Journal of Cleaner Production 194 (2018) 327e341

Contents lists available at ScienceDirect

Journal of Cleaner Production

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

Comparative material-based life cycle analysis of structural

beam-floor systems
Charlotte Dossche, Veerle Boel*, Wouter De Corte
Research Group Schoonmeersen, Department of Structural Engineering, Faculty of Engineering and Architecture, Ghent University, Valentin Vaerwyckweg 1,
B-9000 Ghent, Belgium

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

Article history: In a residential building, various structural flooring systems are feasible. For moderate spans (4 me6 m)
Received 5 October 2017 and for a predetermined column grid, distinct combinations of wood, steel and concrete elements can be
Received in revised form implemented as beam-floor system. Besides comparing these beam-floor systems based on weight, cost,
1 May 2018
speed of installation etc., a comparison can also be based on their environmental impact. Indeed,
Accepted 7 May 2018
Available online 8 May 2018
whereas some materials appear to be environmentally friendly for many applications, this may not be
true for this particular case. For such a comparison, a life cycle assessment (LCA) according to ISO 14040/
44 has been implemented in this paper. In this study, the functional unit consists of a square meter of a
Keywords:
Life cycle analysis
structural beam-floor system of an arbitrary composition, i.e. three main beam materials (concrete, steel
LCA and wood) have been combined with appropriate structural floor systems. The paper presents the results
ISO 14040/44 of the LCAs for different impact categories, and both midpoint and endpoint assessment methods. This
Construction materials provides a broad idea of the environmental profile of the considered beam-floor systems representative
Beam floor system for current Belgian building practice for the defined functional unit. Furthermore, the influence of the
respective impact assessment methods (ReCiPe World versus Europe) was found to be important,
especially regarding wooden structural elements. Lastly, some sensitivity analyses (use of recycled ag-
gregates, use of aerated concrete waste, transport types and distances, ratio virgin-recycled steel) have
been performed in order to obtain a more nuanced view on the setup and results of this study. These
sensitivity analyses show that mainly the ratio virgin-recycled steel in a beam-floor system can result in
widely varying impacts on the environment and thus give some opportunities for improved environ-
mental impact.
© 2018 Elsevier Ltd. All rights reserved.

1. Introduction
The building industry is responsible for about one third of the
emitted greenhouse gasses worldwide and consumes more than
40% of the world energy (UNEP, 2009; Sierra-Pe
rez et al., 2016).
Moreover, every year, three billion tonnes of raw materials are
being mined for the construction sector (Eaton et al., 2000).
Consequently, it is important to consider ecology in the building
sector and to consciously deal with raw materials. In order to
examine these topics, a frequently used tool is a life cycle analysis or
LCA. This tool can provide insight in processes and services focusing
on the environment (Rashid and Yusoff, 2015). As such LCA has
* Corresponding author. Valentin Vaerwyckweg 1, B-9000 Ghent, Belgium.
E-mail address: veerle.boel@ugent.be (V. Boel).
been widely used already to assess the environmental impact of the
building sector.
In (Sierra-Pe
rez et al., 2016) an environmental assessment is
performed of façade-building systems and thermal insulation ma-
terials for different climatic conditions in Spain. According to the
climate conditions, specific façade systems and insulation materials
are suitable. In (Purnell, 2012; Sierra-Pe
rez et al., 2016) the
importance of embodied CO2 emission (EC) regarding life cycle
assessment is explained. The embodied CO2 emission (EC) is
emitted as a consequence of the manufacture of the component,
generated during mining, processing, and transport of raw mate-
rials, site operations, etc (fixed value). The sum of EC and the
operational CO2 (OC) is the total CO2 emitted during the lifetime of
the component. The operational CO2 emission is the CO2 produced
as a consequence of its ongoing operation, maintenance, and its
eventual disposal (variable value). In the construction sector the EC

