Journal of Cleaner Production 116 (2016) 90e99

Contents lists available at ScienceDirect

Journal of Cleaner Production

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

Climate impact assessment in life cycle assessments of forest

products: implications of method choice for results and decision-
making
Frida Røyne a, b, *, Diego Pen
~ aloza c, d, Gustav Sandin c, Johanna Berlin a,
Magdalena Svanstro €m e

a
SP Technical Research Institute of Sweden, Department of Energy and Bioeconomy, Eklandagatan 86, SE-412 61 Gothenburg, Sweden
b
Umeå University, Department of Chemistry, Umeå, Sweden
c
SP Technical Research Institute of Sweden, Department of Sustainable Built Environment, Borås/Stockholm, Sweden
d
Royal Institute of Technology KTH, Department of Building Materials, Stockholm, Sweden
e
Chalmers University of Technology, Department of Chemistry and Chemical Engineering, Gothenburg, Sweden

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

Article history: As life cycle assessments are often conducted to provide decision support, it is important that impact
Received 21 November 2014 assessment methodology is consistent with the intended decision context. The currently most used
Received in revised form climate impact assessment metric, the global warming potential, and how it is applied in life cycle as-
4 January 2016
sessments, has for example been criticised for insufficiently accounting for carbon sequestration, carbon
Accepted 5 January 2016
stored in long-lived products and timing of emission. The aim of this study is to evaluate how practi-
Available online 13 January 2016
tioners assess the climate impact of forest products and the implications of method choice for results and
decision-making.
Keywords:
LCA
To identify current common practices, we reviewed climate impact assessment practices in 101 life
Wood cycle assessments of forest products. We then applied identified common practices in case studies
Global warming comparing the climate impact of a forest-based and a non-forest-based fuel and building, respectively,
Carbon and compared the outcomes with outcomes of applying alternative, non-established practices.
Fuel Results indicate that current common practices exclude most of the dynamic features of carbon uptake
Building and storage as well as the climate impact from indirect land use change, aerosols and changed albedo.
The case studies demonstrate that the inclusion of such aspects could influence results considerably, both
positively and negatively. Ignoring aspects could thus have important implications for the decision
support. The product life cycle stages with greatest climate impact reduction potential might not be
identified, product comparisons might favour the less preferable product and policy instruments might
support the development and use of inefficient climate impact reduction strategies.
© 2016 Elsevier Ltd. All rights reserved.

1. Introduction to support various types of decision-making in relation to the

Increased manufacturing of products from wood is often seen as
a means of mitigating climate change and society's dependence on
non-renewable resources. Life cycle assessment (LCA) can be used
* Corresponding author. SP Technical Research Institute of Sweden, Department
of Energy and Bioeconomy, Eklandagatan 86, SE-412 61 Gothenburg, Sweden.
Tel.: þ46 10 516 50 00; fax: þ46 33 13 55 02.
E-mail addresses: frida.royne@sp.se (F. Røyne), diego.penaloza@sp.se
(D. Pen~ aloza), gustav.sandin@sp.se (G. Sandin), johanna.berlin@sp.se (J. Berlin),
magdalena.svanstrom@chalmers.se (M. Svanstro €m).
development of new products (Hetherington et al., 2014; Clancy
et al., 2013; Sandin et al., 2014a) or policy-making (Bringezu
et al., 2007; Gustavsson et al., 2006). However, there are short-
comings in established methodology and practices for assessing the
climate impact of forest products in LCAs as they may not, for
example, capture fully the dynamic nature of carbon flows in the
forest from sowing to harvest and in the forest product life cycle
from raw material extraction to disposal (Branda ~o et al., 2013;
Lippke et al., 2011; McKechnie et al., 2011). Consequently, climate
impact assessments of forest products may not be sufficiently ac-
curate and robust to support the decision at hand. The purpose of

