Sei sulla pagina 1di 14

Building and Environment 93 (2015) 165e178

Contents lists available at ScienceDirect

Building and Environment


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

Critical review of life cycle assessment (LCA) for the built environment
at the neighborhood scale
Marc Lotteau a, b, *, Philippe Loubet a, c, Maxime Pousse b, Emmanuel Dufrasnes b, d,
Guido Sonnemann a, e
a
Univeristy of Bordeaux, ISM, UMR 5255, 351 Cours de la Lib eration, F-33400 Talence, France
b
Nobatek, 67 rue de Mirambeau, F-64600 Anglet, France
c
INP Bordeaux, ISM, UMR 5255, 351 Cours de la Lib eration, F-33400 Talence, France
d
ENSAS, 6-8 boulevard du President Wilson, F-67068 Strasbourg, France
e
CNRS, ISM, UMR 5255, 351 Cours de la Liberation, F-33400 Talence, France

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

Article history: The built environment is recognized as a major hotspot of resource use and environmental impacts. Life
Received 22 April 2015 cycle assessment (LCA) has been increasingly used to assess the environmental impacts of construction
Received in revised form products and buildings during the last 25 years. A new trend stems in the application of LCA to larger
23 June 2015
systems such as urban islets or neighborhoods. This review aims at compiling all papers related to LCA of
Accepted 26 June 2015
the built environment at the neighborhood scale. A focus is carried out on 21 existing case studies which
Available online 2 July 2015
are analyzed according to criteria derived from the four phases of LCA international standards. It sums up
current practices in terms of goal and scope definition, life cycle inventory (LCI) and life cycle impact
Keywords:
Life cycle assessment (LCA)
assessment (LCIA). The results show that the case studies pursue different goals. They are either con-
Neighborhood sustainability assessment ducted on existing or model neighborhoods with an aim at building knowledge to feed urban policy
(NSA) making. Or they are conducted on actual urban development projects for eco-design purpose. Studies are
Critical review based on different scopes, resulting in the selection of different functional units and system boundaries.
Neighborhood scale A comparison of data collection strategies is provided as well as a comparison of LCIA results for cu-
District scale mulative energy demand and greenhouse gases emissions. Methodological challenges and research
needs in the field of application of LCA to neighborhood scale assessment are identified, such as the
definition of the functional unit and the need for contextualization methodologies aligned with data
availability at the design stages of a neighborhood development project.
© 2015 Elsevier Ltd. All rights reserved.

1. Introduction is addressed through a wide range of methodologies at both the


individual building scale and the city scale.
More than fifty percent of the world population (2011) lives in At the individual building scale, among various methodologies
urban areas [1] and the built environment, referring to the building (e.g. statistical models, simulation) life cycle assessment (LCA) is
and transportation sectors, is a major hotspot of resource use and the clearly accepted scientific methodology for quantitative
environmental impacts. For instance it accounts for 62% of global assessment of buildings over their entire lifespan accounting for
final energy consumption (2009) [2] and 55% of greenhouse gas upstream impacts [5]. Several reviews have been conducted on LCA
emissions (2004) [3]. In industrialized countries, buildings are in the construction industry [6e9]. These reviews classify LCA
responsible for 42% of the final energy consumption, 35% of studies according to their scale of analysis, from LCA of building
greenhouse gas emissions and more than 50% of all extracted ma- materials and component combinations to LCA of the whole con-
terials [4]. struction process [6]. They all point out the fact that case studies
The analysis of environmental impacts of the built environment found in literature are difficult to compare because of their specific
properties like building type, climate, comfort requirements, local
regulations, etc. Nevertheless trends can be identified such as the
dominance of the use phase (especially due to energy consumption
* Corresponding author. Nobatek, 67 rue de Mirambeau, F-64600 Anglet, France.
E-mail address: mlotteau@nobatek.com (M. Lotteau). for heating and cooling) and the increase of the share and absolute

http://dx.doi.org/10.1016/j.buildenv.2015.06.029
0360-1323/© 2015 Elsevier Ltd. All rights reserved.
166 M. Lotteau et al. / Building and Environment 93 (2015) 165e178

value of embodied energy for low-energy buildings. Regarding A neighborhood, being a complex, multifunctional, long-lasting
methodological aspects, most of LCA of buildings are process-based and dynamic system, raises specific issues regarding for example
LCA and are life cycle energy assessment (LCEA), meaning that they the definition of the functional unit or the consideration of tem-
focus on energy related issues, rather than full LCA (addressing a poral aspects in LCA. This review also summarizes the main
broader range of environmental impacts) [8]. methodological challenges for this new field of application and
At the city scale, assessment methodologies for environmental associated research opportunities.
impacts quantification are numerous as well. Baynes et al. [10]
examine general approaches under three categories; 2. Material and methods
consumption-based approaches (CBA), metabolism-based ap-
proaches (MBA) and complex systems approaches. CBA capture 2.1. Selection of LCA papers dealing with the built environment at
economy-wide impacts as a function of urban consumption and the neighborhood scale
lifestyles. The most commonly used CBA method is environmen-
tally extended inputeoutput analysis (EE-IOA) capable of assessing The present review aims at compiling papers presenting an in-
direct and indirect water, and energy use, ecological footprinting tegrated assessment of the built environment at the neighborhood
and most notably greenhouse gas (GHG) accounting. MBA record scale using LCA approaches. The built environment is the summa-
the material, water, waste and energy flows of an urbanized area. tion of all human-made structures, infrastructure and trans-
Methods are essentially an urban materials and energy flow anal- portation systems [5]. For this review, we propose to differentiate
ysis (MEFA). MBA methods increasingly account for indirect or four fields of the built environment that are derived from Popovici
cross-boundary impacts of production and consumption. Complex (2006) [13]; buildings, open spaces (roads, green spaces), networks
systems approaches attribute causes and effects through quanti- (water, telecommunications, sewage, heating distribution, elec-
fying relationships and feedbacks throughout the urban system. tricity distribution) and mobility. We chose to select case studies
These approaches are more recently developed and include at- dealing with at least two fields of the built environment. Consid-
tempts at modeling the underlying dynamics of cities at various ering the importance of buildings on environmental burdens as
scales. According to Anderson et al. [5] LCA is again a dominant well as the fact that building LCA literature has already been widely
method for urban level analysis and is complementary with CBA investigated, we also propose to classify the papers according to the
and MBA approaches. Indeed hybrid LCA techniques appear as a scale of analysis applied to buildings and we add to our selection,
promising direction for realizing an optimum combination of both papers that address buildings only but with an analysis performed
CBA and MBA methods [10]. In their review of existing tools and at the neighborhood scale (and not at the individual building scale
methods to assess the environmental impacts of a territory, Loiseau to be then extrapolated at the neighborhood scale). Eventually the
et al. [11] studied in detail a set of methods that are based on the present review compiled (i) papers including several fields of the
study of material and energy flows (i.e. Human and Environmental built environment and (ii) papers addressing buildings only but
Risk Assessment, Ecological Footprint, Material and Energy Flow with an analysis performed at the neighborhood scale.
Analysis, Exergy, Emergy and LCA). The authors show that LCA
provides a relevant framework and is the only method which can 2.2. Analysis grid of LCA papers
avoid burden shifting between life cycle stages, environmental
impacts and territories. The case studies analysis follows the four steps of LCA according
There is a growing interest for the neighborhood scale in the to ISO [14]: (phase 1) definition of goal and scope, (phase 2) life
field of urban sustainability assessment. It is a typical operational cycle inventory (LCI), (phase 3) life cycle impact assessment (LCIA)
scale for urban development projects and integrates key levers for and (phase 4) interpretation of the results. For each phase, a set of
urban eco-design. Indeed, this change of scale is driven by the need criteria has been selected from the ISO and ILCD guidelines [15].
to address district scale levers to design buildings and neighbor- The set of criteria is detailed below and a summary is provided in
hoods of higher environmental performance and to address key Table 1.
issues such as bioclimatic design, shared equipment (e.g. district
heating), urban density or mobility issues. For instance, decisions 2.2.1. Criteria for LCA phase 1 e goal and scope
made at the settlement level (orientation, compactness, urban The studies' goals are compared according to their intended
density) largely affect heating/cooling loads, a major contributor to applications and the reasons for carrying out the studies. A focus is
the energy balance of an urban area. Oliver-Sol a et al. [12] state that placed on whether the studies are “neighborhood development
it seems more likely that the environmental and energy concerns sustainability studies” aiming at producing specific knowledge for a
that nowadays focus mainly on buildings will soon be transferred to particular urban development project or “urban sustainability
neighborhood planning. studies” aiming at producing general knowledge for urban sus-
As an answer to this growing interest for sustainability inves- tainability policy making. The analysis of the scope definition in-
tigation and planning at the neighborhood scale, a new trend stems cludes (i) the choice of functional unit (FU); (ii) key information
in the application of LCA to neighborhood projects. Several papers about the system (geographic location, number of inhabitants,
related to the LCA of neighborhoods have already been published area); (iii) the definition of system boundaries; and (iv) the life
but existing literature is still scarce and heterogeneous. This review cycle steps considered. Concerning the boundaries, the analysis
aims at compiling all papers related to this topic. A focus is carried investigates whether or not the case studies include the different
out on 21 existing case studies (from 14 papers) which are analyzed fields of the built environment as defined in Section 2.1: buildings
according to criteria derived from the four phases of LCA interna- (B), open spaces (OS), networks (N) and mobility (M). Concerning
tional standards. It sums up current practices in terms of goal and the life cycle step, the inclusion of Construction (raw materials
scope definition, life cycle inventory (LCI) and life cycle impact extraction, manufacturing, buildings and infrastructures construc-
assessment (LCIA) methodological choices, as well as main results. tion), Operation (operation and maintenance), and Deconstruction
This analysis shows that whereas all studies aim at providing (end of life) is reviewed.
quantitative information to policy makers or designers, they are
based on different scopes and focus on different issues, resulting in 2.2.2. Criteria for LCA phase 2 e life cycle inventory
various functional units and system boundaries. The LCI is the data collection part of LCA. It consists of detailed
M. Lotteau et al. / Building and Environment 93 (2015) 165e178 167

Table 1
Description of criteria taken into account within the review.

