Environ. Res. Lett. 13(2018) 015007

https://doi.org/10.1088/1748-9326/aaa512

LETTER

Does replacing coal with wood lower CO₂ emissions?

OPEN ACCESS

Dynamic lifecycle analysis of wood bioenergy

RECEIVED

29 August 2017

John D Sterman1,4

, Lori Siegel² and Juliette N Rooney-Varga³

REVISED

MIT Sloan School of Management, 100 Main Street, Cambridge, MA 02139, United States of America

1 December 2017

Climate Interactive, 1201 Connecticut Avenue NW, Suite 300, Washington, DC, 20036, United States of America

ACCEPTED FOR PUBLICATION

UMass Lowell Climate Change Initiative and Deptartment of Environmental, Earth, and Atmospheric Sciences, 265 Riverside Street,

4 January 2018

Lowell, MA 01854, United States of America

PUBLISHED

Author to whom any correspondence should be addressed.

18 January 2018

E-mail: jsterman@mit.edu

Original content from

Keywords: bioenergy, biofuels, wood pellets, greenhouse gas emissions, climate change, system dynamics

this work may be used

under the terms of the

Supplementary material for this article is available online

Creative Commons

Attribution 3.0 licence.

Any further distribution

Abstract

of this work must

maintain attribution to

Bioenergy is booming as nations seek to cut their greenhouse gas emissions. The European Union

the author(s) and the

title of the work, journal

declared biofuels to be carbon-neutral, triggering a surge in wood use. But do biofuels actually reduce

citation and DOI.

emissions? A molecule of CO₂ emitted today has the same impact on radiative forcing whether it

comes from coal or biomass. Biofuels can only reduce atmospheric CO₂ over time through

post-harvest increases in net primary production (NPP). The climate impact of biofuels therefore

depends on CO₂ emissions from combustion of biofuels versus fossil fuels, the fate of the harvested

land and dynamics of NPP. Here we develop a model for dynamic bioenergy lifecycle analysis. The

model tracks carbon stocks and fluxes among the atmosphere, biomass, and soils, is extensible to

multiple land types and regions, and runs in ≈1s, enabling rapid, interactive policy design and

sensitivity testing. We simulate substitution of wood for coal in power generation, estimating the

parameters governing NPP and other fluxes using data for forests in the eastern US and using

published estimates for supply chain emissions. Because combustion and processing efficiencies for

wood are less than coal, the immediate impact of substituting wood for coal is an increase in

atmospheric CO₂ relative to coal. The payback time for this carbon debt ranges from 44-104 years

after clearcut, depending on forest type—assuming the land remains forest. Surprisingly, replanting

hardwood forests with fast-growing pine plantations raises the CO₂ impact of wood because the

equilibrium carbon density of plantations is lower than natural forests. Further, projected growth in

wood harvest for bioenergy would increase atmospheric CO₂ for at least a century because new

carbon debt continuously exceeds NPP. Assuming biofuels are carbon neutral may worsen

irreversible impacts of climate change before benefits accrue. Instead, explicit dynamic models should

be used to assess the climate impacts of biofuels.

1. Introduction

Stupak et al 2007). The United Kingdom subsidizes

wood pellets for electric power generation and has

Limiting global warming to no more than 2^◦ C requires

become the world’s largest pellet importer

(Thrän

large, rapid cuts in fossil fuel consumption by mid-

et al 2017). The US federal government and a number

century (Figueres et al 2017, IPCC 2014). In response,

of US states are considering whether to declare wood

governments around the world are promoting biomass

fuels carbon-neutral or to promote their use (Corn-

to reduce their greenhouse gas (GHG) emissions. The

wall 2017), while at COP23 in Bonn ‘China and 18

European Union declared biofuels to be carbon-neutral

other nations representing half the world’s population

to help meet its goal of 20% renewable energy by

said…they planned to increase the use of wood...to gen-

2020, triggering a surge in use of wood for heat and

erate energy as part of efforts to limit climate change’

electricity (European Commission 2003, Leturcq 2014,

(Biofuture Platform 2017, Doyle and Roche 2017).

Environ. Res. Lett. 13 (2018) 015007

But do biofuels actually reduce GHG emissions?

Dynamic analysis is required to answer the second

The appeal is intuitive: fossil fuels inject carbon

question (e.g. Helin et al 2013). The carbon cycle and

sequestered in geological reservoirs for millions of years

climate impacts of bioenergy involve multiple stocks

into the atmosphere, where it accumulates and causes

of carbon (e.g. in biomass, soils and dead organic

global warming (IPCC 2013). In contrast, biofuels recy-

matter, and the atmosphere) and the processes that

cle carbon from the atmosphere, helping to keep fossil

control the flow of carbon among those stocks includ-

carbon in the ground (IPCC 2013).

