Energy and Buildings 127 (2016) 433–441

Contents lists available at ScienceDirect

Energy and Buildings

journal homepage: www.elsevier.com/locate/enbuild

Cost-effective energy saving measures based on BIM technology: Case

study at National Taiwan University
Sy-Jye Guo a,1 , Taibing Wei b,c,∗,1
a
Professor, Department of Civil Engineering, National Taiwan University, Taipei 10617, Taiwan
b
Ph.D. Candidate, Department of Civil Engineering, National Taiwan University, Taipei 10617, Taiwan
c
Lecturer, Department of Civil Engineering and Architecture, Wuyi University, Wuyishan 354300, China

a r t i c l e i n f o a b s t r a c t

Article history: This study utilized BIM technology (eQUEST) to conduct an energy consumption analysis and simulated
Received 16 October 2015 re-design of the Civil Engineering Research Building at National Taiwan University. After obtaining a
Received in revised form 27 May 2016 verifiably accurate simulation, energy-saving effects were investigated by altering the building’s envelop
Accepted 4 June 2016
design and analyzing the results. In addition, cost remains a crucial factor throughout the process in
Available online 5 June 2016
order to reflect the relationship between building expenses and energy efficiency during design scheme
optimization. The results of this study can provide better and more comprehensive choices for building
Keywords:
owners, designers, and developers in future.
Energy consumption
Energy simulation © 2016 Elsevier B.V. All rights reserved.
Economic analysis
Envelop
Orthogonal design

1. Introduction tual construction and operation phases. Utilizing BIM significantly

According to the United Nations Environment Programme
(UNEP) [1], buildings are the largest worldwide consumers of
energy. The impact of energy consumption is particularly signif-
icant in the usage phase, which can last several decades. Most of
the energy used by any building is consumed during the usage
(or operational) stage of the building’s life-cycle. Researchers
and developers have considered methods of energy consumption
reduction for many years, however, reducing energy consump-
tion almost always means increasing cost. For this reason, energy
efficiency is not often a high priority for building owners. It is cru-
cial to secure techniques for successfully optimizing designs while
keeping them cost-effective, as doing so provides more options for
building owners or designers in the design phase.
Building Information Modeling (BIM) is a relatively new tech-
nique that has been developed and applied in the engineering
field. It is a data model that integrates all relevant information
for engineering projects with three-dimensional digital technology.
BIM is mainly used in the design phase, but also during the even-
∗ Corresponding author at: Department of Civil Engineering and Architecture,
Wuyi University, Wuyishan 354300, China.
E-mail address: D02521012@ntu.edu.tw (T. Wei).
1
These authors contributed to the work equally and should be regarded as co-first
authors.
increases design efficiency, and reduces risk throughout the engi-
neering and construction process. Many researchers have explored
the possibility of using BIM to simulate energy consumption [2–6].
Persson et al. [7], in one such study, analyzed a low-energy build-
ing in Sweden using the dynamic energy response of buildings LTH
(Derob-LTH) technique to find that windows have a considerable
impact on the energy consumption and greenhouse gas emis-
sions of residential buildings. Window size, specifically, was shown
to affect the winter thermal loads and summer cooling loads of
buildings, and Derob-LTH analysis successfully provided the most
efficient size of south-facing windows for residential buildings.
There have been numerous studies published previously that
relate to eQUEST application [8–10], specifically in terms of energy
consumption simulation. Zhu [11], for example, used eQUEST sim-
ulation models to analyze the effect of various energy-saving
measures on building energy consumption conditions. With office
buildings in Hangzhou area as example, Hu et al. [12] compared the
actual monthly power consumption of office buildings in Hangzhou
against their consumption as-simulated by eQUEST, and found that
eQUEST is indeed able to quickly and reliably predict the energy
consumption of buildings. In a study by Kensek [13], compared
actual monthly electricity usage results against results provided
by eQUEST, DesignBuilder, and Vasari simulation, then simulated
four alternative energy conservation schemes for three different
climate conditions in California by targeting the building envelop;