http://dx.doi.org/10.1016/j.jclepro.2016.01.009
0959-6526/© 2016 Elsevier Ltd. All rights reserved.
 F. Røyne et al. / Journal of Cleaner Production 116 (2016) 90e99
this paper is to explore this problem. This is done by means of two
research questions:
1. What are the current LCA practices for assessing the climate
impact of forest products?
2. How do the results from LCAs of forest products change if the
impact assessment accounts for more climate impact aspects
than is the case in current common practices?
The approach used for answering the research questions is for
the first one an analysis of LCAs of forest products published
1997e2013, and LCA case studies for the second one. We also
discuss how different decision-making contexts of relevance to
forest products are affected by potential shortcomings of current
common practices.
Several analyses of methodology and practices in climate impact
assessment have been published in recent years (Branda ~o et al.,
2013; Pawelzik et al., 2013; Helin et al., 2013; Cherubini and
Strømman, 2011; Cherubini et al., 2009; Petersen and Solberg,
2005). This study differs from these as the focus is on quantita-
tive assessment of implications for LCA results and as it discusses
implications of practices in different decision-making contexts,
recognising that in many published LCAs there is a weak link be-
tween the decision-making context and methodological choices.
Before each research question is dealt with in separate sections,
we first provide an overview of important aspects of climate impact
assessment of forest products.
2. Background
An important aspect of climate impact assessment of forest
products is the time perspective of the climate metric, i.e. the time
period applied when calculating the effect of each emission pulse in
terms of radiative forcing. The most commonly used metric to
assess the contribution of greenhouse gases (GHG) is Global
Warming Potential (GWP), and default time periods in different
characterisation methods are 20, 100 and 500 years. The choice of
time period influences the relative importance of different types of
GHG emissions.
Another important aspect related to the temporal dimension of
climate impact is the timing of GHG emissions and carbon
sequestration. How this aspect should be handled is an issue for
LCA in general but may be particularly pronounced for forest
products since both GHG emissions (if forest biomass is used for
long-lived products such as buildings, thereby delaying emissions)
and carbon sequestration (as forests are relatively slow-growing)
may occur over a long period of time. The potential risk of pass-
ing critical tipping points in the climate system, and the urgent
impact mitigation this calls for also emphasises the need to
somehow account for the timing of the climate impact (Jørgensen
et al., 2014; Helin et al., 2013; Levasseur et al., 2010). The urgency
can to some extent be addressed by selecting a short time period for
the climate metric, such as 20 years, effectively ignoring the radi-
ative forcing of GHGs 20 years after their release and assigning a
relatively high importance to shorter-lived GHGs (e.g. methane).
However, this still does not account for the timing of emissions.
Discounting future emissions is one means of accounting for the
timing, but then there is the challenge of determining a proper
discount rate. Another means to account for the timing is to use
dynamic characterisation methods, where the time period over
which each emission pulse is integrated depends on when it occurs
in the product life cycle (Levasseur et al., 2010).
Another time-related aspect of climate impact is the temporary
storage of carbon in forest products, which prevents the carbon
from being emitted as carbon dioxide to the atmosphere, while
91
allowing regrowth of forest biomass (i.e. carbon sequestration) in
the forest. It is argued that this causes a temporary reduction in
radiative forcing, and there are several proposals on how this aspect
can be captured in LCAs (Vogtla €nder et al., 2014; Levasseur et al.,
2010; Moura-Costa and Wilson, 2000). On the other hand,
research has also suggested that temporary carbon storage may not
reduce climate impact as it lowers the carbon dioxide gradient
between the atmosphere and potential carbon reservoirs (e.g. the
oceans), thus reducing carbon dioxide removal from the atmo-
sphere. When the temporarily stored carbon is released once again,
the atmospheric carbon dioxide concentration is therefore, it is
argued, higher than would have been the case without temporary
storage (Kirschbaum, 2006). This view of temporary carbon storage
has been criticised as it disregards the cumulative climate impact
(Dornburg and Marland, 2008).
A related climate impact assessment aspect is the question of
whether or not biogenic carbon dioxide emissions should be
considered climate neutral and thus omitted when calculating
climate impact potentials (Vogtla €nder et al., 2014; Garcia and
Freire, 2014; Sjølie and Solberg, 2011). The climate neutrality of
biogenic carbon dioxide is based on the assumption that forest
products (and other bio-based products) are carbon neutral, i.e.
that there is a balance between carbon sequestration at the forest
level and the re-emission of this carbon at the product's end of life
(EoL). This assumption has been questioned by some authors as the
fate of carbon dioxide molecules emitted into the atmosphere is
indifferent to its source (Gunn et al., 2012) or because excessive
biomass harvesting may reduce carbon stocks (McKechnie et al.,
2011; Johnson, 2009). Furthermore, even in cases where the car-
bon neutrality assumption is valid, this does not automatically
imply climate neutrality as a temporal shift between emitted and
sequestered carbon may contribute to a temporary increase in
radiative forcing (Helin et al., 2013; Cherubini et al., 2011), just as an
overlap of carbon stored in products and carbon sequestered at the
forest level may reduce the radiative forcing (as discussed in the
previous paragraph).
Another important aspect influencing the climate impact
assessment of forest products is the handling of multifunctional EoL
processes. For example, this issue arises because the energy content
of non-energy forest products (e.g. building materials) is often
utilised at the products' EoL for heat and/or power production. This
creates a multifunctional process (waste handling and energy
production) and thus an allocation problem, which can be resolved
in many different ways (Sandin et al., 2015; International
Organisation for Standardisation, 2006). In LCAs of buildings, the
allocation problem can be solved by expanding the system being
studied to encompass the avoided emissions of the displaced
alternative energy system, often termed system expansion with
substitution. The inclusion of a credit for such avoided emissions
has been shown to significantly influence the climate impact of
forest products (Sandin et al., 2014b; Perez-Garcia et al., 2005). The
potential significance of the EoL credits for the forest products'
climate impact is the reasons for why we, in the present paper,
include it as an aspect of the climate impact assessment, although it
is rather an aspect of the product system modelling (set in the goal
and scope definition and affecting the life cycle inventory phase of
an LCA).
LCAs that account for carbon exchanges in the forest have
focused mostly on above-ground pools, while less attention has
been paid to changes in the below-ground carbon stored in soil, due
to land use (forestry) or land use change (afforestation, deforesta-
tion) (Helin et al., 2013). Carbon pools in the soil are especially large
in boreal forests (Liski et al., 2006), and several authors have
attempted to include the climate impact of soil carbon disturbances
in LCA studies (e.g. Branda ~o et al., 2011; Repo et al., 2011;
92 F. Røyne et al. / Journal of Cleaner Production 116 (2016) 90e99
Stephenson et al., 2010). They all identify the climate impact of soil
disturbances as significant but come to different conclusions
regarding whether the impact is positive or negative as different
wood species and forest management practises were considered in
the different studies. Repo et al. (2011) and Stephenson et al. (2010)
only considered carbon dioxide, whereas Brand~ ao et al. (2011)
considered several GHGs. In another study, methane emissions
from soil disturbances were shown to be insignificant compared to
other climate impacts in the value chains of forest products
(Delucchi, 2003). Nitrous oxide emissions resulting from soil dis-
turbances can be important, but mostly for annual biofuel crops
where fertilisation rates are high (Wrage et al., 2005).
An indirect effect of forestry can be increased competition of
land with expanding or intensified land use elsewhere. The effect
of displaced land use change is referred to as indirect land use
change (ILUC). ILUC can result in positive or negative climate
impact, which can be significant compared to the direct climate
impact of bio-based products (Searchinger et al., 2008). The
climate impact of ILUC has been in focus in recent years particu-
larly in discussions about energy products, since a shift from fossil-
based to bio-based energy could increase the competition for land
and thus cause ILUC (Berndes et al., 2013). Just as was the case for
the EoL credits discussed above, ILUC can be argued to be an aspect
of the product system modelling, rather than of the climate impact
assessment methodology. However, because of its potential influ-
ence on the climate impact of forest products, we have, in the
present study, chosen to include it as an aspect of the climate
impact assessment.
The capacity of the Earth's surface to reflect sunlight back into
space, its albedo, is another potentially important climate impact
Fig. 1. A fuel product system (short-lived product) and a building product system (long-lived product) illustrate aspects that could be of relevance in climate impact assessment of
forest products. It should be noted that a real forest product system may include additional biogenic and non-biogenic GHG emissions, e.g. from forestry, production and use
processes. These were omitted for clarity reasons.
aspect of forest products as it is influenced by land use (i.e. forestry)
and land use change (i.e. afforestation and deforestation). The al-
bedo effect is greater when the surface is white (e.g. snow-covered)
and smooth (e.g. clear-cut forest). Cherubini et al. (2012) showed
that it could be important to address changes in albedo in the
climate impact assessment, especially in regions with frequent
snow cover. An additional aspect that contributes to a changed
reflection of solar radiation is the formation of organic vapours by
forests. The organic vapours create aerosols, which reflect sunlight
