MULTI-CRITERIA DECISION ANALYSIS TECHNIQUES APPLIED TO

THE CONSTRUCTION OF A COMPOSITE BOX-GIRDER BRIDGE

V. Yepes, D. Martínez-Muñoz, N. Ata-Ali, J.V. Martí
ICITECH. Dept. Ingeniería de la Construcción y Proyectos de Ingeniería Civil. Universitat
Politècnica de València (SPAIN)

Abstract
This work is based on the case study of a multiple-criteria decision analysis, aimed for the students of
the Predictive and optimization models for concrete structures subject, which is taught in the master’s
degree in Concrete Engineering at the Universitat Politècnica de València. This course features
different blocks, among which the optimization of structural and constructive elements by different
techniques stands out. Additionally, different techniques of multiple-criteria decision analysis are
studied. This case study allows the comparison between two different evaluation methods; the direct
comparison method SAW, and the pairwise comparison method AHP. The object of this study is to
carry out an analysis of the different types of structural alternatives of a bridge. In this case, four
possibilities are considered: Lower Arch, bowstring, cable-stayed and box-girder bridges; for a span
length of 100 m, without any supports due to the environmental conditions. The student will acquire
different competencies that will allow him, not only to obtain the technical knowledge, but also the
capacity to analyze, and solve problems, creativity, innovation, as well as an ethical, professional and
environmental awareness. The result of this article opens a field for the investigation of these type of
processes through other methods of multiple-criteria decision analysis.
Keywords: multi-criteria, decision making, composite structures, steel-concrete, construction, bridge,
postgraduate education, structural design.

1 INTRODUCTION

1.1 Postgraduate studies in the engineering field

Training and education in engineering is a fundamental part of the development of any country. Taking
courses and postgraduate Master’s helps the students to obtain the necessary knowledge to be able
to specialize in a specific scientific field. In Spain there are many universities that offer various masters
and courses. The Universitat Politècnica de València, offers several of them among which is the
Master's Degree in Concrete Engineering. It has the EUR-ACE International Seal of Excellence, which
certifies that these studies have a broad compliance with the professional requirements of
engineering. The purpose is to perform a work within the field of "Predictive and optimization models
for concrete structures" where our team of researchers has developed different works [1-12].

1.2 Background
For years, construction projects have had a very high magnitude and complexity that require
specialized engineers with more experience. In addition, these designs must have the quality of being
sustainable, a concept that is now booming, defined as the "Development that meets the needs of the
present without compromising the ability of future generations to meet their own needs" [13]. With this
long-term vision, the study of the life cycle has become one of the basic tools to be able to make
decisions in the design of any construction. It is precisely in this sector, where reaching an agreement
between the social, environmental and economic aspect makes possible the sustainability of the
design [14-16] to achieve this consensus, the use of decision-making methods is necessary [17-21].
In decision making we find different criteria and different alternatives to choose from [21-24]. These
criteria are of different importance and are those who incline us for one alternative or another. To be
able to safely take the most appropriate option it is necessary to have a tool that allows us to measure
the importance by means of a barometer based on different scales [25, 26].
Making an appropriate decision-making is a complicated task if there are many alternatives and if the
criteria used are contradictory [27]. For this purpose, support tools are used to facilitate the choice [20,
28]. These tools are analytical techniques that allow us to determine some weighting factors of the

Proceedings of INTED2019 Conference ISBN: 978-84-09-08619-1

1458
11th-13th March 2019, Valencia, Spain
criteria that have been considered [29, 30]. Some researchers have developed these methods
combining them with others and have developed tools that could be coupled to other organizational
and activity needs [31, 32]. In this paper, we are going to focus on two methods for decision making,
and with them we will assess which one of the four bridges analyzed is the one that has the greatest
possibility to be built in a real sustainable design.

