary energy consumption. When using time-sharing control and self-contained gas boiler instead of district heating, an amount of almost 27.8 kWh/(m²·a) and 77.5 kWh/(m²·a) can be saved respectively. Through extensive statistical analysis of the data collected, this paper demonstrated that gross floor area, heating energy source and canteen had a close relationship with the total primary energy consumption regarding complete schools. Eventually, a linear regression equation was established to make a simple prediction about the total energy consumption of existing complete schools and to estimate the energy consumption of complete schools to be built.

Graphical abstract

Keywords

schools / energy consumption index / primary energy / energy saving / regression analysis

1 Introduction

With resources and environmental problems becoming more and more serious, many countries, including China, have made great effort to save energy and reduce emissions. Energy problem has become the focus of global attention. In 2011, building energy consumption (not containing biomass energy) accounted for approximately 19.74% of the total energy consumption in China [1]. As an important part of civil buildings, public buildings have a significant impact on building energy consumption. In 2011, public buildings amounted to nearly 7.97 billion square meters, accounting for 17% of the total building area in China. However, the energy consumption of public buildings (excluding heating energy in northern China) accounted for 24.8% of the total building energy consumption [1]. School buildings account for a considerable proportion in all kinds of public buildings and play a momentous role in energy saving. The Chinese government has set up several Schools Partnerships, such as NUECA (Nation University Energy Conservation Alliance), UAIEE (University Alliance for Industrial Energy Efficiency), and CGUN (China Green University Network) for colleges and universities to improve energy conservation [2]. However, little attention has been paid to buildings in primary and middle schools whose energy consumption also account for a certain proportion, resulting in a huge waste of energy in China.

Many developed countries have conducted studies on energy consumption of school buildings and established corresponding benchmarks. The UK have been producing energy benchmarks and performance guides for almost 30 years, known as Good Practice Guide 343 [3], which includes typical and best practice values for primary schools, respectively. Benchmarking standards for the energy performance of schools have existed in many EU countries. Several studies have been conducted concerning the typical annual heating use in school buildings. Desideri and Proietti [4] have analyzed the thermal and electric energy consumption of 29 high schools in a province in Italy, concluding that thermal energy is the main consumption in all kinds of schools, accounting for approximately 80% of the total annual energy consumption and that the energy consumption indexes are different when considering the division of building and school typologies. The energy efficiency and gas commission performance of 15 schools in Argentina have been reported [5], which demonstrate that the average energy consumption of these schools is 123 kWh/ (m²·a) while 87% of the primary and secondary schools are characterized as “low emission buildings.” Corgnati et al. [6] have proposed a methodology for heating energy assessment concerning 140 existing school buildings in Italy and established a benchmark, demonstrating that the indicator is suitable for long-term assessment on large building stocks. Through detailed analysis of electricity, water, and gas consumption quantities in public schools in Toronto, Issa et al. [7] have concluded that improving the energy efficiency in conventional buildings have a very significant impact on energy consumption both in the short-term and the long-term by retrofitting. Kim et al. [8] have analyzed the energy consumption characteristics of 10 elementary schools in South Korea, determining that the average energy consumption index in 2010 is 1040 MJ/m² per year for electricity, 92 MJ/m² per year for oil, and 325 MJ/m² per year for gas, and in terms of energy use, heating consumes most of the energy, followed by cooling and lighting. Based on the surveying data of 320 school buildings from different regions in Greece and after comprehensive analysis, Santamouris et al. [9] have demonstrated that the average heating and electricity energy consumption values are 68 kWh/(m²·a) and 27 kWh/(m²·a), respectively. However, a field study by Thewes et al. [10] has determined an average value of 93 kWh/ (m²·a) for the thermal end energy consumption in Luxembourg. The difference in energy consumption values can be explained, on the one hand, by the sample size, and on the other hand, by the selection of buildings. This shows that many conclusions on energy consumption should be considered with caution. British studies have analyzed the dependence of special-use areas (e.g. cafeteria) and found that consumption have increased by up to 7%–10%. Energy consumption is likely to increase by as much as 20% if a gym is taken into consideration [11]. Katafygiotou and Serghides [12] have made a detailed analysis of structural elements and energy consumption of 24 secondary schools in Cyprus. The school sector consumes large amount of energy for heating and electricity and therefore the energy saving measures are vital. An average of 13% of the total energy use in the US, 4% in Spain and 10% in the UK is consumed by schools. In Slovenia the average heating energy consumption in schools is 100 kWh/ (m²·a) and the average total energy consumption is 192 kWh/ (m²·a) [13]. In Ireland, investigations in schools have shown that the consumption for electricity is between 5 and 35  kWh/ (m²·a) and for heating is between 50 and 200 kWh/ (m²·a) with an average consumption for heating at 96 kWh/ (m²·a) [14]. Several studies have focused on the thermal behavior of school buildings in Northern Greece [15,16], which indicate that the most efficient energy saving scenarios are thermal insulation for heating, night ventilation and ceiling fans for cooling. In China, Zhou et al. [2] have made a survey of the energy consumption and energy conservation measures in 98 colleges and universities in Guangdong province, finding that there exist great differences in per unit energy consumption between different types of universities classified by the discipline, nature, and level of the schools. In addition, several studies [17–20] have been conducted on the prediction of energy consumption, mainly using the multiple regression analysis based on the statistical relationship between dependent and independent variables. With the help of computer simulation techniques and comparative tests, Perez et al. [21] have modified these variables one by one and determined the variables having the greatest impact on the energy consumption for school buildings in the hot-humid climate.