or form particles that catalyse the formation of clouds, which also
reflect sunlight (Spracklen et al., 2008).
Fig. 1 summarises the above-mentioned aspects of the climate
impact of forest products. The figure shows where and when the
different aspects occur in two product systems: for a fuel and for a
building. These product systems were selected because they
represent two extremes in terms of service life (less than a year for
a fuel, a hundred years or more for many buildings) and the service
life influences several aspects. It should be noted that the handling
of some of the discussed aspects in LCA may overlap; credit given to
carbon sequestered in a product (aspect D in Fig. 1), for example, is
equivalent to taking (some) account of the timing of emissions
(aspect A). Regarding biogenic carbon dioxide emissions as climate
neutral (aspect C) is equivalent to disregarding the timing of
emissions and carbon sequestration (aspect A). It should also be
noted that the classification and definition of climate impact as-
pects is based on the scope of this particular study. Other classifi-
cations and definitions are possible in other contexts. Further issues
of relevance for the climate impact assessment are assumptions
related to the area that biomass is harvested from, so called spatial
system boundaries (Cintas et al., 2015; Cherubini et al., 2013) and
 F. Røyne et al. / Journal of Cleaner Production 116 (2016) 90e99
what one assumes happens with the forest land in the absence of
harvesting, the so-called baseline or reference situation (Helin
et al., 2015; Soimakallio et al., 2015; Ter-Mikaelian et al., 2015).
These issues are not further discussed in this paper for clarity
reasons, as they would introduce additional dimensions.
3. Current practices in the climate impact assessment of
forest products
3.1. Method
To identify current practices in climate impact assessment in
LCAs of forest products, a literature search was conducted among
peer-reviewed papers, published before January 1, 2014, in the
Scopus database using the Boolean search string “(‘forest’ OR
‘wood’) AND (‘life cycle analysis’ OR ‘life cycle assessment’ OR
‘LCA’)”. The selection criteria for including a paper in the literature
review were that the raw material for the studied product(s) had to
be (fully or partly) woody biomass (virgin, residue or waste) and
that the prime focus of each LCA had to be on environmental
assessment of products (i.e., a case study) and not on method
development. These criteria led to the inclusion of 101 papers
published 1997e2013. All categories of products were investigated:
energy, fuels, construction materials, plastics, chemicals and other
commodities. Forest biomass types included solid timber, forest
residues and industrial waste wood, from slow-grown boreal for-
ests to short-rotation wood crops. To reveal methodological
choices, we analysed how the climate impact aspects of Fig. 1 were
handled in the climate impact assessment of each paper.
3.2. Results and discussion
Fig. 2 shows the findings from the analysis of the 101 papers
included in the literature review (a list of the papers and the results
from the analysis per product type can be found in the
Supplementary Material).
Fig. 2 shows that the timing of climate impact (aspect A) is only
considered in 3% of the papers and that a climate impact metric
with a 100-year time perspective is what is used most often
Fig. 2. Methodological choices for climate impact assessment in the 101 LCA studies included in the literature review. The letters (AeI) refer to the aspects in Fig. 1.
93
(aspect B). LCA practitioners thus usually ignore the fact that GHG
emissions and carbon sequestration in forest products may be
highly distributed in time. In the analysis of aspect C, we only
considered papers that include EoL processes (60%). Among these
papers, as many as 87% consider biogenic carbon dioxide emissions
as climate neutral, which in practice means that the time lag be-
tween emissions and uptake is disregarded or that the chosen
spatial system boundaries support such an assumption (e.g. a
landscape approach instead of a single-stand approach, see
Cherubini et al. (2013)). No papers credit the avoided radiative
forcing from the storage of carbon in products over time (aspect D).
It is not surprising that studies on short-lived products such as
fuels do not include such a credit, but neither do studies on long-
lived products such as buildings. In the analysis of aspect E, we
only considered papers that include EoL processes and where the
EoL processes are multifunctional (38%). We can see that the
practice of crediting does not dominate among the papers
considered (34%). One reason for this could be that there are un-
certainties regarding the EoL fate of a product, especially for long-
lived products such as buildings (Sandin et al., 2014b). Further-
more, we can see that aspects F, G, H and I are seldom accounted
for (in 7%, 0%, 0% and 1% of the studies, respectively), which is not
surprising considering the lack of standards or universally agreed
procedures for calculating the impact of these aspects. An addi-
tional finding is that in approximately 20% of the studies, the
methodological choices concerning the time perspective of the
climate impact metric, biogenic carbon dioxide neutrality, carbon
sequestration in the product and crediting of avoided EoL emis-
sions are not clearly articulated.
Fig. 3 shows the distribution in time of the papers included in
the literature review. A steady increase from 2008 and on can be
observed. Although exceptions from the common practice become
more frequent with time, they are still rare in relation to the
number of published papers. We can thus see that there is an
increasing trend in conducting forest product LCA. The trend is
however not followed by a change in climate impact assessment
practices. Although climate impact aspects discussed in this paper
have been addressed in recent years, they are included in very few
studies.
94 F. Røyne et al. / Journal of Cleaner Production 116 (2016) 90e99
Fig. 3. Time distribution of published forest product LCAs 1997e2013, with occurrences of exceptions from the common practice indicated. The letters (AeI) refer to the aspects in
Fig. 1. Aspects D (credit for carbon stored in the product), G (quantitative consideration of climate impact from ILUC) and H (quantitative consideration of the albedo effect) are not
included in the figure as they have not been accounted for in any of the papers.
4. Implications for results if additional aspects are included
4.1. Method
Two case studies were employed to demonstrate the implica-
tions of different methodological choices in the climate impact
assessment in LCAs of forest products. We selected a short-lived
product (a wood-based fuel: ethanol) and a long-lived product (a
building with a solid wood structure and wood materials in the
building envelope). These were selected for the same reason that a
fuel and a building product system were shown in Fig. 1, i.e. they
represent extremes in terms of service life, and differing aspects of
climate impact can therefore be expected to be of relevance for
their aggregated life cycle impact. The products were compared
with functionally equivalent non-forest products in order to study
how comparisons of forest and non-forest products may depend on
climate impact assessment practices. Ethanol was compared with
petrol while the building with a wood structure was compared with
a building with a concrete structure (but with a similar building
envelope with a high wood material content). Details on system
boundaries and inventory data can be found in the Supplementary
Material.
When calculating the climate impact in each case study, we first
applied current common practices, as identified in the literature
review (section 3), and then alternative practices for each climate
impact aspect, as listed in Table 1. A detailed explanation of the
methods used and the reasoning behind the choice of methodology
is presented in the Supplementary Material. As no standard or
universally agreed procedure exists for calculating the climate
impact from soil disturbances (aspect F) or the albedo effect (aspect
H), we used numbers from other LCA studies as estimates. As
research on, and the understanding of, ILUC from forestry (aspect
G) and the climate effect of aerosols (aspect I) is still limited, these
aspects were not considered in the calculations. All calculations
were made to address the research questions in this paper and the
results should not be used outside of this context.
4.2. Results and discussion
Figs. 4 and 5 show results from the fuel and the building case
study, respectively. The figures show results using current common
practices as well as the lowest and highest results possible with any
combination of the alternative practices listed in Table 1. It should
be stressed that some of the aspects considered with the alternative
practices are associated with very large uncertainties, e.g. the
climate impact of soil disturbances and changed albedo, and that
the results thus should be seen as order-of-magnitude indications
of possible results of using the alternative practices.
In the fuel case study, the lowest and highest results for ethanol
differ considerably (Fig. 4). The inclusion of biogenic carbon emis-
sions (aspect C) has the largest impact as such emissions from the
production phase and the fuel combustion in the vehicle are not
taken into account in current common practices. The result for
petrol, however, is barely affected as the fuel is not derived from
biomass (petrol is only affected by the different time perspectives of
the climate metric and by a very small share of biogenic carbon in
the product system). Consequently, the comparison between
ethanol and petrol is influenced considerably by the choice of
climate impact assessment methodology.
The lowest and highest results for the wood building also differ
(Fig. 5), but not to the same extent as for ethanol. The comparison
between the wood and concrete buildings is also less influenced by
the choice of methodology than for the fuels since the operational
energy dominates the results, and the buildings have similar
heating and electricity use in the operation phase. The only dif-
ference in the operation phase is that the concrete building has a
slightly higher capacity to store heat than the wood building, which
influences the heating demand (also discussed by Heeren et al.
(2015)). Since the heating demand is assumed to be supplied
mainly by fossil energy, the difference does not significantly in-
fluence the outcome of different climate impact assessment prac-
tices. However, the case study considers buildings with low energy
use and the results are therefore more sensitive to choices in the
 F. Røyne et al. / Journal of Cleaner Production 116 (2016) 90e99 95

