Research Article

Advances in Mechanical Engineering

2017, Vol. 9(11) 1–21
Ó The Author(s) 2017
Study and analysis of office building DOI: 10.1177/1687814017734110
journals.sagepub.com/home/ade
energy consumption performance in
severe cold and cold region, China

Zequn Ding, Hongqing Zhu, Yang Wang and Xiaoling Ge

Abstract
Large office building energy consumption in China is several times that of ordinary buildings. Therefore, it is necessary
to study the characteristics of large office buildings to determine their energy consumption situations and the important
factors affecting their energy consumption. Using typical public buildings in Severe Cold and Cold Region as examples,
this work will analyze and study the energy consumption situation and influencing factors of office building energy con-
sumption using an orthogonal experimental method. In this article, the basic information and energy consumption data
of 56 typical office buildings in Beijing, Tianjin, Dalian, Ji’nan, and Harbin are discussed, including 19 non-government
office buildings and 37 government office buildings. According to the investigation data, the energy consumption situation
and some energy consumption influencing factors are analyzed using software such as SPSS20.0. Through an orthogonal
experiment, the author selects the exterior walls of the buildings’ heat transfer coefficient, the heat transfer coefficient
of the external windows, the lighting equipment, the power density, and the power density of the air-conditioning system
as the five factors that have the greatest impact on building energy consumption for further analysis and study. The simu-
lation software eQUEST is employed in the study. Finally, energy saving measures are developed according to the results
of the analysis and study.

Keywords
Office building, energy audit, energy consumption, influence factor

Date received: 8 February 2017; accepted: 5 September 2017

Handling Editor: Oronzio Manca

Introduction emissions increasing from 14.24 3 108 ton in 1978 to

Over the past century, the global climate has changed
perceptibly, and over the last three decades, global
warming has become the most significant climatic con-
cern. Substantial carbon dioxide emissions have
resulted in a series of environmental issues.1 Our econ-
omy has made significant progress since we adopted a
policy of reform and openness, and China is now the
world’s second largest economy.2 However, as the
world’s largest energy consumer,3 China’s consumption
of coal increased from 5.86 3 108 tce in 1980 to
36.2 3 108 tce in 2012.4 During this time, China has
also become the world’s largest CO2 emitter,5 with CO2
79.55 3 108 ton in 2011.6
Building energy consumption has become the one of
the three major energy consuming industries, in addi-
tion to the industrial and transportation industries.
Office buildings are a symbol of the post-industrial era
State Nuclear Electric Power Planning Design & Research Institute,
Beijing, P.R. China
Corresponding author:
Zequn Ding, State Nuclear Electric Power Planning Design & Research
Institute, Beijing 100095, P.R. China.
Email: dingzequn@tju.edu.cn

Creative Commons CC-BY: This article is distributed under the terms of the Creative Commons Attribution 4.0 License
(http://www.creativecommons.org/licenses/by/4.0/) which permits any use, reproduction and distribution of the work without
further permission provided the original work is attributed as specified on the SAGE and Open Access pages (https://us.sagepub.com/en-us/nam/
open-access-at-sage).
2 Advances in Mechanical Engineering
Figure 1. Energy consumption intensity of public buildings in
some country of the world, 2011.
and of today’s global knowledge economy. After more
than a century of development, office buildings contain
more than half of the working population in cities and
have become one of the most important architectural
forms of 21st century.7 Large office building energy
consumption in China is several times that of ordinary
buildings, although the overall energy consumption
(except in northern regions) level of office buildings in
China is lower than those of the United States, Japan,
or other developed countries, but the former’s develop-
ment speed was faster.
From Figure 1, it can be seen clearly that in China,
both energy consumption per unit area and energy con-
sumption per capita of office buildings are far lower
than those in developed countries. At the same time,
Chinese office building has a bivariate distribution
structure, which is different from developed countries
in Europe and America.
There are two major methods for analyzing building
energy consumption: one is investigation and research
and the other is energy consumption simulation. When
finding building energy consumption data, investigation
and research is the most direct and reliable method.
However, the main shortcomings of the survey research
are a heavy workload and long cycle. As a result, many
scholars choose simulation software to analyze build-
ings’ energy consumption situation.
As the energy crisis occurred earlier in the United
States, Canada, and other developed countries, they
started their building energy analyses earlier. From the
beginning of the last century into the 1980s, American
scholars conducted research on the air conditioning
used for hotels, schools, hospitals, office buildings, and
other public buildings, and these basic research data are
the foundation for constructing energy saving works.8
In 1998, Robert Tamblyn and other scholars researched
the energy consumption for approximately 80 buildings
in the Canadian capital of Toronto; the construction
unit area total energy consumption index was used to
strengthen energy use intensity (EUI) and the energy
consumption structure was obtained.9
In the early 1990s, the Greek University of Athens,
Greece, set up a building energy consumption survey
group in support of the Ministry of Industry, spon-
sored by the Greek Technology Department, research
department, Ministry of Commerce, the Greek manu-
facturing center and the European Commission Energy
Management Center, aim to study, in the European
(mainly Greek) context, more than 1200 hotels, office
buildings, shopping malls, schools and hospitals, and
other public buildings for a period of 5 years.10
In 2004, Tsinghua University, based on measured
data of building energy consumption in hotels and shop-
ping malls, discussed the classification of office buildings
with three characteristics; after analyzing the energy sav-
ing potential of large public buildings, they also sug-
gested that the comprehensive utilization of all kinds of
building energy saving technologies could increase large
public buildings’ energy savings by 30%–50%.11 Zhang
Huan of Tianjin University investigated typical office
building annual energy consumption data in 2010 for 4–
6 and 8–11 months; for the Tianjin area, the two time
periods showed a sample building electricity consump-
tion per unit area of 26.79–125.45 kW h/(m2 a), with
average value of 64.25 kW h/(m2 a), respectively.12
Energy consumption simulation software has been
used in some areas for theoretical analysis and in the
analysis of energy consumption by many scholars both
at home and abroad. W Zhang13 of the Harbin
Institute of Technology used the eQUEST tool to simu-
late a building and a new model that had applied
energy-efficient measures to reduce annual energy con-
sumption. Q Li14 of Hunan University used the energy
consumption simulation analysis and simulation soft-
ware DeST to analyze a complex commercial building,
including a detailed analysis of the load characteristics,
construction and influencing factors, and typical room
load; also according to the above analysis, they also
put forward measures for saving energy in the typical
room. X Tan15 and G Chen16 of Chongqing University
used DeST to simulate the energy consumption of a
typical office building and analyzed the influence fac-
tors to propose energy saving measures.
SE Chidiac et al.17 used energy consumption simula-
tion software for architectural studies on aspects such
as location, size, operation, building envelope, electri-
cal, air-conditioning, and ventilation system perfor-
mance, as both single and multiple factors, for the
construction of energy saving effects. N Fumo et al.18
used the EnergyPlus benchmark model for simulation
to obtain a coefficient and proposed a simple method
that determines a series of predetermined coefficients
for the monthly electricity consumption and fuel bills,
Ding et al. 3