http://dx.doi.org/10.1016/j.enbuild.2016.06.015
0378-7788/© 2016 Elsevier B.V. All rights reserved.
434 S.-J. Guo, T. Wei / Energy and Buildings 127 (2016) 433–441
the primary goal of the study was to determine if the simulated
retrofitting options produced similar trends in all three programs.
There have also been valuable studies on field energy con-
sumption of building envelops. Nikoofard et al. [14], for example,
evaluated the economic feasibility of window modifications using
energy simulations built based on a Canadian hybrid residential
end-use energy model, plus greenhouse gas emissions informa-
tion. Their results showed that thermally improved windows can
substantially reduce energy consumption and greenhouse gas
emissions from the Canadian residential sector. In another study,
Nikoofard [15] quantified the magnitude of the effect of site shad-
ing on the energy requirements of residential buildings in Canada,
and found that external shading should indeed be given due con-
sideration in energy simulations. Kim et al. [16] studied the impact
of different exterior shading devices on air conditioning loads in
residential buildings, demonstrating that exterior shading devices
can effectively reduce air conditioning energy consumption. Liu
et al. [17] found that retrofitted external building walls consistently
make the largest contribution to ECER in high energy-consuming
buildings compared to other parameters, followed by retrofitted
external windows.
A cost-benefit analysis framework was proposed by Goodacre
et al. [18] to assess the potential scale of the benefits of upgrading
heating and hot water energy efficiency in English stock buildings.
Ge et al. [19] studied the effects of clay-brick-powder on concrete
mechanical properties, and applied their orthogonal experimen-
tal design to study the significance sequence of all influencing
factors; the optimal proportions were determined based on exper-
imental and orthogonal analysis. Kaklauskas et al. [20] developed a
multivariant design and multiple criteria method for building refur-
bishment analysis that can be applied to determine the significance,
priorities, and utility degree of building refurbishment alternatives
to select the optimal variant. There have been several previous
researchers who have studied the economic aspect of this technol-
ogy. For example, Mehmet et al. [21] used an annualized Life-Cycle
cost method for economic analysis. Which conducted performance
experiments and economic analysis on a horizontal ground-source
heat pump (GSHP) system. The GSHP system has been shown to
offer unique economic advantages compared to the five most com-
mon conventional heating methods. Another study by Mehmet
[22] featured a techno-economic comparison between the ground-
coupled heat pump (GCHP) system and air-coupled heat pump
(ACHP) system. The results indicated that system parameters have
a substantial effect on performance, and that GCHP systems are
preferable to ACHP systems for the purposes of cooling spaces eco-
nomically.
2. Methodology
The BIM technique is relatively new and has been developed
extensively in recent years. It is now commonly applied in the
engineering field. BIM is best described as a data model built
by integrating three-dimensional digital techniques with vari-
ous manner of relevant information about a given engineering
project. BIM is mainly used in the design stage, construction stage,
and late operation and management stages. It can considerably
improve efficiency and reduce risk throughout the entire construc-
tional engineering process. In this study, eQUEST was applied to
conduct an energy consumption simulation. The eQUEST energy
consumption simulation model was established based on the Civil
Engineering Research Building of Taiwan University, then the outer
wall, roof, and extended length of sunshades were analyzed via
the simulation (in regards to both energy efficiency and cost-
effectiveness,) to determine the effects of different energy-saving
design schemes.
Typically, to conduct a test with this large amount of test factors
that must be considered simultaneously would be highly complex.
The number of independent factors that must be accounted for
makes it difficult to ensure a fully comprehensive test. Orthog-
onal testing is a highly efficient test method which solves the
multi-factor test problem by applying an orthogonal table that is
organized according to different parameters [23]. The orthogonal
table contains representative horizontal parameter combinations
determined after comprehensive testing to determine orthogonal-
ity, so the table can effectively be used to replace comprehensive
tests through facilitating analysis of only the most important con-
ditions. An orthogonal test design was applied in this study to find
the optimal building design scheme among all combinations with
fewer simulation iterations.
The economical index Net present value (NPV) is the differ-
ence between the present values of cash inflows and outflows. NPV
was also utilized to analyze the economic feasibility of the design
schemes, in addition to sensitivity analysis to determine the influ-
ence of cost variations on the index. It is important to note that there
may have been some differences in the cost analysis results due to
different brands and manufacturers of building materials, though
the typical Taiwanese building structure was the main source of
information.
3. Case-study at National Taiwan University
3.1. Characterization of the Civil Engineering Research Building
The Civil Engineering Research Building (Fig. 1), the newest
teaching building in the Civil Engineering Department of National
Taiwan University, was completed on June 23, 2008. It is located
in No. 188, Section Three, Xinhai Road, Taipei City. The building
includes a basement floor and nine above-ground floors (the sec-
ond floor is a shock insulation floor,) and the total floor area is about
9886 m2 . The main structural sections of the building were con-
structed using a pre-casting method. The total cost of the building
was about 8,133,641 USD. The majority of the department’s testing
equipment is still housed in the old Civil Engineering building, and
the new building contains mostly classrooms and office space for
professors and other department personnel.
The new building consumes a considerable amount of energy
overall, and though it is convenient for professors and students,
the financial burden of the building weighs rather large on the Civil
Engineering Department. Energy consumption statistics (Table 1)
of the Civil Engineering Research Building were gathered over the
past three years to inform this study.
The data gathered for this study shows low power consumption
due to the holiday season in February, during which time the build-
ing was more likely to be unoccupied. The peak season for power
consumption was June to October, during which time the temper-
ature is relatively high in Taiwan, so air conditioning consumption
was relatively high.
The Energy Use Intensity (EUI) is the ratio of the total annual
energy consumption to the total floor area (kWh/m2 a). The EUI
of the Civil Engineering Research Building is 121.86 kWh/m2 a
(Table 1) according to actual energy consumption, which is quite
close to the maximum value, 123.4 kWh/m2 a [25], of common
buildings in each college of Taiwan University. This indicates that
the building consumes energy at a significantly high level.
3.2. Taiwanese climate conditions and original design
The northern area of Taiwan features a subtropical climate,
while the southern area has a tropical climate. Taiwan as a whole
possesses a typical oceanic climate, with high-temperature and
 S.-J. Guo, T. Wei / Energy and Buildings 127 (2016) 433–441 435

Fig. 1. The Civil Engineering Research Building.

Table 1
Energy consumption of Civil Engineering Research Building (kWh).