https://doi.org/10.1016/j.jclepro.2018.05.062
0959-6526/© 2018 Elsevier Ltd. All rights reserved.
328 C. Dossche et al. / Journal of Cleaner Production 194 (2018) 327e341
can be seen as the utmost important part of the total CO2 emission,
due to the millions of tonnes of materials used in construction
projects and the dominance of EC by emissions associated with
materials. Furthermore, the analysis of Purnell (2012) demonstrates
that determining the environmental credentials of structural ma-
terials on the basis of their EC per unit mass or volume is useless.
Rather, the environmental impact should be based on the impact
per unit of structural performance, per unit component length
(functional unit). In (Vilches et al., 2017) a review is conducted on
research related to the environmental evaluation of building
refurbishment and renovation using the lifecycle assessment (LCA)
methodology. One of the most important overall conclusions was
that system boundaries (end-of-use and expected service life) and
functional units (1 m2 or the entire building) have a decisive in-
fluence on the LCA results.
Oregi et al. (2017) indicate for building assessment a relatively
minor importance of the transport (A4) and end of life (C1e4)
stages in the large majority of cases, both in economic and envi-
ronmental terms. It was found that the omission of those stages
would not generally lead to deviations of more than 1% on the total
life-cycle economic or environmental impacts.
In (Hoxha et al., 2017) reference is made to (Leung et al., 2015) in
order to highlight the need for uncertainty quantification. Leung
et al. (2015) performed a review of 134 journal articles dealing
with the uncertainties associated with buildings’ environmental
impacts. They discovered that the majority of the papers reviewed
aimed to develop methods for better modelling and prediction of
uncertainties, but not to identify construction material un-
certainties (impact coefficient, density, mass, and life time) at the
building scale. In order to increase the reliability of LCA results,
Hoxha et al. (2017) identified in their study the building materials
that have the largest relative contribution to buildings’ impacts and
uncertainties by studying single-family houses and multi-family
buildings. In order to do so they studied a functional unit of 1 m2
of net floor area (NFA) and a design life of 50 years. The largest
contributor to the environmental impact of both building types was
found to be reinforced concrete. Insulation materials and non-
structural wood on the other hand, were found to be the largest
contributors to the uncertainties of the final results for respectively
single-family houses and multi-family buildings. Besides, it was
stated in (Hoxha et al., 2017) that, when considering uncertainties
in the impacts of building materials, the uncertainty of the LCA
calculation itself could be only slightly smaller than the difference
between building projects. Further, it was shown that LCA will be
able to distinguish significantly between two projects only if their
difference is higher than approximately 20%.
In (Monteiro and Freire, 2012) a comparison of three impact
assessment methods (cumulative energy demand; CML 2001; Eco-
indicator’99) is made by means of the life-cycle assessment of a
Portuguese single-family house with seven alternative exterior wall
solutions. It was found that the operational patterns associated
with reduced HVAC levels specific for Portuguese houses signifi-
cantly influence the LCA results. Based on the comparison of the
three mentioned assessment methods, LCA studies of dwellings
should include multiple environmental impacts, but should not
address toxicity categories. Inconsistent results for the different
methods, namely a different ranking of the alternative walls, were
found regarding the toxicity categories.
Induced by the growing number of databases and methods in
LCA, Ha €fliger et al. (2017) performed a study in which they assess
the results’ sensitivity to the choice of EPD datasets between
various construction materials of four buildings. They assumed that
a construction material is sensitive if it is associated with a vari-
ability of more than 20% in between different datasets. As such it
was found that “Windows & doors”, “Insulation materials” and
“Wood and wood products” are associated with strong sensitivity
to the choice of datasets while “Cement based products” and the
other materials, such as sealings and coverings, show only minor
sensitivity. The effect of different modelling choices, i.e. “Cradle-to-
Gate”, “Cradle-to-Grave”, and “Cradle-to-Cradle” is investigated in
their study as well. For the evaluation of the system boundary
sensitivities, each material has a different material service life, but
the reference study period (RSP) for buildings is set at 60 respec-
tively 120 years. “Wood and wood products”, “Insulation materials”
and “Windows and doors” show a strong sensitivity to most
modelling choices. “Cement based products” and “Masonry” on the
contrary are mainly affected by the building RSP. If a long RSP is
chosen, insulation materials as well as doors and windows should
be focused on. If a reduced RSP is chosen, the buildings’ environ-
mental impact can be reduced by actions taken on concrete and
masonry materials. For the other materials, such as sealings and
coverings, changes in the modelling choices do not affect the
overall LCA results of the buildings in a significant way. This is
mainly due to the fact that their contribution to the total building
impact is not significant (contribution less than 10%).
In (Chastas et al., 2018) a range of embodied emissions for
reinforced concrete is mentioned between 505.7 kg/m2 and
1050 kg/m2. Based on the analysis of different building structures it
was found that reinforced concrete has a higher global warming
potential (GWP) impact in reference to all other analyzed structures
(wood, steel, concrete, masonry) and combinations with secondary
materials that have a high mass contribution to the building’s
overall weight. According to Buyle et al. (2018) concrete has the
largest environmental impact, mainly due to the production of
cement. Cement can be manufactured from clinker (Portland
cement) or from ground granulated blast-furnace slag (GGBFS), a
dependent and therefore constrained by-product of steel produc-
tion. Also according to (U.S. Environmental Protection Agency,
2007) mentioned in (Takano et al., 2014) the cement production
process is dominant in terms of energy consumption. The most
energy intensive unit process in the production is the firing of
clinker, which is a globally equal production process. Consequently,
it is concluded by Takano et al. (2014) that not much variation in the
greenhouse gas (GHG) emissions arising from concrete production
can be expected because of the unified production system and
energy profile. In (Knoeri et al., 2013) the life impacts of 12 recycled
concrete (RC) mixtures with two different cement types were
investigated and compared to corresponding conventional con-
cretes (CC) for three structural applications (i.e. lean, indoor and
outdoor structural concrete). The effect of cement content an
transport distances are analyzed. For the RC mixtures, environ-
mental benefits from co-products of the recycling operation (steel
scrap recovering) were considered as avoided impacts. The results
of this study show about 30% environmental benefits for all RC
options at endpoint level (Ecoindicator’99 and ecological scarcity)
mainly attributed to the avoided burdens associated with rein-
forcing steel recycling and avoided disposal of construction and
demolition waste. When the GWP is considered, the results are
more balanced as they primarily depend on the additional amount
of cement needed to produce the recycled concrete mixtures
compared to the conventional concrete mixtures. It was found that
limiting the additional amount of cement to 10% keeps the impact
in a comparable range. It also should be kept in mind that addi-
tional transport distances above 15 km concerning the RC mixtures
result in higher environmental impacts. In (Ghisellini et al., 2018) a
review is made on the reuse/recycling of construction and demo-
lition waste (C&DW) at the end-of-life of a building as well as on
the production of recycled products. For instance the use of recy-
cled coarse aggregates (RCA) for concrete production is studied.
Depending on the local situation (transport distances in relation to
 C. Dossche et al. / Journal of Cleaner Production 194 (2018) 327e341
modes of transport) limit distances for transport can be set in order
to obtain a better environmental profile in case RCA is used. Values
of 30 km as well as 100 km are mentioned. Also an important
parameter is the cement content used in the mix with RCA
compared to content used in the mix with NCA (normal coarse
aggregates).
Additional important specific conclusions concerning structural
elements are made in the studies mentioned in (Purnell, 2012) and
(Purnell and Black, 2012) are (1) the variations induced by the
changes in section capacity or component length are at least as
significant as the variations between the materials themselves; (2)
as load capacity (and thus size) of structural elements increases, the
EC per unit of structural performance decreases: ‘heavy’ compo-
nents are more efficient than ‘light’ components; (3) concerning
concrete the CO2 emissions associated with aggregates, recycled or
virgin, are negligible, changing the aggregate type from uncrushed
to crushed, or the cement strength class from 42.5 MPa to 52.5 MPa,
both have a relatively small effect on EC (savings of 9 ± 1% and
7 ± 1% respectively), the EC of concrete is however overwhelmingly
dominated by that of cement (>95% in most cases); (4) the opti-
mum concrete grade for most applications with respect to mini-
mizing EC is 50 MPa (RC 50/60), using the optimal strength
concrete will result in EC reductions of up to 40% for any given
concrete family; (5) use of a superplasticiser was found to reduce
overall EC by ±26%, this can be attributed to a reduced water and
binder content keeping the w/b ratio constant, the EC imparted by
the superplasticiser itself was found to be negligible, also
mentioned in (Black and Purnell, 2016); (6) use of steel for light
duty columns should be avoided; (7) glulam timber is the most
carbon-efficient solution for very light-duty columns and long,
light-duty beams.
In (Vitale et al., 2017) the role of recycling of reinforcing steel is
studied in depth and appears to play a paramount role (65% of the
total avoided impacts related to respiratory inorganics, 89% for
global warming and 73% for mineral extraction). It is interesting to
note that steel provides the largest contributions to avoided im-
pacts since it accounts according to their study for less than 3% on
mass basis of the total C&DW. It is concluded that the overall
environmental performance of the end-of-life phase of the studied
building is strongly improved by the increase of the quantity and
quality of residues sent to recycling, but the dominant contribution
in all the impact categories is attributed to the steel recycling
process. However, according to the authors it has to be emphasized
that the recovery of inert material is not negligible, and is high-
lighted in their study in terms of land occupation.
In (Buyle et al., 2018) structural steel is considered as having the
second most environmental impact, after concrete. Recycling rate
(41%) vs. recycling potential (95%) turns out to be the main reason
for the deviation in results between consequential LCA and attri-
butional LCA. Besides some small losses, steel can be recycled
completely with minimal loss of quality and therefore reduces the
demand for virgin steel. On the other hand the demand for steel is
higher than the supply of iron scrap, resulting in a recycled content
in the attributional approach that is well below the recycling po-
tential in a consequential LCA.
Lacovidou and Purnell (2016) state that the EC of new con-
struction components is important in ensuring the long-term
sustainability of the construction sector. The design parameters
(dimensions, section choice, and load capacity), energy mix, ma-
terial design and recycled content have to be taken into account for
determining the EC of a construction component.
Takano et al. (2014) point out that LCA is a data-intensive
method, and that it is not always easy for LCA practitioners to ac-
cess all primary data due to confidentiality. For this reason inter-
nationally several databases have been developed, but each of them
329
could lead to different conclusions concerning the LCA results for a
given product. In their study five LCA databases were compared by
calculating GHG emission values in the material production phase
(cradle-to-gate) of three reference buildings. The outcomes indi-
cated that for these five generic databases the results were of the
same magnitude and showed similar trends, but “quite large” nu-
merical variations were noticed. They also concluded that the
development of a national open database would be required to
provide adequate information. Even simple national open data
would be very helpful especially for practical users (e.g. architects
or constructors) because of its user friendliness and locality. A
Flemish-Belgian database with this purpose was developed by the
Public Waste Agency of Flanders (OVAM) in cooperation with
research institutes (BBRI, KUL, VITO), regional governments (Flan-
ders, Brussels, Wallonia), the federal Belgian government, and ex-
perts from the construction sector and various material producers,
which all supplied manufacturers’ and sector specific information
(Allacker et al., 2013). In order to possess sufficient generic envi-
ronmental data, the extensive Swiss LCI database Ecoinvent was
harmonised to the Flemish-Belgian context and resulted in envi-
ronmental profiles of 115 variants of building components repre-
sentative for the Flemish-Belgian building context, i.e.
“Environmental Profile of Building Elements: towards an integrated
environmental assessment of the use of materials in buildings”
(EPBE). This database is as such adapted to the Flemish-Belgian
building context to a very high degree, with adequate scientific
background. The study ‘Environmental profile of building elements’
(EPBE), published by the Public Waste Agency of Flanders, is an
attributional study according to Buyle et al. (2018) and mainly
serves to assess the current practice. If, however, the aim is to assist
decision makers at different levels at making choices affecting the
future situation, the authors suggest to complement the EPBE study
with a consequential model, as demonstrated in their publication.
The combination of two subjects, i.e. the building industry and
LCAs, is the topic of this paper. More specifically, this study zooms
in on important structural elements in buildings, i.e. structural
beam-floor systems. The aim is to derive insight in the environ-
mental profile of these subsystems, as it will serve further insight in
the environmental impact of buildings later on. In order to be more
complete in investigating this subject, several case studies with
practical beam-floor combinations, as custom in Belgian building
practice, have been used. As such, the functional unit considered
consists of a square meter of the specified structural beam-floor
system. This way more meaningful comparisons can be made be-
tween the different systems. One should keep in mind that the type
of floor will have an influence on the design of the beam(s) sup-
porting it. Finally, this study shows which aspects of the different
floor-beam combinations seem most important in an ecological
manner.
In this paper the goal and scope of the study are first presented.
Secondly, the basic design analysis of beams and floors is explained,
after which the inventory of the life cycle assessment is presented
quite profoundly. Thirdly, the impact assessment, or the results of
the LCA, are presented and discussed. Finally, sensitivity analyses
have been performed in order to obtain a more nuanced view on
the results.
2. Scope and calculations
2.1. Scope of the study
This paper will focus on the environmental impact of structural
beam-floor combinations, frequently used in Belgium, according to
the specifications of ISO 14040/44 (ISO 14040/44, 2006). As such,
the study concerns beams made of concrete, steel or wood,
330 C. Dossche et al. / Journal of Cleaner Production 194 (2018) 327e341
Table 1
Beam-floor combinations.
Thin solid Precast Pretensioned
slab þ in-situ reinforced hollow core
concrete topping hollow core slab slab (PHCS)
(SS) (RHCS)
Concrete beam (CB) x x x
Steel beam (IPE) x x x x
Wooden beam e e e x
(Glulam)
x ¼ This beam-floor combination has been incorporated in the study.
- ¼ This beam-floor combination has not been incorporated in the study.
combined with ten types of floors. Technical websites of Belgian
contractors and producers of structural building components are
consulted to select the beams and floors relevant for the Belgian
building practice (e.g. Paulibeton, Doubeton, Fingo, Vanthuyne-ide,
Xella, Eurabo, Megaton, …). Only the beam-floor combinations
which are regularly implemented in true practice have been
observed. Table 1 summarizes the types of beams and floors
considered in this study, and indicates which combinations have
been included.
Concerning the concrete elements it needs to be specified if
prefabricated or in situ concrete is used. In this study, the following
fabrication techniques are assumed based on current practice. all
reinforced concrete beams (CB) are prefabricated. The mentioned
concrete floors are also prefabricated. In the case of the thin solid
slab an in-situ concrete topping is supplied (SS). In the case of the
reinforced hollow core slab (RHCS), pretensioned hollow core slab
(PHCS), and the aerated concrete floor mortar is used to fill up the
joints in between the elements. In the case of the beam and block
floor system (BB) and steel deck floor (SD) in situ concrete is needed
on top. An in-situ screed material is needed to ensure a plane
surface on the floor type with dovetailed sheet (FJDS). Finally, all
wooden en steel elements are prefabricated.
The system boundary of the life cycle analysis (LCA) corresponds
to a “cradle to grave” analysis. In fact, as mentioned in the CEN/TC
350 standard for the sustainability assessment of buildings (EN
15978) (AFNOR, 2016), the life cycle of a building exists of five main
stages which are subdivided into smaller modules. Very little
studies incorporate all the stages of a full LCA, such studies are
“cradle to cradle” studies. More often, studies only observe the life
cycle stages that are important for that specific study. These studies
are called “cradle to gate”, “gate to gate” or “cradle to grave” studies.
In this beam-floor study, the first stage (modules A1 to A3),
which is called the product stage, has been fully included.
Furthermore, materials transport from the factory to the con-
struction site (Module A4) has also been taken into account.
Conversely, the study does not cover the building activities at the
construction site itself (Module A5), the use phase (Modules B1 to
B7), nor the demolition phase (Module C1), due to the absence of
data needed to model these phases accurately. However, the
transport to the waste processing site (Module C2), the waste
processing (Module C3) and the disposal (Module C4) have been
included in the study. The recycling potential of materials (Module
D) has been taken into account to a limited extent, by using the
principle of ‘avoided impact’. In agreement with (Purnell, 2012;
Sierra-Pe
rez et al., 2016) it is assumed that the considered mod-
ules have the largest environmental impact.
2.2. Calculations
The structural design analysis of the different beams and floors
is based on the beam-floor construction as presented in Fig. 1. Each
Aerated Beam and Light steel floor Light steel Steel I- Softwood þ OSB
concrete block floor joist þ dovetailed floor deck joist þ OSB (SWOSB)
floor (AC) system (BB) sheet (FJDS) joist þ OSB (SD) (IJOSB)
(FJOSB)
x x e e e e e
x x x x e e
e e e e x x
beam-floor system is subjected to a permanent load of 1.5 kN/m2
and an imposed load of 2.0 kN/m2 (normal use by persons
considered) according to the European design code EN 1991-1-1:
Eurocode 1 - Actions on structures - Part 1-1: General actions -
Densities, self-weight, imposed loads for buildings (EN 1991-1-
1:2002). Moreover, the beam span is set at 6 m, the floor span at
4 m, and where feasible, a floor span of 6 m is considered as well.
The beams are considered simply supported over columns or walls
which are not included in this study. In this way the structural
performance of each alternative beam-floor system is the same. A
functional unit of a square meter of the specified structural beam-
floor system can be considered for the LCA study. A such a repre-
sentative structural subsystem is studied of residential buildings
(normal use by persons).
Firstly, specific information regarding the floors has been
extracted from technical documentation provided by several spe-
cific suppliers (e.g. Paulibeton, Doubeton, Fingo, Vanthuyne-ide,
xella, Eurabo, Megaton, …). Herein, the weight is the guiding fac-
tor in selecting out of this documentation a specific floor for each of
the types mentioned in Table 1. The reason for this choice can be
explained by the fact that both the glulam and IPE beam only
contain one main load bearing material, i.e. respectively wood and
steel. In this way, the lightest element will also contain the least
material, which is in most cases beneficial in the context of ecology.
This in turn leads to the decision to implement weight as the
guiding factor in selecting floors out of documentation. However,
this line of thought does not apply to the reinforced concrete beam,
which can appeal to both the concrete and the steel to bear a
specific load. In order to draw conclusions in the same way for each
type of floor, weight will remain the criterion in selecting a concrete
floor out of different possibilities with corresponding load
resistance.
Secondly, the design of the reinforced concrete (rectangular
section), steel (IPE) and wooden (rectangular section) beams have
been designed According the structural European design codes for
reinforced concrete (Eurocode 2), prestressed concrete (Eurocode
2), steel (Eurocode 3) and timber (Eurocode 5) (EN 1992-1-1:2004;
EN 1993-1-1:2005; EN 1995-1-1:2004). Due account is taken of the
ultimate limite state (ULS) and the service limite state (SLS). The
concluding beam and floor dimensions are summarized in Table 2.
Fig. 1. Beam-floor situation.
 C. Dossche et al. / Journal of Cleaner Production 194 (2018) 327e341 331