Table 1
Methodological practices applied in the case studies (see the Supplementary Material for further details).

Aspect of the climate impact Current common Alternative practices

practices

A. Timing of GHG emissions and carbon
sequestration.
B. Time perspective of the climate impact metric.
C. Climate neutrality assumption of biogenic
carbon dioxide sequestration and emissions.
D. Credit for carbon stored in product.
E. Credit for avoided climate impact due to
multifunctional EoL processes.
F. Climate impact from soil disturbances.
G. Climate impact from ILUC.
H. Climate impact from changes in the albedo
effect.
I: Climate impact from aerosols.
Not considered Discounting of emissions and dynamic LCA (Levasseur et al., 2010). Only applied for the
building case study.
100 years 20 years or 500 years (Goedkoop et al., 2013).
Biogenic carbon dioxide Biogenic and fossil carbon dioxide emissions are assumed to have the same climate impact.
emissions are excluded.
No credit Carbon storage is credited using the ILCD Handbook method (European Commission, 2010)
and the method proposed by Cherubini et al. (2011). Only applied for the building case study.
No credit Credit for replacing energy from solar power, natural gas or coal. Only applied for the building
case study.
Not considered Both negative and positive climate impact from soil disturbances due to forestry activities is
considered. The calculation is based on the model developed by Kilpela €inen et al. (2011) for
negative impact and the results from Stephenson et al. (2010) for positive impact. For the
building case study, only negative impact is considered as the Stephenson et al. model is for a
short-rotation crop intended for energy use.
Excluded due to lack of knowledge and methodology.
Not considered Considered using results from Cherubini et al. (2012).
Excluded due to lack of knowledge and methodology.