2 MULTI-CRITERIA DECISION-MAKING METHOD

Decision making is an action that appears daily, many situations require the need to choose one
alternative among others. Commonly, that decision has an easy solution because only it only
considers one criterion. But in the professional trajectory of an engineer, the decisions that have to be
made carry an important impact, in terms of cost, or environmental and social impact among others.
Because of this, the decision-making process takes an important role to assess the different
alternatives and decide which one is the most suitable one according to various criteria.
When the problem only depends on one variable, the optimum solution is the one that minimized or
maximizes that criteria. Real problems usually have lots of aspects that could be contradictories
between them and come into conflict. Problems that take into account more than one criteria are
difficult to evaluate directly, because of this, the multi-criteria decision making methods were
developed to allow the assessment of a number of alternatives or solutions Aj (i=1, 2,..., n) based in
the score of each alternatives rij in relation to a number of criteria Ci (j=1,2,…, m). The interaction
between the solutions and the criteria could be expressed as the decision matrix Mmn:

Figure 1. Decision matrix

The punctuations vary depending on the criteria. If the criteria are quantitative (e.g. cost), the value of
the punctuation is a number, whereas if it is qualitative (e.g. aesthetic), to assign a value of
punctuation is more difficult, usually the alternatives are classified according to an assessment scale.
Subsequently, that classification is transformed in numerical values.
We have seen that criteria could be qualitative or quantitative, furthermore, the range of values of the
assessment of each criteria could be different. Because of this, the punctuations need to be
normalized and the original values rij are transformed to r’ij normalized. Concurrently, the weight wi of
the criteria Ci depending on the importance of that criteria to the final goal must be obtained.
Therefore, the decision matrix, is converted in another matrix before being evaluated, where the initial
scores rij have been in normalized with associated weights vij:

(1)

Weights are associated with the relative importance of the criteria to achieve the final score of the
solution. The weight allocation is an important factor, because slight variations of these weights could
mean a change in the selected alternative. There are a lot of methods for weight assigning, which
could be objective or subjective. In this case, two subjective weight assigning method have been used:
A direct comparison method and a pairwise comparison one.

2.1 Direct scoring methods

Decision makers assign directly the criteria weights, they must rate different criteria considering a
scale, usually distributing 100 points between the different criteria assigning them a percentage.
These are the simplest methods, based on evaluating the different alternatives through basic
arithmetic operations.

1459
The Simple Additive Weighting (SAW) method is the oldest, it assesses the different alternatives by
making the sum of the normalized value of each criterion by its corresponding weight.

2.1.1 Simple Additive Weighting (SAW)

This is the simplest and the most used multi-criteria decision-making method. It assesses the different
alternatives by making the sum for each alternative (Ai) of the normalized value of each criterion (r’ij)
multiplied by its corresponding weight (wi).

(2)

Once the sum is obtained for all the alternatives, they are compared and ranked with an ascending
order. The alternative whose value is the greatest will be the best. The allocation of the relative weight
of each criterion is directly assigned by experts. The normalization of an alternative criterion is
performed by the division of the criterion value of that alternative between the maximum value of that
criterion for all alternatives:

(3)

At first SAW was thought only to maximize criteria that are positive, therefore, it has two limitations
that should be taken into account. On the one hand, to normalize the criteria to be minimized, a
conversion of this minimization is made by a maximization, so that the normalization of an alternative
criterion is made by the division of the minimum value of that criterion for all alternatives between the
value of the criterion of that alternative:

(4)

2.2 Pairwise comparison methods

Pairwise comparison methods are very useful to obtain the weights of the different criteria and to
evaluate subjective criteria by compared the alternatives to each other. The Analytical Hierarchy
Process (AHP) method was the first one to be developed, and one of the most used methods in
decision making problems.

2.2.1 Analytical Hierarchy Process (AHP)

The Analytical Hierarchy Process (AHP) method was the first to be developed by Thomas L. Saaty,
which coupled with its simplicity has made it a widely used and popular decision-making method. To
this end, a decision-making problem must be structured in a hierarchical way where the final objective
is at the highest level, and the criteria and sub-criteria at the lower levels as shown in Fig. 2 it is very
important the right choice of criteria and sub-criteria, which must be very well defined, relevant and
mutually exclusive.
When defining the hierarchical structure, it should be taken into account that the number of criteria
should not be excessive. For example, according to Bahurmoz [25], the number of criteria and sub-
criteria in each level should not be greater than 7, in order to avoid an excessive pairing, and Miller
[33] says that the number of criteria assimilated by the people is 7±2.
After defining the hierarchical structure, the comparison of the criteria of each group of the same
hierarchical level is made, and the direct comparison by pairs of the alternatives with respect to the
criteria of the lower level. This process will be systematically repeated in ascending direction until the
final objective is assessed. These valuations are carried out by means of the fundamental scale
proposed by Saaty [26] of Table 1.

