Applied Energy 114 (2014) 663675

Contents lists available at ScienceDirect

Applied Energy
journal homepage: www.elsevier.com/locate/apenergy

Benchmarks as a tool for free allocation through comparison with similar

projects: Focused on multi-family housing complex
Taehoon Hong , Choongwan Koo, Sungug Lee
Department of Architectural Engineering, Yonsei University, Seoul 120-749, Republic of Korea

h i g h l i g h t s
 We propose the model for establishing benchmarks for free allowance allocation.
 The model can preliminarily estimate the amount of allowances in construction site.
 The prediction performance of the proposed model is superior in all classication.
 For the concrete, prediction accuracy and standard deviation are 93.45% and 6.01.
 For the steel bar (94.20%; 4.34) and for the formwork (94.28%; 4.67), respectively.

a r t i c l e

i n f o

Article history:
Received 14 May 2012
Received in revised form 12 September 2012
Accepted 13 October 2013

Keywords:
Emission trading scheme
Process-based LCA
Product-level LCA
Multi-family housings
Reinforced concrete frame
Case based reasoning

a b s t r a c t
A multilateral effort to reduce greenhouse gas (GHG) emissions has been implemented around the world.
In particular, the Emissions Trading Scheme (ETS) emerged as a market-based approach used to control
GHG emissions by providing carbon credits (or allowances). One of the most controversial issues in the
ETS is the question of how the allowances will be distributed. Therefore, this research aimed to develop a
decision support model for establishing benchmarks as a tool for free allocation in the construction
industry. It can be used in the pre-design phase to estimate the amount of allowances in a given construction site. In this study, a total of 147 types of data on the reinforced concrete frame in multi-family housing projects in South Korea were collected and used to develop the advanced Case-Based Reasoning
(CBR), which can be used to establish benchmarks as a tool for free allocation.
Results showed that the prediction performance of the advanced CBR model was superior (prediction
accuracy; standard deviation) in all classications: concrete (93.45%; 6.01), steel bar (94.20%; 4.34), and
formwork (94.28%; 4.67). In the case study, a total of 60 possible combinations were evaluated in terms of
the economic and environmental impact simultaneously with the retrieved cases. The results of this
study could be expanded into other areas including new renewable energy, rehabilitation projects, and
demolition projects.
2013 Elsevier Ltd. All rights reserved.

1. Introduction
The rapid growth of cities and subsequent industrialization has
led to the rise of various environmental issues, such as global
warming and depletion of resources. With the Kyoto Protocol in
1997, however, a multilateral effort to reduce greenhouse gas
(GHG) emissions has been implemented around the world [1,2].
Under the treaty, Annex I Parties (which consist of 37 industrialized countries and the European Community) commit themselves
to binding targets for GHG emissions. Toward this end, the protocol
denes three exibility mechanisms that can be used by Annex I
Parties [3]. The three exibility mechanisms are Emissions Trading
Scheme (ETS), Clean Development Mechanism (CDM), and Joint
Corresponding author. Tel.: +82 2 2123 5788; fax: +82 2 2248 0382.
E-mail address: hong7@yonsei.ac.kr (T. Hong).
0306-2619/$ - see front matter 2013 Elsevier Ltd. All rights reserved.
http://dx.doi.org/10.1016/j.apenergy.2013.10.035

Implementation (JI). Among these, the ETS (or cap-and-trade) is a

market-based approach used to control GHG emissions by providing carbon credits (or allowances) as economic incentives for
achieving the emissions reduction target. That is to say, nations
that emit less than their quota will be able to sell the emission
credits to nations whose emissions exceed their quota [4,5].
One of the most controversial issues in the ETS is the question of
how the allowances will be distributed. Since the ETS creates a signicant value, decisions about the allocation of allowances in essence result in arguable issues. It involves whether or not to
freely allocate the allowances, whether or not to auction the allowances, or whether or not to use a combination of free allocation and
auctioning [6]. Emerging programs have changed in the transition
from free allocation to auction over time. A combination of both
free allocation and partial auction offers exibility in order to
achieve environmental and economic objectives [7,8]. For example,

664

T. Hong et al. / Applied Energy 114 (2014) 663675

in the rst and second trading periods of the European Union Emissions Trading Scheme (EU-ETS), the majority of emission allowances have been freely given to entities covered under the
program, according to historical emissions. In the third trading period of the EU-ETS, free allocation of emission allowances will be
progressively replaced by auctioning of the allowances by 2020.
Yet, free allocation will surely continue to play a signicant
role up to 2020. The proposal is being negotiated in the European
Parliament [9].
Within a free allocation, there may be a variety of acceptable
ways to distribute allowances: (i) grandfathering, allowances
based on historical emissions; and (ii) benchmarking, allowances
based on energy input or product output [9]. When allowances are
freely given to entities, the following requirements should be met.
They should be allocated in a manner that is fair, transparent, and
ambitious. In other words, the allocation approach needs to allow
entities getting a strong incentive for the achieving GHG emissions
reduction target. In this regard, free allocation based on historical
emissions called grandfathering is potentially problematic. Under the free allocation method of grandfathering, most allowances
are assigned to the entities that have emitted most. To make the
ETS more efcient and effective, however, free allocation should
levy penalty on those who have emitted most. This can be achieved
through free allocation based on energy input or product output
called benchmarking. Theoretically, this can nd the optimal
solution for allowance allocation that is fair, transparent, and
ambitious. Yet there remain considerable challenges in designing
an allocation scheme and in determining concrete values as the actual benchmarks [9,10].
South Koreas National Assembly passed legislation, The Act on
Allocation and Trading of Greenhouse Gas Emissions Allowances,
which would set the GHG emissions reduction target starting in
2015. It applies both to entities that emit more than 125,000 tCO2eq./yr and to factories or buildings that produce more than 25,000
tCO2eq./yr. About 95% of allowances will be allocated for free to
companies, factories, or buildings for the rst period (2015
2017) and the second period (20182020) [11]. To keep pace with
the current trend, the construction industry has taken various actions to reduce GHG emissions in buildings. The South Korean government has conducted a variety of research to establish the
allocation methods that are appropriate for the characteristics of
the construction industry. In particular, due to the uniqueness of
the construction site, which is substantially different from the
characteristics of the manufacturing industry, both policymakers
and construction entities are becoming more interested in the
benchmarking approach for allowance allocation.
Therefore, this research aimed to develop a decision support
model for establishing benchmarks as a tool for free allocation in
the construction industry. It can be used in the pre-design phase
to estimate the amount of allowances for each product (e.g., concrete, steel bar, or formwork) that is produced, transported, and
constructed in a given construction site. Along with this, it can preliminarily estimate the construction cost that is required to achieve
the level of benchmark for allowance allocated to a given
construction site. Using the model developed in this study, both
policymakers and construction entities can establish in advance
the level of benchmark for allowance allocation specied to a given
construction site and negotiate it with each other. Also, construction entities can assess eco-friendly technologies under budget
constraints.
The scope of this study is limited to conduct the economic and
environmental impact assessment at the sites of construction projects, especially the collection of materials, which are assembled
into a reinforced concrete frame in multi-family housing complex
projects. Toward this end, the product-level LCA method was
adopted to conduct environmental impact assessment. The

product-level LCA method is one of the four-level methods (i.e.,

material-level, product-level, building-level, and industry-level)
to conduct an LCA, and is calculated as a collection of materials,
which are assembled into a nal product. After a quantity takeoff
of the product is completed, the amount of the emissions from
each component of the product is determined. The detailed information on the product-level LCA method can be founded in [12].
It has a limitation in analyzing all materials, and, thus, the main
materials that occupy a considerable amount of the total environmental load should be determined. As proposed by [13], the
environmental load evaluation of a standard apartment unit in
Korea shows that the total ratio of CO2 emissions by concrete, steel
bar, and formwork accounts for 70.12% of total CO2 emissions
generated during the construction phase of a reinforced concrete
frame in a multi-family housing complex. As provided by [12],