LCA phase Qualitative criteria Quantitative criteria

Phase 1 e Goal and scope definition - Goal - Functional unit


- System boundaries - Geographic location
- Life cycle steps considered - Number of inhabitants
- Size
- Lifetime or period of analysis
Phase 2 e LCI - Source of foreground data
- Source of background data
Phase 3 e LCIA - LCIA method selected - Primary energy consumption data
- Impacts and damages taken into account - Climate change impact data
- Normalization (yes/no?)
- Weighting (yes/no?)
Phase 4 e interpretation - Mono or multi-criteria - Contribution analysis, identification of hotspots
- Sensitivity check

tracking of all the flows in and out of the system, including raw The characterization step consists in converting the emissions
resources or materials, energy by type, water, and emissions to air, and resources that are assigned to each of these impact categories
water and land by specific substance. into indicators using impact assessment models [18], also called
The analyzed system is typically differentiated into the pro- characterization models or LCIA methodologies. These models aim
cesses of the foreground system and those of the background sys- to connect, each LCI to its potential environmental impacts or
tem [15]. The foreground system is defined as those processes of damages thanks to characterization factors [19]. A characterization
the system that are under control of the decision maker involved in factor is a substance-specific factor for expressing the impact from
the study [16]. The background system comprises those processes the particular elementary flow in terms of the common unit of the
that are operated as part of the system but that are not under direct category indicator [18]. These characterization factors are based on
control of decisions analyzed in the study. environmental mechanisms that link (man-made) interventions to
There are two main approaches to data collection: process- a set of areas of protection. The end of the environmental mecha-
based LCA and Input-Output LCA (IO-LCA). Process-based LCA is a nism is called the endpoint. A point positioned halfway along the
bottom-up process analysis where the system is modelled by environmental mechanism can be chosen as an indicator, often
means of its specific information whereas IO-LCA uses economy or referred to as the midpoint [20]. Thus there are different types of
industry sector wide inventory data and break these down in a top- impact indicators; (i) inventory indicators that are direct LCI results,
down process. A third approach, called hybrid-LCA consists in a (ii) midpoint and endpoint indicators that are assessed through
combination of the two former approaches. characterization models.
Two main modelling principles are used in LCA practice that The criteria used for analyzing the LCIA phase of the case studies
influence the data required to perform the analysis: attributional include the chosen LCIA methodology, the list of selected impact
and consequential modelling. Attributional LCA depicts the po- and damage categories and the presence or absence of normaliza-
tential environmental impacts that can be attributed to the tion and weighting, which are optional elements. Specific LCIA
analyzed system over its life cycle, and this system is embedded results are collected and compared among the studies for relevant
into a static technosphere (i.e. the foreground system do not modify and available impact categories, i.e., primary energy consumption
the background system in which it is embedded). Consequential (PEC) and climate change (CC).
modelling aims at identifying the consequences that a decision in
the foreground system has for other processes and systems of the 3. Results
economy, both in the analyzed system's background system and on
other systems. In consequential LCA, the analyzed system is Following the selection process we came to a corpus of 14 pa-
embedded in a dynamic technosphere [15]. pers accounting for 21 case studies. The key points of analysis of the
The analysis of the Life Cycle Inventory (LCI) phase of reviewed reviewed papers are summarized in Table 2.
papers deals with the procedures used to collect foreground and
background data (i.e., source of data) and the completeness of the
inventories. 3.1. Phase 1: goal and scope definition

3.1.1. Goal of the studies


2.2.3. Criteria for LCA phases 3 and 4 e life cycle impact assessment Reviewed papers are almost equally split in two main types of
and interpretation studies that we propose to name “neighborhood development
According to ISO [17], Life Cycle Impact Assessment (LCIA) sustainability studies” and “urban sustainability studies”. “Neigh-
proceeds through two mandatory steps; (i) selection of impact borhood development sustainability studies” are LCA of real
categories and classification, and (ii) characterization. neighborhood development or renovation projects that are con-
The selection of impact categories and classification consists in ducted either for decision-support or eco-design. On one hand,
selecting the impact categories which are of relevance to the study, decision-support studies are conducted at an early stage of the
and assigning the elementary flows from the inventory to these development of a project and aim at assessing its environmental
impact categories according to the substances‘ ability to contribute impacts in order to define the project major orientations. For
to different environmental problems [18]. Impact categories instance, in order to provide decision-support for urban planning
considered in a LCIA include generally climate change, ozone Riera Perez and Rey [21] assessed two densification scenarios for an
depletion, eutrophication, acidification, human toxicity, respiratory existing neighborhood to investigate the dynamism between sus-
inorganics, ionizing radiation, ecotoxicity, photochemical ozone tainability and local urban planning regulations that determine
formation, land use, water deprivation and resource depletion. densification opportunities. On the other hand eco-design studies
168
Table 2
Key points of analysis of reviewed papers.

Reference Country City/Region Case study Scenarios Size (km2) Type of Number of inhabitants Residential FU Boundaries LC steps
per cases neighborhood and jobs density
(inhab./km2)

Stephan et al., 2013 Australia Melbourne 1 21 1.4 100% residential 700 inhab. 500 - km2 neighborhood - B - Cons.
- inhabitant - OS (roads) - Op.
- N (power lines, water
distribution, gas distribution,
sewage)
- M (passenger car, train)
Nichols and USA Austin 4 1 13.1 91% residential 4865 inhab. 370 - inhabitant - B - Cons.
Kockelman, 2014 1.7 91% residential 3394 inhab. 2040 - OS (roadways, sidewalks, - Op.
2.2 71% residential 4939 inhab. 2200 driveways, parkings)

M. Lotteau et al. / Building and Environment 93 (2015) 165e178


1.3 90% residential 7728 inhab. 5940 - N (water & wastewater
systems)
- M (passenger car, bus)
Norman et al., 2006 Canada Toronto 2 1 NA 100% residential NA 27 000 - inhabitant - B - Cons.
5700 - m2 of living - OS (roads) - Op.
space/inhabitant - N (utilities)
- M (private automobiles and
public transit)
2
Riera-Perez et al., 2013 Switzerland Lausanne 1 4 0.07 97% residential 1029 inhab. 14 700 - m ERA (Energy - B - Cons.
28 jobs Reference Area*) - M - Op
- Decons.
Davila and Reinhart, USA Cambridge 1 9 NA 100% residential NA NA - m2 floor area - B - Cons.
2013 - Op
- Decons.
Cherqui, 2005 France La Rochelle 1 2 0.02 100% residential 650 inhab. 32 500 - neighborhood - B - Cons.
- Op
- Decons.
Li and Wang, 2009 China Beijing 1 1 4.42 100% residential 152 000 inhab. 34 400 - household - B - Cons.
- M - Op
- Decons.
Trigaux et al., 2014 Belgium NA 1 4 NA 100% residential NA NA - m2 floor area - B - Cons.
- OS (roads) - Op
- Decons.
Forsberg, 2003 Sweden Hammarby 1 2 NA 100% residential NA NA - inhabitant - B - Cons.
€ stad
Sjo 767 apartments - OS (roads, parks) - Op
- N (energy production plants,
sewage treatment plants)
- M
Colombert et al., 2011 France Paris 1 4 0.15 Mixt NA NA - neighborhood - B - Cons.
36 670m2 dwellings - OS (roads, parkings, - Op
40 000m2 offices green spaces) - Decons.
6 500m2 activities - N (water & wastewater,
6000m2 retirement home district heating)
1 000m2 shops
15 000m2 public equip.
M. Lotteau et al. / Building and Environment 93 (2015) 165e178 169

aim at assessing and comparing several scenarios in terms of


- Decons.

- Decons.

- Decons.

- Decons.
masterplanning [22,23] or other design alternatives in terms of
- Cons.

- Cons.

- Cons.