January February March April May June

2011 87,049 72,757 84,046 84,973 98,056 105,620

2012 85,778 79,239 110,778 111,493 113,442 115,722
2013 98,850 81,515 100,792 99,523 114,599 115,796
Avg EC 90,559 77,837 98,539 98,663 108,699 115,759

July August September October November December TEC

2011 95,223 89,529 96,783 101,907 99,184 85,574 1,100,701

2012 110,941 111,746 106,101 104,544 96,712 98,641 1,245,137
2013 109,996 109,631 108,077 99,342 102,356 102,617 1,243,094
Avg EC 105,387 103,635 103,654 101,931 99,417 100,629 1,204,709

Note: Avg EC = Average energy consumption; TEC = Total energy consumption.

Energy consumption is extracted from Power Supervisory Control and Data Acquisition, NTU [24].

rainy characteristics; its annual average temperature is 23.6 ◦ C, its
summer is long, and its winter is short [26]. The Civil Engineer-
ing Research Building is located in Taipei, in the northern part of
Taiwan, which is perennially rainy, hot, and humid in the summer.
The primary consideration for the majority of Taiwan’s archi-
tectural designers is heat-proofing. The basic requirements [27]
for energy-saving design in Taiwan are: 3.5 W/m2 ◦ C outer walls,
1 W/m2 ◦ C roof, and no U value requirement for windows. The
design of the Civil Engineering Research Building as follows
(Table 2).
The building contains a central, water-cooled air conditioning
system. There are four water chiller units, with individual cooling

Fig. 2. Model established by eQUEST.

436 S.-J. Guo, T. Wei / Energy and Buildings 127 (2016) 433–441

Table 2
Design scheme of main building envelop.

Scheme Main Material Structure U Value (W/m2 K) Unit Cost

(USD/m2 )

Roof Steel plate PU: 2 mm, Light concrete: 100 mm, Steel plate: 1.5 mm, Mineral wool 0.94 71
panel: 2 mm, Air layer, Mineral wool acoustic panel: 15 mm
External wall Prefabricated wall Porcelain tiles: 8 mm, Precast reinforced concrete slab: 180 mm, Air 2.16 189
layer, Plasterboard: 25 mm
Windows Aluminum alloy frameLow-E glass Low-E 8 + Air + 8, ␩i = 0.33 – –
Sunshade Aluminum louver panel Horizontal louvers 0.65 m – 224

Note: unit cost are gathered from local companies.

capacities of 698 kW, 174 kW, and 52 kW for two of the sets. 698 kW Table 3
Energy consumption values (kWh).
of the compressors are screw type, and the other three sets are
reciprocating compressors. January February March April May June

Avg EC 90,559 77,837 98,539 98,663 108,699 115,759

4. Energy consumption simulation
4.1. eQUEST simulation of Civil Engineering Research Building
The model (Fig. 2) used here was established by applying
eQUEST to the building’s design scheme, as-is. Energy consumption
analysis was then conducted.
Architectural energy consumption simulation was conducted
by inputting Taipei climate data into the eQUEST model. The main
parameter settings of the model are as follows. As far as load con-
ditions in the building, the density of personnel is 10 m2 /person,
total heat gain of the human body is considered 132 W/person.
The density of equipment capacity is 5.4 W/m2 . Set temperature
for air conditioning operation is based on actual operation condi-
tions. The temperature scope is 18–26 ◦ C, and the operation time is
from 9:00 a.m. to 8:00 p.m. on weekends, and 8:00 a.m. to 9:00 p.m.
on weekdays. The EER of air conditioning to refrigeration is 3.4.
In order to avoid any influence of energy consumption in
extreme months on the true values, this study adopted an average
energy consumption (Avg EC) of three years as a practical value
(Table 1) to compare against eQUEST simulation results and ana-
lyzed any discrepancies between the two datasets (Table 3 and
Fig. 3).
As opposed to office buildings, which have fixed commuter
time, there exists uncertainty in the operating frequency of the
teaching building which created relatively large differences month-
by-month. Comparison shows that April is the month with the
maximum difference, 9.06%, and that the difference of total energy
consumption is 1.97% (Fig. 3). Overall, the simulated results are
relatively close to the measured values.
SC 97,300 78,300 94,500 107,600 107,000 120,700
July August September October November December TEC
Avg EC 105,387 103,635 103,654 101,931 99,417 100,629 1,204,709
SC 104,900 103,900 108,000 105,300 99,100 101,900 1,228,400
Note: SC = Simulation consumption; TEC = Total energy consumption.
4.2. Simulated improvement plan
In respect to design, there still exists significant potential to
save energy in the Civil Engineering Research Building. In addition
to adjusting usage habits, the most important parameter is opti-
mizing a specific plan for energy efficiency, containing two major
parameters [28]: switching to high-efficiency equipment, such as
an inverter air conditioner; and a re-design of the building envelop
to facilitate energy efficiency, including better walls, roofing mate-
rials, and effective use of sunshades, ventilation, and lighting, which
combined will help the building to conserve energy.
Vasari [29] is a building design and analysis tool that lets the
designer focus on the conceptual design phase. With built in fea-
tures such as energy analysis, solar radiation analysis.
The model (Fig. 4) was established by applying Vasari with the
original address on Google Earth to analyze solar radiation. The sim-
ulation utilizes the accumulated solar radiation of one year, from
Jan 1, 2014 to Dec 31, 2014.
The section that received maximum sunlight in the Civil Engi-
neering Research Building was the roof, with accumulated solar
radiation of about 893.3 kWh/m2 . The west wall, which accounts
for the largest area, received the second largest amount of accumu-
lated radiation, exceeding 446.6 kWh/m2 (Fig. 4). For overall energy

