Content uploaded by Bjart Holtsmark
Author content
All content in this area was uploaded by Bjart Holtsmark
Content may be subject to copyright.
Harvesting in boreal forests and the biofuel carbon debt
Bjart Holtsmark
Received: 5 January 2011 / Accepted: 9 August 2011 / Published online: 26 August 2011
#
Springer Science+Business Media B.V. 2011
Abstract Owing to the extensive critique of food-crop-based biofuels, attention has
turned toward second-generation wood-based biofuels. A question is therefore whether
timber taken from the vast boreal forests on an increasing scale should serve as a
source of wood-based biofuels and whether this will be effective climate policy. In a
typical boreal forest, it takes 70–120 years before a stand of trees is mature. When this
time lag and the dynamics of boreal forests more generally are taken into account, it
follows that a high level of harvest means that the carbon stock in the forest stabilizes
at a lower level. Therefore, wood harvesting is not a carbon-neutral activity. Through
model simulations, it is estimated that an increased harvest of a boreal forest will create
a b iofuel carbon debt that takes 190–340 years to repay. The length of the payback time
is sensitive to the type of fossil fuels that wood energy replaces
1 Introduction
Carbon neutrality of bioenergy combustion is incorporated into most countries’
climate policies. No country imposes taxes on CO
2
emissions from the combustion of
bioenergy. Moreover, the European Union emissions trading scheme incorporates the
assumption that bioenergy is a carbon-neutral fuel; firms included in this market are not
committed to acquiring and surrendering allowances for emissions from the combustion
of bioenergy.
The reasoning behind the carbon neutrality assumption is that the harvest of one
crop is replaced by the growth of a new crop, which reabsorbs the quantity of carbon
that was released by burning the first crop. This is a reasonable argument in the case
of food-crop-based biofuels, as new crops replace those that are harvested, usually
within 1 year. However, the carbon neutrality of food-crop-based biofuels has recently
Climatic Change (2012) 112:415–428
DOI 10.1007/s10584-011-0222-6
Electronic supplementary material The online version of this article (doi:10.1007/s10584-011-0222-6)
contains supplementary material, which is available to authorized users.
B. Holtsmark (*)
Statistics Norway, PO box 8131 Dep, 0033 Oslo, Norway
e-mail: Bjart.Holtsmark@ssb.no
been questioned. Fargione et al. (2008) found that converting native habitats to cropland
releases CO
2
from both existing vegetation and carbon stored in soils. Fargione et al.
(2008) therefore concluded t hat production of food-crop-based biofuels may create a
biofuel carbon debt by releasing CO
2
at a level that is many times the level of annual
greenhouse gas reductions that these biofuels would provide by displacing fossil fuels.
Searchinger et al. (2008) analyzed the global effects of using grain or existing cropland
for biofuel production. They argued that most previous analyses failed to take account of
the carbon emissions that occur as farmers worldwide respond to higher crop prices and
convert forest and grassland to new cropland to replace the grain or cropland diverted to
biofuels; see also Gibbs et al. ( 2010 ); Gurgel et al. (2007); Lapola et al. (2010), and Melillo
et al. (2009), among others.
More generally, Wise et al. (2009) and Searchinger et al. (2009) underlined that the
current practice of accounting for CO
2
emissions from combustion of bioenergy as zero
means there are strong incentives to clear land, thus releasing large amounts of greenhouse
gases.
The criticism of food-crop-based biofuels has not been directed toward wood-based
biofuels to the same degree, at least not wood fuels from boreal forests. Even within the
research community it has been common to consider timber from boreal forests as a carbon-
neutral energy source; see for example Bright and Strømman (2009); Petersen and Solberg
(2005); Raymer (2006); Sjølie et al. (2010); Sjølie and Solberg (2009), and Zhang et al.
(2010). Especially, the possibility of producing liquid biofuels from cellulosic biomass
(second-generation biofuels) is considered a promising alternative to using food crops (Hill
et al. 2006).
This article therefore analyzes the effects of wood fuels from boreal forests with regard
to the release of CO
2
into the atmosphere. It would be reasonable to argue that wood fuels
are carbon neutral if new trees grew so fast that they replaced those that are felled a year
later or at least after only a few years. However, this is not the case in a boreal forest. Even
after 10 or 20 years, new trees are still only saplings. In typical boreal-forested areas, it
usually takes 70–120 years before a stand of trees is mature (Storaunet and Rolstad 2002).
As will be shown in Section 3, this long growth period implies that a higher level of harvest
entails a lower stock of carbon stored in the forest. Hence, the assumption that wood fuels
from boreal forests are climate neutral should be replaced with realistic assumptions about
the dynamic consequences of harvest in a boreal forest.
In this article, I use model simulations to study how increasing the harvest from the
Norwegian forest by 30%, starting in 2010 will influence the net release of CO
2
into the
atmosphere. With regard to the use of the wood harvested, two cases are considered. In the
first case, I assume that the wood is used as the raw material for manufacturing pellets. The
pellets then replace coal in power plants. This is a relevant example because power
producers in Europe are committed to acquiring and surrendering allowances for emissions
resulting from fossil-fuel combustion only, not for emissions resulting from the combustion
of bioenergy. In the case of Norway the example has special relevance, as the world’s
second-largest wood-pellet production plant (BioWood Norway) h as recently been
established on the west coast of Norway, and it will manufacture pellets on a large scale
for this purpose.