climate impact assessment methodology than conventional build-
ings where the operational energy would be more dominant. If a
high proportion of operational energy would come from biomass
sources, the results would be more sensitive to the choice of
methodology. Finally, it should be noted that the building envelope
in the concrete building is very similar to the envelope in the wood
building, with a high content of wood materials. If the wood
building would be compared with a building made entirely of
concrete, the effects of changes in climate assessment methodology
would be more significant.
In Figs. 4 and 5, in addition to results from using current com-
mon practices, only the most extreme (lowest and highest) results
are presented. A number of additional methods were, however,
tested. More details are shown for the building case study in Figs. 6
and 7. The building case study is shown here as more climate
impact aspects are relevant for the building than for the fuel case
study.
Fig. 6 shows different results of five different practices of ac-
counting for aspects A and D. Apart from the current common
practices, four alternative methods were applied to account for the
dynamic nature of carbon sequestration and storage and biogenic
carbon dioxide emissions: discounting of future emissions (dis-
count rate 1% per year), the GWPbio method, the ILCD method, and
Dynamic LCA (for further details on each method, see the
Supplementary Material). Results for the alternative methods span
from being similar to current common practices (GWPbio) to
considerably lower. The differences in results reflect the un-
certainties and disagreements regarding how to model the

Fig. 4. LCA results from the fuel case study, for driving 1 km with an ethanol fuel (100% forest-based) and petrol, respectively. ‘Result with current common practices’ is the most
common way of calculating climate impact, namely using GWP100 and not including biogenic carbon dioxide emissions or climate impact from soil disturbance, ILUC, aerosols or
the albedo effect. ‘Lowest result’ is the result of using GWP500, accounting for reduced climate impact due to soil disturbances, including climate impact from the albedo effect and
not regarding biogenic carbon dioxide emissions as climate neutral. ‘Highest result’ is the result of using GWP20, accounting for increased climate impact due to soil disturbances,
excluding climate impact from the albedo effect, and not regarding biogenic carbon dioxide emissions as climate neutral.
96 F. Røyne et al. / Journal of Cleaner Production 116 (2016) 90e99

Fig. 5. LCA results from the building case study, for one square metre of a multi-dwelling building with a wood or concrete structure, respectively. ‘Result with current common
practices’ is the most common way of calculating climate impact, namely using GWP100 and not including biogenic carbon dioxide emissions or climate impact from soil
disturbance, ILUC, aerosols or the albedo effect. ‘Lowest possible result’ derives from using GWP500, discounting future emissions, giving credit for carbon storage, including albedo
effects (estimate), not taking into account biogenic carbon dioxide emissions and including EoL credits for replacing energy production from coal. ‘Highest possible result’ derives
from using GWP20, taking into account increased climate impact due to soil disturbances, excluding albedo effects, including biogenic carbon dioxide emissions and excluding EoL
credits.

dynamics of carbon flows in forest product systems. The dis-
agreements are further emphasised in a recent publication pre-
senting an alternative way of generating GWPbio values, with a
different choice of baseline (Helin et al., 2015). This underlines the
importance of a consensus process (possibly combined with further
method development) for agreeing on the methodology for how to
account for carbon dynamics in climate impact assessment in LCAs
of forest products.
Fig. 7 shows results for different EoL substitution scenarios for
the building case study, reflecting different ways of accounting for
aspect E. Since the EoL phase is assumed to take place in 100 years,
it may not be valid to assume today's disposal practices. Also,