- Cons.
buildings and public spaces features [24e26]. Furthermore Her-
- Op

- Op

- Op

- Op
fray et al. [27] proposes the LCA of two reference eco-districts in

heating, waste management)


Freiburg, Germany, in order to provide best practice references for
eco-design. “Urban sustainability studies” are LCA of real existing

- N (water, sewage, district

- N (water, sewage, district


- OS (roads, green spaces)

- OS (roads, green spaces)

- OS (roads, green spaces)


neighborhoods or model neighborhoods derived from real exist-
- OS (roads, parkings,

ing buildings [28,29] that aim at studying some key urban sus-
- N (water, sewage,

- N (water, sewage,
district heating)

district heating)

tainability issues and build knowledge to feed urban policy


green spaces)

making. The issue of urban density and urban sprawl is the main
topic of interest [21,28e32].
heating) Among reviewed papers only two papers [21,23] use LCA to
address environmental issues in the frame of a more compre-
- B

- B

- B

- B

hensive sustainability assessment (i.e. addressing the three pillars


of sustainability). Other papers focus on environmental issues and
- neighborhood

- neighborhood

- neighborhood

especially on energy issues. 35% of papers deal with energy and


GHG only [21,28e31].

3.1.2. Scope and functional unit


NA

Reviewed case studies are highly heterogeneous. The neigh-


borhood scale is not well-defined in terms of size and case studies
range from 0.02 km2 to 4.42 km2. Case studies account for
16 400
10 100

different types of neighborhoods. Some cases are a hundred


9340

NA

NA

percent residential neighborhoods with both individual houses


and collective housing, others include offices and shops (mixt
240 occupants in a school
734 occupants in offices
432 occupants in shops

neighborhoods). The range of residential densities is extremely


broad as well, with densities ranging from 370 to 34 400 inhab./
60 000m2 dwellings

km2. Almost all the studies are located in developed countries,


45 000m2 offices
15 000m2 office

600 dwellings

except one in China.


394 inhab.

887 inhab.

The choice of the functional unit (FU) is heterogeneous as well


125 jobs

and most of the time the FU is not clearly defined. Five papers
clearly define the FU as “the neighborhood” as a whole with its
NA

NA

specific features and services provided (such as housing provision


63% residential

39% residential

or job provision) [22,23,25e27]. For instance in Herfray (2011)


[26] the FU is « a neighborhood with 887 inhabitants, 734 em-
ployees in offices, 432 employees in shops and 240 students,
Mixt

Mixt

Mixt

Mixt

located near Paris and considered over a 80 years lifespan”. Other


papers use either per capita or spatial functional units. As for per
capita FUs, three papers define the FU as “the inhabitant”
[28,30,31] and Li and Wang [32] define the FU as “the household”.
0.024
0.039

0.095

0.08
NA

Four studies use spatial FUs such as m2 of neighborhood [30], m2


of living space per inhabitant [28], m2 of Energy Reference Area
(i.e. m2 of heated floor) [21] or m2 of floor area [29]. Eventually
Stephan et al.[30] and Norman et al. [28] use both per capita and
1

spatial FUs to improve comparability between case studies or


scenarios. Stephan et al. [30] state that the simultaneous use of
spatial, per capita and absolute units is highly relevant to a full
understanding of urban density effects. For instance [28], found
2

^t 1

that a low-density neighborhood used around 2 to 2.5 times more


Meudon La Fore
e
Marne La Valle

energy than a high-density neighborhood on a per capita basis,


but only 1 to 1.5 as much energy on a per “unit of living space”
Freiburg

basis. Indeed the use of spatial unit measures the intensity of


Lyon

resources use and the impact intensity of an urban project.


Whereas the use of per capita units captures social differences
such as the house size and allows for the assessment of the effi-
Germany

France

France

France

ciency of use of resources and the impacts intensity of the


population.
Peuportier et al., 2006

 and Lespinoy,

3.1.3. Scope: boundaries, life cycle steps


Herfray et al., 2011

and Roux, 2013

The LCA model of a neighborhood is complex. Popovici (2006)


Herfray, 2011 þ

[13] for instance proposes a comprehensive model in which the


Peuportier

neighborhood is composed of three elements; buildings, open


2011
Trocme

spaces and networks and the life cycle steps are the construction
phase including the production of the neighborhood constitutive
elements and the construction of the neighborhood, the use
170 M. Lotteau et al. / Building and Environment 93 (2015) 165e178

Fig. 1. Neighborhood life cycle steps and associated contributors at both building and neighborhood scale.

phase, the renovation phase including the maintenance of the mobility field (i.e. passenger car, train and bus).
neighborhood and the demolition phase including the demolition This heterogeneity in the definition of the neighborhood system
of the neighborhood and management of demolition waste. In this to be assessed is further strengthened by a third level of hetero-
framework, the daily mobility is included in the use phase of the geneity regarding the contributors that are taken into account for
neighborhood on the same level as energy consumption for heating each physical element. The building field presents the highest
or water use. For the purpose of this review we propose a slight number of possible contributors and these contributors are not
adaptation of this model. To define system boundaries we use the evenly accounted for in the reviewed papers. For the use phase
four fields of the built environment as defined in Section 2.1: only, buildings analysis can include various energy consumption
buildings (B), open spaces (OS), networks (N) and mobility (M). As items (heating/cooling, ventilation, domestic hot water, specific
for life cycle steps, we define i) the Construction phase (raw ma- electricity, appliances, cooking, etc.) but also other items such as
terials extraction, manufacturing, buildings and infrastructures water use or domestic waste production.
construction), ii) the Use phase (operation and maintenance), and On the other side there are some shared choices regarding
iii) the Deconstruction phase (end of life). For each life cycle phase system boundaries selection. For instance, utility networks are al-
there are building scale contributors and neighborhood scale con- ways modelled as “passive elements” meaning that network ma-
tributors (i.e. related to OS, N and M). The neighborhood life cycle terials are the only contributors to this field. This comes from the
steps and main associated processes encountered in reviewed pa- fact that utility consumption (electricity, water, gas) are contribu-
pers are summed up in Fig. 1, adapted from Ref. [30] for the purpose tors to the buildings and open spaces fields. Another common point
of this review. is that expenditures such as households’ activities outside the area
of the neighborhood (work, school etc.) or households' consump-
tion that is not directly related to the buildings or the daily mobility
3.1.3.1. System boundaries. Within the reviewed papers there is a
(e.g. food, clothing, leisure or travel) are not taken into account.
great heterogeneity in the selection of system boundaries which is
Indeed these expenditures are expected to be quite comparable
always linked to the specific goal of the study. First of all, the four
across the same households independently of the neighborhood
fields of the built environment are not evenly taken into account.
they are living in Ref. [31].
Only four papers account for the four fields of the neighborhood
The Physical elements and associated contributors that are
system. Five papers deal with buildings, open spaces and networks,
encountered within the reviewed papers are summarized in
two papers focus on buildings and mobility, one paper focuses on
Table 3.
buildings and open spaces and two papers address buildings only.
Furthermore each field is covered in different ways from one
study to another. Depending of the studies different physical ele- 3.1.3.2. Life cycle steps considered. Concerning life cycle steps, all
ments are taken into account or neglected for each field of the built the studies include the operational phase. Seventy percent of pa-
environment; i) different types of roads, sidewalks, parking and pers take into account all three phases, i.e. construction, operation
green spaces for the open spaces field, ii) electricity, water, and deconstruction, whereas remaining 30% do not account for
wastewater, gas, district heating or waste management networks deconstruction. However, construction and operation phases can
for the Network field, and iii) different transportation means for the be accounted for in different ways, for instance Norman et al. [28]
M. Lotteau et al. / Building and Environment 93 (2015) 165e178 171

Table 3
Possible physical elements and associated contributors for the different fields of the built environment.

Built environment fields Buildings Open spaces Networks Mobility

Physical Elements - buildings - roads - electricity network - passenger car


- sidewalks - water network - train
- parking - wastewater network - bus
- green spaces - gas network
- district heating network

Contributors - construction materials - construction materials - construction materials - vehicle manufacturing


- construction waste - construction waste - construction waste - daily mobility
- maintenance materials - maintenance materials - maintenance materials
- maintenance waste - maintenance waste - maintenance waste
- heating - public lighting - demolition waste
- cooling - water for maintenance of public spaces
- domestic hot water - public wastes
- ventilation - demolition wastes
- specific electricity
- cooking
- drinking water
- domestic wastes
- wastewater
- demolition waste