the ATHENA Impact Estimator covers around 1200 assemblies,

consisting mainly of concrete, steel, and wood products used in
foundations and structural assemblies. Accordingly, this study selected concrete, steel bar, and formwork as the main materials
for the reinforced concrete frame of a multi-family housing project.
Also, the process-based LCA method was implemented as a cradle-to-gate approach for assessing the environmental load from
the material manufacturing through the on-site construction of
the building project. The process-based LCA method is one of the
two methods to conduct an LCA, and focuses on a specic product
rather than a sector. Accordingly, the major advantage of this
method is the ability to compare two products that have the same
function. The detailed information on the process-based LCA method can be founded in [12]. Since collection of the detailed design
information for performing energy simulation is limited in the
pre-design phase, an analysis of the operational environmental
load during the operation and maintenance phases was excluded
from this study.
Meanwhile, Case-Based Reasoning (CBR), one of the data-mining methods, was adopted to establish the level of benchmark for
allowance allocation specied to a given construction site. CBR
has a powerful advantage because it cannot only present the predicted value, but also historical data as references. Based on this
feature, policymakers or construction entities can estimate the level of benchmark for a given project by comparison with similar
projects that are retrieved through the CBR algorithm. In other
words, the CBR is characterized by suggesting the prediction results with a high explanatory power based on historical data. Despite such advantages of CBR, its prediction accuracy is inferior
to that of the other methodologies, such as Multiple Regression
Analysis (MRA) and Articial Neural Network (ANN). To improve
prediction accuracy, MRA and ANN were integrated to ltering
the prediction results generated by CBR. Also, Genetic Algorithm
(GA) was used to apply the concept of optimization. The research
team names a series of processes in the advanced CBR model.
Additional information on the advanced CBR model can be found
in previous studies conducted by the research team [14,15].
In this study, a total of 147 project characteristics and quantity
data were collected on the reinforced concrete frame in multi-family housing projects in South Korea. This study was carried out in
three steps: (i) the collected data were analyzed at the level of
the main materials (i.e., concrete, steel bar, and formwork) to
establish the case base; then, by using the advanced CBR model,
the quantity of the main materials is estimated; (ii) using the estimated quantity, the construction costs and CO2 emissions in the
material manufacturing through on-site construction were estimated; and (iii) based on the estimated construction costs and
CO2 emissions, the study proposed possible combinations on
which the economic and environmental impact assessment was
performed. The detailed input data can be found in Table S1 of
the supplementary data.

T. Hong et al. / Applied Energy 114 (2014) 663675

The advanced CBR model for economic and environmental

impact assessment in the pre-design phase, which this study proposes, can be characterized as follows:
 Using this model, both policymakers and construction entities can establish in advance the level of benchmark for
allowance allocation specied to a given construction site
and negotiate it with each other.
 This model estimates the construction costs and CO2 emissions based on the available information at the pre-design
phase, so it can save time and effort in terms of efciency.
 This model improves prediction accuracy on material quantity by using the advanced CBR model with a combination
of various methodologies, including MRA, ANN, and GA,
as well as CBR in terms of effectiveness.
 The proposed model suggests not only the quantity as the
result of a simple prediction, but also shows the characteristics of the projects performed in the past, because users
can compare several design options by changing the project
characteristics or selecting more optimal options among
the retrieved cases. For example, if high-strength concrete
is used, its CO2 emissions per unit quantity can be applied
to analyze the change in the nal result.
 By using a Microsoft Excel-based VBA, the proposed model
systemizes the complicated process and equation to calculate CO2 emissions; and nally.
 It is expected that more products or assemblies can be evaluated using the proposed model.
2. Literature review
2.1. Life cycle assessment for buildings
Many studies have assessed the environmental load to whole
buildings and assemblies by using the internationally recognized
Life Cycle Assessment (LCA) methodology. First, various studies
have been conducted to assess CO2 emissions from buildings.
These assessments evaluated greenhouse gas emissions based on
classication standards, such as structural type, plan type, activity,
construction method, or structural strength. A detailed examination of the energy and greenhouse emissions associated with the
construction of alternative structural systems, such as wood, steel,
and concrete structural assemblies, was analyzed [16]. The comparison of the environmental effects of steel- and concrete-framed
buildings has been analyzed according to all life-cycle phases [17].
The environmental performance of high-strength concrete used in
super tall buildings was evaluated using the suitable LCA method
by life-cycle phases [18]. The economic and environmental assessment of green roof systems or energy-saving techniques for buildings was conducted by using life cycle cost (LCC) and life cycle CO2
(LCCO2) analysis [1922]. Several studies on the effect of energysaving and CO2 emissions reduction in the operation and maintenance phase were carried out by using LCC, LCCO2, or LCA [2328].
Second, there have been several studies that were focused on
developing the model for conducting LCA. A simple life cycle CO2
assessment system was proposed to assess GHG emissions in the
life cycle phases of a standard Korean apartment [13,29]. A hybrid
approach, combining both process-based LCA (P-LCA) and economic inputoutput LCA (EIO-LCA), has also been proposed
[30,31]. A new LCA approach for buildings, called region-based life
cycle impact assessment (R-LCIA), has likewise been proposed,
which consists of the local environmental burden (EB) and attached EB [32]. The relevance of simplied LCA of building components, which aims at providing results of similar quality as
comprehensive assessments with less effort, has been analyzed
[33]. The LCI model was formed as a global methodology that

665

combined advanced optimization techniques, LCI, and cost-benet

assessment, including boundary conditions for thermal comfort,
indoor air quality, and legal requirements for energy performance
[34]. The integrated model for assessing the cost and CO2 emission
(IMACC) was developed for sustainable structural design in readymix concrete [35].
Third, the environmental assessment tools were examined and
analyzed in terms of their characteristics and limitations in conducting building environmental assessment. The role and limitations of current environmental building assessment methods in
identifying building sustainability were analyzed from the perspective of different countries [36]. The foundations for the development of an LCA program for buildings were established with a
comparison of domestically and foreign designed programs [37].
A review of recent developments of LCA methods was carried
out, focusing on areas where there has been signicant methodological development in recent years [38]. The eld of building environmental assessment tools was claried and discussed in terms of
the differences of the tools as a group rather than as individual factors. For example, a classication system, Assessment Tool Typology, was introduced by the ATHENA Institute, which has three
levels: (i) Level 1 was dened as product comparison tools and

information sources (e.g., BEES ); (ii) Level 2, whole building design

or decision support tools (e.g., ATHENA); and (iii) Level 3, whole

building assessment frameworks or systems (e.g., LEED ) [39]. The

detailed analysis on the tools can be found in Table S2 of the
supplementary data.
As mentioned above, most of the previous studies have focused
on assessing the environmental load of entire buildings or assemblies, but not on estimating it in the early stage of a project. In
other words, the current environmental assessment tools for buildings could be used after completing the detailed design, based on
the bill of quantity. Accordingly, it is determined that they have
shortcomings that could demand a considerable amount of time
and effort in assessing the environmental load, and could not be
also used in the early phase of a project. Additionally, they could
not be used to establish benchmarks as a tool for free allocations
of GHG emissions permits in the construction industry.
2.2. Preliminary estimation in the early stages of the construction
project
There are many studies on preliminary estimation in the early
stages of the construction project. Most of the previous studies
have focused on cost estimation or energy consumption prediction.
Detailed literature reviews on cost estimation can be found in the
previous studies conducted by the research team [15,40] and those
on energy consumption prediction have been also conducted
[14,25,26]. The estimation of the material quantities in the early
stages was often carried out by using data-mining methodologies,
such as MRA, ANN, and CBR in previous studies. First, the MRA
method was used to estimate material quantities by parametric
statistical equations. Then [41] developed the early cost estimating
models for road construction projects and [42] proposed a conceptual cost-estimate model for bridge foundations based on the estimation of materials quantities.
Second, the ANN method was used to estimate the material
quantities by repetitively performing machine learning. [43] developed a system to assist in the early cost estimation of road tunnels;
[44] established the relationship between the quantities of concrete and form-work required for the structural elements of
high-rise commercial buildings; [45] proposed the logarithm-neuron network to improve its efciency and accuracy in quantity estimation of steel and RC buildings; and [46] developed the model for
identifying and controlling the variances in the quantity of any
work package of building construction projects.