ing NPP, transfer of carbon from biomass to soil,

However, a molecule of CO₂ added to the atmo-

decomposition of organic matter, consumption and

sphere today has the same impact on radiative forcing

respiration of carbon in biomass and soils, etc. Tools

and warming whether it came from coal millions of

are needed to assess the dynamic climate impact of

years old or biomass grown last year. Biofuels can only

bioenergy over policy-relevant time horizons. Because

reduce atmospheric CO₂ over time by increasing net

of the uncertainty and debate over the impacts of biofu-

primary production (NPP) above what it otherwise

els, such tools should allow users to examine alternative

would have been (DeCicco 2013). Assessing the climate

assumptions and scenarios easily and quickly, and

impact of wood and other biofuels therefore depends

would avoid the need to use static summary metrics

on two critical questions: first, at the point of combus-

such as global warming potentials (GWP) and con-

tion, do biofuels generate more or less CO₂ per unit of

tentious debate over the appropriate time horizon for

end-use energy than fossil fuels? Second, what are the

these approximations, e.g. whether to use GWP20 or

dynamics of biomass (re)growth and how do NPP and

GWP100 (Ocko et al 2017).

carbon fluxes from biomass and soils depend on the

To address this need we developed an interactive

fate of the harvested land?

decision-support model that enables policymakers and

Confusion over these questions has caused the sci-

other stakeholders to explore the dynamic impact of

entific debate over the climate impact of bioenergy and,

biofuels on carbon emissions and climate. The model

especially wood, to remain ‘contentious’ (Creutzig et al

is fully documented, freely available, runs in about a

2015, Ter-Mikaelian et al 2015). The wood industry

second on ordinary laptops and is extensible to any

and many governments promote wood as a renewable,

number of land use categories and spatial scales. Users

carbon-neutral fuel, while many environmental groups

receive immediate feedback on the impacts of their sce-

oppose wood bioenergy because it causes deforesta-

narios and assumptions. Here we describe the model

tion, harming natural carbon sinks, ecosystems, and

and use it to explore the dynamics of substituting

biodiversity (Cornwall 2017). Advocates emphasize a

wood for coal in electric power production, using wood

long time horizon to evaluate the impact of biofu-

sourced from a range of forest types in the US to esti-

els, a century or more, by which time it is assumed

mate model parameters governing NPP and carbon

forests will regrow, offsetting initial emissions. Oppo-

fluxes.

nents point to the potential for wood energy to increase

CO₂ levels in the short run, incurring a ‘carbon debt’

that can only be paid off slowly, and worry that the

2. Methods

resulting increase in atmospheric CO₂

will worsen

global warming and lead to irreversible impacts before

2.1. Model structure

the benefits of new growth can occur (Brack 2017,

We build on the widely-used C-ROADS climate policy

Buchholz et al 2016, Cornwall 2017).

model (Sterman et al 2012, Sterman et al 2013), devel-

Life cycle analysis is commonly used to answer

oping a more detailed representation of land use, the

the first question. Results vary with the assumed sys-

carbon stocks associated with different types of land

tem boundary and biofuel harvesting, processing and

and the fluxes arising from them. C-ROADS is a mem-

transport methods (e.g. Buchholz et al 2016). How-

ber of the family of simple climate models, consisting

ever, although wood has approximately the same

of a system of differential equations representing the

carbon intensity as coal (0.027 vs. 0.025 tC GJ⁻¹ of pri-

carbon cycle, budgets and stocks of GHGs, radiative

mary energy; see supplementary material), combustion

forcing and the heat balance of the Earth. C-ROADS

efficiency of wood and wood pellets is lower (Nether-

closely replicates GHG concentrations, global mean

lands Enterprise Agency; IEA 2016). Estimates also

surface temperature, and other climate metrics from

suggest higher processing losses in the wood supply

1850, and matches CMIP5 model projections through

chain (Röder et al 2015). Consequently, wood-fired

2100 across a wide range of Representative Concen-

power plants generate more CO₂ per kWh than coal

tration Pathways (RCPs) (Knutti and Sedlacek 2013,

(supplementary table S5 available at stacks.iop.org/

Vuuren et al 2011). C-ROADS has been used by pol-

ERL/13/015007/mmedia ). Burning wood instead of

icymakers (Sterman et al 2012) and is freely available

coal therefore creates a carbon debt—an immedi-

(www.climateinteractive.org ).

ate increase in atmospheric CO₂ compared to fossil

The carbon cycle in the original C-ROADS model

energy—that can be repaid over time only as—and

includes globally aggregated stocks of carbon in fossil

if— NPP rises above the flux of carbon from biomass

fuels, the atmosphere, terrestrial biomass and soils, and

and soils to the atmosphere on the harvested lands.

a four-layer ocean. Here we disaggregate the treatment

Environ. Res. Lett. 13(2018) 015007

C in Atmosphere

Net Flux C

CH4 From

CO2 From

C to Atm

CH4 From

CO2 From

from Atm to

Net Primary

CO2 from

Respiration &

from Fossil

Respiration &

Ocean

Production

Bioenergy

Fire, Soils to

Fuels

Fire, Biomass

u,r

Atm u,r

to Atm u,r

C in Fossil

Fuels

C in Biomass u,r

C in Soils/Dead Organic Matter u,r

C from

Biomass to

C in Ocean (5

Soil u,r

Layers)

C in Lumber & Structures

u,r

C Biomass to

C Lumber &

Lumber &

Structures to Atm

Structures u,r

u,r

Figure 1. Modified carbon cycle in extended C-ROADS model. Carbon in biomass, soils, and structures (e.g. lumber in buildings),

and fluxes among these compartments, are disaggregated by land type, u, and region, r. Carbon can flow from biomass and soils

from each patch, u, r, to the atmosphere as CO₂ or CH₄. In addition, bioenergy harvest and combustion generate CO₂ . CO₂ and