indicate that building and infrastructure construction and main- analysis of plans. It can be noticed that no paper mentions fore-
tenance are not included in the study boundary due to a lack of ground data related to the building construction works and civil
available data and their relative insignificance to the overall life works. In most of the studies, materials LCI data are not available.
cycle as reported by previous buildings LCA studies. Davila and
Reinhart [29] focuses on embodied energy across the entire life
3.2.1.2. Buildings' operation. Regarding buildings' operation, two
cycle and do not consider operational requirements. Furthermore,
types of energy consumption data are collected. On one hand all the
the authors also neglect both materials transportation and con-
reviewed papers take into account thermal energy requirements,
struction. Regarding the mobility field, only the operation phase is
i.e. energy consumption for either heating or cooling and domestic
accounted for in the reviewed studies, except in Ref. [30] that takes
hot water production, and provide explanation from where data is
into account vehicle manufacturing.
sourced or estimated On the other hand, there are all other (mostly
Life cycle phases taken into account for each study, as well as
electricity-based) energy uses that are not evenly accounted for
system boundaries in terms of fields of the built environment are
across the different studies and for which fewer data sources are
summarized in Table 2.
available.
Thermal energy requirements are either obtained through field
3.2. Phase 2: life cycle inventory (LCI)
surveys [32,35], through average national data [28] or most of the
time through modelling. Different types of modelling approaches
3.2.1. Foreground data
are possible. A majority of papers rely on buildings thermal simu-
When analyzing the built environment at neighborhood scale,
lation; Peuportier et al. [25], Herfray et al. [27], Colombert et al.
the major foreground processes (i.e. specific to the studied system)
[22], Cherqui (2005) [23] and Trocme  and Le pinoy [33] use the
are; (i) consumption of construction materials for buildings, open
same building LCA tool called EQUER, that relies on dynamic
spaces and networks, (ii) energy and water consumption in build-
thermal simulation, whereas in Stephan et al. [30] and Trigaux et al.
ings, networks and open spaces, and (iii) transport requirements
[34] building energy simulation is based on a static heat transfer
for daily mobility. Most of case studies are either large existing
model. Other non-simulation modelling approaches are also used.
neighborhoods [30e32], neighborhood projects [22,23,25,26,33] or
Nichols and Kockelman [31] use a statistical model of Tir-
model neighborhoods [28,29,34], what makes really difficult or
umalachetty et al. [36] that controls for a number of climatic, de-
prevents the systematic collection of detailed site-specific data (e.g.
mographic, and built environment explanatory variables. Riera
detailed composition for each building or actual energy consump-
Perez and Rey [21] calculate a normative thermal energy demand
tion data for each building).
(Swiss society of Architects and engineers norm 380/1) which is
then adapted depending on the energy standard targeted.
3.2.1.1. Construction materials. Construction materials quantities
are most of the time obtained through direct analysis of the
neighborhood plan or masterplan (for neighborhood projects). In 3.2.1.3. Daily mobility. Transport requirements for daily mobility
some studies, the authors also used the detailed plans of few typical are also sourced differently. Norman et al. [28] and Stephan et al.
buildings. In Norman et al. [28], Stephan et al. [30] and Nichols and [30] respectively rely on regional and local averages. Riera Pe rez
Kockelman [31], which study large existing neighborhoods, build- and Rey [21] apply a normative method (based on Swiss Society of
ings are represented by few typical buildings for which detailed Architects and Engineers norm 2039) that weights the Swiss
data was available. For smaller projected-neighborhoods, specific average energy consumption for mobility by several correction
data for each buildings can be obtained from buildings' plans factors related to the context and facilities. Nichols and Kockelman
(depending the progress of the project). Nevertheless the analysis [31] developed statistical models that rely on household de-
of plans only gives access to square meters of buildings of each type, mographics and physical characteristics to estimate the daily
then the studies have to infer the choice of materials in accordance mobility requirements.
with the targeted energy efficiency for projected-neighborhoods, Foreground data sources for materials quantities, energy and
or from statistical data for existing neighborhoods. Open spaces water use in buildings as well as transport requirements for daily
and networks materials quantities are also obtained through the mobility are presented in Table A-1 in supplementary material.
172 M. Lotteau et al. / Building and Environment 93 (2015) 165e178

3.2.2. Background data presence of additional “land use” indicators in Refs. [22,23] that
Background processes are processes that are not specific to the reflect damages to ecosystems due to land transformation and land
studied system, such as extraction of raw materials and production occupation.
of construction products, or production of water and electricity. In Trigaux et al. [34] also selected a broad range of impact and
the reviewed papers few information is available regarding back- damage categories; global warming, depletion of the stratospheric
ground data and their sources. ozone layer, acidification of land and water sources, eutrophication
For construction materials, 66% of the papers use process-based freshwater and marine, photochemical oxidant formation, abiotic
LCA databases. Seven papers use the ecoinvent database through depletion of non-fossil resources, abiotic resource depletion po-
the use of building-LCA tools or neighborhood-LCA tools such as tential, fossil resources, human toxicity, particulate matter forma-
EQUER (and latest version novaEQUER) in Refs. [22,23,25e27,33] or tion, ionizing radiation, ecotoxicity, land use.
Lesosai and Eco-Bat in Ref. [21]. Davila and Reinhart [29] use the ICE Forsberg (2003) [35] analysis is based on a tool called the
(Inventory of Carbon Emission) database from the University of Environmental Load Profile and that focuses on fewer impact cat-
Bath, as a source for embodied energy data. Two papers rely on egories; extraction of non-renewable resources, water use, global
Input-Output (IO) analysis to estimate the energy use and GHG warming potential, acidification potential, eutrophication potential,
emissions associated with the manufacture of construction mate- photochemical ozone depletion potential, radioactive waste and
rials; Norman et al. [28] uses an economic inputeoutput life-cycle use of hazardous compounds.
assessment (EIO-LCA) model that employs the 1997 (most recent Li and Wang [32] focus on five impact categories; global
year available) EIO-LCA model for the United States and Stephan warming potential, acidification potential, nutrient and eutrophi-
et al. [30] uses an IO-based hybrid analysis technique developed by cation potential, photochemical oxidant creation potential and
Ref. [37] with related energy coefficient for Australia. Eventually, in solid waste.
Nichols and Kockelman [31] embodied energy is estimated for Most of the time, the underlying LCIA methods are not docu-
buildings and infrastructure materials based on a meta-analysis of mented in reviewed papers. In Ref. [25], LCIA methods used are; (i)
detailed life-cycle analyses. CML-IA [39] for climate change, depletion of abiotic resources,
As for energy vectors related background data, the papers using acidification potential, eutrophication potential, photochemical
the EQUER tool rely on ecoinvent database. In Riera Pe rez and Rey oxidation (summer smog) and malodorous air, (ii) Eco-Indicator 99
[21] the conversion factors from final energy to non-renewable [40] for ecosystem quality, aquatic ecotoxicity and human health,
energy (NRE) and global warming potential (GWP) are drawn and (iii) Swiss Ecoscarcity [41] for cumulative energy demand,
from KBOB LCA data (KBOB, 2012), which is a database of LCA of water consumption, waste creation, radioactive waste creation.
construction related materials and processes, based on the ecoin- Herfray (2011) [26]uses the same methods except that the author
vent data as well. In Refs. [28e30], conversion factors from final refers to CML-IA for cumulative energy demand, waste creation,
energy to primary energy and/or GHG emissions are sourced from radioactive waste creation, and does not indicate the method used
local reports. Specifically for transportation, Stephan et al. [30] for water consumption.
sourced energy and GHG intensities of the transport modes from In Trigaux et al. [34], LCIA methods are; (i) ReCiPe [20] Midpoint
IO analysis data [38]. for global warming, depletion of the stratospheric ozone layer,
Background data sources for materials production, energy vec- acidification of land and water sources, and ecotoxicity, (ii) ReCiPe
tors and transportation processes are presented in Table A-2. Endpoint for human toxicity, particulate matter formation, ionizing
radiation, and land use, and (iii) CML-IA for eutrophication,
3.3. Phase 3. life cycle impact assessment (LCIA) photochemical oxidant formation, and abiotic depletion of non-
fossil resources.
As previously explained, in a LCA, the emissions and resources This analysis of the LCIA step of the reviewed papers highlights
consumed that can be attributed to a specific product or activity are several points. First of all, 35% of the studies focus on energy related
compiled and documented in a LCI. An impact assessment is then criteria (primary energy consumption and climate change)
performed on the basis of the LCI. although LCA is a recognized multicriteria approach. Furthermore
some impact categories, such as air quality, mineral resources
3.3.1. Impacts and damages depletion or land use which are subjects of high concern for urban
Thirty-five percent of reviewed papers focus on primary energy sustainability, are seldom addressed in reviewed papers. Eventu-
consumption and/or climate change and exclude other impact ally, LCIA methods are not systematically documented and selected
categories [21,28,30]; account for both, whereas [29,31] account for LCIA methodologies are rather old. Only Trigaux et al. [34] rely on
primary energy consumption only. more recent LCIA developments such as ReCiPe or ILCD and no
Other studies address a broader range of impact categories. In paper uses regionalized LCIA methods.
case studies using the EQUER software [22,23,25e27] the following
impacts and damage categories are assessed; 3.3.2. Facultative steps; normalization, weighting
In addition to LCIA mandatory steps, ISO [17] also indicates two
- Inventory indicators: cumulative energy demand (GJ), water facultative steps; normalization and weighting. In the normaliza-
consumption (m3), waste creation (teq), radioactive waste cre- tion step, the different characterized impact scores are related to a
ation (dm3) common reference, e.g. the impacts caused by one person during
- Impact categories assessed with midpoint indicators: climate one year. In the weighting step, a ranking and/or weighting is
change (teqCO2), depletion of abiotic resources (kgSb or performed of the different environmental impact categories.
dimensionless), acidification potential (kgeqSO2), eutrophication Among reviewed papers, two papers only go through a
potential (kgeqPO4), oxidation (summer smog) (kgeqC2H4) and normalization step. In Herfray et al. [27] impact indicators are
malodorous air (Mm3) normalized in equivalent inhabitants-year, using French references
- Damage categories (endpoint indicators): ecosystem quality (e.g. 8.7 teqCO2 emissions per person and per year). In Li and Wang
(PDF.m2.year) and human health (DALY) [32] results are normalized with following average values for
Beijing.
Nevertheless, there are some variations as for instance the Two papers go through a weighting step. Li and Wang [32]
Table 4
LCIA results e primary energy consumption (PEC) and climate change (CC).