666

T. Hong et al. / Applied Energy 114 (2014) 663675

Third, the CBR method was used to estimate the material

quantities by analyzing the similarity with historical cases. [47]
developed the construction cost prediction model based on the
material quantities of representative items that have high correlations with the total construction cost; [48] proposed the
web-based approximate quantities prediction model by types of
construction work in road construction projects; and [49]
developed a two-staged cost estimation model based on the material quantities for the preliminary design stage of the highway
construction project.

construction. Additionally, the product-level LCA method was

adopted to calculate the collection of materials, which are assembled into the reinforced concrete frame in multi-family housing
projects [12]. As mentioned in Section 1, concrete, steel bar, and
formwork were selected as the main materials [13,29].
Finally, a scatter diagram of the estimated construction costs
and CO2 emissions was made. Based on the scatter diagram, the
economic and environmental impact assessment of the new product was conducted.

4. Estimation of material quantities using the advanced CBR

model

3. Research framework
This study aimed to develop the model that could conduct economic and environmental impact assessment of a project during
the pre-design phase, where information might not be adequately
available. The case study focused on the reinforced concrete frame
of a multi-family housing project in South Korea. Fig. 1 shows the
detailed process of this study.
First, the construction costs and CO2 emissions should be estimated in order to conduct the economic and environmental impact
assessment on the reinforced concrete frame of a multi-family
housing project in the pre-design phase. Toward this end, material
quantities on the reinforced concrete frame should be estimated.
Then, a project bearing similar characteristics as that of the new
project should be retrieved by using the advanced CBR model, followed by the estimation of the quantity of materials for the new
project, based on the quantity of materials used in similar projects.
Next, using the estimated quantity of the reinforced concrete
frame of a multi-family housing project, the construction costs
and CO2 emissions are estimated. First, the construction cost is
calculated by multiplying the estimated quantity by its unit cost.
Second, the process-based LCA method was implemented as a
cradle-to-gate approach to assess the environmental load from
the material manufacturing phase to the on-site construction
phase of the building project [12]. CO2 emissions are calculated
accordingly into three phases based on the materials life cycle:
material manufacturing, material transportation, and on-site

4.1. Establishment of database

This study aims at estimating the quantity of the reinforced
concrete frame of a multi-family housing project. Based on the
estimated quantity, the construction costs and CO2 emissions from
the material manufacturing phase to the on-site construction
phase were calculated.
The research team conducted an extensive literature review of
the quantity estimation of the reinforced concrete frame of a multi-family housing project. Based on such preliminary understanding, they conducted an interview with the quantity surveyors
who were highly experienced in multi-family housing projects.
The study selected concrete, steel bar, and formwork as the
main materials [13,29]. It identied nine impact factors on the
quantity of the main materials, and dened them as independent
variables: (i) total oor area; (ii) exclusive use area; (iii) common
use area; (iv) building area; (v) ground oor area; (vi) standard
oor area; (vii) the number of stories; (viii) the number of households; and (ix) size of household. The quantity per unit area was
dened as a dependent variable. Additional information on these
impact factors can be found in the previous studies conducted by
the research team [15,40,47].
As shown in Table 1, the study dened appropriate scales for
these variables. For example, the total oor area is dened as m2
under the ratio scale, and the quantity per unit area is dened as

Fig. 1. Research framework.

667

T. Hong et al. / Applied Energy 114 (2014) 663675

Table 1
Impact factors affecting the quantity of reinforced concrete frame.
Variables

Attributes

Independent variable

Target variable

Unit

Total oor area

Exclusive use area
Common use area
Building area
Ground oor area
Standard oor area
No. of stories
No. of households
Size of household
Quantity per unit area

Type of scale

() m
() m2
() m2
() m2
() m2
() m2
() story
() household
() m2
() ton/m2

Ratio
Ratio
Ratio
Ratio
Ratio
Ratio
Ratio
Ratio
Ratio
Ratio


fSER x

4.2. Retrieval of similar cases

There were various methodologiesCBR, MRA, ANN, or decision
treethat could be applied to the quantity estimation of the reinforced concrete frame of a multi-family housing project during the
pre-design phase. However, the construction materials and methods can be changed, depending on the characteristics of a project.
Moreover, if new techniques or methods are applied to a project, it
will be difcult to produce an accurate result with a simple estimation. In other words, even with an identical material quantity, the
construction costs and CO2 emissions may differ. To reect such
changes in the characteristics of a project, it is necessary to propose
a model that can present not only the prediction result, but also the
characteristics of previous projects. This study selected CBR as a
methodology to satisfy such requirements. To improve prediction
accuracy, the study also used the advanced CBR model that combined various methodologies including MRA, ANN, and GA.
The CBR methodology consists of three phases: the attribute
similarity, the case similarity, and the prediction performance.
The attribute similarity can be calculated by the differences in
the independent variables between the test and the retrieved
cases, which can be shown in Eq. (1). The case similarity can be calculated by using the attribute similarity and attribute weight of all
attributes, which can also be shown in Eq. (2). Finally, the prediction performance can be calculated by the differences in the
dependent variables between the test and retrieved cases, as
shown in Eqs. (3) and (4).

fAS x

jAV Test

Case AV Retriev ed Case j

AV Test

Case

 100

if f AS x P MCAS
if f AS x < MCAS
1

where fAS is the function for calculating the attribute similarity,

AVtest-case is the attribute value of the test case, AVretrieved-case is the
attribute value of the retrieved case, and MCAS is the minimum
criterion for scoring the attribute similarity. Generally, 10% is
recommended as MCAS by a software program called Esteem. If
an attribute is dened under a ratio scale, its attribute similarity
score is calculated using Eq. (1) if the attribute similarity is more
than MCAS; otherwise, 0. On the other hand, if the attribute is
dened under a nominal scale, when the value of the attribute is
the same, the attribute similarity is given as 1; otherwise 0.

fCS x

Pn
i1 fAW i  fASi
Pn
i1 fAW i

p2
AV i PV i
AV i

 100

fPA x 100  fSER x

ton/m under the ratio scale. Then, the study collected related data
on the project characteristics and the quantity of the reinforced
concrete frame of 147 multi-family housing projects. Based on
the collected data, a database was established.