Fig. 3. Comparison between monthly measurements and those predicted via simulation.
 S.-J. Guo, T. Wei / Energy and Buildings 127 (2016) 433–441 437

Fig. 4. Solar radiation analysis in Vasari.

efficiency within the building envelop, then, sunshades would be each of the four sections of the grid protruded were optimized, as
most effective along the west wall of the building. well, at 0.7 m, 0.8 m, 0.9 m, and 1.0 m, respectively).
To this effect, adjusting the building envelop is the first step
to improving energy efficiency. The energy-saving effect of the 4.3. Orthogonal design of civil engineering research building
building’s windows is already quite successful as-designed, leav-
ing less room for improvement; this study thus focuses primarily Experiments on a given structural scheme must be con-
on improving the outer wall, roof, and sunshades. ducted 53 = 125 times each to ensure comprehensive combination
In this study, various roof and outer wall parameters typical of simulation—clearly, a difficult and time-consuming endeavor. To
buildings in Taiwan (Tables 4 and 5) were used to form alternative remedy this, orthogonal testing, which identifies the most influen-
design schemes for improving energy consumption. During eQUEST tial factors of a process by selecting appropriate test conditions and
simulation, the original design was replaced with these schemes combinations of conditions, was utilized in this study.
and various combinations of them to analyze their effects. The study selected the wall, roof, and sunshade as improvement
A horizontal sunshade was adopted in the original design. simulation factor and five value levels (Table 6) to represent differ-
To suit the energy-saving simulation, solely the sunshade design
was adjusted and other structures remained unchanged. The
horizontal sunshade was replaced by a grid sunshade, improv- Table 6
Test factors and levels in orthogonal test.
ing energy efficiency from the original 1228,400 kWh/year up
to 1221,400 kWh/year, thus saving, remarkably, up to 7000 kWh Factor A-Wall U (W/m2 K) B-Roof U (W/m2 K) C-Sunshade (m)
power over the course of one year. (The impact of a grid sunshade Level1 2.160 0.940 0.65
as opposed to a horizontal sunshade was analyzed only insofar as Level2 2.144 0.916 0.70
the shape of the shade, not the shade’s material.) The distance that Level3 1.251 0.795 0.80
Level4 0.981 0.745 0.90
Level5 0.711 0.308 1.00

Table 4
Typical design of external walls in Taiwan [27].

Main material Structure U (W/m2 K) Unit cost (USD)

Brick Porcelain tiles: 10 mm, Mortar: 15 mm, Brick: 230 mm, Mortar: 10 mm 2.144 132
Aluminum plate Aluminum plate: 6 mm, Mineral wool panel: 20 mm, Air layer, Calcium silicate: 25 mm 1.251 205
Glass curtain Toughened glass: 8 mm, Air layer, Fiber surfboard: 4 mm, Mineral wool insulation board: 0.981 268
32 mm, Fiber surfboard: 4 mm
Steel curtain Enamel: 6 mm, Steel plate: 3 mm, Mineral wool panel: 20 mm, Air layer, Fiber surfboard: 0.711 363
4 mm, Mineral wool insulation: 32 mm, Fiber surfboard: 4 mm

Table 5
Typical design of roofs in Taiwan [27].

Main material Structure U (W/m2 K) Unit cost (USD)

Steel plate PU: 2 mm, Light concrete: 100 mm, Steel plate: 1.5 mm, Mineral wool panel: 25 mm, PS: 10 mm 0.916 73
Skewback Skewback: 50 mm, Light concrete: 70 mm, Asphaltic felt: 10 mm, Mortar: 20 mm, Reinforced 0.795 102
Light concrete concrete: 150 mm, Mortar: 15 mm
Reinforced concrete Concrete: 50 mm, Insulation board: 25 mm, PU: 5 mm, Mortar: 15 mm, Reinforced Concrete: 0.745 69
150 mm, Mortar: 15 mm
Steel plate SBS: 1.5 mm, PS: 75 mm, Steel plate: 2.3 mm, Mineral wool fireproof: 3 mm, Air layer, Mineral 0.308 92
fiber plate: 12 mm
438 S.-J. Guo, T. Wei / Energy and Buildings 127 (2016) 433–441

Table 7
Orthogonal design and energy consumption.

Factor 1 2 3 4 5 6 Energy Consumption (kWh)