The studies conducted in advanced countries on the energy consumption of school buildings are viewed as a part of efforts to reduce the energy consumption. School buildings, due to their specialized characteristics and their social/educational character, have special requirements for environmental design for heating, cooling, day lighting and ventilation. To promote energy conservation of primary and secondary schools, the survey in the present paper has been conducted by Tianjin Urban-Rural Construction & Transport Commission and Tianjin Municipal Education Commission. The collection of the reliable data can facilitate the creation of a representative database that is useful to support the reliable energy certification of school buildings. In this paper, the energy consumption of 270 schools (excluding colleges and universities) in Tianjin has been analyzed based on the surveying data from January 2010 to December 2012, to present several specific indexes for the energy-saving goals of schools in Tianjin. Moreover, the relationship between energy consumption and the main influencing factors has also been investigated, and a linear regression equation for complete schools has been established.

2 Research methodology

2.1 Data collection

To rate the energy performance and the energy efficiency of each school, compared to other same-category schools, it is important and necessary to establish a complete database. Few schools in Tianjin have ever been monitored and analyzed to identify the energy consumption. Thus, the first step is to make a systematic data collection of the schools in Tianjin, which can eventually constitute a representative database and make a statistical evaluation.

Originally, the study was undertaken to investigate the general information and energy consumption of schools in the form of questionnaire. The questionnaire was sent to 300 schools, except for universities and colleges, in Tianjin, to collect detailed information on construction, activities and energy use in each school. With the help of logistics staff of each school and coordinating with the relevant staff of Tianjin Municipal Education Commission, a total of 283 feedback questionnaires was received. A careful analysis of the questionnaire indicated that some schools did not have adequate information about the energy consumption and some schools had incorrect monthly energy data. Therefore, these data were excluded by using the simple statistical method (i.e. 3σ). Consequently, the information of 270 schools was found to be valid for the analysis. The information includes the basic construction information (including school typology, gross floor area, age of building, building characteristics, number of students and teachers, class number), the monthly utility bills (including electricity consumption, natural gas consumption and raw coal consumption of 3 years), the heating energy source and school heating hours, the list of mechanical equipment and the operating schedule, and the fact that if the school have canteen or not.

There is a total floor area of approximately 2135849 m² with 212699 students and 23873 teachers and administrative staff in the 270 schools. There are 5633 classes with an average number of 38 students in each class. Only the school buildings that are directly related to educational purposes were included in the investigation. Therefore, such buildings as swimming pools and gymnasiums were excluded. The energy consumption directly influences the ambient comfort and eventually has a greater impact on the energy benchmarks. The deviations of different energy forms among the sample schools are not obvious as they are all used for the same purpose, i.e. education.