CH₄ fluxes associated with changes in land use, e.g. from forest to pasture, cropland or developed land are included in the model but

not shown here. On the policy-relevant time scale (e.g. through 2100), creation of new fossil fuels from terrestrial or oceanic carbon

sources assumed to be negligible. Note: as described in the text and supplementary material, CH₄ fluxes from biomass and soils are

set to zero for forest scenarios considered here to isolate the impact of bioenergy in the scenarios tested.

of terrestrial carbon stocks both geographically and by

Although the model can be configured for any

land type (e.g. forest, pasture, cropland, developed land,

number of land types and uses, here we focus on wood

etc.). For each region, the model represents the area

harvested for electricity generation. For simplicity, we

of each type of land and changes in land use result-

configure the model to represent one region with three

ing from natural processes and human activity, along

categories of land: unmanaged forest, recently har-

with the carbon stocks and fluxes associated with each.

vested forest, and ‘other,’ which includes all other land

The model is extensible to any number of land/land

use categories (cropland, pasture, developed land, etc.).

use categories and geographic areas. For example, one

could configure the model to represent different types

2.2. Parameter estimation

of forests, with similar disaggregation for other land

Each unit of end-use bioenergy displaces the same end-

types, and at geographic scales from regions to nations

use energy generated from fossil fuels, so net CO₂

to, if data are available, even smaller areas.

emissions from biomass at the point of combustion

Figure 1 shows an overview of the carbon cycle

depend on which energy source is more efficient overall,

in the extended model. As in the original model,

given fuel carbon intensity, combustion efficiency, pro-

combustion of fossil fuels injects carbon into the atmo-

cessing losses, and emissions from their supply chains.

sphere. Unlike the original model, carbon stocks in

Typical combustion efficiencies for wood are approx-

biomass and soil are now represented for each cate-

imately 25%, compared to 35% for coal (Netherlands

gory of land and geographical area. The model also

Enterprise Agency 2011, IEA 2016). Published esti-

includes a compartment for carbon stored in lumber

mates vary with the process examined and the system

and structures. Consistent with reporting approaches

boundary considered, but processing losses (in energy

for the IPCC, FAO, and US Forest Service (FAO 2016,

content) for the wood pellet supply chain are on the

Penman et al 2003, Smith et al 2006), biomass in forest

order of approximately 27% if biomass is used in the

land includes living trees, including stems, branches,

drying process (Röder et al 2015), compared to losses

foliage, and coarse roots in both mature and under-

of approximately 11% for coal (IEA 2016). Differences

story trees; the stock denoted ‘soil carbon’ includes soil

in supply chain emissions from extraction/harvest,

organic matter, dead roots, litter (dead foliage, dead

and transportation are uncertain but relatively small

branches, etc), downed and standing dead trees, and liv-

compared to the large differences in combustion and

ing fine roots (Woodall et al 2015). Biomass is increased

processing efficiencies (e.g. Odeh and Cockerill 2008,

by net primary production. Carbon in biomass can

Röder et al 2015). Consequently, wood pellets emit

return to the atmosphere as CO₂ or CH₄ and is trans-

approximately 0.071 tC more CO₂ per GJ of end-use

ferred to the soil stock via litterfall and tree mortality.

energy than coal (see supplementary material).

Carbon is also lost from both biomass and dead organic

The determinants of NPP and carbon fluxes from

matter by fire. Carbon in the soil stock is transferred

biomass and soil to the atmosphere are therefore critical

to the atmosphere through the activity of decom-

toassessing thedynamic impact of bioenergy including

posers and other heterotrophs (Fahey et al 2005). The

the carbon debt payback period and long-run reduc-

supplementary material provides full documentation.

tion in atmospheric CO₂. To estimate the parameters

Environ. Res. Lett. 13 (2018) 015007

200

South Central Oak-Hickory

150

Biomass

100

Soils and Dead Organic Matter

100

Years

200

South Central Managed Shortleaf Loblolly

150

Biomass

100

Soils and Dead Organic Matter

100

Years

Figure 2. Growth curves showing carbon density (tC ha⁻¹) for oak-hickory (top) and managed shortleaf loblolly pine plantations

(bottom) in the south-central US, comparing Smith et al (2006) growth curves (dashed lines with data points) to the model (solid

lines), with best-fit parameters. Supplementary figure S2 and tables S2-S3 show results for all forest types estimated.

governing NPP and these fluxes we use the post-

curves closely: the mean absolute error relative to the

harvest growth curves in Smith et al (2006), which

mean ranges from 0.008%-0.065% for biomass and

span many regions and species in US forests. To illus-

from 0.006%-0.074% for soils (figure 2, table S2).

trate, figure 2 shows the Smith et al growth curves

for south-central US oak-hickory forest and managed

shortleaf loblolly pine plantations. The growth patterns

3. Results

differ markedly in both their shape and time required to

reach maximum biomass. After harvest, the managed

In the scenarios below, we adopt assumptions that

loblolly plantation regrows quickly, following a classic

favor bioenergy. Specifically, we assume bioenergy

S-shaped curve and reaching maximum biomass after

from wood pellets is used to offset coal, the most

about three decades, while the hardwood forest grows

carbon intensive fossil fuel; if wood offsets power gen-

roughly linearly for about 50 years and is still growing

erated from natural gas its carbon debt would be much

after a century. Note that in both cases, soil carbon

larger. Estimates of net CH₄ fluxes from forest biomass

declines for several decades after harvest because the C

and soils are poorly constrained and considered to

flux from biomass to soils is cut while heterotrophic

be insignificant in most global methane budgets (e.g.