Reference FU neighborhood.year FU inhabitant.year FU m2(neigh.).year FU m2 (floor area).year


2
Location Case study Size (km ) Residential Boundaries LC steps Analysis Type of PEC TJ CC kteqCO2 PEC GJ CC teqCO2 PEC MJ CC kgeqCO2 PEC MJ CC kgeqCO2
density duration neighborhood
(inhab./km2)

Stephan et al., 2013 Melbourne 1 1,5 500 - B - Cons. 100 100% residential 73,0 5,4 104,3 7,8 48,7 3,6 1664,8 123,8
(Australia) - OS - Op.
- N
- M
Nichols and Austin (USA) 4 13.1 370 - B - Cons. NA 91% residential 559,5 115,0 42,7
Kockelman, 2014 1.7 2040 - OS - Op. 91% residential 373,3 110,0 219,6
2,2 2200 - N 71% residential 479,1 97,0 217,8
1,3 5940 - M 90% residential 556,4 72,0 428,0

M. Lotteau et al. / Building and Environment 93 (2015) 165e178


Norman et al., 2006 Toronto (Canada) 2 NA 27 000 - B - Cons. 50 100% residential 40,1 3,3 936,4 77,7
NA 5700 - OS - Op. 86,0 8,6 1068,0 107,3
- N
- M
Forsberg, 2003 € stad 1
Hammarby Sjo NA NA - B - Cons. NA 100% residential 2,8
(Sweden) - OS - Op
- N
- M
Riera-Perez et al., 2013 Lausanne 1 0,07 14 700 - B - Cons. 60 97% residential 71,9 4,1 69,8 4,0 1026,5 59,2 1216,8 70,2
(Switzerland) - M - Op
- Decons.
Davila and Cambridge (USA) 1 NA NA - B - Cons. 100 100% residential 531,7
Reinhart, 2013 - Op
- Decons.
Cherqui, 2005 La Rochelle (France) 1 0,02 32 500 - B - Cons. 80 100% residential 8,6 0,4 13,2 0,6 427,5 19,4 438,5 19,9
- Op
- Decons.
Colombert et al., 2011 Paris (France) 1 0,15 NA - B - Cons. 80 Mixt 103,4 2,3 689,1 15,3 1159,2 25,8
- OS - Op 36 670 m2 dwellings
- N - Decons. 40 000 m2 offices
6 500 m2 activities
1 000 m2 shops
15 000 m2 public
equipment
Herfray et al., 2011 Freiburg 2 0,024 16 400 - B - Cons. 80 Mixt 5,7 0,4 14,4 1,1 236,8 17,6
(Germany) 0,039 10 100 - OS - Op 63% residents 0,0 0,5 0,0 1,2 0,0 12,5
- N - Decons. 20% employees
17% students
Herfray, 2011 Marne La Vallee 1 0,095 9340 - B - Cons. 80 Mixt 15,1 17,0 158,0 281,3
(þPeuportier (France) - OS - Op 39% residents
and Roux, 2013) Freiburg 2 - N - Decons. 51% employees 18,6 21,0 550,8 346,9
(Germany) 10% students 5,0 5,7 140,1 74,0
Peuportier et al., 2006 Lyon (France) 1 NA NA - B - Cons. 80 Mixt 63,0 0,8 840,0 10,8
- OS - Op 60 000 m2 dwellings
- N - Decons. 15 000 m2 offices

173
174 M. Lotteau et al. / Building and Environment 93 (2015) 165e178

Fig. 2. Primary energy consumption and climate change; min, Q1, median, Q3 and max values.

applied weighting factors to the five impact categories assessed, boundaries in terms of both life cycle steps (Cons. and Op.) and
but did not discuss the choice of the weighting factors. In Trigaux fields of the built environment (B., OS, N. and M.). For these resi-
et al. [34] a weighting is proposed by means of monetary valuation. dential case studies, the primary energy consumption ranges from
The characterization values for each individual environmental in- 40 to 115 GJ/inhab./year with an average value of 89 GJ/inhab./year
dicator were aggregated by means of the environmental cost (sample made of 7 values). And GHG emissions range from 3.3 to
method. For each individual environmental indicator, the charac- 8.6 teqCO2/inhab./year with an average value of 6.6 teqCO2/inhab./
terization values are multiplied by a monetization factor (e.g.: X kg year (sample made of 3 values). However analysis duration are
CO2 equivalents times Y EURO/kg CO2 equivalents) so as to come to really different; 100 years in Ref. [30], 50 years in Ref. [28] and not
a single environmental score. communicated in Ref. [31].
In the case of mixt neighborhoods, 7 case studies from 5 papers
3.3.3. LCIA results [22,25e27] account for the same life cycle steps (Cons., Op. and
Because of the diversity of the case studies (in terms of neigh- Decons.), the same fields of the built environment (B., OS. and N.)
borhood typology) and the variety of functional units used in the and use the same analysis duration (80 years). For these case
reviewed papers it is not straightforward to provide a summary of studies, the primary energy consumption ranges from 0 to 21 GJ/
LCIA results in a format allowing comparison between case studies. inhab./year with a median value of 12 GJ/inhab./year (sample made
As an attempt to do so, whenever it was possible with the data of 5 values).
available we calculated both primary energy consumption (PEC)
and climate change (CC) indicators for the entire neighborhood as 3.4. Phase 4: interpretation
well as per inhabitant, per m2 floor area and per m2 of neighbor-
hood. These results are provided in Table 4 for residential (or mostly 3.4.1. Contribution analysis
residential) case studies (upper part of the table) and for mixt case A total of nine studies provide a contribution analysis either
studies (lower part of the table). among different fields of the built environment (B., OS., N. and M.)
When interpreting these figures, one has to keep in mind that or among life cycle steps (Cons., Op., Decons.). Considering the
the system boundaries and analysis duration vary across case heterogeneity in the choice of system boundaries as well as in the
studies. When considering the entire sample of case studies for display of results, it is not possible to sum up these contribution
which LCIA results are available, primary energy consumption analysis in one consistent way and draw general conclusions.
ranges from 0 to 115 GJ/inhab./year with a median value of 55 GJ/ Instead and for illustration purpose, we propose hereunder a brief
inhab/year (sample made of 14 values). And GHG emissions range summary of the most complete contribution analysis among
from 0.6 to 8.6 teqCO2/inhab./year with a median value of 3.1 reviewed papers.
teqCO2/inhab./year (sample made of 8 values). Thus correspond- Stephan et al. [30] demonstrate that neighborhood operational
ingly to the heterogeneity of the case studies and of their modelling energy is the major contributor to the entire lifecycle and accounts
in reviewed papers, results covers three orders of magnitude for for 39.4% of primary energy consumption (PEC) and 42.4% GHG
primary energy consumption and two orders of magnitude for emissions. The second contributor is mobility, accounting for 33.6%
climate change on a per capita basis. PEC and 36% GHG emissions, followed by embodied energy ac-
These results, as well as primary energy consumption and GHG counting for 26.0% PEC and 21.6% GHG emissions. Regarding
emissions per (m2 floor area$year) are summarized in Fig. 2 for the embodied energy, it is to be noticed that in this study, the embodied
whole sample of case studies. energy of transport infrastructure (roads) is included in the
Seven residential case studies from 4 papers [28,30,31] are “Mobility” contributor. This study also underlines the important
somewhat comparable because they use the same system contribution of infrastructure (roads and networks) to the
M. Lotteau et al. / Building and Environment 93 (2015) 165e178 175