estimated an hour of electricity and fuel energy con-

sumption. The method has since been applied to
hypothetical buildings located in Atlanta and Meili
Dean, and in these two cases, the error in estimating
the energy consumption per hour was 10%. Fumo
et al.19 also used the EnergyPlus software to analyze
the energy consumption of combined heat and power
(CHP) systems and create economic models. Ke et al.20
examined the Energy Saving Performance Contract
(ESPC) of an office building by applying International
Performance Measurement and Verification Protocol
(IPMVP) Option D, in combination with the energy
analysis model established for the building by the
eQUEST simulation software, to calibrate energy con-
sumption simulation results using actual electricity bill-
ing data.
However, few reports are available concerning the
energy consumption situation and influencing factors
in the office buildings of China. Using the typical pub-
lic buildings in Severe Cold and Cold Region as an
example, this work analyses the current energy con-
sumption situation and influence factors for office
building energy consumption using an orthogonal
experiment method. First, according to the investiga-
tion data, the energy consumption situation and some
energy consumption influencing factor are analyzed.
Then, through an orthogonal experiment, the author Figure 2. Two main sources for data access: (a) public building
selects the heat transfer coefficient of the exterior wall energy consumption and (b) questionnaire monitoring system.
of the building, the heat transfer coefficient of the exter-
nal windows, the energy use of the lighting equipment,
overall power density, and the power density of air-
conditioning system, five of greatest impact factors in buildings used in the study was 19, and the number of
building energy consumption, for further analysis and government office buildings was 37.
study. Finally, the current work puts forward energy Data access for the buildings in this study was gained
saving measures according to the results of the analysis in the following three ways: (1) obtained from the pub-
and study. lic building energy consumption monitoring system, (2)
obtained from reports of the statistical data for energy
consumption compiled by the user, and (3) obtained
from organization personnel who did field research.
Method and data Among these, sources 2 and 3 were the main sources of
Energy consumption audit data for this study (Figure 2).
According to the Energy Audit Guidelines for
Government Office Buildings and Large-Scale Public
Buildings21 and the Environmental Protection Public
Energy consumption simulation and orthogonal
Welfare Project, a detailed energy audit was conducted experiment
in Severely Cold Regions. In this research article, basic eQUEST is a user-friendly building energy simulation
information and energy consumption data were investi- tool that provides high-quality results by combining a
gated for 56 typical office buildings in Beijing, Tianjin, building creation wizard, an energy efficiency measure
Dalian, Ji’nan, and Harbin. The information used in wizard, and a graphical results display module.22 An
the study included the building design specifications, eQUEST building simulation was used to perform
design drawings, calculations, drawings, energy sav- detailed analyses and study the impact factor of office
ings, energy consumption, and data from the comple- building energy consumption in Severely Cold Regions.
tion acceptance report. We divided the buildings into Orthogonal experimental design is a multi-factorial,
two types: government buildings and non-government multi-level experimental design method. It is based on
buildings. The number of non-governmental office orthogonality selected from comprehensive tests of
4 Advances in Mechanical Engineering
Figure 3. Age distribution of the buildings.
certain representative points; these representative points
have ‘‘evenly dispersed, neat features.’’ Orthogonal
experimental design is the main method used for frac-
tional factorial design. It is an efficient, rapid, and eco-
nomical method for the design of experiments. Using
the typical public building in the Severely Cold Regions
as an example, this work will analyze the influence of
various factors on office building energy consumption
with the orthogonal experimental method. In section
‘‘Case study analysis of office buildings’ energy con-
sumption influencing factor,’’ to provide a more intui-
tive and effective analysis of the data, the equivalent
electrical method23 is used to convert the energy con-
sumption into total energy consumption.
Static data analysis
Basic information for the buildings
As shown in Table 8 in Appendix 2, the 56 buildings
comprise a total construction area of 118.95 3 104 m2.
The total area of the government office buildings is
49.9 3 104 m2, and the total area of the non-
governmental office buildings is 69.05 3 104 m2. The
power consumption per unit of construction for the 56
buildings (without heating consumption) is 23.89–
153.13 kW h/(m2 a). The average power consumption per
unit of construction area (without heating) is 71.76 kW h/
(m2 a), and the comprehensive energy consumption per
unit of construction area is 33.86 kgce/(m2 a).
The age distribution of the sample of office buildings
used in this research, as measured by the completion
time shown in Figure 3, is as follows. The period before
1990 (including 1990) has four buildings, representing
7.14% of the survey. The period from 1990 to 1995 has
two buildings, accounting for 3.57% of the survey, and
the period from 1995 to 2000 has 13 buildings, account-
ing for 23.21% of the survey. For the period from 2000
to 2005, there are 11 buildings, representing 19.64% of
the survey, and for the period from 2005 to 2010, there
are 22 buildings, accounting for 39.29% of the survey.
Figure 4. Area distribution of the buildings.
The period since 2010 has four buildings, accounting
for 7.14% of the survey.
As shown in Figure 4, eight buildings have a construc-
tion area below 5000 m2, which accounts for 14.29% of
the total area of the buildings being researched. Eighteen
buildings have an area from 5000 to 10,000 m2, at
32.14% of the total. Eight buildings have an area from
10,000 to 15,000 m2, at 14.29% of the total. Four build-
ings have an area from 15,000 to 20,000 m2, representing
7.14% of the total, and four buildings have an area from
20,000 to 30,000 m2, accounting for 7.14% of the total.
Fourteen buildings have an area of more than 30,000 m2,
for 25% of the total construction area. Although the dis-
tribution of the construction area is uneven between
non-governmental and government office buildings, the
following analysis does not take this into account.
Electricity consumption of the buildings
The principal energy type used in the buildings under
investigation is electricity. The measurements and sta-
tistics associated with the total amount of electricity
consumption are relatively complete, but the data for
the power consumption at the sub-meter scale are not
complete. The annual energy consumption index of the
different building units can be obtained from a statisti-
cal analysis of the provided power data.
After completing research on the survey data, the
construction unit area of annual electricity consump-
tion is in the range of 23.89–153.13 kW h/(m2 a). The
electricity consumption for the average government
office building is 31.41–122.42 kW h/(m2 a), with an
average value of 76.56 kW h/(m2 a). The electricity con-
sumption for the average non-governmental office
building is 23.89–153.13 kW h/(m2 a), with an average
of 68.14 kW h/(m2 a). In addition, the energy intensity
frequency distribution is shown in Figure 5.
The non-governmental office buildings’ electricity con-
sumption is mainly concentrated in the range of 40–
100 kW h/(m2 a), and that of government office buildings
Ding et al.
Figure 5. Electricity consumption distribution of the buildings: (a) non-government office buildings and (b) government office
buildings.
Figure 6. Electricity consumption distribution for different
ages.
is concentrated in the range of 40–80 kW h/(m2 a).
Because date of construction and building energy con-
sumption are closely related, the electricity consumption
of the different non-governmental and government office
buildings of different ages are analyzed in terms of the dis-
tributed construction; the results are shown in Figure 6.
As shown in Figure 6, the average energy consump-
tion for non-governmental office buildings built prior
to 2000 was much higher than those built after 2000.
The annual average electricity consumption per average
unit construction area was 73.53 kW h/(m2 a) before
2000 and 68.14 kW h/(m2 a) after 2000, which is a
decrease of 7.3%. However, the difference in the energy
consumption of government office buildings from 1995
5
to 2005 was minimal. The annual average electricity
consumption for the average unit construction area was
65.31 kW h/(m2 a) before 2000 and 74.57 kW h/(m2 a)
after 2000, which is an increase of 14.2%. The main
reasons for the decline in energy consumption for non-
governmental office buildings after 2000 were probably
the improvement in thermal insulation performance of
the building envelope and the application of energy-
efficient equipment. The main reason for the increase in
building energy consumption for government office
buildings after 2000 was likely related to the significant
increase in office equipment, along with improved
requirements for indoor air quality. In general, public
service facilities have been greatly improved; therefore,
although the government office building maintenance
structure and heat preservation performance continue
to improve, the total energy consumption has still
increased relative to the period before 2000.
In Figure 7, it can be seen that the average electricity
consumption per unit area for buildings where the total
area is below 5000 m2 is the highest, but overall con-
sumption increases and then declines as the total area
grows from 5000 to 30,000 m2. Most of the buildings
below 5000 m2 are government office buildings, which
also tend to be older construction with poor thermal
performance in the building envelope. As the building
area increases, the services needed by the building will
increase, and the electricity consumption will grow.
Heat supply energy consumption of the buildings
According to the research, in addition to electricity
energy consumption, other major sources of energy
consumption include heating, kitchens, dining rooms,
6 Advances in Mechanical Engineering