1460
 Figure 2. AHP hierarchical structure

Table 1. Saaty Scale

Intensity Definition Explanation

1 Equal importance Two activities contribute equally to the objective
3 moderate Importance Experience and judgement strongly favour one activity over another
of one over another
5 Essential or strong Experience and judgement strongly favour one activity over another
importance
7 Very strong An activity is strongly favoured and its dominance demonstrated in
importance practice
9 Extreme importance The evidence favouring one activity over another is of the highest
possible order of affirmation
2,4,6,8 Intermediate values When compromise is needed

In both, the comparison of the different alternatives to the criterion of the level inferior of the
hierarchical structure, as well as the comparison of the different criteria of the same hierarchical level
a matrix called the decision matrix is used. Each time an array is generated, the consistency of this is
evaluated. This is done to detect contradictions in the valuation of decision makers. This consistency
is obtained by the Consistency Index (CI) where λmax is the maximum eigenvalue and n is the
dimension of the decision matrix. A consistency index equal to 0 means that the consistency is
complete. Once the consistency index is obtained, the consistency ratio (CR) is obtained by being
accepted as long as it does not exceed 10%.

(5)

(6)

Once the consistency is verified, weights are obtained, which represent the relative importance of
each criterion or the priorities of the different alternatives with respect to a certain criterion. To do this,
the original AHP uses the method of the eigenvalues, where the following equation must be resolved:

(7)

1461
Where A represents the comparison matrix, w the eigenvector or vector of preference, and λmax the
eigenvalue. [34]

2.3 Problem approach, bridge alternatives and criteria

The case study is to solve a section of viaduct located in the municipal term of Morella (Castellón,
Spain). The track alignment is curved and also has both a longitudinal and transverse slope,
furthermore the environmental conditions impose that a 100 m length span has to be solved without
supports.
To solve the problem exposed, the following design alternatives have been proposed: Lower Arch,
bowstring, cable-stayed and box-girder bridges. These alternatives have been designed and
calculated before in Martínez-Muñoz master’s degree in civil engineering’s thesis [35].

2.3.1 Bridge alternatives

For the 100 m span length there are three typologies that could be competitive: Arch, cable-stayed
and girder bridge. In addition, the construction process must be taken into account, because of the
unusual conditions of that structure.
The first alternative is to build a continuous box-girder steel-concrete composite bridge, which requires
a constructive process that allows to push the beam along the bridge trace, enabling the deflection to
recover when the front part of the beam goes into contact with the stacks. For this, a staying system is
designed, it is composed by a steel staying tower that allows workers to increase the stays tension to
recover the deflection produced by the bending moment. At the final phase, the stating tower is
removed, and the bridge remain in this final position as seen in Fig. 3.

Figure 3. Box-girder bridge alternative

The second bridge alternative is to save the 100 m span length with a Bowstring bridge (Fig. 4). The
deck is made with steel due to the functioning of the structural system that compresses the arch and
tensions the deck, because of this, the tension capacity of the steel makes it the best material to
conform the deck section. The constructive process consists in building the shorter span length
sections and pushing the complete bowstring arch held by cranes arranged on the top of the end
stack.

Figure 4. Bowstring bridge alternative

The third alternative is a low-arch bridge (Fig. 5), the constructions of the substructure allow to the
deck to have an intermediate support reducing the span length to 100 m to half. The construction
process starts with the construction of the lower-arch elements that subsequently are departed to join,
forming the intermediate support, then the deck is pushed above the props.

1462
 Figure 5. Lower arch bridge alternative

The last alternative is a cable-stayed bridge composed of a reinforced concrete pylon, on which seven
pairs of stacks have been anchored. These stacks supports the steel deck that are constructed span
by span in a progressive cantilever assembly.