100 

Pm

i1

calculating the attribute similarity, and n is the number of attributes. By multiplying the attribute similarity by the attribute
weight, the weighted-attribute similarity was derived. Then its
accumulated sum was divided by the accumulated sum of the attribute weight.

where fSER is the function for calculating the standard error rate, AV
is the actual value of the dependent variable, PV is the predicted value of the dependent variable, m is the number of cases, and fPA is
the function for calculating the prediction accuracy.
4.3. Improvement of prediction accuracy
As previously mentioned, the CBR model developed in this
study can present prediction results along with the historical cases,
and its prediction performance can be improved with the continuous accumulation of the case base. According to previous research
[14,15,25,26], however, its prediction performance is somewhat
inferior to that of ANN or MRA. To overcome this disadvantage
and improve its prediction performance, the study used an advanced CBR model that combined various methodologies including
MRA, ANN, and GA.
The advanced CBR model consists of two phases (the ltering
engine development and the genetic algorithm application): (i)
the ltering engine uses the prediction result of ANN and MRA to
calculate the cross range of the results, which can be shown in
Eqs. (5)(8); and (ii) the generic algorithm denes chromosomes
as a group of optimization parameters that affect the objective
function under a series of processes dened above. These chromosomes are composed of genes, each of which signies each optimization parameter. Through various combinations of these genes,
the optimal solution is produced [14]. As shown in Fig. 2, the genes
are classied into four categories in this research (MCAS, RAW,
TRCRMA; and RCS). As mentioned in Section 4.2, MCAS and the
range of attribute weight (RAW) are the critical factors in calculating case similarity from the perspective of prediction performance.
The tolerance range of CRMA (TRCRMA) is applied to derive the ltering engine to improve the prediction performance. Since the
predicted result is presented based on historical cases, the number
of cases selected as similar cases is important. This study used the
range of case selection (RCS) as the optimization parameter.





SERMRA
SERMRA
6 PRMRA 6 PV MRA  1
PV MRA  1 
100
100





SERANN
SERANN
6 PRANN 6 PV ANN  1
PV ANN  1 
100
100

where PRMRA is the predicted range of the MRA model, PVMRA is the
predicted value of the MRA model, SERMRA is the standard error rate
of the MRA model, PRANN is the predicted range of the ANN model,

where fCS is the function for calculating the case similarity, fAW is the
function for calculating the attribute weight, fAS is the function for

Fig. 2. Description of chromosome in the genetic algorithm.

668

T. Hong et al. / Applied Energy 114 (2014) 663675

PVANN is the predicted value of the ANN model, and SERANN is the
standard error rate of the ANN model.

equipment fuel efciency, CCF is the carbon dioxide conversion factor, and n is the number of materials.

MaxMinPRMRA ; MinPRANN 6 CRMA

6 MinMaxPRMRA ; MaxPRANN
7


TRCRMA
6 CRMA
MinCRMA  1 
100


TRCRMA
6 MaxCRMA  1
100

where CRMA is the cross-range between the predicted value of the

MRA and ANN models, TRCRMA is the tolerance range of CRMA, and
CRMA is the ltering range in which TRCRMA was applied to CRMA.
5. Calculation of construction costs and CO2 emissions
5.1. Calculation of construction costs
Construction cost can be calculated by multiplying the estimated quantity by unit cost of each main material. This study selected concrete, steel bar, and formwork as the main materials
[13,29]. With the process in Section 4, the study estimated the
quantity of the main materials as well as the construction cost
by using the unit cost of each main material, as shown in following
equation:

Construction Cost

n
X
EMQ i  UC i

i1

where EMQ is the estimated material quantity, UC is the unit cost,

and n is the number of main materials including concrete, steel
bar, and formwork in this research.
5.2. Calculation of CO2 emissions
CO2 emissions can be calculated as material manufacturing,
material transportation, and on-site construction according to the
material life cycle [12]. (i) In the material manufacturing phase,
CO2 emissions are calculated by using the estimated quantity and
Life Cycle Inventory (LCI), as shown in Eq. (10). In the material
transportation phase, CO2 emissions are calculated by using the
estimated quantity, the load capacity and the fuel efciency of
the transportation vehicle, transportation distance, and the CO2
conversion factor, as shown in Eq. (11). In the on-site construction
phase, CO2 emissions are calculated by the estimated quantity, the
energy consumption per unit quantity of the construction equipment, and the CO2 conversion factor, as shown in Eq. (12). A Microsoft Excel-based VBA was used in this series of processes to
produce the nal result in a systematic, accurate, easy and quick
manner.

CM

n
X
EMQ i  CCF i

10

CC

n
X
EMQ i  ECUQ i  CCF i

where CC is the carbon dioxide emission in on-site construction,

EMQ is the estimated material quantity, ECUQ is the energy consumption per unit quantity, CCF is the carbon dioxide conversion
factor, and n is the number of materials.

6. Prediction performance
Table 2 shows that the prediction performance of the advanced
CBR model is superior (prediction accuracy; standard deviation) in
all classications: concrete (93.45%; 6.01); steel bar (94.20%; 4.34);
and formwork (94.28%; 4.67). In the case of the steel bar, the prediction accuracy of the advanced CBR was somewhat smaller
(94.20%) than that of the ANN model (94.60%). However, this remains an excellent result, an improvement over the prediction
accuracy of the CBR model. This result is identical to the results
veried in previous studies that used the advanced CBR model
[14,15,25,26]. The advanced CBR model is a sophisticated model
that offers both higher explanatory power, which is an advantage
of the CBR methodology, and higher prediction accuracy, which
is an advantage of MRA, ANN, and other methodologies.
The proposed advanced CBR model is based on the CBR model,
and since more cases are accumulated in the case base, prediction
performance will be more accurate.

7. Case study
A case study was conducted to verify the reliability and applicability of the proposed advanced CBR model for the economic and
environmental impact assessment in the pre-design phase.
To acquire the representability of the case study, the study selected a case that is close to the average quantity per unit area of
each main materialconcrete, steel bar, and formworkas its
test case. The test case has the following characteristics: reinforced concrete as the type of structure; 0.6407 m3/m2 as the
quantity of concrete per unit area (14.25% less than 0.7472 on
average); 0.0795 ton/m2 as the quantity of steel bar per unit area
(1.23% more than 0.0785 on average); 5.3885 m2/m2 as the quantity of formwork per unit area (8.85% less than 5.9116 on average); 4251 m2 of the total oor area (17.21% less than 5135 m2
on average); and 56 households (4.66% less than 58.74 on
average).

Table 2
Prediction performance by model.
Classication

Methodology

Prediction accuracy

Standard deviation

Concrete

MRA
ANN
CBR
Advanced CBR

89.40
91.75
91.97
93.45

6.27
5.95
9.57
6.01

Steel bar

MRA
ANN
CBR
Advanced CBR

93.03
94.60
92.69
94.20

5.23
4.37
7.02
4.34

Formwork

MRA
ANN
CBR
Advanced CBR

90.95
94.18
93.07
94.28

4.91
4.15
7.59
4.67

i1

where CM is the carbon dioxide emission in material manufacturing, EMQ is the estimated material quantity, CCF is the carbon
dioxide conversion factor, and n is the number of materials.

CT

n
X
EMQ i
i1

ELC i

TDi  2
 CCF i
EFEi

11

where CT is the carbon dioxide emission in material transportation,

EMQ is the estimated material quantity, ELC is the equipment load
capacity, TD is the one-way transportation distance, EFE is the

12

i1

T. Hong et al. / Applied Energy 114 (2014) 663675

In this research, the streamlined LCA process was conducted by

following four steps stipulated by ISO 14040: (1) goal and scope
denition; (2) inventory analysis; (3) impact assessment; and (4)
results and interpretations [12].
 Step 1. Goal and scope denition: based on the information
available in the pre-design phase, the CO2 emissions generated from the material manufacturing phase to the on-site
construction phase are calculated and analyzed along with
the estimated construction costs. An economic and environmental impact assessment is subsequently performed.
 Step 2. Inventory analysis: the CO2 emissions per unit
quantity of the main materials of the reinforced concrete
frame of a multi-family housing are calculated using an
inter-industry analysis (domestic and overseas).
 Step 3. Impact assessment: to assess global warming
potential, CO2 emissions produced in step 2 are used as a
representative index. Other greenhouse gases, such as
methane (CH4), nitrous oxide (N2O), are included in this
category, thus the term CO2 equivalent is implemented
to express an impact, and not an emission.
 Step 4. Results and interpretations: using the estimated
construction cost and CO2 emissions, an economic and
environmental impact assessment is performed.
Fig. 3 shows the diagram of the scenario for the use of LCA,
based on the above four-phase LCA process in this study. The diagram signies a system boundary that dictates the breadth and
depth of the LCA process. It consists of ve categories, including life
cycle phase, building systems, life cycle inventory, life cycle impact, and the phase during which LCA is conducted.