Figure 7. Electricity consumption distribution for different

areas.

and hot water, but the heating energy consumption is Figure 8. Frequency distribution of heating energy
the largest factor. Construction in the cold region of consumption per unit area.
China, which is commonly known as the northern heat-
ing region and represents approximately 70% of
China’s land area, also accounts for approximately
40% of total construction for the country. Northern
heating energy consumption accounts for more than
24% of the total energy consumption in China, and
achieving energy savings here is the key to building
energy savings in China.
The consumption of natural gas is analyzed for eight
of the 56 buildings in this article that have direct-fired
machine systems. The gas consumption per unit area
varies widely from 0.52 to 11.16 m3/(m2 a), with an
average value of 4.98 m3/(m2 a). Because the ratio of
buildings analyzed to total building sample is small,
results cannot reliably reflect gas heating in the severely
cold area of China.
There are 31 buildings with central heating, which
account for 55.36% of the total building sample and, to
a certain extent, reflect the energy consumption result-
ing from winter heating. The annual heat consumption
per unit construction area varies widely from 0.01 to Figure 9. Heating consumption distribution for different
0.91 GJ/(m2 a), with an average value of 0.46 GJ/(m2 a). building ages.
From the survey results, the high values for the heating
energy consumption index are mainly due to the central those buildings built from 1995 to 2005 having high-
heating construction for government office buildings, energy consumption per unit construction area because
which were built in an earlier construction period, as they were built before the introduction of the ‘‘design
well as the poor thermal performance of the building standard for energy efficiency of public buildings
envelopes; in addition, many buildings do not have heat 2005.’’24 With the development of the economy, and
meters installed and only estimate the data (Figure 8). because people have increasingly demanded a comfor-
As shown in Figure 9, the heating energy consump- table office environment, energy consumption for cen-
tion for central heating shows a downward trend based trally heated office buildings experienced a growth
on how recently the building was completed, which is trend in the years 1995–2005. With the ‘‘design stan-
mainly due to upgrades to enclosure structures and cen- dard for energy efficiency of public buildings 2005,’’24
tral heating efficiency. However, although the overall the introduction of coal gas and the implementation of
trend is downward, there are big fluctuations, with a large number of small boilers, heating energy
Ding et al.
Figure 10. Heating consumption distribution in different areas.
consumption for buildings built after 2005 dropped sig-
nificantly compared with those built before 2005.
As shown in Figure 10, with the growth of the con-
struction area, the heating energy consumption for cen-
tral heating shows a downward trend, which is also
mainly due to upgrades to the enclosure structure and
improvements in the efficiency of central heating.
Although the overall trend is downward, there is a fluc-
tuation for buildings whose area is in the range of
10,000–15,000 m2. This is because there is only one
building in that area range that uses central heating; as
such, this is an outlier.
Comprehensive energy consumption of the buildings
From the survey data, the overall comprehensive energy
consumption during construction is in the range of
Figure 11. Comprehensive energy consumption distribution of the buildings: (a) non-government office buildings and
(b) government office buildings.
7
7.94–58.41 kgce/(m2 a), the comprehensive energy con-
sumption for government office buildings is between
7.94 and 57.23 kgce/(m2 a), and the average value is
39.10 kgce/(m2 a); comprehensive energy consumption
for non-governmental office buildings is 16.14–
58.41 kgce/(m2 a), with an average of 29.16 kgce/(m2 a).
The intensity frequency distribution for comprehensive
energy consumption is shown in Figure 11. The com-
prehensive energy consumption for non-governmental
office buildings is mainly concentrated in the range of
10–40 kgce/(m2 a), and that of government office build-
ings is concentrated in the range of 30–60 kgce/(m2 a).
In Figure 12, it can be clearly seen that the compre-
hensive energy consumption per unit area for non-
governmental office buildings and government office
buildings has decreased over time. For non-
governmental office buildings built after 2010, the com-
prehensive energy consumption per unit area compared
to the same for the period 1990–1995 fell by 58.34%,
whereas the energy consumption per unit area for gov-
ernment office buildings built after 2010 was down by
75.37% compared to 1990. The energy saving effect is
obvious. The relationship between improved electricity
consumption per unit area and building age is less obvi-
ous. This is mainly because although the improved
thermal performance of the building envelope and
energy saving appliances improved the overall energy
efficiency, people’s requirements in terms of indoor
environmental quality and comfort also increased.
Consequently, the rapid increase in the types and quan-
tities of electric equipment used has resulted in no sig-
nificant decline in the total energy consumption per
unit building area. It can be seen from Figure 13 that
the comprehensive energy consumption also shows a
8 Advances in Mechanical Engineering

Figure 13. Comprehensive energy distribution for different

Figure 12. Comprehensive energy consumption consumption
distribution for different ages.
downward trend with an increased building area.
Although electricity consumption grows as the building
area increases, energy consumption for heating
decreases with the increased area, and that reduction is
larger than the increase in energy consumption due to
the increased use of electricity.
Energy consumption level of the buildings
A box-whisker plot was used to reflect the level of
energy consumption for the research buildings. Before
using statistical methods to calculate the datum line or
level, data selection and processing should be done first
to remove the abnormal values. The methods that are
used to judge abnormal values assume a normal distri-
bution based on the mean and standard deviation of
the data. The commonly used methods for a normal dis-
tribution are Grubbs outliers and a Z score. Most of
the data, however, do not strictly obey the normal dis-
tribution, so the effectiveness of these methods is lim-
ited. Instead, a box line is applied; a box line (Box plot),
areas.
also known as the box (Box–whisker plot), is a kind of
statistical plot that shows data dispersion. To describe
the distribution of the data, the box line graph uses five
statistics from the data: the minimum value, the first
four points (Q1), the median (Q2), the first three or four
points (Q3), and the maximum value. The degree of dis-
persion of the data and outliers in the data can be
broadly seen to determine the bias of the data. A case
processing summary and a schematic diagram of the
box line plot are shown in Table 1 and Figure 14.
The center position of the box plot data shows the med-
ian; the upper and lower case on the data of the four digits
(Q3) and four digits (Q1) between Q3 and Q1 shows a dis-
tance called the four points distance (interquartile range
(IQR)); the two ends should show the maximum and mini-
mum values of the data. The standard for judging whether
a value in the box line is abnormal is whether it is less than
Q1–1.5IQR or more than Q3 + 1.5IQR. Test results can
be seen in the figure as less-abnormal data in the box line
graph, except for the comprehensive energy consumption
value of a non-governmental office building and excluding
a data label for 14.
At the same time, it can be seen from Figure 14
that non-governmental office buildings’ electricity

Table 1. Case processing summary.