Figure 6. Cable stayed bridge alternative

2.3.2 Selected criteria to the assessment

The main criteria selected to the assessment of each of the alternatives have been chosen taking into
account the three sustainability Indicators: Environmental, Social and Economic. The environmental
indicator has been assessed directly, but the other two indicators have been assessed taking into
account and evaluating other two sub-criteria for each one. The social criteria have been represented
by the aesthetic and the necessity of conservation of the bridge, while the cost and the constructive
ease have been assessed to obtain the economic sustainability indicator. The hierarchical structure of
the criteria is showed in Fig. 7.

Figure 7. Selected criteria hierarchical structure

3 RESULTS

3.1 Analytical Hierarchy Process (AHP)

The four alternatives: Box-girder (BG), Bowstring (BS), Lower Arch (LA), Cable Stayed (CS) have
been compared according to the different sub-criteria: Aesthetic (A), Conservation (CO), Cost (C) and

1463
Constructive Ease (CE) to assess the main criteria: Environmental (E), Social (S) and Economic (EC)
showed in Fig. 7.
The comparison matrix (CM) values and the eigenvectors obtained are shown below: (8) Main criteria,
(9) Social Sub-criteria, (10) Economic sub-criteria, (11) Environmental sub-criteria, (12) Aesthetic, (13)
Conservation, (14) Constructive Ease, and (15) Cost.

(8)

(9)

(10)

(11)

(12)

(13)

(14)

(15)

The cost eigenvector has been obtained directly, because is an objective criterion. To obtain the
weighting of each alternative value, they have been divided by the sum of the inverse of all
alternatives. Once the values of the weights are obtained for each alternative for each criteria (11-15),
these values are multiplied by the weighting of the criteria (8-10).

1464
 (16)

The BG alternative has obtained the highest score with a 0.3530 score and it is the best alternative
according to this decision-making method (15).

3.2 Simple Additive Weighting (SAW)

To carry out the SAW method, the punctuation values for each alternative according to each criterion
and the sub-criteria weighting have been obtained with the AHP method comparison, because of its
ability to transform subjective values into objective ones for their assessment.

Table 2. SAW method summary

Alternatives
Criteria max/min Optimum Weight BG BS LA CS
E max 0.6541 0.2426 0.6541 0.0772 0.0772 0.1915
A max 0.5651 0.0220 0.0393 0.1260 0.2696 0.5651
CO max 0.5481 0.0660 0.5481 0.1301 0.2785 0.0433
C min 4,165,267.73 0.5021 4,165,267.73 4,379,966.64 4,442,678.10 4,683,410.74
CE max 0.6681 0.1674 0.1458 0.0465 0.1396 0.6681

Score 0.8487 0.5383 0.5783 0.7121

As it is shown in Tab. 2 the BG alternative has obtained a score of 0.8487 and it is the best alternative
according to this decision-making method.

3.3 Method comparison

The selected alternative for both methods is the BG bridge, and the differences obtained between the
assessment of both methods are small, but the difference between the four alternatives is accentuated
by the AHP method.

Table 3. Worsening percentage in SAW and AHP method

Alternatives SAW AHP

BG 0.00% 0.00%
BS 36.57% 53.47%
LA 31.86% 45.90%
CS 16.10% 17.36%

4 CONCLUSIONS
This paper has described how decision-making methods allow the engineers to take the best decision
according to different criteria, evaluating their importance and weighting them in accordance. That
knowledge is included in the subject called predictive and optimization models for concrete structures
that belongs to the MSc course of concrete engineering of the Polytechnic University of Valencia. In
that course many decision-making methods are introduced to the students, but in this work we have
focused on two: Simple Additive Weighting (SAW) and Analytical Hierarchy Process (AHP), a direct
comparison method and a pairwise comparison one respectively. The comparison between four
different structural systems and construction methods have been done, obtaining the box-girder bridge
the highest score with the two methods. Furthermore, the comparison between the two decision-

1465
making procedures used have been studied giving the student, not only the ability of using that
methods, but also the implications and differences between them. This study sensitizes the MSc
student with the importance of the decision-making processes in their professional toolpath and the
possibilities to assess the alternatives for obtaining the best one according to the importance of each
criteria.

ACKNOWLEDGEMENTS
The authors acknowledge the support for the Ministry of Economy and Company and FEDER funding
(Project BIA2017-85098-R).