669

0.6956 m3/m2; (ii) in the case of the steel bar (refer to Table 4), ve
cases were retrieved. The average prediction accuracy was 96.78%,
at which the average quantity per unit area was 0.0791 ton/m2;
and (iii) in the case of the formwork (refer to Table 5), six cases
were retrieved. The average prediction accuracy was 95.43%, at
which the average quantity per unit area was 5.6314 m2/m2. Tables
35 show the detailed description of each retrieved case.
7.2. Calculation of construction costs
Construction costs can be calculated by multiplying the
estimated quantity by the unit cost (refer to Eq. (9)). Table 6 shows
the detailed explanation of the description, unit, and unit cost
of each main material [50]. For example, the quantity of concrete
is 2724.06 m3. Considering that the unit cost of concrete is
US$52.49/m3, and then the construction cost of concrete is
US$142997.85. The same process can be applied to the steel bar
and the formwork, so the construction cost for each of the two
materials is US$227308.86 and US$590983.74, respectively.
7.3. Calculation of CO2 emissions
7.3.1. Material manufacturing phase
In the material manufacturing phase, the CO2 emissions for the
estimated quantity based on the advanced CBR model can be calculated using the LCI (refer to Eq. (10)). Table 7 shows the carbon
dioxide conversion factor of the main materialsconcrete, steel
bar, and formwork. This study established the CO2 emissions per
unit quantity by main material through an inter-industry analysis
(domestic and overseas) [18] and a detailed description by main
material based on the data applied to the test case.

7.1. Estimation of material quantities using the advanced CBR model

Tables 35 show the estimations of material quantities using
the advanced CBR model: (i) in the case of concrete (refer to
Table 3), two cases were retrieved. The average prediction accuracy
was 91.43%, at which the average quantity per unit area was

7.3.2. Material transportation phase

In the material transportation phase, the CO2 emissions for the
estimated quantity based on the advanced CBR model can be
calculated using the characteristics of each transportation vehicle
(e.g., load capacity, fuel efciency, transportation distance, and

Fig. 3. Scenario for the use of LCA.

670

T. Hong et al. / Applied Energy 114 (2014) 663675

Table 3
The retrieved cases using the advanced CBR model (concrete).
Variable

Case
No.

Test case
84
Retrieved
6
case 1
Retrieved
37
case 2
Average quantity per

Total
oor
area
(m2)

Exclusive
use area
(m2)

Common
use area
(m2)

Building
area
(m2)

Ground
oor
area
(m2)

Standard
oor area
(m2)

No. of
stories

No. of
households

Size of
households

Case
similarity
score

Quantity
per unit
area (m3/
m2)

Prediction
accuracy
(%)

4251.78
4282.81

3349.75
3349.64

966.56
1095.92

340.22
391.82

173.80
173.74

299.12
298.09

15
15

56
57

59
59

0.96

0.6407
0.6916

92.05

3979.24

3110.49

897.52

340.22

37.49

299.12

15

52

59

0.96

0.6996

90.81

0.6956

91.43

unit area

Table 4
The retrieved cases using the advanced CBR model (steel bar).
Variable

Case
No.

Test case
84
Retrieved
85
case 1
Retrieved
103
case 2
Retrieved
102
case 3
Retrieved
108
case 4
Retrieved
82
case 5
Average quantity per

Building
area
(m2)

Ground
oor
area
(m2)

Standard
oor area
(m2)

No. of
stories

No. of
households

Size of
households

Case
similarity
score

Quantity
per unit
area
(ton/m2)

Prediction
accuracy
(%)

966.56
964.88

340.22
380.32

173.80
210.22

299.12
329.48

15
15

56
56

59
59

0.92

0.0795
0.0773

97.23

3470.72

1133.32

376.20

211.18

329.48

15

58

59

0.90

0.0760

95.56

4823.90

3470.72

1133.32

376.20

211.18

329.48

15

58

59

0.90

0.0779

97.98

3807.54

2692.80

879.30

376.20

211.18

265.27

15

45

59

0.87

0.0828

95.81

4005.64

3110.82

897.52

466.28

248.69

299.12

15

52

59

0.83

0.0816

97.31

0.0791

96.78

Total
oor
area
(m2)

Exclusive
use area
(m2)

Common
use area
(m2)

4251.78
4692.62

3349.75
3351.04

4823.90

unit area

Table 5
The retrieved cases using the advanced CBR model (formwork).
Variable

Case
No.

Test case
84
Retrieved
105
case 1
Retrieved
85
case 2
Retrieved
99
case 3
Retrieved
103
case 4
Retrieved
102
case 5
Retrieved
108
case 6
Average quantity per

Total
oor
area
(m2)

Exclusive
use area
(m2)

Common
use area
(m2)

Building
area
(m2)

Ground
oor
area
(m2)

Standard
oor area
(m2)

No. of
stories

No. of
households

Size of
households

Case
similarity
score

Quantity
per unit
area (m2/
m2)

Prediction
accuracy
(%)

4.251.78
4282.81

3.349.75
3.349.64

966.56
1.095.92

340.22
391.82

173.80
173.74

299.12
298.09

15
15

56
57

59
59

0.98

5.3885
5.8145

92.10

4.692.62

3.351.04

964.88

380.32

210.22

329.48

15

56

59

0.95

5.3780

99.81

4347.00

3.469.27

1.135.06

427.32

173.74

298.09

15

58

59

0.94

5.7707

92.91

4.823.90

3.470.72

1.133.32

376.20

211.18

329.48

15

58

59

0.92

5.4436

98.98

4.823.90

3.470.72

1.133.32

376.20

211.18

329.48

15

58

59

0.92

5.6421

95.29

3.807.54

2.692.80

879.30

376.20

211.18

265.27

15

45

59

0.90

5.7397

93.48

5.6314

95.43

unit area

Table 6
Unit cost by main material.
Main materials

Description

Unit

Quantity

Unit cost (US$)

Construction cost (US$)

Concrete
Steel bar
Formwork

Ready mixed concrete

Deformed steel bar
Plywood

m3
ton
m2

2724.06
337.94
22910.88

52.49
672.64
25.79

142997.85
227308.86
590983.74

Note: The exchange rate (KRW/USD) is 1116.5 won to a US dollar (as of 5 March 2012). The quantity is based on the actual value of the main materials of Case No. 84, which is
the test case of the case study.

CO2 conversion factor (diesel, 2.5841 kgCO2/)) [17], [29] and [32]
(refer to Eq. (11)). Table 8 shows the characteristics of the transportation vehicle used in this study. Based on the data applied to

the test case, this study established the characteristics of the transportation vehicles (such as the type of transportation vehicles or
transportation distances).