Case
Valid Missing Total
N Percent (%) N Percent (%) N Percent (%)

Non-government office electricity consumption 19 51.4 18 48.6 37 100.0

Non-government office comprehensive consumption 19 51.4 18 48.6 37 100.0
Government office electricity consumption 19 51.4 18 48.6 37 100.0
Government office comprehensive consumption 19 51.4 18 48.6 37 100.0
Heating consumption 31 83.8 6 16.2 37 100.0
Ding et al. 9

Figure 14. Box–whisker plot of the buildings.

Table 2. Case processing result.

Heating Non-government Non-government Government Government

consumption building electricity building comprehensive building electricity building
(kgce/(m2 a)) consumption consumption (removing consumption comprehensive
(kW h/(m2 a)) the abnormal value) (kW h/(m2 a)) consumption
(kgce/(m2 a)) (kgce/(m2 a))

N
Valid 31 19 18 37 37
Missing 6 18 19 0 0
Average 14.37 68.14 28.22 74.57 39.10
Median 17.28 68.07 26.77 69.68 42.85
Mode 0.31a 31.41a 16.14a 23.89a 7.94a
Sum 445.54 1294.72 507.92 2758.90 1446.81
Percentage (%)
25 7.54 49.36 19.85 51.81 33.79
50 17.28 68.07 26.77 69.68 42.85
75 19.79 82.84 32.31 99.81 49.38
a
The existence of multiple modes. Display minimum.

consumption, comprehensive energy consumption, and
government office building electricity consumption,
which are the three indicators in the abdomen, are
more uniform. Additionally, the two indexes for com-
prehensive energy consumption and heating energy
consumption for government office buildings are small,
between the median and 75% (fourth quartile). This
result suggests that for the investigation of the building,
these two indicators constitute more than 25%–50% in
this range, whereas the other three indicators are more
evenly distributed in the range of 0%–100%. At the
same time, from Figure 14 and Table 2, it can be seen
that the mean of these two indicators for electricity
consumption per unit area and heating energy con-
sumption for government office buildings exhibits a
large difference between the mean and the average. The
other three indicators have a smaller gap between the
average and the median value.
Uncertainty factor analysis
1. Level of management is different
The building energy consumption is different for differ-
ent system management levels. Research on buildings
with a relatively high level of preservation of the build-
ing’s original information is robust, as building person-
nel typically have a good mastery of basic building
information, such as the operation of the building,
equipment, and number of personnel. However, the
actual operations and management systems of some
construction are relatively weak, and the relevant
10 Advances in Mechanical Engineering

Table 3. Building basic information.

Building types Location Area (m2) Story height (m) Layer number

Non-governmental office buildings Peking 35,872.56 3.5 3 layers underground, 12 layers on the ground

documented information is poorly arranged, resulting

Table 4. Building envelope information.
in data loss and reduced data reliability.
2. Data collection is different Type of envelope U-values (W/(m2 k))
Basic data are required to carry out any work, but sys- Exterior wall 0.538
tem monitoring, measurement, statistics, and other Roof 0.420
aspects of energy consumption are relatively weak in Floor 0.443
China. Whether for statistical methods or for time Glasses 2.6
interval statistics, it is very difficult to find a unified
standard. This results in a lack of unified data for
energy consumption, while enhancing the contradictory
nature of energy consumption statistics.

3. Limited number of samples, building information,

and energy consumption data

Due to the limited number of samples and the lack of

basic information and energy consumption data, some
data cannot fully reflect office building energy con-
sumption for the cold areas of China.

Case study analysis of office buildings’

energy consumption influencing factor
Figure 15. Building models of 3D.
To further analyze the energy consumption influence
factor for office buildings, the paper choose one typical
building to research. The simulation software eQUEST literature,25–27 a common method to verify the model is
was used in the study. to compare it with the actual energy consumption data
of a building. The calculation formula and the accepta-
ble error range of month error Em, year error Ey as well
Target buildings and building description as the variance coefficient of variation dm are given in
In Model 1 architecture, three floors underground are the previous study.25–27 The smaller these three num-
used for parking garages, levels 1–2 are mainly for com- bers are, the more reliable the model is. The calculation
mercial use, and levels 3–12 are mainly for office use. methods of the three coefficients are as follows
The basic information for the buildings can be seen in
Table 3 and the 3D models in Figure 15. Mt, i  Mc, i
Em = 3 100% ð1Þ
The parameters of the building envelope are shown Mt, i
in Table 4.
In addition to indoor light, equipment, and other P
12
Em
initial conditions are parameters that affect the build- i=1
Ey = ð2Þ
ing’s energy consumption. Indoor light, equipment, 12
and other initial conditions are shown in Table 5. sffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi
The parameters of building Heating,Ventilating and P12
(Mt, i  Mc, i )2 =12
Air Conditioning (HVAC) system are shown in Table 6. i=1
dm = ð3Þ
Mt
Model calibration where Mt,i represents building I’s months of actual
Calibrating the model is an important part of building energy consumption, Mc,i represents building I’s
an energy consumption simulation. According to the months of simulation of energy consumption, and Mt
Ding et al. 11

Table 5. Indoor light, equipment and other initial conditions.

Type of room Occupancy Lighting power Equipment power Fresh air

density (m2/person) density (W/m2) density (W/m2) volume (m3/(h person))

Commercial 5 15 13 30
General office 8 9 15 30
Restroom 10 5 0 30
Stores 10 9 0 0
Exhibition 5 11 5 30
Corridor 50 9 0 20
Staircase 50 9 0 20
Air-conditioning facilities 500 9 0 20
Garage 50 10 0 20
Retiring room 7 9 0 20

Table 6. Air-conditioning system information.

Parameter Model
1

Cooling set point (°C) 26

Heating set point (°C) 18
Cool air supply temperature 8
difference (°C)
Heating air supply temperature 15
difference (°C)
Air-conditioning system CAV
HVAC operation time 8:00–17:00
Refrigerating type Centrifugal chiller
COP of refrigerating unit 5.2
Chilled water supply/return 7/12
temperature (°C) Figure 16. Comparison between simulated and actual energy
Cooling water supply/return 30/35
consumption.
temperature (°C)

CAV: constant air volume; COP: coefficient of performance.

Orthogonal experiment
While there are many influence factors for the energy
represents the average value of the building’s real
consumption of office buildings, they can generally be
energy consumption.
divided into two categories: external and internal
The verification results of the energy consumption
causes. External causes refer to various kinds of inter-
models are shown in Figure 16. The short line in the
ference with the indoor thermal environment, which
graph is the limit error of the line 15% between simu-
includes two parts of the inner disturbance and outside
lated and actual energy consumption error provisions
disturbance. Inner disturbance refers to the changes of
in the document (Figure 16).26
the human body and equipment heat, moisture dissipa-
According to formulas (1)–(3), the calculation results
tion, and lighting heat dissipation of an inside room.
are shown in Table 7: the maximum deviation of the Em
Outside disturbance refers to weather factors, such as
month of the building as it appeared in July is 8.26%,
changes in outdoor air temperature, humidity, solar
the year deviation coefficient is Ey = 4.40%, and the
radiation intensity, and wind speed and direction.
variance coefficient variation is dm = 5.07% months.
Internal causes mainly refer to the condition of the
The deviation coefficient table is within the acceptable
building itself, including building orientation, and to
error range in the literature25–27 compared with the
palisade structure parameters such as structure and
model. Although the model does not meet the delta dm
shape coefficients.
within the26 literature limits, the other limits can satisfy
Office building energy consumption composition
other values in the literature. This shows that the typi-
mainly includes the energy consumption of air condi-
cal building energy models can be used to represent the
tioning, lighting, and equipment; the influence factors
actual situation, and that the actual deviation is smaller,
corresponding to that are air conditioning, lighting,
with good reliability.
and equipment energy consumption.
12 Advances in Mechanical Engineering

Table 7. Comparison of two building simulation errors and the range of acceptable error.