Table 2
Dimensions of the beams and floors.

Beam Floor S [m] Hfloor [m] Hbeam [m] Htot [m] Wbeam [m]

Concrete (dlim) Thin solid slab þ in-situ concrete topping 4 0.13 0.45 0.58 0.35
6 0.13 0.50 0.63 0.35
Reinforced hollow core slab 4 0.13 0.40 0.53 0.35
6 0.20 0.55 0.75 0.35
Pretensioned hollow core slab 4 0.13 0.40 0.53 0.35
6 0.16 0.50 0.66 0.35
Beam and block floor system 4 0.15 0.40 0.55 0.35
6 0.24 0.50 0.74 0.35
Aerated concrete floor 4 0.20 0.40 0.60 0.35
Concrete (decon) Thin solid slab þ in-situ concrete topping 4 0.13 0.45 0.58 0.35
6 0.13 0.55 0.68 0.35
Reinforced hollow core slab 4 0.13 0.40 0.53 0.35
6 0.20 0.55 0.75 0.35
Pretensioned hollow core slab 4 0.13 0.40 0.53 0.35
6 0.16 0.50 0.66 0.35
Beam and block floor system 4 0.15 0.40 0.55 0.35
6 0.24 0.50 0.74 0.35
Aerated concrete floor 4 0.20 0.40 0.60 0.35
Steel (IPE) Thin solid slab þ in-situ concrete topping 4 0.13 0.36 0.49 0.17
6 0.13 0.40 0.53 0.18
Reinforced hollow core slab 4 0.13 0.33 0.46 0.16
6 0.20 0.40 0.60 0.18
Pretensioned hollow core slab 4 0.12 0.33 0.45 0.16
6 0.16 0.40 0.56 0.16
Beam and block floor system 4 0.15 0.33 0.48 0.16
6 0.21 0.40 0.61 0.18
Aerated concrete floor 4 0.20 0.30 0.50 0.15
Steeldeck 4 0.12 0.33 0.45 0.16
6 0.18 0.40 0.58 0.18
Light steel floor joist þ dovetailed sheet 4 0.23 0.30 0.53 0.15
Light steel floor joist þ OSB 4 0.23 0.30 0.53 0.15
Wood (Glulam) Softwood þ OSB 4 0.24 0.50 0.50 0.25
I-joist þ OSB 4 0.26 0.50 0.50 0.25
6 0.38 0.60 0.60 0.25
Aerated concrete floor 4 0.20 0.50 0.50 0.25

S ¼ span of the floor. This can be 4 or 6 m. The length of the beam is in each case 6 m.
H ¼ height.
W ¼ width.
dlim ¼ dimensions of the beam are calculated in order to obtain a minimal height.
decon ¼ dimensions of the beam are calculated in order to obtain an economical height.

3. The LCA model
3.1. Background
All results mentioned further are based on the use of SimaPro 8
software. Most processes are based on the Ecoinvent 3.1 database
(Ecoinvent Centre, 2016) using the “allocation, default-unit” pro-
cesses. Furthermore, where possible, market processes have been
used as a starting point. However, the materials LVL and I-joist have
been based on processes in the USLCI database, since these were
not available in the Ecoinvent database (National Renewable
Energy Laboratory, 2012).
In a nutshell, for each type of beam and floor in the SimaPro
model, a combination is made of the production phase and the
transport to the building site. Furthermore, the beams and floors
are combined according to Table 1, to which a waste scenario is
added in order to eventually obtain complete life cycles. As
explained in section 2.2 the activities at the construction site, the
use phase, and the demolition phase are excluded.
3.2. Production
The different structural beam-floor systems contain several
materials. For each individual beam and floor, proper quantities of
the constituent materials are adopted in the SimaPro model. These
quantities are derived from the calculations mentioned in section
2.2. For each of these materials, a corresponding market production
process in the Ecoinvent 3.1 (and USLCI) database has been used as
a basis.
These market processes have been adjusted to the Belgian/Eu-
ropean context, which involves firstly changing the transport, and
secondly updating the transformation process in the market pro-
cess, both to a more Belgian/European context where possible. The
database in (Allacker et al., 2013) of environmental profiles of 115
variants of building components representative for the Flemish-
Belgian building context is used as a basis.
This first change, namely the transport in the market process,
includes the substitution of each transport process by the same
type of transport for a Belgian/European context. To elucidate this, a
specific exemplary application of these substitutions is shown in
Table 3 for the transport in the “Concrete, normal {GLO}| market for
| Alloc Def, U” process. Nevertheless, some remarks have to be
made. Firstly, some of the replacing processes are transformation
processes, rather than market processes. This is a result of the
original market processes consisting of only a combination of such
transformation processes. Secondly, the tonne-kilometers of
different markets of the same type (this is in Table 3 the freight
train transport) have been summed up and assigned to one Belgian/
European market. Thirdly, the “Transport, freight, train {Europe
without Switzerland}| market for | Alloc Def, U” has been further
adjusted to suit the Belgian context better, by summation of the
processing processes.
332 C. Dossche et al. / Journal of Cleaner Production 194 (2018) 327e341

Table 3
Specific example of changes to the transport processes.