In the second example, I look at the use of wood to produce second-generation liquid
biofuels. This example is relevant as NCPA (2010) presented ambitious scenarios for the
production of second-generation liquid biofuels based on wood.
The simulations show that increasing the harvest in a boreal forest will cause a
significant initial release of CO
2
. Along the lines of Fargione et al. (2008), I label this
416 Climatic Change (2012) 112:415–428
release of CO
2
from the forest as a carbon debt. Over time, the carbon debt could be repaid
through regrowth in the harvested area and through replacement of fossil fuels as an energy
source. In this study, the payback time is found to be within the interval of 190–340 years,
depending on the type of fossil fuel that the wood fuel replaces.
The model simulations in this article are based on the properties of a Norwegian forest.
However, the conclusions are relevant to boreal forests more generally. Additionally, it
should be kept in mind that boreal forests store almost twice as much carbon as tropical
forests (Kasischke 2000, p. 20).
A closely related study is that conducted by McKechnie et al. (2011), although they
studied a temperate forest in Ontario with faster regrowth. Their model simulations have a
timeframe of 100 years, while I present simulations 400 years into the future. In the case in
which pellets replace coal in power plants, McKechnie et al. (2011) deduced a payback time
shorter than that determined in this article. However, this is as expected after taking the
more rapid growth in temperate forests into account. McKechnie et al. (2011) did not
specify the payback time when wood fuels replace liquid fossil fuels, as this was beyond
their simulation timeframe.
Another closely related report is “Biomass Sustainability and Carbon Policy Study”
published by the Manomet Center for Conservation Sciences (2010). They calculated a
payback time shorter than that determined in this study. However, the Manomet report
considered single-harvest events only. As discussed in Section 3.1 and in the supplemental
online material, this makes a large difference.
It should also be mentioned that this article analyzes the consequences of taking
bioenergy from boreal forests when the supply of bioenergy is generated through increased
harvest. Using by-products from the forest industry as bioenergy is a different and less
controversial matter, and is not discussed in this article. One should, however, be aware that
increased demand for by-products from the forest industry could increase the harvest
through indirect market effects.
The next section presents the model and crucial assumptions made. Section 3 discusses a
number of model simulations. First, Section 3.1 considers the relationship between the
harvested volume and the forest’s carbon stock in long-term steady states. Second,
Section 3.2 describes the consequences in both the short and long terms of increasing the
annual harvested volume. Third, Section 3.3 studies the net effects on CO
2
emissions when
the increased harvest considered in Section 3.2 is used as bioenergy and replaces fossil
fuels. The findings are discussed in Section 4. The supplemental online material provides
firstly a detailed model description with all parameter values. Furthermore, the online
material (1) clarifies the importance of considering not only a single-harvest event in the
case where a permanently higher harvest level is to be analyzed, and (2) presents two
scenarios where greater harvest is achieved through expansion of the harvested area, rather
than adjusting the rotation length as considered in Sections 3.2 and 3.3.
2 Model and methods
The model contains a fixed set of 75 000 parcels, each parcel covering an area of
1km
2
, and all having identical dynamic properties with regard to accumulation of dead
and living biomass. However, the time since the last clear-cutting in the parcel (i.e., the
parcels’ stand age) varies. The Norwegian forest has a high proportion of youn g stands;
see Larsson and Hylen (2007). On the basis of their work an age structure in the starting
year of the simulations is assumed such that the stand age for 37% of parcels is less than
Climatic Change (2012) 112:415–428 417
30 years, while that for 21% is more than 80 years. Hence, the stand age for 42% of
parcels is between 30 and 80 years. For further details, see the supplemental online
material.
Essential for the dynamics of the model is that immediately after clear-cutting has taken
place in a parcel, the parcel’s volume of living biomass drops to zero, and thereafter, the
growth path described in Fig. 1 begins again. The volume of living biomass in a single
parcel depends solely on the parcel’s stand age. The parcel’s productivity is fairly normal
for a boreal forest and is close to the growth path of productivity class 14 defined by
Braastad (1975 ).
An important part of the model is the module taking care of carbon stored in deadwood.
The trunks are assumed to constitute 48% of the living biomass. Hence, after clear-cutting a
significant amount of harvest residues is left on the parcel (see Fig. 1). Different “cohorts”
of natural deadwood and harvest residues are treated separately. With regard to the speed of
decomposition of deadwood, there is uncertainty. On the basis of the discussion presented
by Liski et al. (2005), I assumed that 75% of harvest residues and 70% of natural deadwood
decomposed in 50 years. After 100 years, both types of deadwood are assumed to have
decomposed completely. Owing to the uncertainty of these points, however, the
supplemental online material presents a number of sensitivity simulations with regard to
these assumptions.
The accumulation of natural dead wood in each parcel of forest after felling and
replanting is also shown in Fig. 1. Natural losses are low until the trees are 80–90 years old,
increase sharply until the trees are 150 years old, and then gradually stabilize. Note that the
periodization (time unit) in the model is 5 years.