Em,max Ey dm

ASHRAE Guideline 14-200225 6 5% – 6 15%

DOE/GO-102002-155426 6 20% – 6 5%
US Department of Energy27 6 15% 6 10% 6 10%
The model 8.26% 4.4% 5.07%

ASHRAE: American Society of Heating, Refrigerating and Air-Conditioning Engineers.

Figure 17. Variables of energy consumption, electricity consumption, and gas consumption for heating in different exterior wall U-
values.

The paper mainly studies the following factors that
affect the energy consumption of office buildings:
window-wall ratio, wall type, window type, occupancy
density, lighting equipment load, roof type, shading
coefficient and indoor set temperature, air index and
air-conditioning system forms, and chiller coefficient of
performance (COP). To confirm the influence degree of
various factors on the results, two representative
indexes and two typical buildings were selected. The
design and results of orthogonal test table can be seen
in Table 9 in Appendix 2.
According to the range of relative size of influence,
the order of factor importance is air-conditioning
system . lighting density . indoor design temperature
. exterior window type . outside shading . fresh air
volume . personnel density . roof type . COP of
refrigerating unit . window-wall ratio . exterior wall
type. The analysis results show that the influence of the
air-conditioning system is the largest, followed by light-
ing density and the building envelope. Thus, the paper
analyzes these factors.
Enclosure structure
Exterior wall. The wall is a very important part of the
building envelope. The heat enters the room through
Ding et al.
the wall structure through two major paths: the convec-
tion heat transfer between the outdoor air and the
enclosure structure and the solar radiation heat transfer
through the wall.
According to the needs of the research, five different
heat transfer coefficients for the external walls of the
model were selected for simulation and analysis. The influ-
ence on the total energy consumption is shown below.
With an increase in the exterior wall heat transfer
coefficient, the total annual energy consumption, elec-
tricity consumption, and gas consumption for heating
all increased basically linearly. The effect of different
exterior wall heat transfer coefficients is obvious on the
total energy consumption, electricity, and gas con-
sumption. As shown in Figure 17, the increase in gas
consumption is the most obvious. The total energy con-
sumption of the two models increased most obviously
during winter, which also means that the exterior wall
heat transfer coefficient change has a huge effect on the
heating energy consumption.
Some things can be calculated using the data in
Figure 17. In the model, when the exterior wall heat
transfer coefficient rises from 0.5 to 2.5 W/(m2 k), the
total energy consumption grows from 3933.01 3 103 to
Figure 18. Variables of energy consumption, electricity consumption, and gas consumption in different glass U-values.
13
4251.73 3 103 kW h, increasing by 8.1%. Thus, when
the heat transfer coefficient of the exterior wall of the
first modeled building increases 1 W/(m2/K), the total
energy consumption per unit area of the building rises
4.44 kW h(m2 a). Electricity consumption grows from
3748.6 3 103 to 3974 3 103 kW h, increasing by
6.01%. Gas consumption rises from 629.17 3 106 to
947.56 3 106 btu, growing by 50.6%; this growth is
very obvious. The change in the exterior wall heat
transfer coefficient has the greatest impact on gas con-
sumption for heating.
Exterior window. For building energy savings, doors and
windows are the key. In summer, blocking the heat of
the outdoor to indoor conduction can maintain an
indoor refrigeration effect, whereas in winter, this can
reduce the loss of indoor heat. In this process, doors
and windows play a key role.
According to the needs of the research, five groups
of different heat transfer coefficients for the external
windows were selected for simulation and analysis. As
seen from Figure 18, the influence of exterior window
heat transfer coefficient changes for total energy
14
Figure 19. Variables of energy consumption, electricity consumption, and gas consumption in different lighting power densities.
consumption, building energy consumption, and power
consumption of heating gas is obvious. The pictures
show that with an increase in the exterior window, heat
transfer coefficient, the total energy consumption, elec-
tricity consumption, and gas heating energy consump-
tion of the two models decrease during May–September
but increase in November–January. This is mainly
because May–September belongs to the cooling season
for two models. Furthermore, because thermal inertia
exists in the building materials, the outdoor tempera-
ture is lower than the indoor temperature at night, so a
smaller coefficient of thermal conductivity is not condu-
cive to heat release from indoors to outdoors. This
requires interior air conditioning to release the
increased heat during the day and causes a greater heat
increase than the reduced daytime heat gain with an
improved exterior glass thermal insulation perfor-
mance; thus, the energy consumption increases.
With an increase in the exterior window heat trans-
fer coefficient, the total energy consumption and gas
consumption increased basically linearly, and the elec-
tricity consumption showed a trend of decline. The
results can be calculated using the data in Figure 18. In
Advances in Mechanical Engineering
the model, when the exterior window glass heat transfer
coefficient increased from 1.8 to 5.5 W/(m2 k), the total
energy consumption grew from 3903.391 3 103 to
3967.999 3 103 kW h, an increase of 1.66%, represent-
ing limited growth. The effect of the exterior window
heat transfer coefficient on total annual energy con-
sumption is limited, and economies should be fully con-
sidered while saving energy. Electricity consumption
fell from 3736.6 3 103 to 3705.9 3 103 kW h, a
decrease of 0.82%, whereas gas consumption increased
from 569.06 3 106 to 894.23 3 106 btu, with a very
obvious growth of 57.14%.
Lighting and equipment
Lighting power density. According to the results of the
orthogonal test for the model in section ‘‘Orthogonal
experiment,’’ the lighting power density factor is the
second most influential factor, following the air-
conditioning system factors.
The paper uses five different lighting power densities
for the simulation analysis. In the model, total energy
consumption and power consumption present basic
Ding et al.
Figure 20. Variables of energy consumption, electricity consumption, and gas consumption in different equipment power densities.
linear growth with respect to increments of lighting
power density, and the gas consumption presents a
straight decline. This is mainly due to the heating sea-
son: with an increase in the power density of the light-
ing, the heat capacity of the equipment increases to
reduce gas consumption and achieve the required
design temperature. Conversely, during the cooling sea-
son, there is a need for greater power consumption,
which increases the electricity consumption, to achieve
colder summer indoor design temperatures. During the
transition season, lighting power density also increases
the electricity consumption of the building itself.
Further results can be calculated using the data in
Figure 19. For the model, when the lighting power den-
sity rises from 6 to 14 W/m2, the total energy consump-
tion increases from 3697.402 3 103 to 4308.087 3 103
kW h, an increase of 16.52%. Electricity consumption,
which went from 3495.6 3 103 to 4148.6 3 103 kW h,
grew by 18.68%. Gas consumption dropped from
688.51 3 106 to 544.14 3 106 btu, falling by 20.97%.
Equipment power density. The effect of office equipment
power density on the energy consumption of buildings
15
is similar to that of lighting density. Along with an
increase in device power density, the total energy con-
sumption and electricity consumption present basic lin-