A-Wall B-Roof C-Sunshade

Test1 1 1 1 1 1 1 1,221,400
Test 2 1 2 2 2 2 2 1,219,400
Test 3 1 3 3 3 3 3 1,214,800
Test 4 1 4 4 4 4 4 1,211,000
Test 5 1 5 5 5 5 5 1,200,200
Test 6 2 1 2 3 4 5 1,219,600
Test 7 2 2 3 4 5 1 1,216,900
Test 8 2 3 4 5 1 2 1,208,700
Test 9 2 4 5 1 2 3 1,208,900
Test 10 2 5 1 2 3 4 1,208,400
Test 11 3 1 3 5 2 4 1,207,800
Test 12 3 2 4 1 3 5 1,205,900
Test 13 3 3 5 2 4 1 1,202,100
Test 14 3 4 1 3 5 2 1,206,800
Test 15 3 5 2 4 1 3 1,196,900
Test 16 4 1 4 2 5 3 1,206,900
Test 17 4 2 5 3 1 4 1,205,200
Test 18 4 3 1 4 2 5 1,207,600
Test 19 4 4 2 5 3 1 1,205,500
Test 20 4 5 3 1 4 3 1,195,400
Test 21 5 1 5 4 3 2 1,209,100
Test 22 5 2 1 5 4 3 1,212,300
Test 23 5 3 2 1 5 4 1,208,900
Test 24 5 4 3 2 1 5 1,206,600
Test 25 5 5 4 3 2 1 1,197,700
K1 1,213,360 1,212,960 1,211,300 1,208,100 1,207,760 1,208,720
K2 1,212,500 1,211,940 1,210,060 1,208,680 1,208,280 1,211,000
K3 1,203,900 1,208,420 1,208,300 1,208,820 1,208,740 1,205,866
K4 1,204,120 1,207,760 1,206,040 1,208,300 1,208,080 1,208,260
K5 1,206,920 1,199,720 1,205,100 1,206,900 1,207,940 1,207,980
Range 9460 13,240 6200 1920 980 5133

Note: “1” signifies the experimental wall factor, “2” the experimental roof factor, “3” the experimental sunshade factor, and “4”, “5”, and “6” are blank columns.

Table 8 tions and successfully determine A3 + B5 + C5 as that which shows

Energy consumption comparison.
minimum energy consumption.
Factor A Wall U B Roof U C Sunshade Energy Consumption Overall, orthogonal testing was proven reliable, and the combi-
(W/m2 K) (W/m2 K) (m) (kWh) nation A3 + B5 + C5 (Test 27) is definitively the most energy efficient
Test26 0.711 0.308 1 1,196,800 scheme.
Test27 1.251 0.308 1 1,192,300

ent improvement schemes, including wall and roof U values and
extended sunshade width.
Orthogonal design does not require that reciprocal effect be con-
sidered. Orthogonal table structure L25 (56 ) was selected according
to the amount of factors and levels, and the following intuitive
analysis table was created (Table 7):
The various levels of energy consumption simulated by differ-
ent combinations of design scheme parameters are shown in the
orthogonal table, where numbers 1–5 represent different wall, roof,
and sunshade parameters (Table 6), and K is the average value of the
horizontal simulation result corresponding to different schemes;
the smaller the K value, the lower the average energy consumption
(i.e., the better the scheme.) The “range”, accordingly, is K1–K5; the
larger the range, the larger the influence of the factor.
The roof displays the largest impact on energy efficiency as it
shows the most obvious effects on energy consumption, demon-
strated by the simulation results. As indicated by the average value
of K1–K5 (Table 7), the combination that showed the most favor-
able energy-saving effects was A3 + B5 + C5 (where wall U value was
1.251, roof U value was 0.308, and extended sunshade width was
1.0 m.)
4.4. Verification of simulation results
Because A3 + B5 + C5 is not included in the 25 combinations,
further verification (Table 8) was required to compare all combina-
5. Economic analysis
5.1. Economic feasibility analysis
Though the most energy efficient combination of factors was
determined above, construction and power consumption costs still
must be accounted for to form a comprehensive optimization of the
building’s design scheme. Once cost is introduced to the previous
27 combinations, the results, of course, are slightly different and
must be re-analyzed.
This study calculated power consumption cost according to
Taiwan University policy, which is 0.094 USD/kWh in practice
(obtained from the Construction and Maintenance Division, Office
of General Affairs, NTU). Construction cost of the redesigned build-
ing envelop was calculated based on the unit cost of the sunshade,
multiplied by area. Net present value (NPV) was used to build an
economic evaluation index. The results (Table 9) shown below were
obtained by referring to the Taiwan private investment discount
rate of 7.9% over the service life of the building (30 years.)
The economic schemes analyzed above facilitated comparison
against NPV as shown in Table 9. Schemes with positive NPV results
included Test 1, Test 2, Test 4, Test 6, Test 7, Test 8, Test 9, and Test
10. Test 6, at 177,394, showed the maximum NPV value among
all schemes followed by Test 10 (173,941). The original optimal
energy-saving scheme (Test 27) was proven unacceptable after
considering its cost.
 S.-J. Guo, T. Wei / Energy and Buildings 127 (2016) 433–441 439

Fig. 5. Economic benefits of various design schemes.

Table 9 Table 10
Economic analysis (USD). Ranking by effectiveness.