Changes in the stock of biomass in old forests are uncertain (Carey et al. 2001; Pregitzer
and Euskirchen 2004; Seely et al. 2002). It is here assumed that there is a constant volume
biomass of older stands; see Fig. 1. As more recent data suggest that old growth forests are
significant carbon sinks (see Luyssaert et al. 2008), this might be considered conservative.
Assuming that older stands continue to accumulate carbon would imply a longer payback
0
10
20
30
40
50
60
70
80
0
2
4
6
8
10
12
14
16
0 50 100 150 200 250 300 350 400
1000 m3 wood
Stock of dead and living wood (1000 tonnes carbon)
Stand a
g
e
Natural
dead
wood
Harvest
residues
Living
wood
Fig. 1 A single parcel. The development of the volumes of living wood, harvest residues and natural
deadwood after clear-cutting and replanting in the standard parcel. Stand age at time of last felling was
95 years
418 Climatic Change (2012) 112:415–428
time of the carbon debt than found in this article. However, owing to the uncertainty of this
point, and claims that the probability of carbon loss due to different types of disturbances is
greater in older forests, Holtsmark (2010) assumed, as a very conservative estimate, that the
stock of wood and thus the carbon stored in the biomass decline substantially as a parcel
ages. Holtsmark (2010) therefore deduced a payback time shorter than that determined in
this article.
An important point is the role of soil as a carbon sink. In a boreal forest, in contrast
to a tropical rain forest, a large share of the carbon is stored in the soil. According to
Kjønaas et al. (2000), more than 80% of carbon in Norwegian forests is stored in the soil.
An important question is therefore whether harvest is likely to t rigger the r elease of
carbon from soil. As underlined by Fontaine et al. (2007); Friedland and Gillingham
(2010), and Nilsen et al. (2008), the accumulation and possible release of carbon from the
soil are complicated processes that are not easily modeled. I have therefore chosen to
ignore the possibility that harvesting may reduce the capacity of the soil as a carbon sink,
and have considered only how harvesting influ ences the stock of carbon stored in dead
and living biomass, although Nakane and L ee (1995) and Nilsen et al. (2008) suggest that
clear-cutting might trigger the release of carbon from soil, especially if tops and branches
are harvested in addition to trunks. This last point is further underlined by Holmgren et al.
(2007); Kujanpää et al. (2010 ); Kirkinen et al. ( 2008); Palosuo et al. (2001); Repo et al.
(2011), and Schlamadinger et al. (1995). Hence, the estimated payback time of the carbon
debt would probably have been longer if I had relied on less conservative assumptions at
this point.
The carbon content of a cubic meter of biomass depends on the wood’s density. I assume
throughout a density of 423 kg/m
3
, and that half of the mass is carbon. This gives 0.211
tonnes of carbon per m
3
, or 0.774 tonnes CO
2
per m
3
wood used as fuel. For further details,
see the supplemental online material.
3 Results
3.1 Relationship between the length of the rotation cycle and the carbon stock in different
steady states
Figure 2 provides relevant information for a discussion on carbon neutrality. It shows the
volume of timber felled annually (curve) and the entire forest’s stock of carbon in dead and
living wood (columns) for rotation cycles of different lengths (horizontal axis).
It should be noted that Fig. 2 considers the entire forest area of 75000 km
2
and shows
the forest in different steady states, in the sense that the length of the rotation cycle is
constant and has been constant for so long that both the harvest and standing volume are
also constant over time.
Figure 2 confirms that the maximum harvest (volume felled) is obtained with a rotation
cycle of 90 years. The carbon stock of dead and living biomass, on the other hand,
monotonously increases as the rotation period is extended.
The stock of carbon is almost twice as large with a 250-year rotation cycle as with a 90-year
rotation cycle. This may seem difficult to reconcile with Fig. 1, which shows that the stock of
carbon in a single parcel increases by less than 25% as the stand age increases from 90 to
250 years. The explanation is that as the length of the rotation cycle increases, there are fewer
and fewer parcels that have recently been clear-cut. In other words, if the rotation cycle is
long, a large share of the parcels will at any point in time carry a large stock of wood. If the
Climatic Change (2012) 112:415–428 419
rotation cycle is short, on the other hand, a large proportion of the forest will at any point in
time be relatively recently felled, and its stock of wood will be correspondingly small.
Figure 2 indicates that it is misleading to claim that wood provides carbon-neutral
bioenergy, even in the long term. It shows that if the harvest is permanently large, which
requires short rotation cycles, the carbon pool of the forest will be permanently small. In
contrast, if the annual harvest is small, with correspondingly long rotation cycles, the
carbon pool will be permanently large. With a 90-year rotation cycle, for example, an area
of 833 km
2
can be felled each year, giving an annual harvest of 22.5 million cubic meters
(Mm
3
) of timber and 467 million tonnes of carbon (MtC) stored in dead and living wood.
With a 250-year rotation cycle, 300 km
2
can be felled annually and the annual harvest is
only 9.5 Mm
3
; the carbon stored in dead and living wood, on the other hand, rises to 933
MtC (see Table 1).
In other words, increasing the harvest to a higher level on a permanent basis does not
merely result in a temporary drop in the forest’s carbon stock that will in the long term be
entirely counterbalanced by CO
2
uptake by the forest. On the contrary, a permanent
increase in the harvest results in a permanently lower forest carbon stock. Hence, increasing
the harvest level is not a carbon-neutral change, either in the short term or in the long term.