respiration continues to release C from soils and dead

Ito and Inatomi 2012, Saunois et al 2016, Shoemaker

organic matter to the atmosphere.

et al 2014); we therefore assume them to be zero. We

To model NPP we specify a variant of the Richards

assume all land harvested for bioenergy is allowed

(1959) growth model, widely used in forest growth

to regrow without any fire (Buchholz et al 2016),

modeling. The US wood pellet industry is growing

erosion, disease, unplanned logging, or other ecolog-

rapidly, and much of the production is exported to

ical disturbances, including climate change impacts,

the EU and UK. We therefore estimate the carbon

that could limit regrowth or inject GHGs into the

cycle parameters from growth curves for temperate US

atmosphere beyond the direct impact of the bioen-

forests reported by Smith et al (2006). We estimate the

ergy harvest. We further assume that the decline in

parameters of NPP jointly with those governing fluxes

coal use resulting from wood does not lower coal

of CO₂ from biomass to soil and from each compart-

prices, increasing coal demand elsewhere, an effect

ment to the atmosphere using nonlinear least squares

estimated to be large (e.g. York 2012).

and Markov Chain Monte Carlo methods (supplemen-

To isolate the dynamic impact of bioenergy on CO₂

tary material). The model fits the Smith et al growth

emissions we run the model from an initial equilibrium

Environ. Res. Lett. 13(2018) 015007

Figure 3. Change in atmospheric CO₂ concentration resulting from displacement of coal by wood. Δ[CO₂] is relative to continued

coal use. All scenarios show the change in atmospheric CO₂ (ppmv) resulting from a single 1 EJ pulse of end-use energy from biomass

used to displace coal in year 0. Top: south-central (SC) oak-hickory forest; bottom: SC managed shortleaf loblolly plantation. The

bioenergy pulse causes an immediate increase in CO₂ concentration (the initial carbon debt) in scenarios 2-5 due to lower combustion

and processing efficiencies for wood compared to coal. The year in which Δ[CO₂] falls below zero is the carbon debt payback time.

Supplement figure S3 shows the results for all eight forest types examined. S0: Benchmark showing impact of 1 EJ pulse of zero carbon

energy. S1: Bioenergy assumed to have the same combustion and processing efficiency as coal, and the same supply chain emissions;

with 25% of biomass removed from the land harvested through thinning. S2: Actual efficiencies and supply chain emissions for wood

pellets; 25% of biomass harvested through thinning. S3: S2 with 95% of biomass harvested (clear cut). S4: S3 with clear cut and no

regrowth of harvested land and no C released from soil stocks. S5: S4 with C released from soil stocks at the estimated fractional rate.

in which the carbon fluxes from biomass and soils to

electricity generated from coal (total world energy use

the atmosphere are balanced by NPP, and in which net

exceeds 550 EJ yr⁻¹, US EIA 2016).

CO₂ flux to the ocean is zero throughout, identifying

Scenario 0 provides a benchmark showing how

the impacts of bioenergy separate from other sources

atmospheric CO₂ would change if 1 EJ of end-use

of disequilibrium, e.g. prior logging and marine uptake

energy from coal were offset by a zero-carbon energy

of CO₂. Including ocean CO₂ uptake would moder-

source, such as solar or wind (and assuming zero

ate increases in atmospheric CO₂ from bioenergy but

emissions from the supply chain). Displacing 1 EJ of

worsen ocean acidification and other impacts. These

end-use energy from coal with a zero C alternative

effects are left for future work.

keeps 0.07 GtC of fossil carbon in the ground, imme-

Figure

shows the results for a set of sce-

diately and permanently lowering atmospheric CO₂ by

narios using parameters estimated for oak-hickory

approximately 0.04 ppm relative to continued coal use.

forest in the south-central US (supplementary fig-

Scenario 1 simulates the counterfactual case in

ure S3, table S7 provide results for all eight forest

which bioenergy is assumed to have the same carbon

types we estimated). All scenarios examine a 1 exa-

emissions per EJ of end-use energy as coal, includ-

joule (EJ) pulse of end-use electric energy generated

ing the same combustion and processing efficiency

from wood pellets in year 0, offsetting 1 EJ of end-use

and supply chain emissions. We assume that 25%

Environ. Res. Lett. 13 (2018) 015007

of the biomass is removed from each hectare of the

should speed the repayment of the carbon debt. As

harvested forest by thinning, not clear cutting, that

expected, atmospheric CO₂ initially falls faster in the

the forest is allowed to regrow with no subsequent

plantation case compared to regrowth of the oak-

harvest, fire, disease, or other disturbances. Because

hickory forest. However, the concentration bottoms

emissions are counterfactually assumed to be the same

out after approximately 20 years and then starts to rise,

as coal, there is no immediate change in atmospheric

exceeding the CO₂ level when the forest is allowed

CO₂. However, as the forest grows back, carbon is

to regrow. The explanation lies in the different maxi-

gradually removed from the atmosphere to biomass

mum carbon densities of the two forest types: loblolly

and soils. After 100 years, the forest has recovered

plantation grows faster but reaches a lower equilib-

enough to lower atmospheric CO₂ by 0.026 ppm, still

rium carbon density compared to the unmanaged

34% above the zero C case.