EneC EneCS ElpC ElpCS EcoB Scheme NO1 NO2 NO3 NO4 NO5 NO6

Test1 114,812 658 668,166 −395 7874 Priority Test6 Test10 Test7 Test9 Test8 Test1
Test2 114,624 846 673,637 5075 4540 Acceptable Test2 Test4
Test3 114,191 1279 707,902 39,341 −24,812 Unacceptable The remainder
Test4 113,834 1636 685,074 16,512 2076
Test5 112,819 2651 713,421 44,860 −14,734
Test6 114,642 827 500,569 −167,993 177,394
Test7 114,389 1081 509,769 −158,793 171,078 It can also be divided into three distinct categories (Table 10)
Test8 113,618 1852 544,035 −124,527 145,573 when subdividing these feasible schemes: favorable cost reduc-
Test9 113,637 1833 521,206 −147,355 168,187 tion/power consumption reduction schemes, considered the main
Test10 113,590 1880 515,987 −152,575 173,941
Test11 113,533 1937 726,946 58,384 −36,378
priority; cost-effectiveness schemes found acceptable over the
Test12 113,355 2115 953,316 284,754 −260,718 building’s 30-year service life; and unacceptable cost-effectiveness
Test13 112,997 2473 770,412 101,850 −73,754 schemes, in which the original investment is not recovered over 30
Test14 113,439 2031 714,016 45,455 −22,380 years, which were discarded. Basically, the first and second cate-
Test15 112,509 2962 738,634 70,072 −36,421
gories contain the only schemes considered successful or effective.
Test16 113,449 2021 924,769 256,207 −233,239
Test17 113,289 2181 933,969 265,407 −240,623 Taken together, the results showed that after considering cost,
Test18 113,514 1956 934,667 266,106 −243,885 Test 6 scheme (wall U value = 2.144, roof U value = 0.940, and
Test19 113,317 2153 908,110 239,548 −215,084 extended sunshade width = 0.70) is the optimal design scheme.
Test20 112,368 3103 936,457 267,895 −232,641
Test21 113,655 1815 1,217,773 549,212 −528,594
Test22 113,956 1514 1,193,406 524,845 −507,645 5.2. Sensitivity analysis
Test23 113,637 1833 1,223,942 555,381 −534,549
Test24 113,420 2050 1,201,114 532,552 −509,264
Test25 112,584 2886 1,229,461 560,900 −528,103
Construction costs were obtained through both government-
Test26 112,499 2971 1,236,921 568,359 −534,601 sponsored public information and private market inquiry, so there
Test27 112,076 3394 761,012 92,451 −53,885 may have been inconsistencies or omissions between the two sepa-
Note: EneC = Energy Cost = Energy Consumption × 0.094 USD/kWh. rate sources. In addition, material specifications that differ by brand
EneCS = Energy Cost Savings = Original Energy have an influence on actual price, affecting the practical conditions
cost − EneC = 1228,400kWh/year × 0.094 USD/kW − EneC. described in this study. Each improvement scheme is accompanied
ElpC = Envelop Cost = (Unit cost × Total Unit).
by respective price changes. Test 1 was chosen as an example of
ElpCS = Envelop Cost Savings = ElpC − Original Envelop Cost.
EcoB = Economic Benefits = NPV = A(P/A, i, n) − ElpCS = EnvCS(P/EnvCS,0.079,30) − ElpCS. critical conditions among priority improvement schemes, and con-
ducted analyses on single-factor sensitivity and the influence of
cost changes on economic benefit.
A bar graph (Fig. 5) was drawn to visualize the NPV values listed Set the percentages change of walls, roof, and sunshades as
in the table. The graph demonstrates that an acceptable scheme has an “x” reflects the influence of cost change on economic benefit
NPV greater than zero. (Table 11 and Fig. 6).
440 S.-J. Guo, T. Wei / Energy and Buildings 127 (2016) 433–441

Table 11
Sensitivity analysis of Test 1.

−0.15 −0.1 −0.05 0 0.05 0.1 0.15

wall 161 170 180 189 199 208 218

roof 60 64 67 71 74 78 81
sunshade 190 201 213 224 235 246 257
ElpCFwall $582,503 $611,057 $639,612 $668,166 $696,721 $725,275 $753,830
ElpCFroof $658,506 $661,726 $664,946 $668,166 $671,387 $674,607 $677,827
ElpCFsunshade $663,266 $664,899 $666,533 $668,166 $669,800 $671,433 $673,067
CVwall −$86,059 −$57,504 −$28,950 −$395 $28,159 $56,714 $85,268
CVroof −$10,056 −$6836 −$3616 −$395 $2825 $6045 $9265
CVsunshade −$5296 −$3663 −$2029 −$395 $1238 $2872 $4505
NPVwall $95,462 $66,907 $38,353 $9798 −$18,756 −$47,311 −$75,865
NPVroof $19,459 $16,239 $13,019 $9798 $6578 $3358 $138
NPVsunshade $14,699 $13,066 $11,432 $9798 $8165 $6531 $4898

ElpCFwall = UC × (1 + x) × TUwall + UC × TUroof + UC × TUsunshade − OEC.

ElpCFroof = UC × (1 + x) × TUroof + UC × TUrwall + UC × TUsunshade − OEC.
ElpCFroof = UC × (1 + x) × TUsunshade + UC × TUroof + UC × TUwall − OEC.
CV = Cost Variances = ElpCF − OEC.
NPV = A(P/A, i, n) − ElpCS = EnvCS(P/EnvCS,0.079,30) − ElpCF.
UC: Unit Cost; TU: Total Unit; OEC: Original Envelop Cost; ElpCF: Envelop Cost Fluctuation.

Fig. 6. Sensitivity analysis of Test 1.