This also shows that the question of carbon neutrality cannot be resolved by studying the
effect of a single harvest in a single year with subsequent planting. Such a single-event
perspective is an oversimplification that does not incorporate the important long-term
0
200
400
600
800
1000
0 50 100 150 200 250 300 350 400
Len
g
th of rotation periods
Stock of dead and living wood (MtC)
0
5
10
15
20
Annual harvest (Mm3/year)
Natural
deadwood
Harvest
residues
Living wood
Volume
felled
(annually)
Fig. 2 The entire forest. Annual volume of timber felled (black curve) and quantity of carbon stored in dead
and living wood (columns) in different steady states for rotation cycles of different lengths
Length of the
rotation cycle
(years)
Annual
harvest
(Mm
3
)
Area
harvested
(km
2
/year)
Carbon stored in dead
and living biomass
(MtC)
90 22.5 833 467
250 9.5 300 933
Table 1 Example of two
different steady states
420 Climatic Change (2012) 112:415–428
dynamic effects on the forest’s carbon stock (see the supplemental online material for
further discussion).
At this point, the report on biomass sustainability presented by the Manomet
Center for Conservation Sciences (2010) should be mentioned. This report considers
only a single-harvest event rather than conducting “…a more complicated series of
repeated harvest entries.” (ibid. p. 85). The M anomet report is important along different
lines, as it contains new information and provides an interesting discussion on the carbon
neutrality of wood energy. However, the report’s relatively optimistic conclusions with
regard to the time lag between harvest, the released volume of carbon dioxide, and the
payback time of the carbon debt reflect the report’s single-harvest approach and therefore
do not take into account important features of the long-term effect of a hig her level of
harvest on the forest’s carbon stock. As shown in the supplemental online material,the
payback time mo re t han doubles if a series of subsequent h arvest events are considered
instead of a single-harvest event.
3.2 Short-term and long-term effects of increasing the harvest
The previous section considered the carbon stock of a forest in a steady state in the sense
that the rotation cycles were constant over time, giving the forest an even age structure.
This section describes the short-term and long-term effects of increasing the harvest in a
forest with an uneven age structure similar to the age structure found for a Norwegian
forest.
In the reference scenario, the annual harvest is 10 Mm
3
and no residues are harvested.
This is compared with a scenario where the annual harvest increases by 3 Mm
3
to 13 Mm
3
,
with 2010 as the first year of increased harvest. In addition, this scenario assumes that
0.6 Mm
3
of residues is harvested annually.
The chosen numerical example has relevance, as the annual harvest from Norwegian
forests has varied around 10 Mm
3
for several decades (NCPA 2010). However, the
Norwegian government wants to increase the harvest to increase the supply of bioenergy,
and an increase in the annual harvest to 13 Mm
3
is frequently discussed; see NCPA (2010).
What effect does a higher harvest level have?
Given the assumed age structure, with a large share of the parcels having a low stand
age, and because the annual harvest is limited to 10 Mm
3
in the reference scenario, the
forest’s carbon stock increases until the end of the 22nd century (see Fig. 3). Moreover,
even if the harvest is increased to 13 Mm
3
, the forest’s carbon stock increases over this
period, although at a lower rate. Recall here that the maximum harvest that could be
sustained in a steady state is 22.5 Mm
3
/year; see Fig. 2.
It is important to note that Fig. 2 shows different steady-state situations for rotation
cycles of different lengths, whereas Fig. 3 shows the transition from the current state of the
forest to towards a new steady state.
For further clarific ation, consider the ref erence (small-har vest) scenario as
illustrated by the upper curve in Fig. 3. The figure shows that b y 2100, the stock of
carbon has more than doub led from the 2005 level to 857 MtC (see a lso Table 2).
However, Figure 3 also shows that in the larger-harvest scenario, the carbon stock is only
775 MtC in 2100. In other wo rds, the increase in the harvest has resulted in a carbon
stock th at is 82 MtC lower in 2100 than it would have been with the lower harvest level.
More generally, the vertical distance between the upper and lower curves in Fig. 3 shows
the n et reduction in the forest’s carbon stock as a result of the greater harvest; see also the
red curve in Fig. 4.
Climatic Change (2012) 112:415–428 421
3.3 Reducing the use of fossil energy by increasing the use of wood energy
The previous section considered how felling affects the stock of carbon stored in the forest’s
dead and living wood. The argument for felling more timber is precisely that using wood as
a source of bioenergy can reduce the use of fossil energy and thus cut CO
2
emissions. In
this section, I consider the extent to which increasing the timber harvest for bioenergy
production can replace the use of fossil energy. Taking account of both the replacement
effect on fossil fuel combustion and the effects on the forest’s carbon stock, it is possible to
calculate the net CO
2
effects of increased logging.
The quantity of fossil energy that wood fuels can replace varies widely depending on
precisely which technologies are involved. Here I discuss two examples that show how two
different types of wood fuels will affect net CO
2
emissions. In the first case, I have assumed
that the wood is used as the raw material for manufacturing pellets. The pellets then replace
coal in power plants. In the second example, I look at the use of wood to produce second-
generation liquid biofuels.