Original process (es) Replaced by this process

Transport, freight, light commercial vehicle {GLO}| market for | Alloc Transport, freight, light commercial vehicle {Europe without Switzerland}| processing | Alloc Def, U
Def, U
Transport, freight, lorry, unspecified {GLO}| market for | Alloc Def, U Transport, freight, lorry, unspecified {RER}| size-specific lorry transport to generic market for lorry
transport | Alloc Def, U
Transport, freight, train {Europe without Switzerland}| market for | Transport, freight, train {Europe without Switzerland}| market for | Alloc Def, U/Belgium
Alloc Def, U [a tkm] [a þ b þ c þ d þ e tkm]
Transport, freight, train {CN}| market for | Alloc Def, U [b tkm]
Transport, freight, train {CH}| market for | Alloc Def, U [c tkm]
Transport, freight, train {US}| market for | Alloc Def, U [d tkm]
Transport, freight, train {RoW}| market for | Alloc Def, U [e tkm]

The second change to the market processes, i.e. the trans-
formation processes, are specific for each individual material.
Table 5 presents for each material the changes made. As a matter of
fact, since the market processes are frequently presenting a global
scale, the transformation processes herein are often subdivided in
Switzerland (CH) þ the rest of the world (RoW), or Europe
(RER) þ RoW. In this case, only the CH or RER process have been
maintained, adding the quantities of the rest of the world (RoW).
Furthermore, the transport of wood from the production site to the
factory has been adjusted according to information given by the
Public Waste Agency of Flanders (OVAM) (Allacker et al., 2013). This
agency summarized the average transport distances from the main
wood producing countries and their share on the Belgian market.
As a result, Table 4 presents these distances together with the
corresponding process in SimaPro 8.
In Table 5 the concrete grades are mentioned and one can
observe that several concrete grades are used. The respective
concrete grades are chosen to be representative for the concrete
structural elements produced in practice. Strength class C30/37
(‘30e32 MPa concrete’ Process in SimaPro) has been used for the
concrete beams, the prefabricated part of thin solid slabs, and the
prefabricated part of the beam and block floor system. Strength
class C35/45 (‘35 MPa concrete’ Process in SimaPro) has been used
for reinforced hollow core slabs and strength class C50/60 (‘50 MPa
concrete’ Process in SimaPro) for pretensioned hollow core slabs.
The authors are aware that according to (Purnell, 2012; Purnell and
Black, 2012) the optimum concrete grade for most applications
with respect to minimizing the embodied CO2 would be 50 MPa.
However, for this study the current building practice is used as
reference.
3.3. Transport to the building site
The transport scenarios for the transport of the prefabricated
beams and floors, poured concrete and loose structural products to
the building site, are based on the Flemish-Belgian study conducted
by OVAM (Allacker et al., 2013). These scenarios are listed in Table 6.
3.4. Combinations
Each beam production and floor production has been paired and
combined with its transport to the building site. In this way,
different beam assemblies and floor assemblies have been
modeled. Furthermore, these assemblies have been combined ac-
cording to Table 1, to obtain the different beam-floor assemblies.
3.5. End of life
The end of life (EOL) scenario consists of three subscenarios,
which are combined into a top scenario. The three subscenarios are
incineration, landfill and recycling, in which the transport of Table 7
is included. The top scenario combines the different subscenarios
according to Table 8. Tables 7 and 8 are based on the earlier
mentioned study of OVAM (Allacker et al., 2013).
3.6. Life cycles
Each beam-floor assembly has been combined with the
mentioned EOL top scenario, in order to obtain a full life cycle for
the beam-floor combinations.
4. Impact assessment
As mentioned in (Purnell, 2012) the functional unit should
emerge from consideration of the purpose of fundamental struc-
tural elements: e.g. to support a load distributed over a span with
sufficient ground clearance. In this study a functional unit of a
square meter of the specified structural beam-floor system is
considered. In the following sections stepwise the assessment is
built up starting from the beams (ReCiPe World or ReCiPe Europe
endpoint method), over the beam-floor combinations per square
meter (ReCiPe World or ReCiPe Europe endpoint method and
ReCiPe midpoint method), until the beam-floor combinations per
square meter and per life phase (ReCiPe Europe endpoint).
4.1. Beams
Figs. 2 and 3 show the impacts (human health, ecosystems, re-
sources) of the different beams, each bearing a specific floor (floor
span 4 m or 6 m). These figures show the impact assessment
respectively calculated with the ReCiPe Europe endpoint method
and the ReCiPe World endpoint method. The different abbrevia-
tions in the figures are explained in Table 1.
Firstly, regardless the endpoint method used, the beams that
have to bear a floor with a span of 6 m have a higher impact (14%e
31%) than their variant that has to bear a floor of 4 m. This is due to

Table 4
Wood transport: production to factory (Allacker et al., 2013).

Transport, freight, lorry 16e32 metric ton, EURO5 {RER}| transport, freight, lorry 16e32 metric ton, EURO5 | Alloc Def, U 450 km
Transport, freight, sea, transoceanic ship {GLO}| market for | Alloc Def, U 830 km
Transport, freight train {Europe without Switzerland}| market for | Alloc Def, U 75 km
 C. Dossche et al. / Journal of Cleaner Production 194 (2018) 327e341 333

Table 5
Changes to market and transformation processes.

Material Process in SimaPro 8 Changes

In-situ concrete Concrete, normal {CH}| production | Alloc Def, U  100% cement subdivided into 35% Portlandcement þ 55% blast furnace slag cement type
A þ 10% blast furnace slag cement type B (Allacker et al., 2013)
 Every {CH} process has been changed to the equivalent {BE} process or {Europe without
Switzerland}
Prefabricated concrete Concrete, 30e32 MPa {RoW}| concrete The tap water processes have been summed up and combined into one tap water process for
a
production 30e32 MPa, RNA only | Alloc Def, U Europe without Switzerland. The same has been done for cement
Mortar Cement mortar {CH}| production | Alloc Def, U Every {CH} process has been changed to the equivalent {BE} process or {Europe without
Switzerland}
Screed Cement cast plaster floor {CH}| production | Every {CH} process has been changed to the equivalent {BE} process or {Europe without
Alloc Def, U Switzerland}
Aerated concrete Autoclaved aerated concrete block {CH}|  Deleted: “Packing, clay product {GLO}| market for | Alloc Def, U00
production | Alloc Def, U  Avoided product sand because of recycling: only 85% of the “Sand {GLO}| market for | Alloc
Def, U” has been taken into account (Vrijders et al., 2011)
 Every {CH} process has been changed to the equivalent {BE} process or {Europe without
Switzerland}
Brick Brick {RER}| production | Alloc Def, U No changes made.
Hot rolled steel Steel, low-alloyed, hot rolled {RER}| production  “Steel, low-alloyed {GLO}| market for | Alloc Def, U”:
| Alloc Def, U - 67,6% OBC (virgin steel) þ 32,4% EF (recycled steel) (Worldsteel Association, 2015)
- Transport in the market process adjusted similar to Table 3
 “Hot rolling steel {GLO}| market for | Alloc Def, U”:
The transformation processes are combined into only one RER transformation process
Cold-formed steel Steel, low-alloyed, hot rolled {RER}| production  “Steel, low-alloyed {GLO}| market for | Alloc Def, U”:
| Alloc Def, U - 67,6% OBC (virgin steel) þ 32,4% EF (recycled steel) (Worldsteel Association, 2015)
- Transport in the market process adjusted similar to Table 3
 Deleted: “Hot rolling steel {GLO}| market for | Alloc Def, U00
Replaced by: “Sheet rolling, steel {GLO}| market for | Alloc Def, U00
 “Sheet rolling, steel {GLO}| market for | Alloc Def, U”:
The transformation processes are combined into only one RER transformation process
Reinforcement steel Reinforcing steel {RER}| production | Alloc Def,  “Steel, unalloyed {GLO}| market for | Alloc Def, U”:
U - The transformation processes are combined into only one RER transformation process
- Transport in the market process adjusted similar to Table 3
 “Steel, low-alloyed {GLO}| market for | Alloc Def, U”:
- The OBC transformation processes are combined into only one RER transformation process,
and the quantity is set to be 12,4% b
- The EF transformation processes are combined into only one RER transformation process, and
the quantity is set to be 87,6% b
- Transport in the market process adjusted similar to Table 3
 “Hot rolling steel {GLO}| market for | Alloc Def, U”:
The transformation processes are combined into only one RER transformation process
Glulam Glued laminated timber, for indoor use {GLO}| The transport by light commercial vehicle has been maintained, but the other transport has been
market for | Alloc Def, U changed according to Table 4, taking into account the mass density of glulam.
OSB Oriented strand board {GLO}| market for | Alloc The transport by light commercial vehicle has been maintained, but the other transport has been
Def, U changed according to Table 4, taking into account the mass density of OSB.
Softwood Sawnwood, softwood, kiln dried, planed {RER}| The transport by light commercial vehicle has been maintained, but the other transport has been
market for | Alloc Def, U changed according to Table 4, taking into account the mass density of softwood.
LVL Laminated veneer lumber, at plant, US SE/kg/US  The transport has been changed according to Table 4, taking into account the mass density of
LVL.
 The exploitation of timber has been roughly taken into account by changing the process “Dry
veneer, at plywood plant, US SE/US”, i.e. “Green veneer, at plywood plant, US SE/kg/US” was
replaced by “Sawnwood, softwood, raw, air dried {RER}| production | Alloc Def, U00
I-joist Composite wood I-joist, at plant, US SE/kg/US  The processes of OSB and LVL have been changed by the adjusted processes mentioned above.
 There are two types of I-joists, one for each floor span of 4 m and 6 m. For each type, the
quantities of OSB and LVL have been calculated taking into account losses of sawdust.