forest

(figure 4), with estimated equilibrium values

Scenario 2 shows the realistic case with the combus-

of 130 tC ha⁻¹ for loblolly plantation vs. 211 tC ha⁻¹

tion efficiency and supply chain emissions estimated for

for oak-hickory. Consequently, although plantations

wood pellets (supplementary table S5), again assuming

grow faster, they do not remove as much C from the

25% of the biomass is harvested by thinning. Because

atmosphere as was lost when the hardwood forest was

production and combustion of wood generate more

harvested, even if allowed to grow to their maximum

CO₂ than coal, the first impact of bioenergy use is

biomass and remain unharvested. In reality, planta-

an increase in atmospheric CO₂. Regrowth gradually

tions are thinned every few years and harvested about

transfers C from the atmosphere to biomass and soil

every decade (US Forest Service 2000), further lower-

C stocks, leading to a carbon debt payback time of

ing their average C density and increasing atmospheric

52 years; after 100 years CO₂ remains 62% above the

CO₂. Furthermore, repeated harvests can degrade the

zero C case.

productivity of the soils, lowering NPP. To compen-

Scenario 3 is the same as S2 except we now assume

sate, managed plantations are typically fertilized several

the land is clear cut instead of thinned, with 95% of

times per rotation, increasing N₂O emissions that

the biomass removed. Near-complete biomass removal

would further worsen the climate impact of Scenario 6

reflects the growing practice of harvesting whole trees

(Schulze et al 2012).

and residues (branches, litter, etc) (Achat et al 2015). A

The supplementary material reports the 95% con-

95% clear cut requires only 26% as much land as in S2,

fidence intervals (CIs) for the estimated parameters

but the carbon debt payback time increases to 82 years;

(table S4), and sensitivity analysis across the eight

after 100 years CO₂ remains 86% above the zero C case.

forest types arising from parameter uncertainty, com-

Scenario 4 shows the impact of assuming that the

puted by Markov Chain Monte Carlo (table S8). The

harvested area is clear cut as in S3 but never allowed to

95% CIs for the carbon debt payback times vary from

regrow, for example, because it is developed, with the

74-110 years for the hardwood species under clear cut

additional assumption that the flux of C from soils and

(Scenario 3) and 11.25-12 years for the managed plan-

dead organic matter to the atmosphere is set to zero.

tations. The supplementary material also reports the

Without regrowth, the carbon debt is never repaid and

long-run CO₂ reductions for Scenarios 1-5 (table S7).

atmospheric CO₂ remains permanently higher.

For Scenario 3, after 100 years CO₂ falls an average of

Scenario 5 is the same as S4 except the flux of C

51% of the maximum possible reduction (the differ-

to the atmosphere from soils and dead organic mat-

ence between the initial carbon debt and the zero-C

ter continues at the original fractional rate. Without

level in Scenario 0) for the forests and 92% for the

regrowth, there is no flux of CO₂ from the atmo-

plantations.

sphere to terrestrial biomass or soils, but continued

The supplementary material also reports sensitivity

C flux from soils to atmosphere, causing CO₂ concen-

analysis of combustion efficiencies and supply chain

trations to rise beyond the immediate impact of the

emissions. Clearly, innovation that improves the com-

bioenergy. After a century atmospheric CO₂ has risen

bustion and processing efficiencies of wood relative

by 0.076 ppm, 2.3 times more than the initial impact.

to coal reduces the initial carbon debt of wood and

The actual impact of converting harvested forests to

reduces the carbon debt payback time and climate

other uses will likely lie between the results of Scenarios

impacts of wood. However, innovations that improve

4 and 5, but could rise further if conversion of for-

the efficiencies of both fuels yield smaller benefits.

est to other uses increases C fluxes from soils above

For example, combined heat and power systems offer

the values estimated from the Smith et al (2006) data.

substantially higher combustion efficiency than con-

Such an outcome could result from disturbances to

ventional boilers, but would still cause an initial carbon

soils from, e.g. plowing, development, fire or increasing

debt since the combustion and processing efficien-

methanogenesis, all of which we assume to be zero.

cies of wood remain lower than coal in such systems

In Scenario 6 (figure 4) oak-hickory forest is clear

(supplementary figures S5−6).

cut and replanted as a shortleaf loblolly pine managed

The wood pellet industry is expanding rapidly and

plantation. Loblolly pine grows faster than hardwoods

many projections call for substantial growth through

(figure 2), so intuitively the conversion from unman-

2030 or beyond (IEA 2012, IRENA 2015). Scenario

aged hardwood forest to managed pine plantation

(figure

5) shows the impact of linear growth in

Environ. Res. Lett. 13(2018) 015007

Figure 4. Scenario 6: replanting harvested oak-hickory forest after clear cut with managed plantation of shortleaf loblolly pine (south-

central US), compared to allowing the oak-hickory forest to regrow (Scenario 3 in figure 2). Top: change in atmospheric CO₂ (ppmv)

resulting from a single 1 EJ pulse of end-use energy from biomass used to displace coal in year. Δ[CO₂] is relative to continued coal

use. Bottom: carbon in biomass (tC ha⁻¹). For the first 20 years, faster-growing loblolly pine lowers atmospheric CO₂ compared to

regrowth of the oak-hickory forest, but the estimated maximum carbon density of oak-hickory forest is larger than the managed

loblolly plantation (211 vs. 131 tC ha⁻¹, respectively; supplementary table S3). Consequently, the carbon debt is never repaid even if

the loblolly plantation is never harvested. Due to CO₂ flux from soils, atmospheric CO₂ rises after approximately 20 years, exceeding

the level from regrowth of oak-hickory after approximately 50 years.