The exact volumes of CO
2
emissions that can be eliminated using wood energy are of
great importance. On the basis of the works of Sjølie and Solberg (2009) and Weisser
(2007), I have assumed that using 1 m
3
of wood, processed to pellets, instead of coal in a
power plant can eliminate 0.5 tonnes of fossil-generated CO
2
emissions. The method used
to calculate this figure is described in further detail in the supplemental online material.
400
500
600
700
800
900
2000 2100 2200 2300 2400
Million tonnes carbon
Reference scenario (low harvest) Hi
g
her harvest level with use of fellin
g
waste
Fig. 3 Carbon stored in wood in the two scenarios considered in Sections 3.2 and 3.3
Table 2 Carbon stock, emission reductions, and remaining carbon debt in the scenarios considered in
Section 3.2. All figures are in millions of tonnes of carbon
2005 2100 2200 2300
Carbon stored in biomass in reference (small-harvest) scenario 417 857 915 918
Carbon stored in biomass in large-harvest scenario 417 775 820 822
Drop in carbon stock due to increased harvest – 82 95 96
Accumulated reductions in fossil carbon emissions—wood fuels replace coal – 46 95 144
Remaining carbon debt—wood fuels replace coal − 36 0 −48
Accumulated reductions in fossil carbon emissions—wood fuels replace oil – 26 54 81
Remaining carbon debt—wood fuels replace oil – 56 42 15
422 Climatic Change (2012) 112:415–428
To calculate the volume of fossil emissions that can be eliminated using second-
generation biofuels, I followed NCPA (2011, p. 32). This report concludes that using 1 m
3
of wood processed to second-generation liquid biofuel can eliminate 0.28 tonnes of CO
2
emissions generated through the combustion of fossil fuels.
We are now ready to calculate the net effect of the increased harvest on
accumulated CO
2
emissions. Recall that the overall increase in the annual harvest is
3.6 Mm
3
when a share of the residues is also harvested. From Fig. 4 it is possible to gain
a visual impression of the results by comparing the lines with the curve. The red curve in
Fig. 4, the elevation of which is equal to the vertical distance between the curves in
Fig. 3, shows the difference in the carbon stock of the forest between the large- and small-
harvest scenarios. The lines show the accumulated fossil CO
2
emissions that can be
eliminated by increasing the volume of timber harvested and using this harvest to replace
fossil fuels in the two ways discussed. The remaining carb on debts in the two cases
considered are equal to the vertical distance between the red curve and the two lines in
Fig. 4. As long as the curve is above the considered l ine, the remaining carbon debt is
positive. When the line is above the curve, the carbon debt has been fully repaid and a
carbon dividend is collected, using the term introduced by Manomet Center for
Conservation Sciences (2010).
Firstly, consider the case where the wood is processed to pellets and replaces coal in a
power plant. In that case, each cubic meter of wood eliminates 0.5 tonnes of fossil-
generated CO
2
emissions. This means that 1.8 MtCO
2
or 0.5 MtC of fossil emissions are
eliminated each year. Hence, by 2100, fossil CO
2
emissions corresponding to 46 MtC have
been eliminated (see Table 2 and the broken line in Fig. 4).
As mentioned, the increased harvest means that the carbon stock of the forest by 2100 is
82 MtC less than it would have been if the annual harvest had been maintained at 10 Mm
3
.
Subtracting 46 MtC (the accumulated drop in fossil carbon emissions), it follows that in the
0
20
40
60
80
100
120
2000 2050 2100
2150
2200 2250 2300 2350 2400
Million tonnes carbon
Drop in the forest carbon stock due to increased logging
Accumulated reduction in carbon emissions from coal combustion due to increased
supply of pellets
Accumulated reduction in carbon emissions from fossil fuels due to increased supply of
liquid biofuels
Fig. 4 The two straight lines show the accumulated reductions in CO
2
emissions from combustion of fossil
energy achieved by increasing the supply of bioenergy through a higher harvest level. The red curve in Fig. 4
shows the difference in the carbon stock between the small- and large-harvest scenarios
Climatic Change (2012) 112:415–428 423
pellets case, the remaining carbon debt in 2100 is 36 MtC. In other words, although
increasing the harvest eliminates fossil CO
2
emissions from coal combustion corresponding
to 46 MtC, the net accumulated release of carbon to the atmosphere will be 36 MtC higher
in the period 2010–2100 in the large-harvest scenario than in the reference (small-harvest)
scenario.
The development o f th e remaining carbon debt is shown in Fig. 5. Note that the
elevations of the curves in Fig. 5 are equal to the vertical distances between the red curve
and the two lines in Fig. 4. Figure 5 shows that i n the pellet case, the remaining carbon
debt is declining in 2100 and becomes negative around 2200, that is, 190 years after the
increase in the harvest. Hence, the analysis suggests that increasing the harvest and the
use of wood fuels to replace coal in power plants could, for a long period of time, result
in significantly greater CO
2
emissions than the combustion of the coal that the increased
harvest replaces.