2.2 Data preprocessing

The 270 schools, of which 45% were located in urban areas and the others in the suburbs, include 25 complete schools (including senior department and junior department), 48 secondary schools (just consisting of senior school or junior school), 145 primary schools, and 52 kindergartens.

A complete school is comprised of senior department and junior department, both of which share the same campus and attend the same campus activities together. Whereas, a secondary school just consists of senior school or junior school.

Based on the data collected, it was discovered that the energy consumed generally included four forms, namely heating, electricity, natural gas and raw coal. Through detailed energy classification and energy form transformation, different energy forms can be collected and gathered.

The total energy consumption was divided into space heating energy consumption and non-heating energy consumption which consists of the electricity energy consumption and natural gas consumption consumed by cooking.

2.3 Energy consumption analysis

The energy consumption in schools is quite typical. Most buildings are usually not used in the evening and at night, neither during the weekends and school holidays. Actually, classes are given more or less 210 days a year, from 8:00 to 18:00, which means around 2100 h a year. Even though these schools serve for the same purpose, different school typologies have their peculiar characteristics. Furthermore, the school buildings of different typologies also have a significant difference. Most of the complete schools are equipped with canteens and dormitories, yet the other types of schools hardly include these. The separate collation of space heating energy consumption and non-heating energy consumption which contains electricity and natural gas consumption makes it possible to have a detailed analysis of specific energy parameters. To make a complete analysis in a field of energy consumption, qualitative and quantitative information on all kinds of energy is necessary.

This paper mainly examined the gross floor area, number of students and teachers and energy consumption values of each scphool from January 2010 to December 2012. For each school the specific energy consumption indexes (SECIs) were calculated, which made it possible to compare the energy consumption of very different schools in terms of typology, floor area as well as total number of students and teachers enrolled. The SECIs used in the analysis are follows:

H_a: Space heating energy consumption per unit gross floor area (kWh/(m²·a)) calculated as the ratio between the annual space heating energy consumption provided by the heating plant and the total gross floor area.

H_p: Space heating energy consumption per total number of students and teachers (kWh/(p·a)) calculated as the ratio between the annual space heating energy consumption provided by the heating plant and the total number of enrolled students and teachers.

NH_a: Non-heating energy consumption per unit gross floor area (kWh/(m²·a)) calculated as the ratio between the annual non-heating energy consumption and the total gross floor area.

NH_p: Non-heating energy consumption per total number of students and teachers (kWh/(p·a)) calculated as the ratio between the non-heating energy consumption and the total number of enrolled students and teachers.

2.3.1 Space heating energy consumption

Of the 270 schools, 168 schools used district heating, 16 schools which are located in urban areas used their own gas boiler, and 86 schools which are located in suburbs used their own coal boiler as the winter heating energy source.

Most of the schools using district heating did not monitor space heating energy consumption. Only 3 schools using district heating were equipped with heat meter. Thus, the total space heating energy consumption of these 3 schools during 3 years can be directly acquired.

Meanwhile, the space heating energy consumption of the schools using their own gas boilers was calculated by

(1)Q_{\mathrm{hc}}={\displaystyle \frac{{\eta }_{\mathrm{1}}{\eta }_{\mathrm{2}}M{q}_{\mathrm{lp}}}{\mathrm{3600}}},

where Q_hc is the total space heating energy consumption of a school (kWh/a), η₁ which is the average efficiency of the testing boiler in one heating season is calculated by field testing, η₂ is the efficiency of secondary pipe heating network (0.97 [22]), M is the total gas consumption consumed by gas boiler in a heating season (Nm³/a), and q_lp is the net calorific value of natural gas in Tianjin (35375 kJ/Nm³) [23]. The ultrasonic flowmeter (Fuji Electric Co., Ltd. S10C1-00C) and HOBO temperature recorder (Onset HOBO Data Loggers) were used to test the related data. The total supplying water flow, the supplying and returning water temperature, the heating room temperature and the average outdoor temperature were tested during testing period. η₁ is calculated by

(2){\eta }_{\mathrm{1}}={\displaystyle \frac{\mathrm{1}}{n}}{\displaystyle \sum _{i=\mathrm{1}}^n{\displaystyle \frac{C{M}_{i-\mathrm{1},i}\left(t_{\mathrm{g}i}-t_{\mathrm{h}i}\right)}{q_{\mathrm{lp}}{G}_{i-\mathrm{1},i}}}},

where C is the specific heat capacity of water (4.18 kJ/(kg·°C)), M_i−1,i (kg)and G_i−1,i (Nm³) are the total circulating water flow of gas boiler and natural gas consumption between two monitors respectively, t_gi is the outlet water temperature of the boiler (°C), t_hi is the inlet water temperature of the boiler (°C), and n is the total monitoring times. Figure 1 shows the value of η₁ ranging from 0.85 to 0.97 for the testing 16 schools.