OBC ¼ Oxygen-blown converter.

EF ¼ Electric furnace.
a
The 30e32 MPa concrete (C30/37) has been used for the concrete beams, the prefabricated part of thin solid slabs, and the prefabricated part of the beam and block floor
system. The same changes as in this exemplary process have been made for 35 MPa concrete (C35/45) (reinforced hollow core slab) and for 50 MPa concrete (C50/60)
(pretensioned hollow core slab).
b
The ratio between unalloyed (63%) and low-alloyed (37%) steel provided by Ecoinvent in the Reinforcing Steel process has been maintained. The ratios with a * result in a
final ratio of respectively 67% and 33% (Worldsteel Association, 2015).

the fact that these beams contain more material and have a higher
impact due to transport due to the higher weight, also associated
with a larger cross section (Table 2). This larger impact might
however be countered by the fact that a larger span induces less
walls and/or columns.
Secondly, the figures indicate that the IPE beams have generally
the largest environmental impact, at least twice as high compared
to the concrete beams for corresponding floor spans. This large
impact is due to the production phase where steel is obtained out of
pig iron.
Thirdly, the glulam beams have a noticeable higher impact on
ecosystems compared to the IPE and reinforced concrete beams.
This is due to the large impact on agricultural land occupation of
pine and spruce in the production of glulam. Furthermore, the
impact on ecosystems appear to be a lot more important in the
assessment method for Europe than in the assessment method for
the World. The reason for this lies in the normalization factors used
in ReCiPe. The normalization factors used for ecosystems in the
ReCiPe endpoint hierarchist method are 1.81E-04 and 9.17E-04 for
European and global scale respectively. In this way, when dividing
334 C. Dossche et al. / Journal of Cleaner Production 194 (2018) 327e341

Table 6
Transport to the building site.

Product group Poured Prefabricated products for structural Loose structural

concrete work products

Arrangement of transportation % directly from factory to site 100% 100% 40%

% via an inter-mediary supplier 0% 0% 60%
Means of transportation form Factory to site  Heavy truck (>16 100% 100% 100%
tonnes)
 Light truck (3.5e16 0% 0% 0%
tonnes)
 Delivery van (<3.5 0% 0% 0%
tonnes)
Factory to  Heavy truck (>16 N.A. N.A. 100%
supplier tonnes)
Supplier to site  Heavy truck (>16 N.A. N.A. 85%
tonnes)
 Light truck (3.5e16 N.A. N.A. 15%
tonnes)
 Delivery van (<3.5 N.A. N.A. 0%
tonnes)
Average transport distance of transportation Factory to site [km] 35 100 100
form Factory to [km] N.A. N.A. 100
supplier
Supplier to site [km] N.A. N.A. 35

N.A. ¼ Not Applicable.

Table 7
EOL transport.

Means of transport Average transport distance

Site to sorting plant or collection Sorting plant or collection point to final Site to sorting plant or collection Sorting plant or collection point to final
point processing point processing

Heavy truck (>16 tonnes) Heavy truck (>16 tonnes) [km] [km]

Landfill 100% 100% 30 50

Incineration 100% 100% 30 100

the characterization numbers by the normalization factors, a m2 of beam-floor combination in Fig. 4, Fig. 5 (endpoint method)
greater value is obtained on European scale than on global scale. and Fig. 6 (midpoint method). These first two figures show the
This is also the reason why the glulam beams have a relatively low impacts of all beam-floor combinations, calculated with the ReCiPe
impact when considering the world method, but an increased endpoint method for respectively European and Global normali-
impact when calculating this with the European method. Never- zation. Both figures show that the conclusions drawn for the
theless, the final summed values of the human health, ecosystems various beams in the previous section cannot be fully adopted for
and resources are higher for the global method. the beam-floor combinations per m2. In some cases the beams that
Lastly, the figures show that, relative to the IPE and glulam have to bear a floor with a span of 6 m also have a higher impact
beams, the differences between the concrete beams with a single than their variant that has to bear a floor of 4 m, but this is certainly
reinforced section calculated in order to obtain an economical not always the case. According to the impact assessment with the
height or a minimal height do not differ too much in impact on the Europe ReCiPe endpoint method (Fig. 4) the lowest impact can be
environment. Nevertheless, the beams with a height in accordance found for the concrete beams combined with pretensioned hollow
with an economical effective depth generally have a lower impact core slabs. Also, reduced impacts can be achieved with concrete
because of a reduced amount of reinforcement. beams combined with precast reinforced hollow core slabs and,
Glulam beams with I-joist þ OSB, and steel beams (IPE) combined
with pretensioned hollow core slabs. According to the impact
4.2. Beam-floor combinations per m2
assessment with the World ReCiPe endpoint method (Fig. 5) the
same conclusions are valid. In addition, however, it should be
In order to be able to compare each beam-floor combination
noticed that the Glulam beams combined with any floor become
mutually, the different beam-floor combinations are visualized per
the most ecological solution. Also the steel beam combined with a
light steel floor joist þ OSB becomes a much more ecological so-
Table 8 lution. These findings are in agreement with the difference be-
Waste scenarios. tween ReCiPe Europe and World method noticed in section 4.1
Product Landfill Incineration Recycling
(agricultural land occupation of pine and spruce). The ecologically
worst beam-floor solution is the steel beam combined with steel
[%] [%] [%]
deck for a floor span of 6 m. In addition, Fig. 6 summarizes the
Inert waste 5 0 95 impacts calculated with the ReCiPe midpoint method for five
Aerated concrete 70 0 30
different midpoint impact categories (climate change, ozone
Metals 5 0 95
Chemically treated wood 5 95 0 depletion, human toxicity, photochemical oxidant formation, agri-
Composite wood products 5 75 20 cultural land occupation). This figure clearly demonstrates that the
 C. Dossche et al. / Journal of Cleaner Production 194 (2018) 327e341 335

200
180
160
140
120
[Pt]

100
80
60
40
20
0

IPE BB 4 m
Glulam IJOSB 4 m
Glulam IJOSB 6 m

IPE AC 4 m
Glulam SWOSB 4 m

IPE RHCS 6 m
IPE SS 4 m

IPE RHCS 4 m
CBmin BB 4 m
CBmin BB 6 m

IPE SS 6 m
CBecon SS 4 m
CBecon SS 6 m

CBecon RHCS 4 m
CBecon RHCS 6 m

CBecon PHCS 4 m

CBmin PHCS 4 m
CBecon PHCS 6 m

CBmin AC 4 m

CBmin PHCS 6 m
Glulam AC 4 m
CBmin RHCS 6 m

IPE BB 6 m
IPE SD 4 m
IPE SD 6 m
Cbecon BB 4 m

CBmin SS 6 m

CBmin RHCS 4 m

IPE FJDS 4 m
IPE PHCS 4 m
IPE PHCS 6 m
CBecon AC 4 m

CBecon BB 6 m

CBmin SS 4 m

IPE FJOSB 4 m
Human Health Ecosystems Resources

Fig. 2. Impact assessment for the beams, each bearing a specific floor, Europe ReCiPe Endpoint H/A.