Secondly, consider the effect of using the extra harvest of wood as raw material in the
production of second-generation liquid biofuels. In that case, each cubic meter of wood
eliminates 0.28 tonnes of fossil-generated CO
2
emissions. This means that 1.01 MtCO
2
or
0.27 MtC of fossil emissions are eliminated each year. Hence, by 2100, fossil CO
2
emissions corresponding to 26 MtC have been eliminated; see the level of the lower line in
2100 in Fig. 4, which shows the accumulated reduction in fossil carbon emissions in this
case.
The lower line in Fig. 4 is below the red c urve through out the 21
st
,22
nd
, and 23
rd
centuries, and crosses the red curve in 2350. Thus, the calculations indicate that using
second-generation liquid biofuels produced from boreal timber rather than continuing to
use fossil diesel may actually increase the net accumulated release of CO
2
for 340 years,
i.e. generate a carbon debt that w ill be repaid only after more than three centuries. This is
in contrast to the results obtained by assuming that wood is a climate-neutral fuel, where
the effect on the forest’s carbon stock is completely ignored and a carbon dividend i s thus
assumed to be generated from day one and to be equal to the elevation of the lines in
Fig. 4.
-100
-75
-50
-25
0
25
50
75
2000 2050 2100 2150 2200 2250 2300 2350 2400
Million tonnes carbon
Remaining carbon debt –second-generation biofuels replace petrol
Remainin
g
carbon debt –pellets replace coal
Fig. 5 Development of the remaining carbon debt due to increased harvest when the increased harvest is
used to replace fossil fuels
424 Climatic Change (2012) 112:415–428
4 Discussion and conclusion
Bioenergy is usually considered carbon neutral and an important part of a strategy for
reducing CO
2
emissions; see for example IPCC (2000). The generation of biomass in boreal
forests is significant and could potentially serve as an important source for an increased
supply of bioenergy.
The traditional assumption that wood fuels are carbon neutral would have been
appropriate if trees harvested were replaced by new trees within a short period of time.
However, the typical life cycle of a spruce tree in boreal forests includes a growth phase
that lasts about 100 years and then a phase in which the mass remains relatively stable for a
further 100 years. The tree then dies, but remains standing for about 30 years before falling
to the ground and gradually decaying over the course of the following 100 years (Storaunet
and Rolstad 2002).
From that point of view, it appears obvious that harvest from boreal forests is not a
climate-neutral activity. However, as should be made clear from the previous sections, forest
dynamics cannot be understood by studying individual trees or a single harvest and
subsequent regeneration of the felled trees. I have therefore described and simulated a
dynamic model of the Norwegian forest.
In the simulations presented in this article, the harvest is increased by 30%, while the
forest increment is still positive over the whole of the 21st century. Nevertheless, the
increase in the harvest means that the carbon stock in the stylized forest stabilizes at a
different level, as would be expected. Hence, even if the forest increment is positive, wood
harvesting and combustion are not carbon-neutral activities.
The article also presents calculations illustrating the net effect on the CO
2
release of
increased logging when the biomass made available replaces fossil fuels as energy sources.
Two examples are described: processing wood to pellets for use in coal-fired power plants,
and processing wood to liquid biofuels, which are used to replace fossil oil.
The article’s first main finding is that increasing the use of wood from a boreal forest to
replace coal in power plants will create a carbon debt that will only be repaid after
approximately 190 years. Secondly, if the wood is used to produce second-generation liquid
biofuels and replaces fossil diesel, the payback time of the carbon debt is estimated to be
340 years.
In addition, it is important to remember that the analyses presented here do not take into
account the effect of providing subsidies for various alternative forms of energy as a means
of reducing the use of fossil energy. Such subsidies tend to increase overall energy use; see
for example Hutchinson et al. (2010). If this is taken into account, the emission-increasing
effect of using wood as energy will become even more pronounced. A complete analysis
should also include such effects.
An uncertain aspect of the parameterization of the model is the determination of changes
in the volume of deadwood over time. Sensitivity calculations were therefore carried out to
test the effect of varying the rate of decay for deadwood. These simulations show that the
parameterization is not a critical factor (see the supplemental online material).
Nevertheless, it should be stressed that the purpose of this article is not to provide
definitive answers but to draw attention to the importance of taking both short- and
long-term dynamic e ffects of increasing the timber harvest more fully into account
when evaluating the effect o n emissions of increasing the use of energy from wood
combustion. The analysis carri ed out here could be improved along several
dimensions, not least considering more heterogeneous forests with areas of different
productivity while taking fu ll account of how harvest in boreal forests influences the
Climatic Change (2012) 112:415–428 425
amount of carbon stored in the soil. Anoth er topic for future research should be the
capacity of old growth forests as carbon stores when the likeliness of disturbance is
taken into account, and how the frequency of disturbance migh t change in a warmer
climate. And finally, it should be noted that this paper calculate estimates of the
effects on accumulated net release of CO
2
into the atmosphere along different time
horizons only. To provide a complete picture of the climatic effects of this would require a
model of the relationship between net release of CO
2
, its persistence in the atmosphere
and the corresponding effects on radiative forcing, a s well as possible albedo effects of
clear-cutting in boreal forests.
It should also be underlined that the analysis presented here does not make arguments
against the use of bioenergy from boreal forests in general. If bioenergy is obtained through
increased use of residues from different forest-related industries, the CO
2
effects are
probably favorable.