671

T. Hong et al. / Applied Energy 114 (2014) 663675

Table 7
CO2 emission per unit quantity by main material in the material manufacturing stage.
Main materials

Detailed description

CO2 emission per unit quantity

Concrete
Steel bar
Formwork

Ready mixed concrete

Deformed steel bar
Waterproof plywood

186.493 kgCO2/m3
3.052 kgCO2/kg
1.516 kgCO2/kg

7.3.3. On-site construction phase

In the on-site construction phase, the CO2 emissions for the estimated quantity based on the advanced CBR model can be calculated using the characteristics of each construction equipment
(e.g., energy consumption per unit quantity, CO2 conversion factor
(electricity, 0.4705 kgCO2/kW h; diesel, 2.5841 kgCO2/)) [29]
(refer to Eq. (12)). Table 9 shows the characteristics of the
construction equipment used in this study, which was limited to
the equipment commonly used in the reinforced concrete frame,
the tower crane for lifting materials, the hoist for lifting workers,
the pump car for pouring concrete, and the vibrator for consolidating concrete.
7.3.4. Total amount of CO2 emissions
Tables 1012 show the total amount of CO2 emissions from the
material manufacturing phase to the on-site construction phase by
using the estimated quantity based on the advanced CBR model.
The results of the analysis on the main materials (concrete, steel
bar, and formwork) are shown in Tables 10, 11 or 12, respectively.
7.4. Economic and environmental impact assessment
As illustrated in Section 7.1, the quantity per unit area for the
main materials (i.e., concrete, steel bar, and formwork) in the reinforced concrete frame of a multi-family housing project was estimated. Based on these results, the construction costs and CO2
emissions by main material were calculated in Sections 7.2 and
7.3. However, when more than one retrieved case is presented
due to the characteristics of the CBR methodology, the nal result
can be presented as a range instead. Table 13 shows the range of
the estimates of the construction cost and CO2 emissions by main
material. Here, the unit of the construction cost is $1K, and that of
the CO2 emissions is tCO2.
As shown in Table 13, the best estimate of the construction cost
for the concrete was 155.26 with a range of 154.37156.14. The
best estimate of the CO2 emissions for the concrete was 567.65
with a range of 564.40570.92. The same type of result was presented for the steel bar and the formwork. Finally, the estimated
results of each main material were summed together to calculate

the total construction cost and CO2 emissions. As mentioned in

Section 1, however, this study analyzed the main materials (i.e.,
concrete, steel bar, and framework) that occupy a considerable
amount of the total environmental load of a standard apartment
unit in Korea, which accounts for 70.12% of total CO2 emissions
generated from a reinforced concrete frame in a multi-family housing complex. Therefore, the results mentioned above have been
converted into a percentage. The best estimate of the total
construction cost resulted in 1424.87 with a range of 1371.11
1469.87. The best estimate of the total CO2 emissions resulted in
3314.97 with a range of 3195.543428.95.
The nal decision-maker (i.e., policymakers or construction
entities) can establish the target construction costs and CO2 emissions of a given project in the pre-design phase by using the result
presented. However, the result may be different depending on the
decision made by the nal decision-maker (refer to Fig. 4). Fig. 4
shows 60 possible combinations of the estimated construction
costs and CO2 emissions in the scatter diagram. In the case study
of this research, two similar cases were retrieved for the concrete,
ve cases for steel bar, and six cases for formwork. Thus, the case
study had 60 possible combinations (2  5  6; refer to Tables 35).
As shown in Fig. 4, the low and high estimates are located in the
bottom left-hand corner and the upper right-hand corner of the
scatter diagram, respectively, while the best estimate is in the center of the scatter diagram because it is the average value. However,
Fig. 4 shows that the actual value of the test case (Case No. 84) is
located very closely to the low estimate. This result may differ on
a case-to-case basis. Thus, depending on the criteria of the nal
decision-maker, the difference between the target value and the
actual value of the construction costs and CO2 emissions may be
changed.
Figs. 5 and 6 show the difference between the target and actual
values (test case) of the construction costs and CO2 emissions,
depending on the nal decision-maker. Figs. 5 and 6 also show
the most similar case proposed by the advanced CBR model: the
case with the highest case similarity score (Scenario 1). In terms
of the economic impact, the construction cost showed an error of
+3.94%, +0.01%, +7.22% and +5.39% for the best estimate, the low
estimate (Scenario 8; here, Scenario 8 means a combination of the
1st ranked retrieved cases for the concrete, the 2nd ranked for the
steel bar, and the 2nd ranked for the formwork, which is abbreviated as C1 + S2 + F2), high estimate (scenarios 49; C2 + S4 + F1),
and the 1st case similarity (scenario 1; C1 + S1 + F1), respectively.
In terms of the environmental impact, CO2 emissions showed an
error of +3.21%, 0.51%, +6.75%, and +3.29%, respectively (refer to
Fig. 6).
Such results can be changed depending on the attributes of various factors, such as estimated quantity, the type of material and

Table 8
The characteristics of transportation vehicle by main material in material transportation stage.
Main materials

Transport vehicle

Load capacity

Fuel efciency (km/)

Transportation distances (km)

Power source

Concrete
Steel bar
Formwork

Ready mixed concrete truck

20 ton trailer
8 ton truck

6 m3 (13.11 ton)
20 ton
8 ton

2.44
3.1
4.5

8.66
82.85
43.92

Diesel
Diesel
Diesel

Table 9
The characteristics of the construction equipment by main material in the on-site construction stage.
Main materials

Concrete
Steel bar
Formwork

Construction equipment
Tower crane (kW h/ton, electricity)

Hoist (kW h/ton, electricity)

Pump car (/ton, diesel)

Vibrator (/ton, diesel)

1.354
1.354

0.553
0.553
0.553

0.238

0.183

672

T. Hong et al. / Applied Energy 114 (2014) 663675

Table 10
Total amount of CO2 emission (concrete).
Variable

Test case
Retrieved case 1
Retrieved case 2
Average

Case
No.

Quantity per unit area (m3/

m2)

Amount of CO2 emission (kgCO2)

Material
manufacturing

Material
transportation

On-site
construction

Total amount of CO2

emission

84
6
37

0.6407
0.6916
0.6996
0.6956

508028.46
548388.46
554731.87
551560.17

8346.0076
9006.3510
9098.0654
9043.0368

6489.31
7004.84
7085.87
7045.36

522863.78
564399.65
570915.81
567648.56

Note: If the quantity per unit area (0.6916 m3/m2) of the concrete in Retrieved Case 1 (Case No. 6) is applied to the total oor area of the test case (Case No. 84) (4251.78 m2)
(refer to Table 3), the total quantity of the concrete will be 2940.53 m3. Applying this result to the CO2 emission per unit quantity (186,493 kg of CO2/m3) in the manufacturing
stage of concrete, which is shown in Table 7, produces 548388.49 kg of CO2. Tables 8 and 9 can be used to calculate the total amount of CO2 emissions in the stages of material
transportation and on-site construction. As a result, the total amount of CO2 emission of Retrieved Case 1 (Case No. 6) is 564399.65 kg of CO2. Such a process can be equally
applied to the steel bar (refer to Table 11) and formwork (refer to Table 12).

Table 11
Total amount of CO2 emission (steel bar).
Variable

Test case
Retrieved case
Retrieved case
Retrieved case
Retrieved case
Retrieved case
Average

1
2
3
4
5

Case
No.

Quantity per unit area

(ton/m2)

Amount of CO2 emission (kgCO2)

Material
manufacturing

Material
transportation

On-site
construction

Total amount of CO2 emission

84
85
103
102
108
82

0.0795
0.0773
0.0760
0.0779
0.0828
0.0816
0.0791

1031626.39
1003078.24
986208.87
1010864.10
1074448.62
1058876.90
1026435.82

2348.1133
2348.1133
2348.1133
2348.1133
2486.2376
2486.2376
2348.1133

239.81
233.18
229.25
234.99
249.77
246.15
238.61

1034214.31
1005659.53
988786.24
1013447.20
1077184.62
1061609.28
1029022.53

Table 12
Total amount of CO2 emission (formwork).
Variable

Test case
Retrieved case
Retrieved case
Retrieved case
Retrieved case
Retrieved case
Retrieved case
Average

1
2
3
4
5
6

Case
No.