Fig. 6. LCA results from the building case study with different methodological choices for aspect A (timing of GHG emissions and carbon sequestration) and aspect D (credit for
carbon stored in the product). The time horizon for dynamic LCA is 100 years.
 F. Røyne et al. / Journal of Cleaner Production 116 (2016) 90e99
Fig. 7. LCA results from the building case study with different methodological choices for aspect E (credit for avoided climate impact due to multifunctional EoL processes).
assuming that a certain technology will be substituted in 100 years
is inherently speculative, which calls for the use of scenario analysis
to test the sensitivity to the range of possible futures (e.g., as done
in Fig. 7 and in Sandin et al. (2014b)). In this case, it can be seen that
the results are sensitive with regard to the substituted technology,
with substitution of heating from coal yielding (not surprisingly)
the greatest EoL credit. However, the results of the comparison of
the timber and concrete building are, just as in Fig. 6, rather
insensitive to the choice of method: the timber building has a clear
advantage in all cases. This insensitivity is, again, primarily due to
the relatively large contribution from an operation phase that is not
influenced much by the choice of climate impact assessment
methodology.
For some climate impact aspects, the assessment methods used
in the case studies are, as already mentioned, not established, and
the results could change if other methods were applied. Further-
more, aspects that were not considered, such as the climate impact
of ILUC and aerosols, could also influence the results considerably.
Nevertheless, our results provide an indication of the considerable
uncertainties in current climate impact assessment practices
applied to forest products and the need for further development.
That results of LCAs of forest products are sensitive to climate
impact assessment practices is consistent with previous research
(Michelsen et al., 2012).
5. Discussion of implications for practice and decision-
making
To explore the implications of different climate impact assess-
ment practices to different decision-making contexts, we discuss
the effect of different practices in relation to four specific goals of
decision-making: (i) to reduce the climate impact of a product, (ii)
to choose the product with the lowest climate impact among
several products, (iii) to obtain long-term climate impact reduction
and (iv) to obtain short-term climate impact reduction. We pur-
posely avoid defining the contexts as attributional or consequential,
since what should be termed consequential and attributional is not
97
clear-cut and it can be disputed when a consequential or attribu-
tional approach is appropriate (Suh and Yang, 2014).
If the goal is to (i) reduce the climate impact of a product, cur-
rent common practices for climate impact assessment do not
appear to take into account certain information that is of consid-
erable interest for the decision at hand. Even if some hot-spots in
the product value chain can be identified, the practices may
disregard life cycle processes that are even more significant in
terms of improvement opportunities. For example, only the life
cycle processes with large non-biogenic carbon dioxide emissions
would be identified as hot-spots. In the (ii) comparison of products,
current common practices would probably lead to the conclusion
that a forest product has a lower climate impact than alternative,
non-biobased products, unless the manufacturing process of the
forest product leads to considerably higher fossil carbon dioxide
emissions (or other GHG emissions) than the manufacturing pro-
cess of the non-forest product. The LCA practitioner may therefore
fail to identify the preferred product. The decision-making conse-
quences of this could, for example, be that a procurer purchases a
less preferable product from a climate impact perspective, or that
policy making aimed at supporting the development of low carbon
technologies supports an inferior technology.
In LCAs aimed at guiding the development of policies dealing
with (iii) long-term climate impact reduction, the current common
practice would be more aligned with the decision at hand than
when (iv) short-term climate impact reduction is the aim. For
short-term climate impact reduction, the urgency implies that it
could be suitable to use climate impact metrics based on a shorter
time perspective (e.g. GWP20), use dynamic metrics or discount
future emissions. If the harvested forest is slow-growing, it could be
correct to account for temporary carbon storage (in the case of
commodities or building products) as well as the climate impact
arising because of the temporal shift between biogenic carbon di-
oxide emissions and carbon uptake due to the forest regrowth.
Albedo effects and soil disturbances should be taken into account as
long as these are short-term effects. EoL substitution credits would,
however, be less important to consider, particularly in studies of
98 F. Røyne et al. / Journal of Cleaner Production 116 (2016) 90e99
long-lived products such as buildings. In contrast, for long-term
climate impact reduction, it could be justifiable to disregard
biogenic carbon dioxide emissions (also if the harvested forest is
slow-growing, if it can be reasonably assumed that the harvested
biomass is carbon-neutral in a long-term perspective). It could also
be justifiable to disregard disturbances of soil carbon (if these can
be assumed to be reversed over time) and the albedo effect (unless
harvesting results in permanent deforestation). However, excluding
these implies that the importance of the cumulative climate impact
of increased temperatures until the chosen point in time (e.g. until
100 years from now), and the rate of temperature increase until
that time, are downplayed in relation to the instantaneous effect of
elevated temperatures at the chosen point in time (see Dornburg
and Marland (2008) for more information). Assuming that one
wants to focus on the instantaneous effect of elevated tempera-
tures, other characterisation factors than GWP may actually be
more suitable, such as the Global Temperature Potential (GTP),
which reflects the warming potential of a GHG in terms of its in-
fluence on the global mean temperature at the end of a given time
period (Shine et al., 2005).
The challenges of climate impact assessment of forest products
are not only related to a lack of understanding of physical mecha-
nisms, but also to the fact that the choice of methodology is a value-
based choice that for example depends on which parts of a product
system that we think fall under our responsibility. One example of a
value-based choice is how our perception of urgency influences our
choice of time period for the climate metric. In general, we need to
be more explicit in how we describe the goal and scope of our
studies and how we motivate the choice of methodology in relation
to this.
An LCA is always delimited in order to become manageable. It is,
however, important that it provides relevant input to the decision it