REFERENCES
[1] V. Yepes, E. Pellicer, and A. J. Ortega, “Designing a Benchmark Indicator for Managerial
Competences in Construction at the Graduate Level,” J. Prof. Issues Eng. Educ. Pract., vol. 138,
no. 1, pp. 48–54, 2012.
[2] C. Torres-Machí, A. Carrión, V. Yepes, and E. Pellicer, “Employability of Graduate Students in
Construction Management,” J. Prof. Issues Eng. Educ. Pract., vol. 139, no. 2, pp. 163–170, 2013.
[3] T. García-Segura, V. Yepes, J. Alcalá, and E. Pérez-López, “Hybrid harmony search for
sustainable design of post-tensioned concrete box-girder pedestrian bridges,” Eng. Struct., vol.
92, pp. 112–122, 2015.
[4] E. Pellicer, L. A. Sierra, and V. Yepes, “Appraisal of infrastructure sustainability by graduate
students using an active-learning method,” J. Clean. Prod., vol. 113, pp. 884–896, 2016.
[5] T. García-Segura and V. Yepes, “Multiobjective optimization of post-tensioned concrete box-girder
road bridges considering cost, CO2 emissions, and safety,” Eng. Struct., vol. 125, pp. 325–336,
2016.
[6] V. Martí, T. García-Segura, and V. Yepes, “Structural design of precast-prestressed concrete U-
beam road bridges based on embodied energy,” J. Clean. Prod., vol. 120, pp. 231–240, 2016.
[7] T. García-Segura, V. Yepes, and D. M. Frangopol, “Multi-objective design of post-tensioned
concrete road bridges using artificial neural networks,” Struct. Multidiscip. Optim., vol. 56, no. 1,
pp. 139–150, 2017.
[8] T. García-Segura, V. Yepes, D.M. Frangopol, and D.Y. Yang, “Lifetime reliability-based
optimization of post-tensioned box-girder bridges,” Eng. Struct., vol. 145, pp. 381–391, 2017.
[9] P. Zastrow, F. Molina-Moreno, T. García-Segura, J. V. Martí, and V. Yepes, “Life cycle
assessment of cost-optimized buttress earth-retaining walls: A parametric study,” J. Clean. Prod.,
vol. 140, pp. 1037–1048, 2017.
[10] J. Salas and V. Yepes, “A discursive, many-objective approach for selecting more-evolved urban
vulnerability assessment models,” J. Clean. Prod., vol. 176, pp. 1231–1244, 2018.
[11] L. A. Sierra, V. Yepes, and E. Pellicer, “A review of multi-criteria assessment of the social
sustainability of infrastructures,” J. Clean. Prod., vol. 187, pp. 496–513, 2018.
[12] I. J. Navarro, J. V Martí, and V. Yepes, “Reliability-based maintenance optimization of corrosion
preventive designs under a life cycle perspective,” Environ. Impact Assess. Rev., vol. 74, pp. 23–
34, 2019.
[13] World Commission, “Report of the World Commission on Environment and Development : Our
Common Future.”
[14] C. Torres-Machi, V. Yepes, A. Chamorro, and E. Pellicer, “Current models and practices of
economic and environmental evaluation for sustainable network-level pavement management,”
Revista de la Construccion, vol. 13, no. 2, pp. 49–56, 2014.
[15] H. Hasheminasab, Y. Gholipour, M. Kharrazi, and D. Streimikiene, “Life cycle approach in
sustainability assessment for petroleum refinery projects with fuzzy-AHP,” Energy Environ., vol.
29, no. 7, pp. 1208–1223, 2018.