ear growth, and the gas consumption presents a
straight decline, as shown in the model. This is mainly
due to the heating season, for which an increase in the
power density of the office equipment causes the heat
capacity of the equipment to increase, thereby reducing
gas consumption and achieving the required design
temperature. However, for the cooling season, to
obtain a cold indoor design temperature in summer,
more power must be consumed, thereby increasing the
electricity consumption. In the transition season, the
lighting power density increases also increases the elec-
tricity consumption of the building itself.
The data in Figure 20 are used for calculation. In the
model, when the office equipment power density rises
from 11 to 19 W/m2, the total energy consumption
increases from 3741.164 3 103 to 4161.837 3 103 kW h,
increasing by 11.24%. Electricity consumption increases
from 3532.7 3 103 to 4002.2 3 103 kW h, growing by
13.29%. Gas consumption decreases from 711.24 3 106
to 544.65 3 106 btu, falling by 23.42%.
16
Figure 21. Variables of energy consumption and electricity
consumption in different air-conditioning systems.
Air-conditioning system
Ground source heat pumps (GSHP) involve the use of
shallow land energy through the input of a small
amount of high-grade energy (such as energy) to achieve
energy from a low-grade to high-grade heat transfer.
The GSHP is one of the most current and mature tech-
nologies for use of renewable energy in air conditioning
and as a heat source. Therefore, by setting a GSHP as a
building’s cold and heat source (with a variable air vol-
ume (VAV) air-conditioning system), and comparing
these results with the energy use of a non-GSHP con-
stant air volume (CAV) air-conditioning system, VAV
and variable refrigerant volume (VRV) air-conditioning
system, the simulation results below can be generated.
For the non-GSHP data, the common centrifugal chiller
is used as the cold source and the gas-fired boiler as the
heat source for CAV and VAV.
Figure 21 clearly shows that the building’s total
energy consumption and electricity consumption are
smaller than those of CAV and VAV but slightly greater
than that of VRV. The CAV all-air system is the largest
energy consumer, which is mainly caused by the too-
large fan energy consumption. The model’s total energy
consumption when using CAV is 3933.01 3 103 kW h,
while VAV is 2750.78 3 103 kW h, GSHP is 2536.40
3 103 kW h, and that of VRV is only 2389.7 3
103 kW h. Relative to the CAV system, the energy sav-
ing rate is 39.24%. Using a GSHP as the cold and heat
source, the total energy consumption is lower than that
of the centrifugal chiller and gas boiler as cold and heat
sources for CAV and VAV but slightly higher than that
of VRV. Different air-conditioning systems’ energy con-
sumption simulation found that the air-conditioning
system had a huge impact on the building energy con-
sumption, and overall, the energy efficiency of the
GSHP system is still quite obvious.
Advances in Mechanical Engineering
Uncertainty factor analysis
Although the overall results of the simulation verify the
accuracy of the model and the typical building basically
meets all of the indicators, there is still some error. The
causes for this deviation include the following:
1. The software master is not very good, with
mobile errors;
2. The running times of various equipment and
personnel in the documentation are inconsistent
with their actual operation;
3. Because the wide variety of equipment often can-
not match a specific system’s parameters, longer
service life and other issues can occur, leading to
decreased equipment performance; and
4. When using meteorological parameters to simu-
late typical parameters, the meteorological para-
meters may exhibit certain discrepancies.
Conclusion and suggestions
Conclusion
Using typical public buildings in severe cold and cold
regions as examples for the work analysis and study of
the energy consumption situation, the influence factors
of office building energy consumption were found using
an orthogonal experiment method. In this research, the
basic information and energy consumption data for 56
typical office buildings are investigated, including 19
non-government office buildings and 37 government
office buildings. According to the investigation data,
the energy consumption situation and some factors that
influence energy consumption were analyzed. Through
an orthogonal experiment, the author selected the exter-
ior wall of the building’s heat transfer coefficient, as
well as the heat transfer coefficients of the external win-
dows and lighting equipment, the power density, and
the power density of the air-conditioning system as five
of greatest impact factors for building energy consump-
tion for further analysis and study. The simulation soft-
ware eQUEST was used in the study. Energy saving
measures were put forward according to the results of
the analysis and study.
According to the survey data, the construction unit
construction area of annual electricity consumption
ranges from 23.89 to 153.13 kW h/(m2 a); the govern-
ment office building electricity consumption is in the
31.41–122.42 kW h/(m2 a) range, with an average value
of 76.56 kW h/(m2 a); non-government office building
electricity consumption is in the 23.89–153.13 kW h/
(m2 a) range, with an average of 68.14 kW h/(m2 a). In
addition, the non-government office buildings’ electric-
ity consumption is mainly concentrated in the 40–
100 kW h/(m2 a) range, whereas that of government
Ding et al.
office buildings is concentrated between 40 and
80 kW h/(m2 a).
In the research of 56 buildings, for 8 buildings with
direct-fired machine systems, the consumption of natu-
ral gas was analyzed, and the gas consumption per unit
area was in the 0.52–11.16 m3/(m2 a) range with huge
differences; the average value was 4.98 m3/(m2 a).
Overall, 31 buildings with central heating accounted
for 55.36% of the total number of samples and, to a
certain extent, reflected the energy consumption of win-
ter heating. The unit construction area of annual heat
consumption is in 0.01–0.91 GJ/(m2 a) with huge differ-
ences, and the average value is 0.46 GJ/(m2 a).
From the survey data, the comprehensive construc-
tion energy consumption ranges from 7.94 to
58.41 kgce/(m2 a); the government office buildings’
comprehensive energy consumption is 7.94–57.23 kgce/
(m2 a) with an average value of 39.10 kgce/(m2 a); and
non-government office buildings’ comprehensive
energy consumption is 16.14–58.41 kgce/(m2 a) with an
average of 29.16 kgce/(m2 a). In addition, the non-
government office buildings’ comprehensive energy
consumption is mainly concentrated in the 20–40 kgce/
(m2 a) range, whereas that of government office build-
ings is concentrated in the 30–60 kgce/(m2 a) range.
The two indexes of government office buildings’
comprehensive energy consumption and heating energy
consumption are smaller between the median and 75%
(fourth percentile). This result suggests that for the
investigation of the building, the influence from these
two indicators is more in the 25%–50% range, whereas
the other three indicators are more evenly distributed in
a range from 0% to 100%. At the same time, when
these results are combined with those of Figure 13 and
Table 3, it becomes clear that the mean of these two
indicators of government office building electricity con-
sumption per unit area and heating energy consumption
have a large difference between the mean and average,
whereas the other three indicators show a smaller gap
between the average and median values.
The paper analyses and studies the factors that affect
the energy consumption of office buildings using an
orthogonal experiment method. A deviation coefficient
table with acceptable error ranges from the literature25–
27
was compared with the model. Although the model
does not meet the delta dm in the26 literature limits, the
other limits satisfy all three values in the literature. This