Nevertheless, the commonly applied assumption that wood fuels are climate neutral is
not tenable. If this assumption is reevaluated, it may also be necessary to reevaluate the
current taxes and subsidies that apply to bioenergy and forestry. It is not at all clear whether
current policy takes sufficient account of the potential of forests as carbon sinks or of the
fact that burning wood results in CO
2
emissions. As highlighted by Searchinger et al.
(2009), for example, putting a high price on CO
2
emissions from fossil energy emissions
while considering bioenergy to be carbon neutral would create strong incentives to clear
land.
The claim that using wood fuels is carbon neutral is based on the approximation that
logging has a negligible effect on the forest’s carbon stock. This would be a reasonable
approximation if there were a negligible time lag between felling and full regrowth. The
carbon-neutrality claim ignores the significance of this time lag, and the dynamics that
follow, not least that there will be a permanent reduction in the stock of both dead and
living biomass in the forest if the harvest is permanently increased. Thus, making the
common assumption that using wood as bioenergy is carbon neutral also means that it is
assumed that all the effects on the forest’s carbon stock are so small that they can be
ignored.
Acknowledgements Ketil Flugsrud, Hans Goksøyr, Henrik Lindgaard, Olav Norem, Hans Henrik Ramm,
and Trygve Refsdal provided important assistance and ideas at an early stage of this work. I also wish to
thank four anonymous referees as well as Iulie Aslaksen, Rasmus Astrup, Lise Dalsgaard, Per Arild
Garnåsjordet, Cathrine Hagem, Per Kr. Rørstad, Hanne K. Sjølie, and participants at a seminar held at the
University for Life Sciences in October 2010 for comments and suggestions that improved the manuscript
significantly. The Norwegian Research Council provided funding as part of the project Environmentally
friendly transport: How to design policies for sustainable introduction of biofuels.
References
Braastad H (1975) Yield Tables and Growth Models for Picea abies. Reports from The Norwegian Forest
Research Institute 31.9
Bright RM, Strømman AH (2009) Life cycle assessment of second generation bioethanol produced from
Scandinavian boreal forest resources. J Ind Ecol 13:514–530
Carey EV, Sala A, Keane R, Callaway RM (2001) Are old forests underestimated as global carbon sinks?
Global Change Bio 7:339–344
426 Climatic Change (2012) 112:415–428
Fargione J, Hill J, Tilman D, Polasky S, Hawthorne P (2008) Land clearing and the biofuel carbon debt.
Science 319:1235–1238
Fontaine S, Barot S, Barré P, Bdioui N, Mary B, Rumpel C (2007) Stability of organic carbon in deep soil
layers controlled by fresh carbon supply. Nature 450:277–281
Friedland AJ, Gillingham KT (2010) Carbon accounting a tricky business. Science 327:411–412
Gibbs HK, Ruesch AS, Achard F, Clayton MK, Holmgren P, Ramankutty N, Foley JA (2010) Tropical forests
were the primary sources of new agricultural land in the 1980s and 1990s. P Natl Acad Sci 107:16732–
16737
Gurgel AJ, Reilly M, Paltsev S (2007) Potential land use implications of a global biofuels industry. J Agr
Food Ind Organ 5:1–34
Hill J, Nelson E, Tilman D, Polasky S, Tiffany D (2006) Environmental, economic, and energetic costs and
benefits of biodiesel and ethanol biofuels. P Natl Acad Sci 103:11206–11210
Holmgren K, Eriksson E, Olsson O, Olsson M, Hillring B, Parikka M (2007) Biofuels and climate neutrality—
system analysis of production and utilisation. Elforsk report 07:35
Holtsmark B (2010) Use of wood fuels from boreal forests will create a biofuel carbon debt with along
payback time. Discussion Paper 637. Statistics Norway
Hutchinson E, Kennedy PW, Martinez C (2010) Subsidies for the Production of Cleaner Energy: When do
they Cause Emissions to Rise? The B.E. Journal of Economic Analysis & Policy: Vol. 10, Iss. 1, Article
28
IPCC (2000) Special report on emissions scenarios. Working Group III, Intergovernmental Panel on Climate
Change (IPCC), Cambridge University Press, Cambridge
Kasischke ES (2000) Boreal ecosystems in the global carbon cycle. In: Kasischke ES, Stocks BJ (eds) Fire,
climate and carbon cycling in the boreal forest. Springer, New York
Kirkinen J, Palosuo T, Holmgren K, Savolainen I (2008) Greenhouse impact due to the use of combustible
fuels: life cycle viewpoint and relative radiative forcing commitment. Environ Manage 42:458–469
Kjønaas O, Aalde JH, Dalen LS, de Wit HA, Eldhuset T, Øyen BH (2000) Carbon stocks in Norwegian
forested systems. Biotechnol Agron Soc Environ 4:311–314
Kujanpää M, Eggers J, Verkerk H, Helin T, Lindner M, Wessman H (2010) Carbon balance of forest residue
collection and combustion in Southern Finland. Proceeding paper from the 18th European Biomass
Conference and Exhibition
Lapola D, Schaldach MR, Alcamo J, Bondeau A, Koch J, Koelking C, Priess JA (2010) Indirect land-use
changes can overcome carbon savings from biofuels in Brazil. P Natl Acad Sci 103:11206–11210
Larsson JY, Hylen G (2007) Statistics of forest conditions and forest resources in Norway. Reports from The
Norwegian Forest Research Institute 1/07
Liski TP, Peltoniemi M, Sievänen R (2005) Carbon and decomposition model Yasso for forest soils. Ecol
Model 189:168–182
Luyssaert S, Schulze ED, Börner A, Knohl A, Hessenmöller D, Law BE, Ciais P, Grace J (2008) Old-growth
forests as global carbon sinks. Nature 455:213–215
Manomet Center for Conservation Sciences (2010) Massachusetts Biomass Sustainability and Carbon Policy
Study: Report to the Commonwealth of Massachusetts Department of Energy Resources. In Walker T
(ed) Natural Capital Initiative Report NCI-2010-03. Brunswick, ME, USA
McKechnie J, Colombo S, Chen J, Mabee W, Maclean HL (2011) Forest bioenergy of forest carbon? Assessing
trade-offs in greenhouse gas mitigation with wood-based fuels. Environ Sci Technol 45:789–795
Melillo JM, Reilly JM, Kicklighter DW et al (2009) Indirect emissions from biofuels: how important?