Furthermore, space heating energy consumption varies depending on different variables. The climate plays the most important role. Tianjin is located in the cold region whose heating period lasts from November 1 to March 31. The typical annual daily dry bulb temperature in Tianjin is illustrated in Fig. 2.

To normalize the impact of climate on space heating energy consumption, the use of heating degree days (HDD) technique in the normalization of the space heating energy consumption is considered to be a suitable way to neutralize the effect of climate. Then with the help of HDD, the portion of space heating energy consumption was corrected to the average climate conditions in Tianjin:

(3)Q_{\mathrm{bhc}}=Q_{\mathrm{hc}}{\displaystyle \frac{{\mathrm{HDD}}_{\mathrm{b}}}{\mathrm{HDD}}},

where Q_bhc is the normalized heating energy consumption of a school relative to average climate conditions, Q_hc is the measured heating energy consumption, HDD_b is the heating degree days in Tianjin of the typical meteorological year derived from the software of DeST [24] (2487°C·d/a) and HDD is the heating degree days of 2010–2012 analyzed (2113°C·d/a).

Based on the energy consumption of the 270 schools, the specific heating energy consumption indexes (H_a and H_p) were calculated for each school and the calculated indexes which considered the mean of the 3 years were consequently different in terms of the school typology and kind of fuel of the thermal system (see Fig. 3 and Fig. 4).

Owing to the considerable diversities of the calculated energy consumption of each school, it is essential to make a reasonable classification based on the school typology and heating source (see Table 1). Of the energy consumption analyzed above, both school typology and heating source have a great influence on the space heating energy consumption. In terms of school typology, kindergartens have the highest H_a and the second highest H_p, whereas complete schools have the lowest H_a but the highest H_p. Particularly due to the different building types and indoor temperature for diverse ages of students, the specific space heating energy consumption indexes vary significantly. The indoor temperature for kindergarten is always 1–2 degrees higher than that for other school typologies during the heating season, which leads to the highest H_a. As a consequence of the highest school population density, H_p is lower, instead. Since many buildings such as administration building, library, laboratory, and etc. are not occupied every day in complete schools, H_a is normally the lowest. When considering the heating source, the diverse consumption indexes demonstrate that the energy consumption of schools equipped with gas boilers is lower in terms of both H_a and H_p. The reason is that these schools have professional boiler operators to operate the gas boiler during the heating season. With the change of outdoor temperature, these boiler operators regulate the outlet water temperature of the boiler to achieve the indoor temperature requirement. Changes in water temperature will eventually lead to the changes in gas consumption. Meanwhile, these gas boilers are always shut down at night and weekend if there are no students at schools. However, the schools using district heating as the heating source do not have measures to control the flow rates, the heating water temperature since the heating systems in these schools are traditional ones which do not have a control system.

2.3.2 Non-heating energy consumption

In addition to the space heating energy consumption, the remaining energy consumption is defined as the non-heating energy consumption which mostly contains electric energy consumption and gas consumption consumed by canteen if any. Building services are usually the main consumers of the non-heating energy. Lighting, air-conditioning, ventilation and PC systems are responsible for a considerable amount of electric energy consumed in school buildings. Almost all administrative buildings and teachers’ offices are equipped with split air conditioners whereas classrooms are merely equipped with electric fan. Only 14 schools have installed intelligent lighting with smartsens in classrooms and 68% schools have replaced the conventional lamps with compact fluorescent ones.