Figure 5. Change in atmospheric CO₂ concentration resulting from growth in end-use energy supplied by wood, displacing coal.

Δ[CO₂] is relative to continued coal use. Scenario 7 (solid line): linear growth in end-use energy supplied by US wood pellet production,

from the 2016 value of 0.028 EJ to 0.28 EJ yr⁻¹ by 2050 and continuing linearly thereafter. Parameters estimated for south-central US

oak-hickory forest, with harvest by clearcut. Scenario 8 (dashed line): the same as S7 except growth in end-use energy supplied by

wood ceases in 2050. Supplementary figure S4 reports results for all forest types considered.

Environ. Res. Lett. 13 (2018) 015007

end-use bioenergy; Scenario 8 is the same except growth

The carbon debt incurred when wood displaces coal

ceases in 2050. Growth in wood supply causes steady

may never be repaid if development, unplanned log-

growth in atmospheric CO₂ because more CO₂ is

ging, erosion or increases in extreme temperatures,

added to the atmosphere every year in initial car-

fire, and disease (all worsened by global warming) limit

bon debt than is paid back by regrowth, worsening

regrowth or accelerate the flux of carbon from soils to

global warming and climate change. The qualitative

the atmosphere. Further, lower coal prices caused by

result that growth in bioenergy raises atmospheric

the drop in power sector demand may stimulate coal

CO₂ does not depend on the parameters: as long as

use elsewhere, offsetting even the potential long-run

bioenergy generates an initial carbon debt, increasing

benefits of bioenergy (e.g. York 2012).

harvests mean more is ‘borrowed’ every year than is

Fifth, counter to intuition, harvesting existing

paid back. More precisely, atmospheric CO₂ rises as

forests and replanting with fast-growing species in

long as NPP remains below the initial carbon debt

managed plantations can worsen the climate impact of

incurred each year plus the fluxes of carbon from

wood biofuel. Although managed loblolly pine grows

biomass and soils to the atmosphere. Note further

faster than hardwood, speeding the initial recovery of

that in Scenario 8, CO₂ continues to rise for 56 years

forest biomass, the equilibrium carbon density of man-

after bioenergy production growth stops and only falls

aged plantations is lower than unmanaged forest, so

below initial levels 144 years after growth stops. Results

carbon sequestered in plantations never offsets the car-

for the other forest types are similar (supplementary

bon taken from the original forest. This is true even

figure S4).

if the managed plantation is never reharvested, and

worse if the plantation is periodically reharvested. Fur-

ther, typical plantations require periodic fertilization,

increasing N₂O emissions and worsening their climate

4. Discussionand conclusion

impact beyond what we report here (Schulze et al 2012).

We extended the carbon cycle model in the C-ROADS

Sixth, growth in wood harvest for bioenergy causes

climate policy model to account for different land and

a steady increase in atmospheric CO₂ because the initial

land use types, by region. The model explicitly treats

carbon debt incurred each year exceeds what is repaid.

stocks of carbon in fossil fuels, biomass, soils and

With the US forest parameters used here, growth in the

dead organic matter, the atmosphere, and the fluxes

wood pellet industry to displace coal aggravates global

among them including combustion, supply chain emis-

warming at least through the end of this century, even

sions, and regrowth of harvested lands. The model is

if the industry stops growing by 2050.

extensible to any number of land types and uses, and

Seventh, using wood in electricity generation wors-

geographic scales. To demonstrate the approach, we

ens climate change for decades or more even though

analyzed the dynamic impact of displacing coal with

many of our assumptions favor wood, including: wood

wood in electricity production, finding:

displaces coal (the most carbon intensive fossil fuel);

First, yet contrary to the policies of the EU and

all harvested land is allowed to regrow as forest with

other nations, biomass used to displace fossil fuels

no subsequent conversion to pasture, cropland, devel-

injects CO₂ into the atmosphere at the point of com-

opment or other uses; no subsequent harvest, fire or

bustion and during harvest, processing and transport.

disease; no increase in coal demand resulting from

Reductions in atmospheric CO₂ come only later, and

lower prices induced by the decline in coal use for

only if the harvested land is allowed to regrow.

electric power; no increase in N₂O from fertilization

Second, the combustion and processing efficien-

of managed plantations; and no increase in CO₂ emis-

cies of wood in electricity generation are lower than

sions or methanogenesis from disturbed land. Relaxing

for coal (supplementary material). Consequently, the

any of these assumptions worsens the climate impact

first impact of displacing coal with wood is an increase

of wood bioenergy.

in atmospheric CO₂ relative to continued coal use,

In sum, although bioenergy from wood can lower

creating an initial carbon debt.