200

180

160

140

120
[Pt]

100

80

60

40

20

0
IPE AC 4 m

IPE BB 4 m
Glulam IJOSB 4 m
Glulam IJOSB 6 m

IPE RHCS 6 m
Glulam SWOSB 4 m

IPE RHCS 4 m
CBmin BB 6 m

IPE SS 4 m
IPE SS 6 m
CBmin BB 4 m

CBmin PHCS 6 m
CBmin AC 4 m

CBmin PHCS 4 m

Glulam AC 4 m
CBecon SS 4 m
CBecon SS 6 m

CBecon RHCS 4 m
CBecon RHCS 6 m

CBecon PHCS 4 m
CBecon PHCS 6 m

CBmin RHCS 6 m

IPE BB 6 m
IPE SD 4 m
IPE SD 6 m
Cbecon BB 4 m

CBmin SS 6 m

CBmin RHCS 4 m

IPE PHCS 6 m
CBecon AC 4 m

CBmin SS 4 m
CBecon BB 6 m

IPE FJDS 4 m
IPE PHCS 4 m
IPE FJOSB 4 m

Human Health Ecosystems Resources

Fig. 3. Impact assessment for the beams, each bearing a specific floor, World ReCiPe Endpoint H/A.

beam-floor combinations with wood, especially those with Glulam effect of a choice in materials. The importance of the production of
beams, have a high impact on agricultural land occupation, which is the building materials on the environmental impact is also
responsible for the large impact on ecosystems. For the other cat- mentioned in (Purnell, 2012; Sierra-Pe
rez et al., 2016). Keeping
egories, the glulam beams with an OSB floor have the lowest Figs. 2 and 3 in mind, reinforced concrete and wood emerge out of
impact. this study as the most ecological materials for the beams in the
studied beam-floor systems frequently used in the Belgian building
4.3. Beam-floor combinations per m2 and per life phase practice. Also in this case the beams combined with pretensioned
hollow core slabs show the lowest impact. Further optimization of
When evaluating the results with the Europe ReCiPe Endpoint the concrete mixes of the structural elements according to the
method, and focusing on the different life phases, it becomes clear findings of (Knoeri et al., 2013; Purnell, 2012; Purnell and Black,
that the production phases have the largest impact on the envi- 2012) could even further decrease the environmental impact (us-
ronment. This is demonstrated in Fig. 7. Moreover, the production ing optimum concrete strength class, use of superplasticizer in
of the floors is in most cases more important than the production of order to maintain sufficient workability without the need to in-
the beams with which the floors are combined, especially the crease the water and cement content, restriction of the extra
combinations with concrete beams. Furthermore, the transport and cement content in case recycled aggregates are used, use of sup-
end of life scenario are of lesser importance, which accentuates the plementary cementitious materials).
336 C. Dossche et al. / Journal of Cleaner Production 194 (2018) 327e341

18
16
14
12
10
[Pt]

8
6
4
2
0

Glulam + IJOIOSB 4m
Glulam + IJOIOSB 6m
Glulam + AC 4m

IPE + AC 4m
Glulam + SWOSB 4m
IPE + SS 4m
IPE + SS 6m

IPE + RHCS 4m
CBmin + BB 6m
CBecon + BB 4m
CBecon + BB 6m

CBmin + BB 4m

CBmin + PHCS 6m
CBmin + PHCS 4m
CBmin + AC 4m

CBmin + RHCS 6m

IPE + FJOSB 4m
CBecon + PHCS 4m
CBecon + PHCS 6m

CBmin + RHCS 4m
CBecon + SS 4m
CBecon + SS 6m

CBecon + RHCS 4m
CBecon + RHCS 6m

CBmin + SS 4m

IPE + BB 4m
IPE +BB 6m
IPE + SD 4m
IPE + SD 6m
CBmin + SS 6m

IPE + PHCS 4m
IPE + PHCS 6m
CBecon + AC 4m

IPE + RHCS 6m

IPE + FJDS 4m
Human Health Ecosystems Resources

Fig. 4. Impact assessment for the beam-floor combinations per m2, Europe ReCiPe Endpoint H/A.

18

16

14

12

10
[Pt]

0
Glulam + IJOIOSB 4m
Glulam + IJOIOSB 6m
Glulam + AC 4m

IPE + AC 4m
Glulam + SWOSB 4m

IPE + RHCS 4m
IPE + SS 4m
IPE + SS 6m
CBmin + BB 6m
CBmin + BB 4m

CBmin + PHCS 6m
CBmin + PHCS 4m
CBecon + BB 6m

CBmin + AC 4m
CBecon + BB 4m

CBecon + PHCS 4m
CBecon + PHCS 6m

CBmin + RHCS 4m
CBmin + RHCS 6m

IPE + FJOSB 4m
CBecon + RHCS 4m
CBecon + RHCS 6m
CBecon + SS 4m
CBecon + SS 6m

CBmin + SS 4m
CBmin + SS 6m

IPE + BB 4m
IPE +BB 6m
IPE + SD 4m
IPE + SD 6m

IPE + PHCS 6m
IPE +PHCS 4m
CBecon + AC 4m

IPE + RHCS 6m

IPE + FJDS 4m

Human Health Ecosystems Resources

Fig. 5. Impact assessment for the beam-floor combinations per m2, World ReCiPe Endpoint H/A.

5. Sensitivity analysis of these varying numbers, the LCA has been performed with
different values for the virgin-recycled steel ratio. These different
To eventually obtain the different lifecycles mentioned in 3.6, values are listed in Table 9.
several assumptions have been made. In order to look into some of When analyzing the results of this sensitivity analysis,
these assumptions and their respective influence on the life cycle depicted in Figs. 8 and 9, the following conclusions can be drawn.
impact assessment, different sensitivity analyses have been made Firstly, varying the amount of recycled steel in a product has a
in this study. relatively high impact on the results. Secondly, the magnitude of
this impact is dependent on the type of beam-floor combination
5.1. Composition of steel it is applied on. This is very clear in Figs. 8 and 9, where the
results of the combinations with an IPE beam differ more than
According to the Steel Statistical Yearbook 2015 (Worldsteel those with concrete beams, of which the latter contain less
Association, 2015), steel production in Belgium exists of 67.6% structural steel than IPE beams. Thirdly, the results show that the
newly produced steel and 32.4% recycled steel. This is similar to basic scenario and scenario 4 and 5 have a relatively more similar
values mentioned for UK construction practice, i.e. a recycled impact on the environment than compared to scenarios 1, 2 and
fraction of ±60% (Purnell, 2012). However, these average numbers 3 (100% till 50% recycled). This is particularly apparent in the
differ from one producer to another. In order to analyze the impact combinations with a concrete beam. The more steel can be
 E D C B A
[Pt]
[m2a] [kg NMVOC] [kg 1.4-DBeq] [kg CFC-11eq] [kg CO2 eq]

0
2
4
6
8
10
12
14

floor solutions.
0
150
300
0E+0
1E-5
0
50
100

0
40
80

0
0.2
0.4
CBecon + SS 4m
CBecon + SS 4m
CBecon + SS 6m
CBecon + SS 6m
CBecon + AC 4m
CBecon + AC 4m
CBecon + RHCS 4m
CBecon + RHCS 4m
CBecon + RHCS 6m
CBecon + RHCS 6m

formation, E: Agricultural land occupation.

CBecon + BB 4m
CBecon + BB 4m
CBecon + BB 6m
CBecon + BB 6m
CBecon + PHCS 4m
CBecon + PHCS 4m
CBecon + PHCS 6m
CBecon + PHCS 6m
CBmin + SS 4m

ProducƟon beam
CBmin + SS 4m
CBmin + SS 6m
CBmin + SS 6m
CBmin + AC 4m
CBmin + AC 4m
CBmin + RHCS 4m
CBmin + RHCS 4m
CBmin + RHCS 6m
CBmin + RHCS 6m
CBmin + BB 4m
CBmin + BB 4m

Transport beam

recycled the better. When the nowadays recycled percentage of

combinations could be able to compete with the other beam-

steel could be increased from ±68%e80% the steel beam-floor
CBmin + BB 6m
CBmin + BB 6m
CBmin + PHCS 4m
CBmin + PHCS 4m
CBmin + PHCS 6m
CBmin + PHCS 6m
Glulam + AC 4m
Glulam + AC 4m
Glulam + IJOIOSB 4m
Glulam + IJOIOSB 4m

ProducƟon floor
Glulam + IJOIOSB 6m
Glulam + IJOIOSB 6m
Glulam + SWOSB 4m
Glulam + SWOSB 4m
IPE + SS 4m
IPE + SS 4m
IPE + SS 6m
IPE + SS 6m
C. Dossche et al. / Journal of Cleaner Production 194 (2018) 327e341

IPE + AC 4m
IPE + AC 4m

Transport floor
IPE + RHCS 4m
IPE + RHCS 4m
IPE + RHCS 6m
IPE + RHCS 6m

Fig. 7. Beam-floor combinations per m2 and per life phase, Europe ReCiPe Endpoint H/A.
EOL
IPE + BB 4m IPE + BB 4m
IPE +BB 6m IPE +BB 6m
IPE + SD 4m IPE + SD 4m
IPE + SD 6m IPE + SD 6m
IPE + FJOSB 4m IPE + FJOSB 4m
IPE + FJDS 4m IPE + FJDS 4m
IPE + PHCS 4m IPE +PHCS 4m

5.2. Use of recycled aggregates in ready-mixed concrete

IPE + PHCS 6m IPE + PHCS 6m
Fig. 6. Impact assessment for the beam-floor combinations per m2, ReCiPe Midpoint H/A, A: Climate change, B: Ozone depletion, C: Human toxicity, D: Photochemical oxidant