With regard to the non-heating energy consumption in schools, a British study determined that energy consumption can increase by up to 7%–10% if the canteen is taken into consideration [11]. It will be interesting to see whether the non-heating energy consumption of the schools in Tianjin fit into this index. To compare the consumption of different types of energy in all kinds of schools, it is necessary to convert the gas consumption value to net calorific value using the conversion factor (f = 9.83kWh/Nm³) [23].

59 schools are equipped with canteens, accounting for 22% of the total. To analyze the influence of canteen on non-heating energy consumption, the index was calculated separately. Figures 5–12 show the non-heating energy consumption values of each school regarding school typology and whether the school is equipped with canteen separately.

The general analysis of non-heating energy consumption of the 270 schools indicates that the NH_a and NH_p index are particularly high for complete schools and kindergartens mainly because of a wide variety of electric equipment for complete schools such as beam projector, notebook, desktop computer, TV, cooling/heating devices, and the smaller construction area as well as the single building type for kindergartens. Compared with the schools without a canteen, the NH_a index of the schools with a canteen increases by 8%–37%. The increase of the NH_a index for complete schools, secondary schools, primary schools, and kindergartens is 28%, 14%, 8%, and 37%, respectively. The detailed data analysis of these schools verified the performance of installed canteen equipment and the average operating time. It was found out that canteens have a vital influence on the energy consumption especially for kindergarten and complete school. All this can be connected to the building scale and the annual number of meals.

2.3.3 Primary energy consumption

To calculate the total primary energy consumption of each school, it is essential to convert the consumption of various kinds of energies to the same energy with the help of regional conversion factors which are slightly different in each area regardless of the energy vector. Table 2 lists the consumed energy information and the conversion factors [23].

Considering the heating energy consumption, the present study shows quite significant variations ranging from 21.8 kWh/(m²·a) to as much as 344.6 kWh/(m²·a) and 228.8 kWh/(p·a) to as much as 4123.2 kWh/(p·a) as a result of the difference of energy source (see Table 3). Owing to the fact that the gas boilers and coal boilers are operated by boiler operators whose professional knowledge vary greatly, the fuel consumption value of the schools fluctuated markedly. The schools adopted natural gas as the heating energy source consume the least primary energy while the schools adopted raw coal and district heating consume twice as much as primary energy, which is contrary to the conclusion drawn in Ref. [27]. This demonstrates that heating energy source has a great influence on the heating energy consumption and using natural gas can save approximately 77.5 kWh/(m²·a) or 533.2 kWh/(p·a) compared to using district heating.

With regard to the non-heating energy consumption, Figs. 3–10 have shown the specific indexes regarding the classification of school typology. When synthesizing the energy consumption indexes of the schools, the mean primary energy consumption values of these two indexes were estimated to be approximately 55 kWh/(m²·a) and 397 kWh/(p·a).

Comparing these two primary energy consumption, it is concluded that space heating energy consumption accounts for approximately 64%–79% of the total primary energy consumption (TPEC) which is close to the proportion of 80% in Ref. [13]. Adding up the two values of each school, the total primary energy consumption indexes (see Table 4) can be obtained. Astonishingly, school typology does not influence the average value of TPEC primarily with regard to complete school, secondary school and primary school when considering the gross floor area as the baseline. However, the TPEC index of kindergarten is apparently higher (65 kWh/(m²·a)) than that of the other three types of schools.

2.3.4 Energy saving potential

The analysis indicates that space heating energy consumption accounts for more than 3/4 of the total energy consumption for most schools. Considering the main energy consumption and the fact that the whole society is advocating the energy conservation of buildings, it is important to focus on the space heating energy saving potential. Generally, space heating energy saving potential can be discussed from the following three aspects:

1) Time-sharing control: it is discovered in the survey that most schools adopted district heating in winter. Considering the fact that the heating fee is based on the floor area rather than the actual heat metering for these schools in Tianjin, low-temperature heating is ignored. Especially during the night and vacation when the whole school is occupied with nobody, low-temperature heating is not only economical but also energy efficient in the premise of ensuring that the pipelines are not frozen. Nearly all of the schools which have their own boilers reduced natural gas or coal consumption via low-temperature heating. By lowering the supplying and returning water temperature, the indoor temperature can be decreased naturally. Almost 27.8 kWh/(m²·a) can be saved if their own gas boilers were used and temperature was reasonably controlled according to the weather and building using conditions.