is designed to support. For more robust decision-making, there is a
need for further case studies testing the implications of considering
the climate impact aspects addressed in this paper as well as other
climate impact aspects (e.g. ILUC and aerosol effects, which were
not covered in the present study) for different products and
decision-making contexts. To improve impact assessment prac-
tices, there is a need for better inventory data supporting the use of
non-established impact assessment methodology. Eventually, there
is also a need to build a consensus in the LCA community on which
methodology should be used in the assessment of non-established
aspects of climate impact in different contexts.
Until improved methodology, covering all potentially relevant
climate impact aspects, becomes accessible, we recommend that
LCA practitioners (i) reflect on the decision-making context(s) in
which their LCA will and can be used, (ii) consider which climate
impact aspects are potentially important for the decision-making
context(s), and (iii) calculate or estimate the influence of the as-
pect(s) quantitatively (e.g. by sensitivity analysis) or, if this is not
possible, qualitatively. Such efforts would also serve as case studies
and be useful in developing new methodology and establishing
suitable practices for different contexts.
6. Conclusions
A literature review revealed that current common practices for
climate impact assessment in LCAs of forest products exclude most
of the dynamic features of carbon uptake, storage and release.
Biogenic carbon dioxide emissions are in most studies regarded as
climate neutral, the timing of emissions are seldom considered,
GWP100 is the most common climate impact metric, EoL emissions
are credited in only about one third of the studies, soil disturbances
or the forest's influence on aerosol formation are seldom taken into
account and impacts due to a changed albedo effect or ILUC are not
assessed in any of the studies. The methodological choices are often
not clearly articulated, causing additional uncertainty when inter-
preting results. Although the number of annually published forest
product LCAs increases, the climate impact aspects discussed in this
paper were not increasingly addressed within the reviewed time
period (1997e2013).
Case studies involving two forest products (a building and a
fuel) demonstrated that the inclusion of aspects of climate impact
beyond those included in current common practices for impact
assessment could influence LCA results considerably. This could
have significant implications for the potential of LCA results to
provide decision support as the parts of the product life cycle with
greatest climate impact reduction potential might not be identified,
product comparisons might favour the less preferable product in
climate terms and policy instruments might support the develop-
ment and use of inefficient climate impact reduction strategies.
This paper is an attempt to acquire a better understanding of the
limitations and uncertainties of current practices for climate impact
assessment in LCAs of forest products. This understanding is
required for robust decision-making until improved climate impact
assessment methodology, covering all potentially relevant aspects
of the climate impact of forest products, is available and applicable
in practice.
Acknowledgements
We would like to thank RISE Research Institutes of Sweden and
Bio4Energy, a strategic research environment set up by the Swedish
government, for supporting this work.
Appendix A. Supplementary data
Supplementary data related to this article can be found at http://
dx.doi.org/10.1016/j.jclepro.2016.01.009.
References
Berndes, G., Ahlgren, S., Bo € rjesson, P., Cowie, A., 2013. Bioenergy and land use
change e state of the art. WIREs Energy Environ. 2013 (2), 282e303.
Branda ~o, M., Levasseur, A., Kirschbaum, M., Weidema, B., Cowie, A., Jørgensen, S.,
Hauschild, M., Pennington, D., Chomkhamsri, K., 2013. Key issues and options in
accounting for carbon sequestration and temporary storage in life cycle
assessment and carbon footprinting. Int. J. Life Cycle Assess. 18, 230e240.
Branda ~o, M., Mila i Canals, L., Clift, R., 2011. Soil organic carbon changes in the
cultivation of energy crops: Implications for GHG balances and soil quality for
use in LCA. Biomass Bioenerg. 35 (6), 2323e2336.
Bringezu, S., Ramesohl, S., Arnold, K., Fischedick, M., von Geibler, J., Liedtke, C.,
Schütz, H., 2007. What We Know and What We Should Know towards a Sus-
tainable Biomass Strategy. A Discussion Paper of the Wuppertal Institute, No.
163. Wuppertal Institute for Climate, Environment and Energy, Wuppertal,
Germany.
Cherubini, F., Guest, G., Strømman, A.H., 2013. Bioenergy from forestry and changes
in atmospheric CO2: reconciling single stand and landscape level approaches.
J. Environ. Manag. 129, 292e301.
Cherubini, F., Bright, R.M., Strømman, A.H., 2012. Site-specific global warming po-
tentials of biogenic CO2 for bioenergy: contributions from carbon fluxes and
albedo dynamics. Environ. Res. Lett. 7 (4).
Cherubini, F., Peters, G.P., Berntsen, T., Strømman, A.H., Hertwich, E., 2011. CO2
emissions from biomass combustion for bioenergy: atmospheric decay and
contribution to global warming. GCB Bioenerg. 3 (5), 413e426.
Cherubini, F., Strømman, A.H., 2011. Life cycle assessment of bioenergy systems:
state of the art and future challenges. Bioresour. Technol. 102 (2), 437e451.
Cherubini, F., Bird, N.D., Cowie, A., Jungmeier, G., Schlamadinger, B., Woess-
Gallasch, S., 2009. Energy- and greenhouse gas-based LCA of biofuel and bio-
energy systems: key issues, ranges and recommendations. Resour. Conserv.
Recycl. 53 (8), 434e447.
Cintas, O., Berndes, G., Cowie, A.L., Egnell, G., Holmstro €m, H., Ågren, G., 2015. The
climate effect of increased forest bioenergy use in Sweden: evaluation at
different spatial and temporal scales. WIREs Energy Environ. http://dx.doi.org/
10.1002/wene.178.
Clancy, G., Fro€ling, M., Svanstro€m, M., 2013. Insights from guiding material devel-
opment for more sustainable products. Int. J. Sust. Des. 2, 149e166.
 F. Røyne et al. / Journal of Cleaner Production 116 (2016) 90e99
Delucchi, M.A., 2003. A Lifecycle Emissions Model (LEM): Lifecycle Emissions from
Transportation Fuels, Motor Vehicles, Transportation Modes, Electricity Use,
Heating and Cooking Fuels, and Materials. Main Report. Institute if Trans-
portation Studies, UC Davis, USA.
Dornburg, V., Marland, G., 2008. Temporary storage of carbon in the biosphere does
have value for climate change mitigation: a response to the paper by Miko
Kirschbaum. Mitig. Adapt. Strateg. Glob. Chang. 13 (3), 211e217.
European Commission, 2010. International Reference Life Cycle Data System (ILCD)
Handbookdgeneral Guide for Life Cycle Assessmentddetailed Guidance, first
ed. Publications Office of the European Union, Luxembourg.
Garcia, R., Freire, A., 2014. Carbon footprint of particleboard: a comparison between
ISO/TS 14067, GHG protocol, PAS 2050 and climate declaration. J. Clean. Prod.