According to the Belgian Standard NBN B15-001, the application

337

of recycled aggregates is only permitted when applied in concrete

338 C. Dossche et al. / Journal of Cleaner Production 194 (2018) 327e341

Table 9 is important to notice that the results of the different scenarios

Scenarios for the sensitivity analysis of the composition of steel. vary only to a relatively small extent with respect to the total
Basic scenario 32.4% recycled steel þ 67.6% virgin steel impacts of each scenario.
Scenario 1 100% recycled steel þ 0% virgin steel
Scenario 2 80% recycled steel þ 20% virgin steel 5.3. Use of aerated concrete waste in new aerated concrete
Scenario 3 50% recycled steel þ 50% virgin steel
Scenario 4 20% recycled steel þ 80% virgin steel
Scenario 5 0% recycled steel þ 100% virgin steel In the basic scenario of aerated concrete, 15% of the sand has
been assumed as avoided impact to take into account the recycling
of aerated concrete. This amount is based on a document about
valuable applications of recycled aerated concrete (Vrijders et al.,
with a strength up to C25/30. These recycled aggregates however
2011). However, in Belgium, the largest producer of aerated con-
can also be used in other concrete types, when it is proven that the
crete has a market share of 90% and uses up to a maximum amount
concrete mixes possess the adequate performance. This possibility
of 25% of aerated concrete to replace sand. This secondary aerated
is taken into account in this sensitivity analysis.
concrete originates from production, construction and demolition
In the basic scenario, no recycled aggregates are used. According
waste. Furthermore, since only small amounts of recyclable aerated
to NBN B15-001, a maximum of 20% of recycled aggregates can be
concrete are available, new aerated concrete can not always be
used. Therefore, in this sensitivity analysis a replacement of 20% is
produced with recycled substances.
analyzed, together with an intermediate scenario in which 10% of
Based on these findings, a sensitivity analysis is performed with
the aggregates will be replaced by recycled ones. These percentages
a scenario of an avoided impact of 25% of the sand, and another
are taken into account by means of an avoided impact, it is
scenario without any avoided impact. It is important to notice here,
important to notice here that the recycling of aggregates has only
analogous to 5.2., that the recycling of aerated concrete has only
been taken into account by using this method of avoided impact.
been taken into account by using the method of avoided impact. In
This is a simplified way of modelling recycling, not taking into ac-
the same way, this is a simplified way of modelling recycling, not
count various other allocation possibilities.
taking into account various other allocation possibilities.
When analyzing the results of the sensitivity analysis by
Varying this avoided impact has no significant impact on the
Figs. 10 and 11, it is clear that the scenarios in which a certain
final life cycle results, which can be derived from Figs. 12 and 13. It
amount of the aggregates is recycled result in slightly lower im-
is clear that the scenarios in which an amount of the aerated con-
pacts on the environment. The impacts vary only with 0.1e0.22
crete is recycled result in 0.1e0.14 percent lower impacts on the
percent. The fact that the impacts of scenarios with recycled ag-
environment. This is a logic result since only an avoided impact is
gregates are lower, is a logic result since only an avoided impact is
taken into account to model the recycled aerated concrete. How-
taken into account to model the recycled aggregates. However, it
ever, it is important to notice that the results of the different

22

20

18

16

14

12
[Pt]

10

0
Glulam + IJOIOSB 6m
Glulam + AC 4m
Glulam + IJOIOSB 4m

IPE + AC 4m
Glulam + SWOSB 4m

IPE + RHCS 4m
IPE + SS 4m
IPE + SS 6m
CBmin + BB 6m
CBmin + BB 4m

CBmin + PHCS 6m
CBecon + BB 4m
CBecon + BB 6m

CBmin + PHCS 4m
CBmin + AC 4m

IPE + FJOSB 4m
CBecon + PHCS 6m
CBecon + RHCS 4m
CBecon + RHCS 6m

CBecon + PHCS 4m

CBmin + RHCS 4m
CBmin + RHCS 6m

IPE + PHCS 6m
CBecon + SS 4m

IPE + BB 4m
CBecon + SS 6m

CBmin + SS 6m

IPE +BB 6m
IPE + SD 4m
IPE + SD 6m

IPE + PHCS 4m
CBmin + SS 4m
CBecon + AC 4m

IPE + RHCS 6m

IPE + FJDS 4m

Scenario 1 Scenario 2 Scenario 3 Basic scenario Scenario 4 Scenario 5

2
Fig. 8. Sensitivity analysis on steel composition (beam-floor per m ), Europe ReCiPe Endpoint H/A.
 C. Dossche et al. / Journal of Cleaner Production 194 (2018) 327e341 339

22

20

18

16

14

12
[Pt]

10

Glulam + IJOIOSB 4m
Glulam + AC 4m

Glulam + IJOIOSB 6m

IPE + AC 4m
Glulam + SWOSB 4m
IPE + SS 4m
IPE + SS 6m

IPE + RHCS 4m
CBmin + BB 6m
CBmin + BB 4m

CBmin + PHCS 4m
CBmin + PHCS 6m
CBmin + AC 4m
CBecon + BB 4m
CBecon + BB 6m

CBmin + RHCS 6m
CBecon + PHCS 4m
CBecon + PHCS 6m

CBmin + RHCS 4m

IPE + FJOSB 4m
CBecon + SS 4m
CBecon + SS 6m

CBecon + RHCS 4m
CBecon + RHCS 6m

IPE + BB 4m
CBmin + SS 4m
CBmin + SS 6m

IPE +BB 6m
IPE + SD 4m
IPE + SD 6m
IPE + RHCS 6m

IPE + PHCS 4m
CBecon + AC 4m

IPE + FJDS 4m

IPE + PHCS 6m
Scenario 1 Scenario 2 Scenario 3 Basic scenario Scenario 4 Scenario 5

2
Fig. 9. Sensitivity analysis on steel composition (beam-floor per m ), World ReCiPe Endpoint H/A.

18 18
16 16
14 14
12 12
10 10
[Pt]

[Pt]

8 8
6 6
4 4
2 2
0 0
IPE + SS 4m
IPE + SS 6m
IPE + RHCS 6m
CBmin + SS 4m
CBmin + SS 6m
CBmin + RHCS 6m

CBmin + BB 6m

IPE + BB 4m
CBecon + SS 4m
CBecon + SS 6m
CBecon + RHCS 6m
CBecon + BB 4m
CBecon + BB 6m

CBmin + BB 4m

IPE +BB 6m
IPE + SD 4m
IPE + SD 6m

CBmin + SS 6m

IPE + SS 4m
IPE + SS 6m
IPE + RHCS 6m

IPE +BB 6m
CBmin + RHCS 6m

CBmin + BB 6m

IPE + BB 4m
CBmin + BB 4m

IPE + SD 4m
CBecon + SS 4m
CBecon + SS 6m
CBecon + RHCS 6m
CBecon + BB 4m
CBecon + BB 6m
CBmin + SS 4m

IPE + SD 6m
Basic scenario Scenario 10 Scenario 20 Basic scenario Scenario 10 Scenario 20
Fig. 10. a) Sensitivity analysis on recycled aggregates (beam-floor per m2), Europe
Fig. 11. Sensitivity analysis on recycled aggregates (beam-floor per m2), World ReCiPe
ReCiPe Endpoint H/A.
Endpoint H/A.

scenarios vary only to a relatively small extent with respect to the 6. Conclusions
total impacts of each scenario.
This paper presents the results of a ‘cradle to grave’ LCA for
different structural beam-floor combinations, for various
5.4. Transport types and distances impact categories and both midpoint and endpoint assessment
methods.
In the earlier mentioned Flemish-Belgian study of OVAM Based on the impact assessment of the different beams, each
(Allacker et al., 2013), sensitivity analyses have been performed bearing a specific floor (floor span 4 m or 6 m), the following main
about the transport types and distances of wood to the factory, of conclusions can be made:
beams and floors to the construction site, and of demolition waste
to the waste treatment plant. These sensitivity analyses show that a (1) Regardless the endpoint method used, the beams that have
best and worst case scenario of the different transport processes to bear a floor with a span of 6 m have a higher impact than
each only result in small changes of the aggregated life cycle impact their variant that has to bear a floor of 4 m. This larger impact
assessments. Since the transport processes of the current study are might however be countered by the fact that a larger span
based on the study of OVAM, these sensitivity analyses are not induces less walls and/or columns.
remodeled in the current study.
340 C. Dossche et al. / Journal of Cleaner Production 194 (2018) 327e341

14 (3) According to the impact assessment with the World ReCiPe

endpoint method the same conclusions are valid. In addition,
12
however, it should be noticed that the Glulam beams com-
10 bined with any floor become the most ecological solution.
Also the steel beam combined with a light steel floor
8 joist þ OSB becomes a much more ecological solution. The
[Pt]