neighborhood life cycle embodied energy (16.9%) and points out - the relative insignificance of the deconstruction phase except
the relative insignificance of the construction works with only 1.3% for the “waste creation” impact category
contribution to the embodied energy (0,03% to the total life cycle
energy).
On average for the four neighborhoods studied by Nichols and 3.4.2. Sensitivity check
Kockelman [31], building sources account for 53% of PEC, mobility Only two papers directly address uncertainty issues by
sources (operational and embodied on transport infrastructure) completing a sensitivity analysis. In Stephan et al. [30] uncertainty
account for 45.5% PEC and infrastructure sources (water, waste- and variability in the data are taken into account through interval
water and lighting) account for only 1% PEC. analysis. This technique provides a certain range around the
In a comparison of residential developments of different den- nominal value in which the actual figure may lie. The author set the
sities Norman et al. [28] conclude that mobility contributes far uncertainty on the embodied energy to ±20% and ±50% for process
more significantly to overall energy use and GHG emissions in a data and inputeoutput data respectively. The variability in opera-
low-density development context than in a high-density develop- tional energy figures is set to ±20%. According to this interval
ment context and that relatedly, building operational impacts are analysis, the shares of embodied, operational and transport energy
more significant in the high-density case. The authors also high- lie within, 15.3e39.4%, 28.7e52.5%, and 23.9e46% respectively.
light that material production accounts for only approximately 10% Thus, the uncertainty for embodied, operational, and transport
of total energy use and GHG emissions. On a per capita basis, they energy is estimated to be ±47%, ±29%, and ±32% respectively.
found that the majority of GHG emissions in both cases (low and In Nichols and Kockelman [31] the author computed elasticities
high density) result from mobility while the majority of energy use to estimate the primary energy consumption response to changes
results from building operations. in the built environment or user behavior. Elasticities illustrate the
In a comparison of four scenarios based on different types of relative sensitivity of primary energy consumption to specific var-
residential buildings Trigaux et al. [34] shows that the contribution iables. Among a broad set of variables, elasticities of greatest sig-
of road infrastructure ranges from 1% to 6% of the total impact nificance are elasticities of primary energy consumption to
(roads and buildings). residential unit size, population density, and percentage of single
In studies of low energy neighborhoods focusing on buildings, family houses.
open spaces and networks Herfray et al. [27] and Herfray (2011) In addition of above-mentioned sensitivity analysis, the evalu-
[26] respectively found that in terms of primary energy consump- ation of scenarios (done in 7 case studies) can also be considered as
tion, the construction phase accounts for 7% and 13,5%, the opera- sensitivity analysis. This is done by comparing with basis scenario
tion phase accounts for 92% and 85% and the deconstruction phase values and by showing the increase or decrease for each impact
accounts for 1% and 1.5%. Here again we see the dominance of the category. This review does not collect LCIA results from the pro-
operation phase and the relative insignificance of the deconstruc- spective scenarios.
tion phase. Nevertheless Herfray et al. [27] also analyzes a positive
energy neighborhood and as expected results are really different; 4. Methodological challenges & research needs
the construction phase accounts for 41.5% of PEC what becomes
comparable to the contribution of the operation phase (53.5%). It 4.1. System multifunctionality and functional unit definition
should also be noticed that compared to low energy neighbor-
hoods, the increased contribution of the construction phase for the Because of the heterogeneity, complexity and multi-
positive energy neighborhood is both a relative and absolute in- functionality of neighborhoods, it appears difficult to define a
crease. Regarding the deconstruction phase, it is only a relative unique and comprehensive functional unit (FU) to compare
increase. neighborhoods or scenarios for a same neighborhood. Herfray et al.
Cherqui (2005) [23] conducted a contribution analysis according [27] states that the definition of the FU depends on the goal of the
to life cycle steps for the 11 impact categories he considered. This study but insists on the fact that some common aspects have to be
study also highlights the dominance of the operation phase and the included in order to compare alternatives; the neighborhood
relative insignificance of the deconstruction phase among all functions (residential, work, leisure, etc.), some quantities (m2 of
impact categories except for waste creation for which the major housing, offices, shops, public spaces, number of inhabitants and
contributor is indeed the deconstruction phase. workers, etc.), quality criteria for the neighborhood functions
As it has already been highlighted in critical reviews of LCA for (thermal, visual or acoustic comfort, quality of life, etc.) and a
buildings [6e8], it is not possible from this review for neighbor- duration of analysis.
hoods to extract universal general results from the reviewed papers Herfray (2011) [26] whose objective is to compare neighbor-
because the neighborhoods that have been studied are very hoods development scenarios, go through the adaptation, of
different in terms of size, density, location, etc. and also because of different case studies in order to improve their comparability. In a
the heterogeneity in methodological choices (e.g. system bound- comparison of an actual neighborhood with two reference neigh-
aries, analysis duration). Nevertheless some trends can be identi- borhoods, the author proposes a very detailed functional unit as
fied from the above-mentioned contribution analysis: such; “a neighborhood with 887 inhabitants, 734 tertiary em-
ployees, 423 employees in shops and 240 students, situated in Ile
- buildings, followed by mobility are the major contributors, in de France, and considered on a 80 years lifetime”. And a method-
terms of fields of the built environment, to primary energy ology is proposed to adapt the reference neighborhoods so that
consumption and GHG emissions they match with this FU defined from the actual scenario in terms
- contributions to primary energy consumption and GHG emis- of proportion of users of each type, without changing their initial
sions of buildings' operation, mobility and neighborhood densities.
embodied energy can be of the same order of magnitude This idea of a “multi-dimension” functional unit is further
- the predominance of the operation phase developed by Loiseau (2014) [42] at the scale of the territory who
- the significance of the production phase that even become of the proposes an adaptation of the LCA framework to environmental
same order of magnitude as the operation phase for highly en- assessment in land planning. As in the case of neighborhoods, the
ergy efficient neighborhoods author identifies the territory's multi-functionality as a major
176 M. Lotteau et al. / Building and Environment 93 (2015) 165e178

bottleneck in the application of LCA to its environmental assess- precise enough to support the design process regarding buildings
ment because it makes the functional unit definition step prob- energy efficiency.
lematic. A revised approach to functional unit and reference flow is Thus in the field of decision-support for neighborhood planning,
proposed in order to overcome this issue. In conventional LCA, the there is a clear research need for the development of approaches
reference flow is determined by the choice of the functional unit, that would enable the contextualization of the assessment of en-
whereas in the revised approach the author suggests to consider ergy consumption in buildings while remaining consistent with
the territory with a specific land planning scenario as the reference data availability in a project's early design phase. These approaches
flow and then, to determine the main territory's functions and should ideally take into account the influence of the master-
assess their performance in parallel with LCIA. Consequently two planning stage design levers, such as urban morphology, surface
kind of indicators are determined using this approach, (i) a vector of materials or vegetation, on buildings energy requirements. They
environmental impacts (LCIA phase of conventional LCA) and, (ii) a should enable to assess a masterplan on such issues as the potential
vector of land use functions provided by the territory for different use of incoming solar energy by buildings or the urban heat island
stakeholders. We believe that an adaptation of this approach to (UHI) effect for instance. This assessment should be performed at
neighborhood LCA should be more systematically investigated and the neighborhood scale and not at the individual building scale for
developed, considering functions such as; housing provision, jobs which data is not yet available. One research lead consists in
provision, accessibility to daily services, mobility, quality of life, etc. investigating the possible correlations between buildings' energy
Another approach to partially overcome the impossibility to demand and simple evaluation metrics. For instance, in a review of
define a unique FU and improve comparability among case studies early-design phase evaluation metrics for the solar potential of
consist in combining the use of three types of FUs (i.e. spatial, per neighborhood design, Nault et al. [45] investigate whether
capita and absolute) [28,30,31]. compactness and solar exposure levels can be used as performance
indicators when applied to neighborhoods. Srebric et al. [43] also
4.2. Contextualization identifies urban morphology parameters, such as the urban plan
area density, frontal area density and mean height of the buildings,
The contextualization of the LCA of a neighborhood is another as suitable for the development of generalized energy demand
major challenge in order to provide decision-makers with useful models at the neighborhood scale.
analysis, be it for the design of the neighborhood development Contextualization in the LCIA phase consists in recognizing
project or for urban policy making. The lack of consideration of the which environmental mechanisms are relevant in the local context
specific situation of a neighborhood is not only a limitation of LCA, of a neighborhood. A number of environmental mechanisms have a
but is clearly identified as a limitation of most of the existing ap- global scope, while others have a regional one [20]. This difference
proaches for sustainability assessment at neighborhood scale [21]. means that a particular environmental mechanism can have very
Indeed, a neighborhood, contrary to industrial products for important impacts in one region, but not in another. This can be
instance, is by nature a highly contextualized system. It is taken into account through the use of regionalized LCIA methods.
embedded in a specific environment, is subject to a specific climate, Several local and regional impacts categories are being contextu-
interacts with a specific technosphere, hosts users with culturally- alized: for example water deprivation [46], eutrophication [47],
driven behaviors, and can face local environmental issues (in land use [48]. However, most of the methods are not yet oper-
addition to global environmental issues). Thus the need for con- ationalized and efforts should be made to include them in the LCA
textualization of the analysis occurs at a number of levels in both of neighborhoods. Other impact categories such as those related to
LCI and LCIA. air quality which are particularly relevant when assessing the im-
Contextualization in the LCI phase requires using as much as pacts of mobility need further developments to be regionalized.
possible site-specific data for processes of both foreground and Also, there is need of tools coupling LCA and geographical infor-
background systems. As previously stated, a major contributor to a mation system (GIS) that enable the assessment of neighborhood at
neighborhood's life cycle impacts is the operational energy uses of the local scale.
buildings, which can be largely influenced by the building's urban
environment. Srebric et al. [43] proposes a review of the building 4.3. Temporal aspects in LCA
neighborhood properties influencing the energy and airflows. The
authors stress the fact that energy consumption patterns are highly LCA application to neighborhood environmental assessment
dependent on the surrounding urban neighborhood. For instance raises issues related to temporal aspects.
Gracik et al. [44] found that variations in local climate and airflows A first issue stems in the choice of the life cycle duration. A
around buildings can lead to a 17% reduction in air conditioners COP neighborhood is often being built over decades and is composed of
in an urban context (compared to a rural context) Within the long-lasting elements, with various lifetimes that cannot be pre-
reviewed papers different approaches are proposed to assess en- cisely anticipated. Furthermore, components such as buildings are
ergy consumptions in buildings. These approaches can be split in likely to undergo renovations or changes of use during their life-
two categories; (i) use of statistical data and, (ii) use of thermal time. Thus it is complex to determine a proper neighborhood life-
simulation (static or dynamic) at the building scale. Both ap- time or analysis duration, as well as specific lifetimes of sub-
proaches have their advantages and drawbacks. Thermal simula- components for which maintenance is anticipated, which is
tion at the building scale enables the contextualization of the necessary to conduct a LCA. Today there is no consensus in the
assessment of buildings' energy consumption as it can take into choice of analysis duration and reviewed papers propose lifetimes
account the local climate (irradiation, wind) as well as the in- ranging from 50 to 100 years. Furthermore, over such long periods,
teractions between buildings and the environment (e.g. through technologies are likely to evolve which could possibly modify the
shading or solar reflection). Nevertheless this type of analysis re- neighborhood impacts.
quires a large amount of time and data (regarding buildings and A second issue related to temporal aspects in LCA is the
their surroundings) that are not available at the beginning of a consideration of the dynamic behavior of some foreground and
neighborhood development project. On the other hand using sta- background processes. Considering foreground processes, we
tistical data is more consistent in terms of data availability with the already discussed the use of thermal dynamic simulation to assess
early stage of a neighborhood development project, but is not buildings' energy consumption. Considering background processes
M. Lotteau et al. / Building and Environment 93 (2015) 165e178 177