long-run CO₂ concentrations compared to fossil fuels,

Third, after the carbon debt is repaid, atmospheric

its first impact is an increase in CO₂, worsening global

CO₂ is lower, showing the potential long-run benefits

warming over the critical period through 2100 even if

of bioenergy. However, before breakeven, atmospheric

the wood offsets coal, the most carbon-intensive fos-

CO₂ ishigherthanitwouldhavebeenwithouttheuseof

sil fuel. Declaring that biofuels are carbon neutral as

bioenergy, increasing radiative forcing and global aver-

the EU and others have done, erroneously assumes

age temperatures, worsening climate change, including

forest regrowth quickly and fully offsets the emissions

potentially irreversible impacts that may arise before

from biofuel production and combustion. The neu-

the long-run benefits are realized.

trality assumption is not valid because it ignores the

Fourth, biofuels are only beneficial in the long run

transient, but decades to centuries long, increase in

if the harvested land is allowed to regrow to its pre-

CO₂ caused by biofuels.

harvest biomass and maintained there. Natural forests

Methodologically, we demonstrate the feasibility

have high carbon density compared to pasture, crop-

of integrating static life cycle considerations around

land, developed land and managed tree plantations.

the efficiencies of and emissions from biofuels with

Environ. Res. Lett. 13(2018) 015007

explicit modeling of biomass dynamics in a model that

Helin T, Sokka L, Soimakallio S, Pingoud K and Pajula T 2013

Approaches for inclusion of forest carbon cycle in life cycle

runs fast enough to enable policymakers and other

assessment—a review Glob. Change Biol. Bioenergy 5

stakeholders to design and test their own scenarios.

475-86

Future work will integrate the model into full climate

IEA 2012 Technology Roadmap: Bioenergy for Heat and Power

models such as C-ROADS, creating a fast, interac-

(Paris: International Energy Agency) (www.iea.org/

publications/freepublications/publication/2012_

tive simulator that can model the impacts of different

Bioenergy_Roadmap_2nd_Edition_WEB.pdf )

biofuel technologies and scenarios on CO₂ concentra-

IEA 2016 World Energy Outlook (Paris: International Energy

tions, radiative forcing, warming, ocean acidification,

Agency) (www.iea.org/newsroom/news/2016/november/

sea level rise and other impacts.

world-energy-outlook-2016.html )

IPCC 2013 Working Group I Contribution to the IPCC Fifth

Assessment Report Climate Change 2013: The Physical Science

Acknowledgments

Basis (Cambridge: Cambridge University Press)

(www.ipcc.ch/report/ar5/wg1 )

IPCC 2014 Climate Change 2014: Impacts, Adaptation, and

This work was supported by the National Science Foun-

Vulnerability. Part A: Global and Sectoral Aspects.

dation under grants DUE-124558 and ICER-1701062.

Contribution of Working Group II to the Fifth Assessment

We thank Tom Fiddaman, William Moomaw and

Report of the Intergovernmental Panel on Climate Change

ed C B Field et al (Cambridge: Cambridge University Press)

Hazhir Rahmandad for helpful suggestions.

pp 1-32 (www.ipcc.ch/report/ar5/wg2 )

IRENA 2015 Renewable Power Generation Costs in 2014 (Abu

Dhabi: International Renewable Energy Agency)

ORCID iDs

(www.irena.org/DocumentDownloads/Publications/

IRENA_RE_Power_Costs_2014_report.pdf )

John D Sterman https://orcid.org/0000-0001-7476-

Ito A and Inatomi M 2012 Use of a process-based model for

assessing the methane budgets of global terrestrial ecosystems

6760

and evaluation of uncertainty Biogeosciences 9 759-73

Juliette N Rooney-Varga

https://orcid.org/0000-

Knutti R and Sedlacek J 2013 Robustness and uncertainties in the

0001-7102-6919

new CMIP5 climate model projections Nat. Clim. Change 3

369-73

Leturcq P 2014 Wood preservation (carbon sequestration) or wood

References

burning (fossil-fuel substitution), which is better for

mitigating climate change? Ann. Forest Sci. 71 117-24

Achat D, Deleuze C, Landmann G, Pousse N, Ranger J and Augusto

Netherlands Enterprise Agency 2011 CO₂-tool electricity gas and

L 2015 Quantifying consequences of removing harvesting

heat from biomass—version 1.0.1 (Assen, The Netherlands)

residues on forest soils and tree growth-a meta-analysis Forest

(www.rvo.nl/file/1252 )

Ecol. Manage. 348 124-41

Ocko I B, Hamburg S P, Jacob D J, Keith D W, Keohane N O,

Biofuture Platform 2017 Biofuture@COP23: major countries agree

Oppenheimer M, Roy-Mayhew J D, Schrag D P and Pacala

to scale up the low carbon bioeconomy and develop

S W 2017 Unmask temporal trade-offs in climate policy

sustainable biofuels targets (http://biofutureplatform.org/

debates Science 356 492-3

major-countries-agree-scale-low-carbon-bioeconomy-

Odeh N A and Cockerill T T 2008 Life cycle GHG assessment of

develop-sustainable-biofuels-targets/ )

fossil fuel power plants with carbon capture and storage Energ.