Science 326:1397–1399
Nakane K, Lee NJ (1995) Simulation of soil carbon cycling and carbon balance following clear-
cutting in a mid-temperate forest and contribution to the sink of atmospheric C O
2
. Vegetatio
121:147–156
NCPA (2010) Tiltak og virkemidler for å nå norske klimamål mot 2020 (Efforts and policies for Norwegian
climate goals toward 2020). The Norwegian Climate and Pollution Agency (NCPA), Norwegian Water
Resources and Energy Directorate, The Norwegian Petroleum Directorate, Statistics Norway, The
Norwegian Public Road Administration. Report TA2590
NCPA (2011) Skog som biomasseressurs (Forests as biomass resources). Report from The Norwegian
Climate and Pollution Agency. Report TA 2762
Nilsen P, Hobbelstad K, Clarke N (2008) Opptak og utslipp av CO
2
i skog (Capture and Emission of CO
2
from Norwegian Forests). Norwegian Forest and Landscape Institute, report no. 06/2008
Palosuo T, Wihersaari M, Liski J (2001) Net greenhouse gas emissions due to energy use of forest residues—
impact of soil carbon balance. EFI Proceedings no 39, Wood biomass as an energy source challenge in
Europe. European Forest Institute, Joensuu, 115–130
Climatic Change (2012) 112:415–428 427
Petersen AK, Solberg B (2005) Environmental and economic impacts of substitution between wood products
and alternative materials: a review of micro-level analyses from Norway and Sweden. Forest Policy
Econ 7:249–259
Pregitzer KS, Euskirchen ES (2004) Carbon cycling and storage in world forests: biome patterns related to
forest age. Glob Change Biol 10:2052–2077
Raymer AKP (2006) A comparison of avoided greenhouse gas emissions when using different kinds of wood
energy. Biomass Bioenerg 30:605–617
Repo A, Tuomi M, Liski J (2011) Indirect carbon dioxide emissions from producing bioenergy from forest
harvest residues. GCB Bioenergy 3:107–115
Schlamadinger B, Spitzer J, Kohlmaier GH, Lüdeke M (1995) Carbon balance of bioenergy from logging
residues. Biomass Bioenerg 8:221–234
Searchinger TD, Heimlich TD, Houghton RA et al (2008) Use of US croplands for biofuels increases
greenhouse gas through emissions from land-use change. Science 319:1238–1240
Searchinger TD, Hamburg SP, Melillo J et al (2009) Fixing a critical climate accounting error. Science
326:527–528
Seely B, Welham C, Kimmins H (2002) Carbon sequestration in a boreal forest ecosystem: results from the
ecosystem simulation model, FORECAST. Forest Ecol Manag 169:123–135
Sjølie HK, Solberg B (2009) Greenhouse Gas Implications by Production of Wood Pellets at the BioWood
Norway plant at Averøy, Norway. Report drawn up for BioWood Norway. Department of Ecology and
Natural Resource Management, Norwegian University of Life Sciences
Sjølie HK, Trømborg E, Solberg B, Bolkesjø TF (2010) Effects and costs of policies to increase bioenergy
use and reduce GHG emissions from heating in Norway. Forest Policy Econ 12:57–66
Storaunet KO, Rolstad J (2002) Time since death and fall of Norway spruce logs in old-growth and
selectively cut boreal forest. Can J For Res 32:1801–1812
Weisser D (2007) A guide to life-cycle greenhouse gas (GHG) emissions from electric supply technologies.
Energy 32:1453–1559
Wise M, Calvin K, Thomson A, Clarke L, Bond-Lamberty B, Sands R, Smith SJ, Janetos A, Edmonds J
(2009) Implications of limiting CO
2
concentrations for land use and energy. Science 324:1183–1186
Zhang Y, McKechnie J, Cormier D, Lyng R, Mabee W, Ogino A, MacLean HI (2010) Life cycle emissions
and cost of producing electricity from coal, natural gas, and wood pellets in Ontario, Canada. Environ
Sci Technol 44:538–544
428 Climatic Change (2012) 112:415–428