Quantity per unit area

(m2/m2)

Amount of CO2 emission (kgCO2)

Material
manufacturing

Material
transportation

On-site
construction

Total amount of CO2 emission

84
105
85
99
103
102
108

5.3885
5.8145
5.3780
5.7707
5.4436
5.6421
5.7397
5.6314

691965.38
752680.99
684307.93
748351.68
697713.27
729053.32
734998.11
738331.49

2925.6147
3228.2644
2875.1730
3127.3812
2925.6147
3076.9396
3127.3812
3127.3812

345.41
375.24
342.08
372.98
348.38
363.57
367.05
368.45

695236.41
756284.50
687525.19
751852.04
700987.27
732493.82
738492.55
741827.32

equipment, unit cost, or carbon dioxide conversion factor. Therefore, the economic and environmental assessment should consider
the characteristics of a given project. The result should be used to
establish the target criteria for the construction costs and CO2
emissions of the given project.
8. Conclusions and discussion
The objective of this study is to develop a decision support
model for establishing benchmarks as a tool for free allocation in
the construction industry, which can be used in the pre-design
phase. The scope of this study is limited to conduct the economic
and environmental impact assessment at the sites of construction
projects, especially the collection of materials that are assembled
into a reinforced concrete frame in multi-family housing complex
projects.
The process-based LCA method was implemented as a cradleto-gate approach to assess the environmental load from the material manufacturing phase to the on-site construction phase of a
building project. The product-level LCA method was also adopted

to calculate the collection of materials, which are assembled into

the reinforced concrete frame in a multi-family housing project.
Material quantity was estimated using the advanced CBR model
that was proposed in this study. A total of 147 project characteristics and quantity data were collected on the reinforced concrete
frame in multi-family housing projects in South Korea. This study
was carried out in three steps: (i) the collected data were analyzed
at the level of the main materials (i.e., concrete, steel bar, and formwork) to establish the case base and the quantity of the main materials is estimated using the advanced CBR model; (ii) using the
estimated quantity, the construction costs and CO2 emissions in
the material manufacturing phase to the on-site construction
phase were estimated; and (iii) based on the estimated construction costs and CO2 emissions, the study proposed possible
combinations on which the economic and environmental impact
assessment was performed.
A case study was conducted to verify the reliability and applicability of the advanced CBR model, which was proposed for the economic and environmental impact assessment in the pre-design
phase of the construction project. The results are discussed below.

673

T. Hong et al. / Applied Energy 114 (2014) 663675

Table 13
Range of estimates of the construction cost and CO2 emission by main material.
Main materials

Evaluation index

Concrete

Construction cost
CO2 emission

Steel bar

Estimates range (70.12%)

Best

Estimates range (100%)

Low

High

Best

Low

High

155.26
567.65

154.37
564.4

156.14
570.92

221.42
809.54

220.15
804.91

222.68
814.20

Construction cost
CO2 emission

226.24
1029.02

217.22
988.79

236.83
1077.18

322.65
1467.51

309.78
1410.14

337.75
1536.20

Formwork

Construction cost
CO2 emission

617.62
727.79

589.83
687.53

637.7
756.28

880.80
1037.92

841.17
980.50

909.44
1078.55

Total

Construction cost
CO2 emission

999.12
2324.46

961.42
2240.71

1030.67
2404.38

1424.87
3314.97

1371.11
3195.54

1469.87
3428.95

Note: The unit of the construction cost is US$1K and the unit of the CO2 emissions is tCO2.

Fig. 4. Scatter diagram of estimated construction costs and CO2 emissions.

Fig. 5. Comparison of estimated construction costs in terms of economic impact.

First, the study applied the advanced CBR model to each of the
main materials of the reinforced concrete frame of a multi-family
housing project. The result showed that the prediction performance of the advanced CBR model was superior to that of the other
models (i.e., MRA, ANN, or CBR) (refer to Table 2). The result proved
that the advanced CBR model is a sophisticated model that offers
both higher explanatory power, which is the advantage of the
CBR methodology, and higher prediction accuracy, which is the
advantage of MRA, ANN, and other methodologies.
Second, using the estimated quantity by the advanced CBR model,
the study could estimate the construction costs and CO2 emissions of

each main material. Particularly, CO2 emissions were subdivided

into the material manufacturing phase, the material transportation
phase, and the on-site construction phase so as to systematically
and accurately calculate the result. Toward this end, the Microsoft
Excel-based VBA was used in this series of processes.
Third, possible combinations were constructed based on the
estimated construction costs and CO2 emissions, which were the
object of the economic and environmental impact assessment.
The result was presented both in a scatter diagram and in a bar
chart. In this way, the study offered visible and concrete references
to support nal decisions. In other words, policymakers and

674

T. Hong et al. / Applied Energy 114 (2014) 663675

Fig. 6. Comparison of estimated CO2 emissions in terms of environmental impact.

construction entities can establish in advance the level of benchmark for allowance allocation specied to a given construction site
and negotiate it with each other. Also, it is expected that construction entities can assess the eco-friendly technologies under budget
constraints by changing the project characteristics or selecting
more optimal choices among the retrieved cases.
With the reinforced concrete frame of a multi-family housing
project, this study limited its scope to performing the economic
and environmental impact assessment during the pre-design
phase. To complement this studys limitation, the following research efforts are currently being conducted by the research
team:
 Extended research on nishing and mechanical electrical
plumbing, other than the reinforced concrete frame of a
multi-family housing project.
 Research on the establishment of assembly costs and CO2
emissions data at the level of building components (e.g.,
wooden wall, concrete roof, etc.) to perform economic
and environmental impact assessment in the schematic
design phase.
 Research on performing an economic and environmental
impact assessment according to changes in the structural
type, plan type, activity, construction method, or structural
strength.
 Research on multilateral impact categories based on the
factors affecting life cycle impact assessment, such as global warming potential, acidication potential, eutrophication potential, and ozone depletion potential.
 Research related to expansion into other areas, such as new
renewable energy sources, rehabilitation projects, and
demolition projects.

Acknowledgements
This research was supported by Basic Science Research Program
through the National Research Foundation of Korea (NRF) funded
by the Ministry of Education, Science and Technology (Nos.
2012-004376 and 2012-0001247).

Appendix A. Supplementary material

Supplementary data associated with this article can be found, in
the online version, at http://dx.doi.org/10.1016/j.apenergy.2013.
10.035.