66, 199e200.
Goedkoop, M., Heijungs, R., Huijbergts, M., De Schryver, A., Struijs, J., Van Zelm, R.,
2013. ReCiPe 2008, a Life Cycle Impact Assessment Method Which Comprises
Harmonised Category Indicators at the Midpoint and the Endpoint Level.
Ruimte en Milieu, Ministerie van Volkshuisvesting, Reimtelijke.
Gunn, J.S., Ganz, D., Keeton, W., 2012. Biogenic vs geologic carbon emissions and
forest biomass energy production. GCB Bioenerg. 4, 239e242.
Gustavsson, L., Madlener, R., Hoen, H.-F., Jungmeier, G., Karjalainen, T., Klo, S., 2006.
The role of wood material for greenhouse gas mitigation. Mitig. Adapt. Strateg.
Glob. Chang. 11, 1097e1127.
Heeren, N., Mutel, C.L., Steubing, B., Ostermeyer, Y., Wallbaum, H., Hellweg, S., 2015.
Environmental impact of buildings e what matters? Environ. Sci. Technol. 49
(16), 9832e9841.
Helin, T., Salminen, H., Hynynen, J., Soimakallio, S., Huuskonen, S., Pingous, K., 2015.
Global warming potentials of stemwood used for energy and materials in
Southern Finland: differentiation of impacts based on type of harvest and
product lifetime. GCB Bioenerg. http://dx.doi.org/10.1111/gcbb.12244.
Helin, T., Sokka, L., Soimakallio, S., Pingoud, K., Pajula, T., 2013. Approaches for in-
clusion of forest carbon cycle in life cycle assessment e a review. GCB Bioenerg.
5 (5), 475e486.
Hetherington, A.C., Borrion, A.L., Griffiths, O.G., McManus, M., 2014. Use of LCA as a
development tool within early research: challenges and issues across different
sections. Int. J. Life Cycle Assess. 19, 130e143.
International Organisation for Standardization, 2006. Environmental Management
e Life Cycle Assessment e Requirements and Guidelines (ISO 14044:2006).
Johnson, E., 2009. Goodbye to carbon neutral: getting biomass footprints right.
Environ. Impact Assess. 29, 165e168.
Jørgensen, S.V., Hauschild, M.Z., Nielsen, P.H., 2014. Assessment of urgent impacts of
greenhouse gas emissions - the climate tipping potential (CTP). Int. J. Life Cycle
Assess. 19 (4), 919e930.
Kilpela€inen, A., Alam, A., Strandman, H., Kellom€aki, S., 2011. Life cycle assessment tool
for estimating net CO2 exchange of forest production. GCB Bioenerg. 3, 461e471.
Kirschbaum, M.U.F., 2006. Temporary carbon sequestration cannot prevent climate
change. Mitig. Adapt. Strateg. Glob. Chang. 11 (5e6), 1151e1164.
Levasseur, A., Lesage, P., Margni, M., Deschenes, L., Samson, B., 2010. Considering
time in LCA: dynamic LCA and its application to global warming impact as-
sessments. Environ. Sci. Technol. 44, 3169e3174.
Lippke, B., Oneil, E., Harrison, R., Skog, K., Gustavsson, L., Sathre, R., 2011. Life cycle
impacts of forest management and wood utilization on carbon mitigation:
knowns and unknowns. Carbon Manag. 2, 303e333.
Liski, J., Lehtonen, A., Palosuo, T., Peltoniemi, M., Eggers, T., Muukonen, P.,
Ma €kipa€€
a, R., 2006. Carbon accumulation in Finland's forests 1922e2004 e an
estimate obtained by combination of forest inventory data with modelling of
biomass, litter and soil. Ann. For. Sci. 63, 687e697.
McKechnie, J., Colombo, S., Chen, J., Mabee, W., MacLean, H.L., 2011. Forest bio-
energy or forest carbon? Assessing trade-offs in greenhouse gas mitigation with
wood-based fuels. Environ. Sci. Technol. 45 (2), 789e795.
99
Michelsen, O., Cherubini, F., Strømman, A.H., 2012. Impact assessment of biodi-
versity and carbon pools from land use and land use changes in life cycle
assessment, exemplified with forestry operations in Norway. J. Ind. Ecol. 16 (2),
231e242.
Moura-Costa, P., Wilson, C., 2000. An equivalence factor between CO2 avoided
emissions and sequestration e description and applications in forestry. Mitig.
Adapt. Strateg. Glob. Chang. 5, 51e60.
Pawelzik, P., Carus, M., Hotchkiss, J., Narayan, R., Selke, S., Wellisch, M., Weiss, M.,
Wicke, B., Patel, M.K., 2013. Critical aspects in the life cycle assessment (LCA) of
bio-based materials e reviewing methodologies and deriving recommenda-
tions. Resour. Conserv. Recycl. 73, 211e228.
Perez-Garcia, J., Lippke, B., Comnick, J., Manriquez, C., 2005. An assessment of car-
bon pools, storage, and wood products market substitution using life-cycle
analysis results. Wood Fiber Sci. 37, 140e148.
Petersen, A.K., Solberg, B., 2005. Environmental and economic impacts of substi-
tution between wood products and alternative materials: a review of micro-
level analyses from Norway and Sweden. For. Policy Econ. 7 (3), 249e259.
Repo, A., Tuomi, M., Liski, J., 2011. Indirect carbon dioxide emissions from producing
bioenergy from forest harvest residues. GCB Bioenerg. 3 (2), 107e115.
€m, M., 2015. Allocation in LCAs of
Sandin, G., Røyne, F., Berlin, J., Peters, G., Svanstro
biorefinery products: implications for results and decision making. J. Clean.
Prod. 93, 213e221.
Sandin, G., Clancy, G., Heimersson, S., Peters, G.M., Svanstro €m, M., ten Hoeve, M.,
2014a. Making the most of LCA in technical inter-organisational R&D projects.
J. Clean. Prod. 70, 97e104.
Sandin, G., Peters, G.M., Svanstro € m, M., 2014b. Life cycle assessment of construction
materials: the influence of assumptions in end-of-life modeling. Int. J. Life Cycle
Assess. 19, 723e731.
Searchinger, T., Heimlich, R., Houghton, R.A., Dong, F., Elobeid, A., Fabiosa, J.,
Tokgoz, S., Hayes, D., Tun-Hsiang, Y., 2008. Use of U.S. croplands for biofuels
increases greenhouse gases through emissions from land-use change. Science
319 (5867), 1238e1240.
Sjølie, H.K., Solberg, B., 2011. Greenhouse gas emission impacts of use of Norwegian
wood pellets: a sensitivity analysis. Environ. Sci. Policy 14 (8), 1028e1040.
Shine, K.P., Fuglestvedt, J.S., Hailemariam, K., Stuber, N., 2005. Alternatives to the
global warming potential for comparing climate impacts of emissions of
greenhouse gases. Clim. Change 68, 281e302.
Soimakallio, S., Cowie, A., Brand~ ao, M., Finnveden, G., Ekvall, T., Erlandsson, M.,
Koponen, P., Karlsson, P.-E., 2015. Attributional life cycle assessment: is a land-
use baseline necessary? Int. J. Life Cycle Assess. 20, 1364e1375.
Spracklen, D.V., Bonn, B., Carslaw, K.S., 2008. Boreal forests, aerosols and the im-
pacts on clouds and climate. Philos. Trans. R. Soc. A 366 (1885), 4613e4626.
Stephenson, A.L., Dupree, P., Scott, S.A., Dennis, J.S., 2010. The environmental and
economic sustainability of potential bioethanol from willow in the UK. Bio-
resour. Technol. 101 (24), 9612e9623.
Suh, S., Yang, Y., 2014. On the uncanny capabilities of consequential LCA. Int. J. Life
Cycle Assess. 19 (6), 1179e1184.
Ter-Mikaelian, M.T., Colombo, S.J., Chen, J., 2015. The burning question: does forest
bioenergy reduce carbon emissions? A review of common misconceptions
about forest carbon accounting. J. For. 113 (1), 57e68.
Vogtla€nder, J., Van Der Velden, N., Van Der Lugt, P., 2014. Carbon sequestration in
LCA, a proposal for a new approach based on the global carbon cycle; cases on
wood and on bamboo. Int. J. Life Cycle Assess. 19 (1), 13e23.
Wrage, N., Groeningen, J.W.v., Oenema, O., Baggs, E.M., 2005. A novel dual-isotope
labelling method for distinguishing between soil sources of N2O. Rapid Com-
mun. Mass Spectrom. 19, 3298e3306.