After sensitivity analysis of Test 1, it became clear that the outer
wall was most affected by cost changes. When the cost of outer
wall exceeded 1.72% (obtained from Table 11) the scheme became
unacceptable—at this point, the roof and sunshade would need to
increase to 15.21% and 19.99%, respectively.
6. Discussion
After all energy-saving scheme combinations were re-analyzed
to include cost, the outer wall failed to reach the minimum level
of efficiency (Test 26.) The main cause of this is that a large area
of the outer Civil Engineering Research Building receives western
exposure, and thus a large amount of solar radiation. Reducing the
U value of the outer wall alone is inefficient, however, because
the roof receives the majority of the solar radiation and thus has
the greatest impact on energy efficiency. The U value range of
the optimized alternative roof structures is relatively large—in this
study, energy-saving effects were very pronounced as roof U value
decreased from 0.940 to 0.308. Basically, the outer wall structure,
though a relevant factor during re-design, does not impact energy
efficiency nearly as much as the roof.
In this study, the orthogonal design method was employed to
select representative points for comprehensive testing to identify
the scheme which has minimum energy consumption (A3 + B5 + C5)
while keeping the number of simulation iterations low and the test
precision high. Through orthogonal design, we were able to select
the most typical and representative points from a comprehensive
set of results. The best scheme as-identified by orthogonal testing
should be subjected to additional validation, however, to ensure
that the identification is accurate.
With existing architecture as the research object, we used
eQUEST to build a model based on existing architectural struc-
ture and equipment. Actual energy-consumption data was used for
comparison, simulation, and adjustment to validate the simulation
results.
To date, as discussed above, many researchers [30–33] have
used BIM to model buildings in effort to save resources during
design, planning, and construction stages. The architecture itself
 S.-J. Guo, T. Wei / Energy and Buildings 127 (2016) 433–441
does not actually exist in these stages, of course, so there may be
some disparity between the simulated energy consumption and
the actual, future energy consumption. When selecting the optimal
scheme based on BIM, the simulation is still informative because
it identifies notable advantages and disadvantages in the design.
In other words, the most reasonable scheme is made clear through
BIM despite any differences among simulated and actual energy
consumption.
In this study, we utilized only the cost of material to conduct
economic sensitivity analysis due to space constraints. In addition,
multiple factors affect which scheme is optimal (e.g., service life,
cost of energy). We only considered a building with 30-year ser-
vice life, while the service life of some architectural structures can
reach 50 years or more. Cash flow into the building changes as it
ages, thus the benefits of saving energy also change over the course
of the building’s service life. We also calculated the cost of energy
as the per-colleague actual electric charge collected by the univer-
sity, though an electric energy shortage may occur in Taiwan in
the future which would drive up the price of energy. This is impor-
tant because the greater the increase in energy price, naturally, the
greater the benefit of saving energy.
7. Conclusion
BIM technology can reduce errors at the design stage and
effectively improve the overall efficiency of design. Choosing
appropriate methods based on BIM analysis can facilitate successful
design and provide different design alternatives, ultimately iden-
tifying the most cost-effective and energy-efficient solutions. In
this study, BIM simulation results showed that the optimal energy-
saving design scheme is Test 27, but after the cost factor was
introduced, it became unacceptable. Test 6, which showed accept-
able (but not the most optimal) energy-saving effects, became the
overall best scheme due to its low cost. In short, the fully optimal
design scheme was that which showed the highest energy savings
at the lowest cost.
BIM technology was proven capable of developing a highly accu-
rate simulation platform for the energy consumption of buildings.
This study also considered the feasibility of the proposed scheme
according to cost changes and the economic benefits of energy effi-
ciency, and found that Test 6 is the optimal improvement scheme.
In this study, we utilized BIM to determine the most
energy-efficient building design possible. That said, the most
energy-efficient design may not be the best solution for building
design overall. Energy-saving benefits must be combined with eco-
nomic analysis to identify the optimal design in a comprehensive
manner. BIM is an auxiliary tool, and ultimately, cost is a key factor
in any decision-making process during building design and con-
struction.
References
[1] UNEP, Common carbon metric for Measuring Energy Use & Reporting
Greenhouse Gas Emissions from Building Operations. Available from: http://
www.unep.org/sbci/pdfs/UNEPSBCICarbonMetric.pdf (accessed 2014 15.11).
[2] R. Sullivan, J. Bull, P. Davis, S. Nozaki, Z. Cumali, G. Meixel, Description of an