shows that the typical building energy models can be
used to show the actual situation, and that the actual
deviation is smaller, with good reliability.
According to the range of results for relative size of
influence, the order for the building’s influential factors
is air-conditioning system . lighting density . indoor
design temperature . exterior window type . outside
shading . fresh air volume . personnel density . roof
type . COP of refrigerating unit . window–wall ratio
17
. exterior wall type. Through analysis, it was found
that the air-conditioning system is the most influential
impact factor, followed by the lighting equipment and
the building envelope.
With the increase in the exterior wall heat transfer
coefficient, the total annual energy consumption and
electricity and gas consumption increased nearly linearly;
with an increase in the exterior window heat transfer
coefficient, total annual energy consumption and gas
consumption increased nearly linearly, but the electricity
consumption was decreased; and with an increase in the
lighting power and office equipment density, the total
annual energy consumption and electricity consumption
had a linear growth trend, but the heating gas consump-
tion decreased. The use of GSHP as a cold and heat
source air-conditioning system consumes less energy than
the CAV and VAV and slightly more than the VRV.
Suggestions
An analysis of the results indicates that the main fac-
tors affecting the energy consumption of buildings are
the building envelope, lighting equipment, office equip-
ment, and air-conditioning system. Therefore, the arti-
cle puts forward the following suggestions:
1. To minimize the heat transfer coefficient of the
exterior walls and windows, an appropriate
reduction of the heat transfer between the
indoor environment and the outdoor air is
indicated.
2. Replace lighting with energy saving lamps,
reducing the lamp power density. A large num-
ber of old T8 lamps are currently being used in
some office buildings in China.
3. To improve the energy management system,
office users should form the habit of turning off
the lights when they leave and using fewer lights
to satisfy illumination needs when possible,
reducing the power consumption of lighting
lamps.
4. Turn off the computer, printer, and other office
equipment in a timely manner, in order to
reduce their power consumption.
5. Where there is a choice in the appropriate form
of air-conditioning system, attempt to the limit
possible to avoid using high-energy consuming
air-conditioning systems.
6. To the extent possible, use renewable energy
sources to reduce the pollutant emissions and
energy consumption of buildings.
Acknowledgement
The authors also would like to acknowledge Ministry of
Environment Protection of the People’s Republic of China
18
and School of Environmental Science and Engineering of
Tianjin University in supporting our building energy audit
and providing valuable comments for this research.
Declaration of conflicting interests
The author(s) declared no potential conflicts of interest with
respect to the research, authorship and/or publication of this
article.
Funding
The author(s) disclosed receipt of the following financial sup-
port for the research, authorship and/or publication of this
article: This research was funded by the State Nuclear Electric
Power Planning Design & Research Institute Support Key
Project under grant number 100-KY2017-FYZ-N01 and by
the Beijing Municipal Science and Technology Project under
grant number Z171100000317004.
References
1. Intergovernmental Panel on Climate Change. Climate
change 2007: impacts, adaptation and vulnerability. New
York: Cambridge University Press, 2007.
2. National Bureau of Statistics of China (NBS). Interna-
tional statistical yearbook 2013. Beijing, China: China
Statistic Press, 2013.
3. Zhou S, Tong Q, Yu S, et al. Role of non-fossil energy in
meeting China’s energy and climate target for 2020.
Energ Policy 2012; 51: 14–17.
4. National Bureau of Statistics of China (NBS). Tianjin
statistic yearbook 2013. Beijing, China: China Statistics
Press, 2013.
5. Zeng L, Xu M, Liang S, et al. Revisiting drivers of
energy intensity in China during 1997–2007: a structural
decomposition analysis. Energ Policy 2014; 67: 640–647.
6. International Energy Agency (IEA). CO2 emissions from
fuel combustion 2013, 2013, http://www.doc88.com/p-95
95130594159.html
7. Cheng L. Research on design measures of energy effi-
ciency in office buildings in hot summer and cold winter
zone. Master’s Thesis, Zhejiang University, Hangzhou,
China, 2007.
8. International Energy Agency (IEA). Key world energy
statistics (www.iea.org), 2002, http://alofatuvalu.tv/FR/
12_liens/12_articles_rapports/IEA_rpt_2012_us.pdf
9. Zhu B. Energy saving renovation of office buildings. Sci
Technol Overseas Build Mater 2004; 25: 102–105.
10. Cxakmanus I. Renovation of existing office buildings in
regard to energy economy: an example from Ankara,
Turkey. Build Environ 2007; 42: 1348–1357.
11. Zhifeng X and Yi J. Energy consumption and energy sav-
ing potential analysis for large-scale public buildings in
Beijing. Heat Vent Air Cond 2004; 34: 8–10; 24.
12. Li W. A research on energy conservation method of office
building in Tianjin. Master’s Thesis, Tianjin University,
Tianjin, China, 2013.
13. Zhang W. Analysis of public buildings energy-saving poten-
tial based on energy simulation. Master’s Thesis, Harbin
Institute of Technology, Harbin, China, 2012.
Advances in Mechanical Engineering
14. Li Q. Energy consumption simulation and energy-saving
analysis on public buildings at Changsha district. Master’s
Thesis, Hunan University, Changsha, China, 2013.
15. Tan X. Survey and research on energy consumption of
office buildings in Chongqing. Master’s Thesis, Chongqing
University, Chongqing, China, 2008.
16. Chen G. Research on energy efficiency and energy consump-
tion simulation of public buildings in Chongqing. Master’s
Thesis, Chongqing University, Chongqing, China, 2007.
17. Chidiac SE, Catania EC and Morofsky E. Effectiveness
of single and multiple energy retrofit measures on the
energy consumption of office buildings. Energy 2011; 36:
5637–5652.
18. Fumo N, Magro P and Luck R. Methodology to estimate
building energy consumption using EnergyPlus bench-
mark models. Energ Buildings 2010; 42: 2331–2337.
19. Fumo N, Mago P and Chamra L. Energy and economic
evaluation of cooling, heating, and power systems based
on primary energy. Appl Therm Eng 2009; 29: 2665–2671.
20. Ke M, Yeh C and Jian J. Analysis of building energy con-
sumption parameters and energy savings measurement
and verification by applying eQUEST software. Energ
Buildings 2013; 61: 100–107.
21. Braun MR, Altan H and Beck SM. Using regression
analysis to predict the future energy consumption of a
supermarket in the UK. Appl Energ 2014; 130: 305–313.
22. Zhang D, Zhang X and Cai N. Study on energy saving
possibility of digital variable multiple air conditioning
system in three office buildings in Shanghai. Energ Build-
ings 2014; 75: 23–28.
23. Yi J and Xiu Y. The equivalent electrical method is
adopted in the energy analysis. Energ Chin 2010; 3: 5–11.
24. GB 50189-2015:2005. Design standard for energy effi-
ciency of public building.
25. ASHRAE Guideline 14-2002. Measurement of energy
and demand savings, 2002, http://www.eeperforman-
ce.org/uploads/8/6/5/0/8650231/ashrae_guideline_14-
2002_measurement_of_energy_and_demand_saving.pdf
26. DOE/GO-102002-1554:2007. International Performance
Measurement & Verification Protocol (IPMVP).
27. US Department of Energy. M&V guidelines: measure-
ment and verification for federal energy projects version
3.0, 2007, https://portal.hud.gov/hudportal/documents/
huddoc?id=doc_10604.pdf
Appendix 1
Notation
Em months error
Ey year error
Mc,i building I’s months of simulation of
energy consumption
Mt average value of building real energy
consumption
Mt,I building I’s months of actual energy
consumption
dm variance coefficient of variation
Appendix 2
Ding et al.