Brack D 2017 The Impacts of of the Demand for Woody Biomass for

Policy 36 367-80

Power and Heat on Climate and Forests (London: Chatham

Penman J, Gytarsky M and Hiraishi T 2003 Good Practice

House, the Royal Institute of International Affairs)

Guidance for Land Use, Land Use Change, and Forestry

Buchholz T, Hurteau M D, Gunn J and Saah D 2016 A global

(Hayama, Kanagawa: Institute for Global Environmental

meta-analysis of forest bioenergy greenhouse gas emission

Strategies for the Intergovernmental Panel on Climate

accounting studies Glob. Change Biol. Bioenergy 8 281-9

Change) (www.ipcc-nggip.iges.or.jp/public/gpglulucf.html )

Cornwall W 2017 Is wood a green source of energy? Scientists are

Richards F J 1959 A flexible growth function for empirical use

divided Science 355 18-21

J. Exp. Bot. 10 290-300

Creutzig F et al 2015 Bioenergy and climate change mitigation: an

Röder M, Whittaker C and Thornley P 2015 How certain are

assessment Glob. Change Biol. Bioenergy 7 916-44

greenhouse gas reductions from bioenergy? Life cycle

DeCicco J M 2013 Biofuel’s carbon balance: doubts, certainties and

assessment and uncertainty analysis of wood

implications Clim. Change 121 801-14

pellet-to-electricity supply chains from forest residues Biomass

Doyle A and Roche A 2017 Nineteen nations say they will use more

Bioenergy 79 50-63

bioenergy to slow climate change Reuters (www.reuters.com/

Saunois M et al 2016 The global methane budget 2000-2012 Earth

article/us-climatechange-accord-biofuels/nineteen-nations-

Syst. Sci. Data 8 697-751

say-theyll-use-more-bioenergy-to-slow-climate-change-

Schulze E-D, Körner C, Law B, Haberl H and Luyssaert S 2012

idUSKBN1DG2DO )

Large-scale bioenergy from additional harvest of forest

European Commission 2003 Directive 2003/87/EC of the European

biomass is neither sustainable nor greenhouse gas neutral

Parliament and of the Council of 13 October 2003 (Brussels:

Glob. Change Biol. Bioenergy 4 611-6

European Parliament) pp 32-46 (http://eur-lex.europa.eu/

Shoemaker J K, Keenan T F, Hollinger D Y and Richardson A D

legal-content/EN/TXT/?uri=CELEX:32003L0087 )

2014 Forest ecosystem changes from annual methane source

Fahey T J et al 2005 The biogeochemistry of carbon at Hubbard

to sink depending on late summer water balance Geophys. Res.

Brook Biogeochemistry 75 109-76

Lett. 41 673-9

FAO 2016 Global Forest Resources Assessment 2015: How are the

Smith J E, Heath L S, Skog K E and Birdsey R A 2006 Methods for

World’s Forests Changing? (Rome: Food and Agriculture

calculating forest ecosystem and harvested carbon with

Organization of the United Nations) (www.fao.org/3/

standard estimates for forest types of the United States

a-i4793e.pdf )

(www.fs.usda.gov/treesearch/pubs/22954 )

Figueres C, Schellnhuber H J, Whiteman G, Rockstrom J, Hobley A

Sterman J, Fiddaman T, Franck T, Jones A, McCauley S, Rice P,

and Rahmstorf S 2017 Three years to safeguard our climate

Sawin E and Siegel L 2012 Climate interactive: the C-ROADS

Nature 546 593-5

climate policy model Syst. Dyn. Rev. 28 295-305

Environ. Res. Lett. 13 (2018) 015007

Sterman J D, Fiddaman T, Franck T, Jones A, McCauley S, Rice P,

US Forest Service 2000 Landowners Handbook for Managing

Sawin E and Siegel L 2013 Management flight simulators to

Southern Pines (Atlanta, GA: US Forest Service-Southern

support climate negotiations Environ. Modell. Softw. 44

Region) (http://ncforestry.info/fs/managing_southern_

122-35

pines/ )

Stupak I et al 2007 Sustainable utilisation of forest biomass for

US EIA 2016 International Energy Outlook 2016 (Washington, DC:

energy—possibilities and problems: policy, legislation,

US Energy Information Administration) p 276 (www.eia.gov/

certification, and recommendations and guidelines in the

outlooks/ieo/pdf/0484(2016).pdf )

Nordic, Baltic, and other European countries Biomass

Vuuren D et al 2011 The representative concentration pathways: an

Bioenergy 31 666-84

overview Clim. Change 109 5-31

Ter-Mikaelian M T, Colombo S J and Chen J 2015 The burning

Woodall C W et al 2015 The US Forest Carbon Accounting

question: does forest bioenergy reduce carbon emissions? A

Framework: Stocks and Stock Change 1990-2016 (Newtown

review of common misconceptions about forest carbon

Square, PA: US Forest Service) (www.fs.usda.gov/

accounting J. Forest. 113 57-68

treesearch/pubs/49858 )

Thrän D, Peetz D and Schaubach K 2017 Global Wood Pellet

York R 2012 Do alternative energy sources displace fossil fuels?

Industry and Trade Study 2017 (Paris: IEA Bioenergy)

Nat. Clim. Change 2 441-3

p 243 (http://task40.ieabioenergy.com/wp-content/uploads/

2013/09/IEA-Wood-Pellet-Study_final-2017-06.pdf )