developments are ongoing regarding dynamic LCA aspects in order - When assessing the built environment at the neighborhood
to account for the temporal evolution of the electricity production scale key features of the neighborhood should be systematically
mix [26]. This improves the accuracy of the evaluation and is communicated to improve interpretability of the results and
particularly relevant for low-energy buildings and neighborhoods. comparability between studies; i.e. number of inhabitants,
Peuportier and Roux [24] reports that applying this dynamic LCA number of non-resident users, neighborhood area, total floor
model may lead to 30% difference compared to a standard LCA area and analysis duration.
practice considering an annual average electricity mix. Among - The multi-functional built environment LCAs should take
possible practical outcome are: the study of demand shifting stra- advantage of the adaptation of the LCA framework to territorial
tegies according to the temporal variation of impacts, increasing assessment [42]. This approach would consist in providing a
the passive solar fraction despite higher cooling load in order to new set of indicators in addition to potential environmental
minimize the impacts over one year, improve decision making impacts indicators. This new set of indicators would assess the
process for onsite electricity production [24]. performance of a neighborhood regarding functions to be
The dynamic LCA model developed by Herfray (2011) [26] ac- determined.
counts for the temporal evolution of the electricity production mix - For “neighborhood development sustainability studies”, the
by assessing hourly the electricity production mix and associated influence of the built environment (urban morphology, pres-
impacts related to different uses of electricity (heating, cooling, ence of vegetation, choice of surface materials) on buildings
domestic hot water, …), and including local production of renew- energy consumption should be accounted for. This element of
able energy). This dynamic LCA model enables the author to contextualization should be addressed with methodologies
investigate attributional LCA vs. consequential LCA. In the attribu- aligned with the data available at the masterplanning stage of a
tional methodology, each energy use at time t is related to the neighborhood development project.
global production mix at time t, whereas in the consequential - For such long-lasting and complex “objects” as neighborhoods,
approach, each energy use at time t is related to the production mix efforts should be made to systematically include uncertainty
induced by this particular energy use at time t. Results on a positive analysis related to long-term temporal evolution of key
energy neighborhood show a 20% reduction of life cycle energy parameters.
demand assessed with consequential LCA compared to an assess- - The temporal evolution of the electricity production should be
ment by attributional LCA. accounted for, especially for low energy neighborhoods
including local production of renewable energy.
5. Conclusions - LCIA developments now enable full and comprehensive multi-
criteria assessment of the built environment. Thus mono-
This paper reviews built environment LCAs at the neighborhood criterion approaches should be avoided in order to prevent
scale and provides a synthesis and analysis of their main method- pollution shifting.
ological choices. It shows that existing case study literature is scarce - A minimum set of impact and damage categories for neighbor-
and that methodological choices are highly heterogeneous at each hood assessment should be decided. It could include impact
phases of the LCA. Some papers analyze existing large neighbor- categories such as land use, air quality and non-fossil abiotic
hoods or modeled neighborhoods and use them as prototype resources depletion (in addition to a more classical set of impact
neighborhoods to study urban policy related issues such as envi- categories) that seem highly relevant for the assessment of the
ronmental impacts of residential density. Other papers apply LCA as built environment.
a decision-support for smaller neighborhoods masterplanning - As another dimension of contextualization, recent advances in
processes. As for the goal and scope definition phase, functional regionalized impact assessment models related to local and
unit definition, system boundaries selection and lifetime of the regional impact should be adapted to neighborhood LCA.
neighborhood vary a lot across studies. This lack of methodological
consensus impedes to provide a proper LCIA results comparison Eventually this review paves the way for future research with
among reviewed papers. Because of the scale of the neighborhood the aim of developing a standardized approach for assessing the
system, the LCI phase is a critical point and several strategies are environmental performance of the built environment at the
proposed by the reviewed papers to provide data for the main neighborhood scale, an essential issue as the neighborhood scale is
foreground processes, i.e. (i) construction materials quantities, (ii) the operational scale of urban planning.
buildings operational energy consumption and, (iii) transport re-
quirements. These strategies range from statistical data collection Acknowledgments
or meta-analysis of neighborhood sub-components LCAs, to
detailed simulations based on statistical or physical modelling. In The authors acknowledge the support of Nobatek (Technological
terms of LCIA, 35% of the papers focus on two impact categories Resources Center in the field of sustainable construction), and the
only, i.e. cumulative energy demand and climate change. Other French National Association for Technical Research (CIFRE
studies account for a dozen of classical impact categories. Really Convention 2014/0243). The authors also thank the members of the
few papers address air quality or land use whereas these are highly CyVi-ISM research group (Life Cycle Assessment and Sustainable
relevant impact categories for urban built environment assessment. Chemistry www.ism.u-bordeaux1.fr/spip.php?rubrique1130) for
One paper only uses the most recent LCIA methodologies and no helpful discussions.
paper uses regionalized LCIA methodologies. Regarding the results
interpretation phase it can be noticed that two papers only provide Appendix A. Supplementary data
some kind of sensitivity check.
As all reviewed papers do, this review highlights the fact that Supplementary data related to this article can be found at http://
LCA application to neighborhood assessment is not straightfor- dx.doi.org/10.1016/j.buildenv.2015.06.029.
ward. Each of the reviewed paper proposes its own adaptation of
LCA methodology at the neighborhood scale and this review References
permitted to highlight recommendations and challenges. These are
summarized below: [1] United Nations, Population, Development and the Environment 2013, 2013.
178 M. Lotteau et al. / Building and Environment 93 (2015) 165e178