Table 8. Basic information of audit building.

Number Area (m2) Classification Built time Electricity Gas Central heating Comprehensive Comprehensive
consumption consumption consumption consumption consumption
per unit area per unit area per unit area (tce) per unit area
(kW h/(m2 a)) (m3/m2 a) (GJ/m2 a) (kgce/m2 a)

J1 3467 Non-government building 2011 88.23 – – 107.066 30.88

J2 7712 Non-government building 2001 49.36 5.19 – 181.885 23.58
J3 11,956 Non-government building 2010 77.61 – – 324.758 27.16
J4 12,324 Non-government building 2010 76.83 – – 331.385 26.89
J5 14,186 Non-government building 2011 50.83 – – 252.362 17.79
J6 17,801 Non-government building 2001 79.89 – – 497.767 27.96
J7 22,296 Non-government building 2001 31.41 5.47 – 393.138 17.63
J8 30,191 Non-government building 2008 51.93 – 0.169 722.511 23.93
J9 33,891 Non-government building 2001 44.75 6.29 – 789.548 23.3
J10 38,402 Non-government building 2002 68.07 – – 1405.876 36.61
J11 43,000 Non-government building 1999 61.14 – 0.006 1164.834 27.09
J12 48,500 Non-government building 2000 63.54 – 0.006 1292.263 26.64
J13 51,780 Non-government building 1996 89.83 1.27 0.009 2434.485 47.02
J14 52,000 Non-government building 1991 122.42 3.54 0.012 3037.359 58.41
J15 54,500 Non-government building 2009 46.1 – 0.005 879.401 16.14
J16 56,491 Non-government building 2009 48.35 – 0.005 955.871 16.92
J17 58,000 Non-government building 1998 73.98 11.16 – 2287.481 39.44
J18 65,400 Non-government building 1999 82.84 0.52 – 3010.097 46.03
J19 68,639 Non-government building 2009 87.61 0.17 0.28 1408.96 20.53
J20 4245 Government building 2010 133.08 – – 197.747 46.58
J21 4526 Government building 2009 124.61 – – 197.41 43.61
J22 5174 Government building 2009 123.86 – – 224.329 43.35
J23 7617 Government building 2008 119.06 6.38 – 377.196 49.52
J24 12,096 Government building 2008 153.13 – – 648.283 53.59
J25 19,000 Government building 2010 54.71 – – 425.524 22.4
J26 19,323 Government building 2010 55.24 – – 441.037 22.82
J27 8547.73 Government building 2011 60.09 0.46 – 67.889 7.94
J28 2269 Government building 2011 106.58 1.94 – 35.08 13.62
J29 6437 Government building 1999 23.89 – 0.24 73.67 11.44
J30 8197.28 Government building 2002 38.81 0.87 0.3 131.62 16.06
J31 47,203 Government building 2007 104.35 – 0.62 2701.471 57.23
J32 29,796.48 Government building 2003 95.27 – 0.55 1545.179 51.86
J33 24,589.81 Government building 1988 88.37 – 0.51 1178.978 47.95
J34 22,000 Government building 2006 125.45 – – 957.6916 43.53
J35 16,585 Government building 1989 107.93 – 0.52 913.7773 55.1
19

(continued)
 20
Table 8. Continued

Number Area (m2) Classification Built time Electricity Gas Central heating Comprehensive Comprehensive
consumption consumption consumption consumption consumption
per unit area per unit area per unit area (tce) per unit area
(kW h/(m2 a)) (m3/m2 a) (GJ/m2 a) (kgce/m2 a)

J36 13,000 Government building 2010 61.62 – – 277.9605 21.38

J37 12,762 Government building 1990 29.34 – – 483.3577 37.87
J38 11,796 Government building 2008 26.79 – – 447.3397 37.92
J39 11,000 Government building 2005 73.97 – 0.81 586.8414 53.35
J40 9874.07 Government building 2007 77.83 – 0.63 480.2608 48.64
J41 9397.8 Government building 1999 53.28 – 0.6 365.4676 38.89
J42 9240 Government building 2009 50.06 – 0.69 395.8994 42.85
J43 9000 Government building 1999 41.77 – 0.64 328.2259 36.47
J44 8959 Government building 2004 68.13 – 0.53 375.2972 41.89
J45 8750.68 Government building 2007 72.5 – 0.6 398.2516 45.51
J46 8525 Government building 2003 53.12 – 0.58 364.6065 42.77
J47 8000 Government building 2006 43.86 – 0.53 267.5343 33.44
J48 7875.26 Government building 1998 81.9 – 0.64 395.469 50.22
J49 6400 Government building 1995 77 – 0.67 317.7644 49.65
J50 5570 Government building 1990 50.49 – 0.49 190.1569 34.14
J51 5411 Government building 1996 54.44 – 0.59 211.9963 39.18
J52 5000 Government building 2002 69.68 – 0.57 217.904 43.58
J53 4000 Government building 1999 74.43 – 0.85 218.8967 54.72
J54 3780 Government building 2000 80.88 – 0.62 186.0928 49.23
J55 2400 Government building 2000 54.09 – 0.91 90.71 37.79
J56 100,640 Government building 2010 49.29 12.08 – 2085.41 20.72
Advances in Mechanical Engineering
Table 9. Orthogonal experiment design and result of the model.

1 2 3 4 5 6 7 8 9 10 11 12
Ding et al.

Window– Exterior Exterior Occupancy Lighting Roof External Indoor Fresh air Air- COP of Result
wall wall window density power type sunshade design volume conditioning refrigerating (104
ratio (W/m2 k) (W/m2 k) (m2/person) density (W/m2 k) temperature (m3/(h person)) system unit kW h)
(W/m2) (°C)

Experiment1 0.4 0.538 2.6 8 9 0.42 No 26 30 Constant air 5.2 393.29

volume
Experiment2 0.4 0.538 2.6 8 9 0.326 Yes 24 20 Variable air 4.2 279.07
volume
Experiment3 0.4 0.538 1.8 4 7 0.42 No 26 20 Variable air 4.2 261.34
volume
Experiment4 0.4 0.404 2.6 4 7 0.42 Yes 24 30 Constant air 4.2 388.16
volume
Experiment5 0.4 0.404 1.8 8 7 0.326 No 24 30 Variable air 5.2 272.09
volume
Experiment6 0.4 0.404 1.8 4 9 0.326 Yes 26 20 Constant air 5.2 374.56
volume
Experiment7 0.4 0.538 1.8 4 9 0.42 Yes 24 30 Variable air 5.2 286.29
volume
Experiment8 0.3 0.538 1.8 8 7 0.326 Yes 26 30 Constant air 4.2 358.95
volume
Experiment9 0.3 0.538 2.6 4 7 0.326 No 24 20 Constant air 5.2 388.33
volume
Experiment10 0.3 0.404 1.8 8 9 0.42 No 24 20 Constant air 4.2 387.76
volume
Experiment11 0.3 0.404 2.6 4 9 0.326 No 26 30 Variable air 4.2 285.98
volume
Experiment12 0.3 0.404 2.6 8 7 0.42 Yes 26 20 Variable air 5.2 257.74
volume
Mean1 328.09 327.88 332.09 324.82 334.49 329.10 331.47 321.98 330.79 381.84 328.72
Mean2 327.51 327.72 323.50 330.78 321.10 326.50 324.13 333.62 324.80 273.75 326.876
Range 0.579 0.163 8.595 5.960 13.39 2.602 7.338 11.643 5.995 108.089 1.840

COP: coefficient of performance.

21