1466
[16] L. F. Cabeza, L. Rincón, V. Vilariño, G. Pérez, and A. Castell, “Life cycle assessment (LCA) and
life cycle energy analysis (LCEA) of buildings and the building sector: A review,” Renew. Sustain.
Energy Rev., vol. 29, pp. 394–416, 2014.
[17] L. A. Sierra, E. Pellicer, and V. Yepes, “Social Sustainability in the Lifecycle of Chilean Public
Infrastructure,” J. Constr. Eng. Manag., vol. 142, no. 5, p. 05015020, 2016.
[18] V. Penadés-Plà, T. García-Segura, J. V. Martí, and V. Yepes, “A review of multi-criteria decision-
making methods applied to the sustainable bridge design,” Sustainability, vol. 8, no. 12, p. 1295,
2016.
[19] D. Jato-Espino, E. Castillo-Lopez, J. Rodriguez-Hernandez, and J. C. Canteras-Jordana, “A
review of application of multi-criteria decision making methods in construction,” Autom. Constr.,
vol. 45, pp. 151–162, 2014.
[20] M. Bystrzanowska and M. Tobiszewski, “How can analysts use multicriteria decision analysis?,”
TrAC Trends Anal. Chem., vol. 105, pp. 98–105, 2018.
[21] J.-Z. Wu and G. Beliakov, “Nonadditivity index and capacity identification method in the context of
multicriteria decision making,” Inf. Sci. (Ny)., vol. 467, pp. 398–406, 2018.
[22] L. Rodríguez-Sinobas, S. Zubelzu, S. Perales-Momparler, and S. Canogar, “Techniques and
criteria for sustainable urban stormwater management. The case study of Valdebebas (Madrid,
Spain),” J. Clean. Prod., vol. 172, pp. 402–416, 2018.
[23] C.-S. Chen, Y.-H. Chiu, and L. Tsai, “Evaluating the adaptive reuse of historic buildings through
multicriteria decision-making,” Habitat Int., vol. 81, pp. 12–23, Nov. 2018.
[24] H. Wang, L. Duanmu, R. Lahdelma, and X. Li, “Developing a multicriteria decision support
framework for CHP based combined district heating systems,” Appl. Energy, vol. 205, pp. 345–
368, Nov. 2017.
[25] A. M. A. Bahurmoz, “The Analytic Hierarchy Process: A Methodology for Win-Win Management,”
2006.
[26] R. W. Saaty, “The analytic hierarchy process—what it is and how it is used,” Math. Model., vol. 9,
no. 3–5, pp. 161–176, 1987.
[27] V. Motuziene, A. Rogoža, V. Lapinskiene, and T. Vilutiene, “Construction solutions for energy
efficient single-family house based on its life cycle multi-criteria analysis: A case study,” J. Clean.
Prod., vol. 112, pp. 532–541, 2016.
[28] R. O. Parreiras et al., “A flexible multicriteria decision-making methodology to support the strategic
management of Science, Technology and Innovation research funding programs,” Eur. J. Oper.
Res., vol. 272, no. 2, pp. 725–739, 2019.
[29] E. K. Zavadskas, Z. Turskis, and S. Kildiene, “State of art surveys of overviews on MCDM/MADM
methods,” Technol. Econ. Dev. Econ., vol. 20, no. 1, pp. 165-179, 2014.
[30] I. Lipuscek, M. Bohanec, L. Oblak, and L. Z. Zadnik Stirn, “A multi-criteria decisionmaking model
for classifying wood products with respect to their impact on environment.,” Int. J. Life Cycle
Assess., vol. 15, no. 4, pp. 359-367, 2010.
[31] J. P. Basbagill, F. L. Flager, and M. Lepech, “A multi-objective feedback approach for evaluating
sequential conceptual building design decisions,” Autom. Constr., vol. 45, pp. 136–150, 2014.
[32] B. G. Hermann, C. Kroeze, and W. Jawjit, “Assessing environmental performance by combining
life cycle assessment, multi-criteria analysis and environmental performance indicators,” J. Clean.
Prod., vol. 15, no. 18, pp. 1787–1796, 2007.
[33] G. A. Miller, “The magical number seven, plus or minus two: Some limits on our capacity for
processing information,” Psychol. Rev., vol. 101, no. 2, pp. 343–352, 1994.
[34] T. García-Segura, V. Penadés-Plà, and V. Yepes, “Sustainable bridge design by metamodel-
assisted multiobjective optimization and decision-making under uncertainty,” J. Clean. Prod., vol.
202, pp. 904–915, 2018.
[35] D. Martínez-Muñoz, “Diseño estructuraI del tramo I del viaducto sobre el barranco de la Bota en la
variante de la N-232 (T.M. de Morella, Castellón). Diseño Estructural,” 2018.

1467