6 ecologically worst beam-floor solution is the steel beam

combined with steel deck for a floor span of 6 m.
4 (4) Considering the ReCiPe midpoint method for five different
midpoint impact categories it is demonstrated that the
2
beam-floor combinations with wood, especially those with
0 Glulam beams, have a high impact on agricultural land
CBecon + AC 4m CBmin + AC 4m Glulam + AC 4m IPE + AC 4m occupation, which is responsible for the large impact on
Scenario 0 Basic scenario 0.15 Scenario 0.25 ecosystems. For the other categories, the glulam beams with
an OSB floor have the lowest impact.
Fig. 12. Sensitivity analysis on aerated concrete (beam-floor per m2), Europe ReCiPe (5) When focusing on the different life phases, it becomes clear
Endpoint H/A.
that the production phases have the largest impact on the
environment. Moreover, the production of the floors is in
14 most cases more important than the production of the beams
with which the floors are combined, especially the combi-
12 nations with concrete beams. Reinforced concrete and wood
emerge out of this study as the most ecological materials for
10
the beams in the studied beam-floor systems current in the
8 Belgian building practice. Also in this case the beams com-
[Pt]

bined with pretensioned hollow core slabs show the lowest

6 impact.
4
2 rameters used to model this LCA, have no significant impact when
0 concrete, use of aerated concrete waste in new aerated concrete,
CBecon + AC 4m CBmin + AC 4m Glulam + AC 4m IPE + AC 4m
Scenario 0 Basic scenario 0.15 Scenario 0.25
Fig. 13. Sensitivity analysis on aerated concrete (beam-floor per m2), World ReCiPe
Endpoint H/A.
(2) The IPE beams have generally the largest environmental
impact, at least twice as high compared to the concrete
beams for corresponding floor spans.
(3) The glulam beams have a noticeable higher impact on eco-
systems compared to the steel and reinforced concrete
beams. Furthermore, the impact on ecosystems appear to be
a lot more important in the assessment method for Europe
than in the assessment method for the World.
(4) Relative to the steel and glulam beams, the differences be-
tween the concrete beams with a single reinforced section
calculated in order to obtain an economical height or a
minimal height do not differ too much in impact on the
environment.
Furthermore, when analyzing the beam-floor combinations per
m2, the following main conclusions are made:
(1) In some cases the beams that have to bear a floor with a span
of 6 m also have a higher impact than their variant that has to
bear a floor of 4 m, but this is certainly not always the case.
(2) According to the impact assessment with the Europe ReCiPe
endpoint method the lowest impact can be found for the
concrete beams combined with pretensioned hollow core
slabs. Also reduced impacts can be achieved with concrete
beams combined with precast reinforced hollow core slabs
and, Glulam beams with I-joist þ OSB, and steel beams (IPE)
combined with pretensioned hollow core slabs.
Lastly, several sensitivity analyses show that most of the pa-
their values are changed (use of recycled aggregates in ready-mixed
transport types and distances according the Belgian concrete
practice). However, varying the ratio virgin-recycled steel can result
in quite different scores on environmental impact. When the
nowadays recycled percentage of steel could be increased from
±68%e80% the steel beam-floor combinations could be able to
compete with the other beam-floor solutions.
Acknowledgements
This research was supported by Ghent University. The authors
wish to express their gratitude for the support. Also the authors
thank Jens Baetens and Cedric Vennens for the delivered work
during their master thesis, which contributed to the research re-
ported in this paper.
References
AFNOR Normalisation, 2016. CEN/TC350: Sustainability of Construction Works.
AFNOR Normalisation. Available online at: http://portailgroupe.afnor.fr/public_
espacenormalisation/CENTC350/index.html. (Accessed 30 March 2016).
Allacker, K., Debacker, W., Delem, L., De Nocker, L., De Troyer, F., Janssen, A.,
Peeters, K., Servaes, R., Spirinckx, C., Van Dessel, J., 2013. Environmental Profile
of Building Elements: towards an Integrated Environmental Assessment of the
Use of Materials in Buildings. OVAM, Mechelen: Danny Wille. Available online
at: https://www.vlaanderen.be/nl/publicaties/detail/environmental-profile-of-
building-elements-towards-an-integrated-environmental-assessment-of-the-
use-of-materials-in-buildings.
Black, L., Purnell, P., 2016. Is carbon dioxide pricing a driver in concrete mix design?
Mag. Concr. Res. 68 (11), 561e567.
Buyle, M., Braet, J., Audenaert, A., Debacker, W., 2018. Strategies for optimizing the
environmental profile of dwellings in a Belgian context: a consequential versus
an attributional approach. J. Clean. Prod. 173, 235e244.
Chastas, P., Theodosiou, T., Kontoleon, K.J., Bikas, D., 2018. Normalising and
assessing carbon emissions in the building sector: a review on the embodied
CO2 emissions of residential buildings. Build. Environ. 130, 212e226.
Eaton, D., Jones, S.R., Pennington, D.G., Roberti, D.A., Chandler, R.E., 2000.
A Technical Manual for Material Choices in Sustainable Construction. Integrated
Waste Management Board, Californie €.
Ecoinvent Centre, 2016. The Ecoinvent Database. Ecoinvent Centre. Available online
 C. Dossche et al. / Journal of Cleaner Production 194 (2018) 327e341
at: http://www.ecoinvent.org/database/database.html. (Accessed 31 March
2016).
EN 1991-1-1, 2002. Eurocode 1: Actions on Structures - Part 1-1: General Actions -
Densities, Self-weight, Imposed Loads for Buildings.
EN 1992-1-1, 2004. Eurocode 2: Design of Concrete Structures - Part 1-1: General
Rules and Rules for Buildings.
EN 1993-1-1, 2005. Eurocode 3: Design of Steel Structures - Part 1-1: General Rules
and Rules for Buildings.
EN 1995-1-1, 2004. Eurocode 5: Design of Timber Structures - Part 1-1: General -
Common Rules and Rules for Buildings.
Ghisellini, P., Ripa, M., Ulgiati, S., 2018. Exploring environmental and economic costs
and benefits of a circular economy approach to the construction and demolition
sector. A literature review. J. Clean. Prod. 178, 618e643.
Ha€fliger, I.-F., John, V., Passer, A., Lasvaux, S., Hoxha, E., Ruschi Mendes Saade, M.,
Habert, G., 2017. Buildings environmental impacts’ sensitivity related to LCA
modelling choices of construction materials. J. Clean. Prod. 156, 805e816.
Hoxha, E., Habert, G., Lasvaux, S., Chevalier, J., Le Roy, R., 2017. Influence of con-
struction material uncertainties on residential building LCA reliability. J. Clean.
Prod. 144, 33e47.
ISO 14040, 2006. Environmental Management - Life Cycle Assessment - Principles
and Framework.
Knoeri, C., Sanye
-Mengual, E., Althaus, H.-J., 2013. Comparative LCA of recycled and
conventional conrete for structural applications. Int. J. Life Cycle Assess. 18,
909e918.
Lacovidou, E., Purnell, P., 2016. Mining the physical infrastructure: opportunities,
barriers and interventions in promoting structural components reuse. Sci. Total
Environ. 557e558, 791e807.
Leung, W., Noble, B., Gunn, J., Jaeger, J.A., 2015. A review of uncertainty research in
impact assessment. Environ. Impact Assess. Rev. 50, 116e123.
Monteiro, H., Freire, F., 2012. Life-cycle assessment of a house with alternative
exterior walls: comparison of three impact assessment methods. Energy Build.
47, 572e583.
National Renewable Energy Laboratory, 2012. U.S. Life Cycle Inventory Database.
341
Available online at: https://www.lcacommons.gov/nrel/search. (Accessed 8
August 2016).
Oregi, X., Hernandez, P., Hernandez, R., 2017. Analysis of life-cycle boundaries for
environmental and economic assessment of building energy refurbishment
projects. Energy Build. 136, 12e25.
Purnell, P., 2012. Material nature versus structural nature: the embodied carbon of
fundamental structural elements. Eviron. Sci. Technol. 46, 454e461.
Purnell, P., Black, L., 2012. Embodied carbon dioxide in concrete: variation with
common mix design parameters. Cement Concr. Res. 42, 874e877.
Rashid, A., Yusoff, S., 2015. A review of life cycle assessment method for building
industry. Renew. Sustain. Energy Rev. 45, 244e248.
Sierra-Pe
rez, J., Boschmonart-Rives, J., Gabarrell, X., 2016. Environmental assess-
ment of façade-building systems and thermal insulation materials for different
climatic conditions. J. Clean. Prod. 113, 102e113.
Takano, A., Winter, S., Hughes, M., Linkosalmi, L., 2014. Comparison of life cycle
assessment databases: a case study on building assessment. Build. Environ. 79,
20e30.
United Nations Environment Programme (UNEP), 2009. Buildings and Climate
Change - Summary for Decision-makers.
U.S. Environmental Protection Agency, 2007. Energy Trends in Selected
Manufacturing Sectors: Opportunities and Challenges for Environmentally
Preferable Energy Outcomes. Available online at: http://www.epa.gov.
Vilches, A., Garcia-Martinez, A., Sanchez-Montan ~ es, B., 2017. Life cycle assessment
(LCA) of building refurbishment: a literature review. Energy Build. 135,
286e301.
Vitale, P., Arena, N., Di Gregorio, F., Arena, U., 2017. Life cycle assessment of the end-
of-life phase of a residential building. Waste Manag. 60, 311e321.
Vrijders, J., Nielsen, P., Quaghebeur, M., 2011. Onderzoek Naar Hoogwaardige Toe-
passingen Van Cellenbetonafval. OVAM, Mechelen: Danny Wille. Available on-
line at: http://www.ovam.be/sites/default/files/FILE1305033641683ovor
110510_Onderzoek_hoogwaardige_toep_cellenbetonafval.pdf.
Worldsteel Association, 2015. Steel Statistical Yearbook 2015. Worldsteel Commit-
tee on Economic Studies, Brussels.