2) Heating energy source: in Sections 2.3.1 and 2.3.3, it was concluded the three heating energy sources had a great influence on space heating energy. Comparing the gas boiler with coal boiler, the efficiency and emission refuse differ considerably. The gas boiler efficiency is found to be more than 0.9 in field testing whereas the coal boiler efficiency is basically less than 0.7 and coal combustion will produce a lot of gaseous pollutants and solid waste such as SO_x, NO_x, H₂S and clinker which will cause serious pollution to the atmosphere and environment. However, the gas boiler just produces a little carbon monoxide and oxynitride. In addition, the primary energy consumption of the coal boiler is twice as much as or more than the gas boiler mainly because of the low efficiency of the coal boiler. Considering the energy consumption and environmental pollution, the coal boiler should be replaced by the gas boiler. When comparing the gas boiler with district heating, it is found that the schools using their own gas boilers can save 77.5 kWh/(m²·a) or 533.2 kWh/(p·a) when considering the primary energy and ignoring the initial investment and ongoing management fees. The primary heating energy consumption is affected not only by the actual thermal consumption of buildings but also by the efficiency of heating boiler and transmission efficiency of outdoor pipe network. These two primary energy saving values are the comprehensive effect of low temperature heating at night and high efficiency of gas boiler and pipe network for the schools using their own gas boilers.

3) Heating terminal form: the field investigation demonstrated that nearly all the rooms of these schools were equipped with radiator to warm the indoor air. However, a popular type of heating system, radiant floor heating, is considered as an energy saving method. The indoor room design temperature can be reduced by 2°C, which directly leads to the reduction of the supplying water temperature and returning water temperature and indirectly results in the decrease of the primary energy consumption.

3 Energy consumption prediction model

The regression analysis prediction method is commonly used for energy consumption prediction. Its ultimate target is to establish a regression equation between independent variables and dependent one according to their relationships. The established model can predict the variation of the independent variable based on the changes of dependent variables.

3.1 Selection of variables

From the above energy consumption study of each typology school, it is observed that the energy consumption is affected by many factors. To find out the factors which have significant impacts on the energy consumption, it is necessary to make an in-depth discussion on the influencing factors. The 20 complete schools are randomly selected as the analyzing sample. Relevant variables which may have impacts on the final conclusion should be selected with caution. The dependent variable is the TPEC (kWh/a), and Table 5 shows the selected independent variables. Based on the definition above, the statistic data could be easily classified and listed in Table 6.

3.2 Partial correlation analysis

With the help of the software tool SPSS 19.0, the correlation analysis values of the six independent variables to the total primary energy consumption were obtained and tabulated in Table 7. It is noted that gross floor area has the greatest influence followed by heating energy source and the kitchen on the energy consumption of the schools. In addition, the unrelated significant levels of X₂, X₃ and X₅ are greater than 0.05 which demonstrates that these three factors have no significant correlation with the TPEC.

3.3 Multiple regression analysis

All of the selected independent variables have, more or less, an effect on the total primary energy consumption. To make a detailed analysis of these influencing factors, a regression analysis should be made and a regression equation should be reached ultimately. In most cases, it is assumed that the general multiple regression model is linear. Supposing that there are n independent variables and the dependent variable Y can be written as

(4)Y=b_{\mathrm{0}}+b_{\mathrm{1}}{X}_{\mathrm{1}}+b_{\mathrm{2}}{X}_{\mathrm{2}}+\cdots +b_n{X}_n+\epsilon ,

where Y is the dependent variable, X_n is the independent variable, b_n is the regression parameter and ε is the residual. The next step was to analyze the model which can best characterize the real values of the sample. With the help of t-test, the coefficient of determination R² was calculated to be 94.4% and the significant level was zero which illustrated that the model was reasonable and notable. The calculated coefficient b_n based on the regression model obtained the final regression equation with the independent variables “gross floor area” (X₁), “heating energy source” (X₄), and “canteen” (X₆) (see Table 8). According to F-test, the existing regression mode will become non-significant if an extra independent variable was added. Accordingly, the variables “total number of teachers and students” (X₂), “construction year” (X₃) and “rank of school” (X₅) were eliminated.