[2] IEA, World Energy Outlook 2011, Int Energy Agency, 2011, http://dx.doi.org/ [28] J. Norman, H.L. Maclean, M. Asce, C.A. Kennedy, Comparing high and low
10.1787/weo-2011-en, 1e666. residential density : life-cycle analysis of energy use and greenhouse gas
[3] B. Metz, O. Davidson, P. Bosch, R. Dave, L. Meyer, Contribution of Working emissions, J. Urban Plan. Dev. (2006) 10e21, http://dx.doi.org/10.1061/(ASCE)
Group III to the Fourth Assessment Report of the Intergovernmental Panel on 0733-9488(2006)132:1(10).
Climate Change, 2007. [29] C.C. Davila, C. Reinhart, Urban energy lifecycle : An analyitical framework to
[4] European Commission, Roadmap to a Ressource Efficient Europe, 2011. evaluate the embodied energy use of urban developments, in: Proc. BS2013
[5] J.E. Anderson, G. Wulfhorst, W. Lang, Energy analysis of the built environ- 13th Conf. Int. Build. Perform. Simul. Assoc. Chambe ry, Fr. August 26e28,
mentdA review and outlook, Renew. Sustain Energy Rev. 44 (2015) 149e158, 2013, pp. 1280e1287.
http://dx.doi.org/10.1016/j.rser.2014.12.027. [30] A. Stephan, R.H. Crawford, K. de Myttenaere, Multi-scale life cycle energy
[6] O. Ortiz, F. Castells, G. Sonnemann, Sustainability in the construction industry: analysis of a low-density suburban neighbourhood in Melbourne, Australia,
A review of recent developments based on LCA, Constr. Build. Mater. 23 Build. Environ. 68 (2013) 35e49, http://dx.doi.org/10.1016/
(2009) 28e39, http://dx.doi.org/10.1016/j.conbuildmat.2007.11.012. j.buildenv.2013.06.003.
[7] M. Buyle, J. Braet, A. Audenaert, LCA in the construction industry : a review, [31] B.G. Nichols, K.M. Kockelman, Life-cycle energy implications of different res-
Int. J. Energy Environ. 6 (2012) 397e405. idential settings: recognizing buildings, travel, and public infrastructure, En-
[8] L.F. Cabeza, L. Rinco n, V. Vilarin~ o, G. Perez, A. Castell, Life cycle assessment ergy Policy 68 (2014) 232e242, http://dx.doi.org/10.1016/
(LCA) and life cycle energy analysis (LCEA) of buildings and the building j.enpol.2013.12.062.
sector: A review, Renew. Sustain Energy Rev. 29 (2014) 394e416, http:// [32] D. Li, R. Wang, Hybrid Emergy-LCA (HEML) based metabolic evaluation of
dx.doi.org/10.1016/j.rser.2013.08.037. urban residential areas: The case of Beijing, China. Ecol. Complex 6 (2009)
[9] M. Karimpour, M. Belusko, K. Xing, F. Bruno, Minimising the life cycle energy 484e493, http://dx.doi.org/10.1016/j.ecocom.2009.09.002.
of buildings: Review and analysis, Build. Environ. 73 (2014) 106e114, http:// [33] M. Trocme , O. Le
pinoy, ACV Quartier, application sur un projet de VINCI
dx.doi.org/10.1016/j.buildenv.2013.11.019. Immobilier. Indic. d’eco Concept. pour les quartiers durables, Chaire ParisTech
[10] T.M. Baynes, T. Wiedmann, General approaches for assessing urban environ- Eco-conception des ensembles b^ atis des infrastructures, 2011.
mental sustainability, Curr. Opin. Environ. Sustain 4 (2012) 458e464, http:// [34] D. Trigaux, K. Allacker, F. De Troyer, Model for the Environmental Impact
dx.doi.org/10.1016/j.cosust.2012.09.003. Assessment of Neighbourhoods, 2014, http://dx.doi.org/10.2495/EID140091.
[11] E. Loiseau, G. Junqua, P. Roux, V. Bellon-Maurel, Environmental assessment of [35] A. Forsberg, Environmental Assessment of the Urban Environment e Devel-
a territory: an overview of existing tools and methods, J. Environ. Manage 112 opment and First Application of the Environmental Load Profile for Ham-
(2012) 213e225, http://dx.doi.org/10.1016/j.jenvman.2012.07.024. marby Sjo €stad, Royal Institute of Technology, Stockholm, 2003.
[12] J. Oliver-Sola, A. Josa, A.P. Arena, X. Gabarrell, J. Rieradevall, The GWP-Chart: [36] S. Tirumalachetty, K.M. Kockelman, B.G. Nichols, Forecasting greenhouse gas
an environmental tool for guiding urban planning processes. Application to emissions from urban regions: microsimulation of land use and transport
concrete sidewalks, Cities 28 (2011) 245e250, http://dx.doi.org/10.1016/ patterns in Austin, Tex. J. Transp. Geogr. 33 (2013) 220e229, http://dx.doi.org/
j.cities.2011.01.003. 10.1016/j.jtrangeo.2013.08.002.
[13] E. Popovici, Contribution  a l’analyse de cycle de vie des quartiers, PhD [37] G.J. Treloar, Extracting embodied energy paths from input-output tables :
dissertation, Ecole Nationale supe rieure des Mines de Paris, Paris, 2006. towards and input-output-based hybrid energy analysis method, Econ. Syst.
[14] ISO, 14044: Environmental Management e Life Cycle Assessment e Re- Res. 9 (1997) 375e391, http://dx.doi.org/10.1080/09535319700000032.
quirements & Guidelines, 2006. [38] M. Lenzen, Total requirements of energy and greenhouse gases for Australian
[15] EC JRC-IES, International Reference Life Cycle Data System (ILCD) Handbook e transport, Transp. Res. Part D. Transp. Environ. 4 (1999) 265e290, http://
General Guide for Life Cycle Assessment e Detailed Guidance, 2010, http:// dx.doi.org/10.1016/S1361-9209(99)00009-7.
dx.doi.org/10.2788/38479. [39] J. Guinee, M. Gorre e, R. Heijungs, G. Huppes, R. Kleijn, A. de Koning, et al., Life
[16] R. Frischknecht, Life Cycle Inventory Analysis for Decision-making, Swiss Cycle Assessment an Operational Guide to the ISO Standards, 2001. Leiden.
Federal Institute of Technology Zurich, 1998. [40] M. Goedkoop, R. Spriensma, The Eco-indicator99: a Damage Oriented Method
[17] ISO, 14044: Environmental Management e Life Cycle Assessment e Principles for Life Cycle Impact Assessment: Methodology Report, 2000.
& Framework, 2006. [41] R. Frischknecht, N. Jungbluth, H.-.J. Althaus, G. Doka, T. Heck, S. Hellweg, et al.,
[18] EC JRC-IES, ILCD Handbook: Analysing of Existing Environmental Impact Overview and Methodology, 2004 ecoinvent report No. 1. Dübendorf, Suisse.
Assessment Methodologies for Use in Life Cycle Assessment, 2010. [42] E. Loiseau, Elaboration d’une de marche d'e valuation environnementale d'un
[19] O. Jolliet, A. Brent, M. Goedkoop, N. Itsubo, R. Mueller-Wenk, C. Pen ~ a, et al., territoire basee sur le cadre me thodologique de l'Analyse du Cycle de vie (
Final Report of the LCIA Definition Study, 2003. ACV ) Application au territoire du bassin de Thau, 2014.
[20] M. Goedkoop, R. Heijungs, M. Huijbregts, A De Schryver, J. Struijs, R. Van Zelm, [43] J. Srebric, M. Heidarinejad, J. Liu, Building neighborhood emerging properties
ReCiPe 2008, a Life Cycle Impact Assessment Method Which Comprises and their impacts on multi-scale modeling of building energy and airflows,
Harmonised Category Indicators at the Midpoint and the Endpoint Level, Build. Environ. 91 (2015) 246e262, http://dx.doi.org/10.1016/
2009. j.buildenv.2015.02.031.
[21] M.G. Riera Pe rez, E. Rey, A multi-criteria approach to compare urban renewal [44] S. Gracik, M. Heidarinejad, J. Liu, J. Srebric, Effect of urban neighborhoods on
scenarios for an existing neighborhood. Case study in Lausanne (Switzerland), the performance of building cooling systems, Build. Environ. 90 (2015) 15e29,
Build. Environ. 65 (2013) 58e70, http://dx.doi.org/10.1016/ http://dx.doi.org/10.1016/j.buildenv.2015.02.037.
j.buildenv.2013.03.017. [45] E. Nault, G. Peronato, E. Rey, M. Andersen, Review and critical analysis of
[22] M. Colombert, C. De Chastenet, Y. Diab, C. Gobin, G. Herfray, T. Jarrin, et al., early-design phase evaluation metrics for the solar potential of neighborhood
Analyse de cycle de vie  chelle du quartier: un outil d'aide a
a l’e  la de
cision? Le designs, Build. Environ. 92 (2015) 679e691, http://dx.doi.org/10.1016/
cas de la ZAC Claude Bernard a  Paris (France), Environ. Urbain/Urban Environ. j.buildenv.2015.05.012.
5 (2011) 1e21, http://dx.doi.org/10.7202/1007605ar. [46] A. Kounina, M. Margni, J.-B. Bayart, A.-M. Boulay, M. Berger, C. Bulle, et al.,
[23] F. Cherqui, Me thodologie d’e valuation d'un Projet d'ame nagement Durable Review of methods addressing freshwater use in life cycle inventory and
d'un Quartier, Me thode ADEQUA, PhD dissertation, Universite  de La Rochelle, impact assessment, Int. J. Life Cycle Assess. 18 (2012) 707e721, http://
La Rochelle, 2005. dx.doi.org/10.1007/s11367-012-0519-3.
[24] B. Peuportier, C. Roux, Eco-design of urban sttlements using LCA, in: LCA [47] L.B. Azevedo, A.D. Henderson, R. van Zelm, O. Jolliet, M. a J. Huijbregts,
[avniR] Conf. 2013, Proc. 3rd Int. Conf. Life Cycle Approaches, 2013, pp. 1e6. Assessing the importance of spatial variability versus model choices in life
[25] B. Peuportier, E. Popovici, M. Trocme , Analyse de cycle de vie  al’echelle du cycle impact assessment: the case of freshwater eutrophication in Europe,
quartier e Bilan et perspectives du projet, ADEQUA, 2006. Environ. Sci. Technol. 47 (2013) 13565e13570, http://dx.doi.org/10.1021/
[26] G. Herfray, Contribution a  l’e
valuation des impacts environnementaux des es403422a.
quartiers, PhD dissertation, 2011. [48] T. Koellner, L. Baan, T. Beck, M. Branda~o, B. Civit, M. Margni, et al., UNEP-SETAC
[27] G. Herfray, E. Vorger, B. Peuportier, Life Cycle Assessment Applied to Urban guideline on global land use impact assessment on biodiversity and
Settlements and Urban Morphology Studies. CISBAT 11 Cleantech Sustain, ecosystem services in LCA, Int. J. Life Cycle Assess. 18 (2013) 1188e1202,
2011. Build. From Nano to urban scale, Lausanne, 2011. http://dx.doi.org/10.1007/s11367-013-0579-z.

Potrebbero piacerti anche