References
[1] Sayigh A. Renewable energy the way forward. Appl Energy 1999;64:1530.
[2] rge-Vorsatz D, Harvey LDD, Mirasgedis S, Levine MD. Mitigating CO2
emissions from energy use in the worlds buildings. Build Res Inform
2007;35(4):37998.
[3] Bashmakov I, Jepma C. Climate Change 2001 (6. Policies, Measures, and
Instruments). The IPCC Third Assessment Report (TAR) Work Group III (WG3),
2001.
[4] Bartels M, Msgens F. Is a cap-and-trade system always efcient? The case of
new entrants in the emissions trading system of the EU. J Energy Eng
2008;134:339.
[5] Cl S. The effectiveness of the EU emissions trading scheme. Clim Policy
2009;9:22741.
[6] C2ES. Greenhouse Gas Emissions Allowance Allocations (Preliminary
September 2007 version). Center for Climate and Energu Solutions; 2007.
[7] Sijm JPM, Hers JS, Lise W. The implications of free allocation versus auctioning
of EU ETS allowances for the power sector in the Netherlands. Energy Research
Center of the Netherlands (ECN); 2008.
[8] Wrke M, Myers E, Burtraw D, Mandell S, Holt C. Opportunity cost for free
allocations of emissions permits: an experimental analysis. Environ Resource
Econ 2010;46:3316.
[9] DEHSt. Benchmarks as a Tool for Allocation in the Future EU ETS. German
Emissions Trading Authority (DEHSt) at the Federal Environment Agency,
Germany; 2009.
[10] Groenenberg H, Blok K. Benchmark-based emission allocation in a cap-andtrade system. Clim Policy 2020;2(1):105109.
[11] The 18th Korean Congress. The Act on Allocation and Trading of Greenhouse
Gase Emission Allowances; 2012.
[12] AIA. A Guide to Life Cycle Assessment of Buildings. New York (United States):
The American Institute of Architects; 2010.
[13] Shin S, Tae S, Woo J, Roh S. The development of environmental load evaluation
system of a standard Korean apartment house. Renew Sustain Energy Rev
2011;15(2):123949.
[14] Hong T, Koo C, Jeong K. A decision support model for reducing electric energy
consumption in elementary school facilities. Appl Energy 2012;95(7):25366.
[15] Koo C, Hong T, Hyun C. The development of a construction cost prediction
model with improved prediction capacity using the advanced CBR approach.
Expert Syst Appl 2011;38(7):8597606.
[16] Cole RJ. Energy and greenhouse gas emission associated with the construction
of alternative structural systems. Build Environ 1998;34(3):33548.
[17] Guggemos AA, Horvath A. Comparison of environmental effects of steel- and
concrete-framed buildings. J Infrastructure Syst 2005;11(2):93101.
[18] Tae S, Baek C, Shin S. Life cycle CO2 evaluation on reinforced concrete
structures with high-strength concrete. Environ Impact Assess Rev
2011;31(3):25360.
[19] Hong T, Kim J, Koo C. LCC and LCCO2 analysis of green roofs in elementary
schools with energy saving measures. Energy Build 2012;45(2):22939.
[20] Kim J, Hong T, Koo C. Economic and environmental evaluation model for
selecting the optimum design of green roof systems in elementary schools.
Environ Sci Technol 2012;46(15):847583.
[21] Saiz S, Kennedy C, Bass B, Presssnail K. Comparative life cycle assessment of
standard and green roofs. Environ Sci Technol 2006;40:43126.
[22] Hong T, Kim H, Kwak T. Energy-saving techniques for reducing CO2 Emission in
elementary schools. J Manage Eng 2012;28(1):112.
[23] Ortiz O, Castells F, Sonnemann G. Operational energy in the life cycle of
residential dwellings: the experience of Spain and Colombia. Appl Energy
2010;87(2):67380.
[24] Chantrelle FP, Lahmidi H, Keilholz W, Mankibi ME, Michel P. Development of a
multicriteria tool for optimizing the renovation of buildings. Appl Energy
2011;88(4):138694.

T. Hong et al. / Applied Energy 114 (2014) 663675

[25] Hong T, Koo C, Park S. A decision support model for improving a multi-family
housing complex based on CO2 emission from gas energy consumption. Build
Environ 2012;52(6):14251.
[26] Hong T, Koo C, Kim H. A decision support model for improving a multi-family
housing complex based on CO2 emission from electricity consumption. J
Environ Manage 2012;112(15):6778.
[27] Lee C, Hong T, Lee G, Jeong J. Life cycle cost analysis on glass type of high-rise
buildings for increasing energy efciency and reducing CO2 emissions in
Korea. J Constr Eng Manage 2012. [Available, online 24 October 2011].
[28] Jing YY, Bai H, Wang JJ, Liu L. Life cycle assessment of a solar combined cooling
and heating and power system in different operation strategies. Appl Energy
2012;92:84353.
[29] Tae S, Shin S, Woo J, Roh S. The development of apartment house life cycle CO2
simple assessment system using standard apartment house of South Korea.
Renew Sustain Energy Rev 2011;15(3):145467.
[30] Bilec MM, Ries RJ, Matthews HS, Sharrard AL. Example of a hybrid life-cycle
assessment
of
construction
processes.
J
Infrastructure
Syst
2006;12(4):20715.
[31] Bilec MM, Ries RJ, Matthews HS. Life-cycle assessment modeling of
construction
processes
for
buildings.
J
Infrastructure
Syst
2010;16(3):199205.
[32] Li Z. A new life cycle impact assessment approach for buildings. Build Environ
2006;41(10):141422.
[33] Kellenberger D, Althaus HJ. Relevance of simplications in LCA of building
components. Build Environ 2009;44(4):81825.
[34] Cerbeeck G, Hens H. Life cycle inventory of buildings: a calculation method.
Build Environ 2010;45(4):103741.
[35] Hong T, Ji C, Park H. Integrated Model for Assessing the Cost and CO2 Emission
(IMACC) for sustainable structural design in ready-mix concrete. J Environ Eng
2012;103:18.
[36] Ding GKC. Sustainable constructionthe role of environmental assessment
tools. J Environ Manage 2008;86(3):45164.
[37] Lee K, Tae S, Shin S. Development of a life cycle assessment program for
building (SUSB-LCA) in South Korea. Renew Sustain Energy Rev
2009;13(8):19942002.

675

[38] Finnveden G, Hauschild MZ, Ekvall T, Guinee J, Heijungs R, Ekvall T, et al.

Recent developments in life cycle assessment. J Environ Manage
2009;91:121.
[39] Haapio A, Viitaniemi P. A critical review of building environmental assessment
tools. Environ Impact Assess Rev 2008;28:46982.
[40] Koo C, Hong T, Hyun C, Koo K. A CBR-based hybrid model for prediction a
construction duration and cost based on project characteristics in multi-family
housing projects. Can J Civ Eng 2010;37(5):73952.
[41] Mahamid I. Early cost estimating for road construction projects using multiple
regression techniques. Australasian J Constr Econ Build 2011;11(4):87101.
[42] Fragkakis N, Lambropoulos S, Tsiambaos G. Parametric model for conceptual
cost estimation of concrete bridge foundations. J Infrastructure Syst
2011;17(2):6674.
[43] Petroutsatou K, Georgopoulos E, Lambropoulos S, Pantouvakis JP. Early cost
estimating of road tunnel construction using neural networks. J Constr Eng
Manage 2012;138(6):67987.
[44] Tam CM, Fang CF. Comparative cost analysis of using high-performance
concrete in tall building construction by articial neural networks. ACI Struct J
1999;96(6):92736.
[45] Yeh IC. Quantity estimating of building with logarithm-neuron networks. J
Constr Eng Manage 1998;124(5):37480.
[46] Al-Tabtabai H, Kartam N, Alex AP. Neural networks for the identication and
control of quantity variance in construction projects. Computing in, Civil Eng
1996:22732.
[47] Hong T, Hyun C, Moon H. CBR-based cost prediction model-II of the design
phase
for
multi-family
housing
projects.
Expert
Syst
Appl
2011;38(3):2797808.
[48] Kang T, Park W, Lee Y. Development of CBR-based road construction project
cost estimation system. In: Proceedings of the 28th international symposium
on automation and robotics in construction, ISARC 2011:13149.
[49] Kim D, Kim B, Han S. Two-staged early cost estimation for highway
construction projects. In: Proceedings of the 25th international symposium
on automation and robotics in construction, ISARC 2008:4905.
[50] Construction Association of Korea (CAK). Monthly Construction Market Price:
477 October. Seoul, South Korea: Construction Association of Korea; 2010.