In short, by using the SPSS software and statistics theory, a linear regression equation was obtained with regard to the TPEC (kWh/a) of complete schools in Tianjin, which is expressed as

(5)\mathit{Y}\mathrm{=}-3789026+273{\mathit{X}}_1+2415464{\mathit{X}}_4+752621{\mathit{X}}_6\mathrm{.}

The final regression model not only can make a simple prediction about the total energy consumption of the existing complete schools but also may estimate the energy consumption of a complete school to be built.

3.4 Validation of the model

To estimate the validity of the established model, a validation test with 5 schools which were selected randomly out of the total sample (N = 25) was performed. A comparison between the actual values and the predicted values of the regression model and the validation sample is illustrated in Fig. 13. Compared with a mean relative error (MRE) of 8.7% for the regression model, a specified MRE of 5.9% for the validation sample demonstrates that the established model is valid. In comparison with published prediction relative errors for consumption forecasts by Beusker et al. [27] which are 17% for regression model and 10% for validation sample, the validation test also verifies a suitable quality on the whole.

4 Conclusions

This survey, in the form of questionnaire, investigated the energy consumption of 270 schools in Tianjin based on the data collected from January 2010 to December 2012. Considering the different school typologies and energy sources, several specific energy consumption indexes were calculated to provide some reference for the government to elaborate policies and measures and for schools to save energy. The following conclusions are reached from the field study:

1) School typology had a certain impact on the non-heating energy consumption. Moreover, heating energy source had an important effect on the space heating energy consumption. However, considering the total primary energy consumption, school typology, excluding kindergartens, scarcely influences the value when viewing the gross floor area as the baseline.

2) For the heating energy consumption, kindergartens and complete schools have the highest and lowest value of H_a respectively. However, complete schools and primary schools have the highest and lowest value of H_p respectively. The reason for this is that different types of schools have diverse personnel density and building size. Moreover, schools having their own gas boiler consume the least primary energy while schools having their own coal boiler consume twice as much or more primary energy compared to these schools. Heating energy consumption accounts for approximately 64%–79% of the total primary energy consumption, which indicates that changing heating energy source can have a huge impact on the total energy consumption.

3) For the non-heating energy consumption, kindergartens have the highest value of NH_a followed by complete schools. Moreover, complete schools have the highest value of NH_p followed by kindergartens. Comparing the schools which have a canteen to the ones which do not have a canteen, the NH_a index can increase by 8%–37%, which demonstrates that canteens have a close influence on the energy consumption.

4) Three aspects about space heating energy saving potential were mentioned: First, time-sharing control should be used for the supplying and returning water temperature. Second, the gas boiler is the most economical and environmentally-friendly heating source compared with the coal boiler and district heating. Small coal boilers should be banned according to the national policy. The primary energy consumption can be saved by 77.5 kWh/(m²·a) or 533.2 kWh/(p·a) when transforming district heating into gas boilers of the schools as winter heating source while ignoring the initial investment and ongoing management fees. Third, considering the heating terminal form, the conventional radiator heating should be replaced by radiant floor heating.

5) Extensive statistical analysis indicates that gross floor area, heating energy source and canteen have a greater effect on the total energy consumption regarding complete schools. A linear regression equation is established which cannot only make a simple prediction about the total energy consumption of the existing complete schools but also estimate the energy consumption of the complete schools to be built.

References

[1] Building Energy Research Center of Tsinghua University. 2013 Annual Report on China Building Energy Efficiency. Beijing: China Architecture & Building Press, 2013

[2] Zhou X, Yan J W, Zhu J W, Cai P P. Survey of energy consumption and energy conservation measures for colleges and universities in Guangdong province. Energy and Building, 2013, 66: 112–118

[3] GPG343 Good Practice Guide. Saving energy—a whole school approach. 2014-06

[4] Desideri U, Proietti S. Analysis of energy consumption in the high schools of a province in central Italy. Energy and Building, 2002, 34(10): 1003–1016

[5] Filippín C. Benchmarking the energy efficiency and greenhouse gases emission of school buildings in central Argentina. Building and Environment, 2000, 35(5): 407–414