earth-contact modeling capability in the DOE-2.1B Energy Analysis Program,
Chicago,USA, in: ASHRAE symposium on field measurements of heat transfer
in building envelopes, 91, 1985.
[3] J. Huang, N. Bourassa, F. Buhl, E. Erdem, R. Hitchcock, Using EnergyPlus for
California title-24 compliance calculations, in: Second National IBPSA-USA
Conference, Cambridge, Massachusetts, USA, 2006.
441
[4] M. Krarti, P. Chuangchid, P. Ihm, Foundation heat transfer module for
energy-plus program, in: Seventh International IBPSA Conference, Rio de
Janeiro, 2001.
[5] D.B. Crawley, L.K. Lawrie, C.O. Pedersen, et al., EnergyPlus: new capable, and
linked, J. Archit. Plan. Res. 21 (4) (2004) 292.
[6] D.B. Crawley, L.K. Lawrie, F.C. Winkelmann, W.F. Buhl, Y.J. Huang, C.O.
Pedersen, et al., EnergyPlus: creating a new-generation building energy
simulation program, Energy Build. 33 (2001) 319–331.
[7] M.-L. Persson, A. Roos, M. Wall, Influence of window size on the energy
balance of low energy houses, Energy Build. 38 (3) (2006) 181–188.
[8] M.-T. Ke, C.-H. Yeh, J.-T. Jian, Analysis of building energy consumption
parameters and energy savings measurement and verification by applying
eQUEST software, Energy Build. 61 (2013) 100–107.
[9] H. Kim, A. Stumpf, W. Kim, Analysis of an energy efficient building design
through data mining approach, Autom. Constr. 20 (2011) 37–43.
[10] L. Weytjens, S. Attia, G. Verbeeck, A.D. Herde, A comparative study of the
‘architect-friendliness’ of six building performance simulation tools, Sustain.
Build. Technol. Urban Dev. 2 (2011) 237–244.
[11] Y. Zhu, Applying computer-based simulation to energy auditing: a case study,
Energy Build. 38 (2006) 421–428.
[12] Y.-J. Hu, X.-M. Zhang, P.-Y. Lu, et al., Study on the analysis method of energy
consumption of building with eQUEST software, J. Zhejiang Univ. Technol. 2
(2012) 75–83.
[13] K. Kensek, Comparing Building Performance Trends in Three Energy
Simulation Programs, ASES, Baltimore, 2013.
[14] S. Nikoofard, V. Ugursal, I. Beausoleil-Morrison, Techno economic assessment
of the impact of window improvements on the heating and cooling energy
requirement and greenhouse gas emissions of the canadian housing stock, J.
Energy Eng. (2014) 1943–7897, http://dx.doi.org/10.1061/(ASCE)EY, 0000140.
[15] S. Nikoofard, V.I. Ugursal, I. Beausoleil-Morrison, Effect of external shading on
household energy requirement for heating and cooling in Canada, Energy
Build. 43 (2011) 1627–1635.
[16] G. Kim, H.S. Lim, T.S. Lim, L. Schaefer, J.T. Kim, Comparative advantage of an
exterior shading device in thermal performance for residential buildings,
Energy Build. 44 (2012) 105–111.
[17] Y. Liu, W. Guo, Effects of energy conservation and emission reduction on
energy efficiency retrofit for existing residence: a case from China, Energy
Build. 61 (2013) 61–72.
[18] C. Goodacre, S. Sharples, P. Smith, Integrating energy efficiency with the social
agenda in sustainability, Energy Build. 34 (1) (2002) 53–61.
[19] Z. Ge, Z. Gao, R. Sun, L. Zheng, Mix design of concrete with recycled
clay-brick-powder using the orthogonal design method, Constr. Build. Mater.
31 (2012) 289–293.
[20] A. Kaklauskas, E. Zavadskas, S. Raslanas, Multivariant design and
multiplecriteria analysis of building refurbishments, Energy Build. 37 (2005)
361–372.
[21] H. Esen, M. Inalli, M. Esen, Techno economic appraisal of a ground source heat
pump system for a heating season in eastern Turkey, Energy Convers.
Manage. 47 (9–10) (2006) 1281–1297.
[22] H. Esen, M. Inalli, M. Esen, A techno-economic comparison of ground-coupled
and air-coupled heat pump system for space cooling, Build. Environ. 42 (5)
(2007) 1955–1965.
[23] L. Jiang, Study on the design and analysis methods of orthogonal experiment,
Exp. Technol. Manage. 9 (2010) 52–55.
[24] Office of General Affairs, NTU, Power Supervisory Control And Data
Acquisition, Available from: http://140.112.166.97/power/index.aspx
(accessed 2014 21.11).
[25] Bureau of Energy, Ministry of Economic Affairs of Taiwan Energy-saving
Building Technology Application Manual, 2013, Available from: http://www.
ecct.org.tw/print/files/1021231.pdf (accessed 2014 12.11).
[26] Climate Statistics of Taiwan. Available from: http://www.cwb.gov.tw/V7/
climate/ (accessed 2015 20.11).
[27] Construction and Planning Agency Ministry of the Interior, Building Energy
Effiency Design Specifications for Office Buildings. Available from: http://
www.cpami.gov.tw/chinese/filesys/file/chinese/publication/law/lawdata/
1000810233-1.pdf (accessed 2014 12.11).
[28] Building Envelope Energy Efficiency. Available from: http://www.ecct.org.tw/
event/files/102092403doc/05 (accessed 2015 12.11).
[29] Introduction to Vasari. Available from: http://help.autodesk.com/view/
VASARI/B3/ENU/ (accessed 2015 12.11).
[30] D. Bryde, M. Broquetas, J.M. Volm, The project benefits of building
information modeling (BIM), Int. J. Project Manage. 31 (2013) 971–980.
[31] M. Gray, J. Gray, M. Teo, S. Chi, F. Cheung, Building Information Modeling, An
International Survey, Brisbane, Australia, 2013.
[32] F. Leite, A. Akcamete, B. Akinci, G. Atasoy, S. Kiziltas, Analysis ofmodeling
effort and impact of different levels of detail in building information models,
Autom. Constr. 20 (2011) 601–609.
[33] M. Nepal, S. Staub-French, J. Zhang, M. Lawrence, R. Pottinger, Deriving
construction features from an IFC model, in: Proceedings of the CSCE 2008
Annual Conference, Quebec, Canada, 2008.