[6] Corgnati S P, Corrado V, Filippi M. A method for heating consumption assessment in existing buildings: a field survey concerning 120 Italian schools. Energy and Building, 2008, 40(5): 801–809

[7] Issa M H, Attalla M, Rankin J H, Christian A J. Detailed analysis of electricity, water, and gas consumption quantities and costs in Toronto’s public schools. Canadian Journal of Civil Engineering, 2010, 37(1): 25–36

[8] Kim T W, Lee K G, Hong W H. Energy consumption characteristics of the elementary schools in South Korea. Energy and Building, 2012, 54: 480–489

[9] Santamouris M, Mihalakakou G, Patargias P, Gaitani N, Sfakianaki K, Papaglastra M, Pavlou C, Doukas P, Primikiri E, Geros V, Assimakopoulos M N, Mitoula R, Zerefos S. Using intelligent clustering techniques to classify the energy performance of school buildings. Energy and Building, 2007, 39(1): 45–51

[10] Thewes A, Maas S, Scholzen F, Waldmann D, Zurbes A. Field study on the energy consumption of school buildings in Luxembourg. Energy and Buildings, 2014, 68(Part A): 460–470

[11] Building Research Energy Conservation Support Unit. Saving energy in schools, energy consumption guide 73. Watford, UK, 1998

[12] Katafygiotou M C, Serghides D K. Analysis of structural elements and energy consumption of school building stock in Cyprus: energy simulations and upgrade scenarios of a typical school. Energy and Building, 2014, 72: 8–16

[13] Butala V, Novak P. Energy consumption and potential energy savings in old school buildings. Energy and Building, 1999, 29(3): 241–246

[14] Hernandez P, Burke K, Lewis J O. Development of energy performance benchmarks and building energy ratings for non-domestic buildings: an example for Irish primary schools. Energy and Building, 2008, 40(3): 249–254

[15] Dimoudi A, Kostarela P. Energy monitoring and conservation potential in school buildings in the C’ climatic zone of Greece. Renewable Energy, 2009, 34(1): 289–296

[16] Theodosiou T G, Ordoumpozanis K T. Energy, comfort, and indoor air quality in nursery and elementary school buildings in the cold climatic zone of Greece. Energy and Building, 2008, 40(12): 2207–2214

[17] Hong T, Koo C, Jeong K. A decision support model for reducing electric energy consumption in elementary school facilities. Applied Energy, 2012, 95: 253–266

[18] Gaitani N, Lehmann C, Santamouris M, Mihalakakou G, Patargias P. Using principal component and cluster analysis in the heating evaluation of the school building sector. Applied Energy, 2010, 87(6): 2079–2086

[19] Baker K J, Rylatt R M. Improving the prediction of UK domestic energy-demand using annual consumption-data. Applied Energy, 2008, 85(6): 475–482

[20] Lam J C, Wan K K M, Liu D, Tsang C L. Multiple regression models for energy use in air-conditioned office buildings in different climates. Energy Conversion and Management, 2010, 51(12): 2692–2697

[21] Perez Y V, Capeluto I G. Climatic considerations in school building design in the hot-humid climate for reducing energy consumption. Applied Energy, 2009, 86(3): 340–348

[22] He P. Research on carbon trading methodology in heat metering of residential building. Dissertation for the Master’s Degree. Tianjin: Tianjin University, 2013

[23] National Bureau of Statistics China. China Energy Statistical Yearbook 2012. Beijing: China Statistic Press, 2012

[24] Department of Building Science of Tsinghua University. DeST launcher. 2014-06

[25] State Electricity Regulatory Commission. Electricity regulatory annual report 2010. 2010

[26] Ministry of Housing and Urban-Rural Development of the People’s Republic of China (MOHURD). Design Standard for Energy Efficiency of Civil Buildings. Beijing: China Architecture & Building Press, 1995

[27] Beusker E, Stoy C, Pollalis S N. Estimation model and benchmarks for heating energy consumption of schools and sport facilities in Germany. Energy & Environment, 2012, 49: 324–335

RIGHTS & PERMISSIONS

Higher Education Press and Springer-Verlag Berlin Heidelberg