NRDC: Think Wood Pellets are Green? Think Again. (PDF)
NRDC issue brief
may 2015
ib:15-05-a
Think Wood Pellets are Green? Think Again.
biomass is often described as a clean, renewable fuel and a greener alternative
to coal and other fossil fuels for producing electricity. but recent science shows
that many forms of biomass—especially from forests—produce higher carbon
emissions compared to fossil fuels. in particular, a growing body of peer-reviewed,
scientific studies shows that burning wood from whole trees in power plants to
produce electricity can increase carbon emissions relative to fossil fuels for many
decades—anywhere from 35 to 100 years.1 This time period is significant: climate
policy imperatives require dramatic short-term reductions in greenhouse gases, and
these emissions will persist in the atmosphere well past the time when significant
reductions are needed.
Unfortunately, the biomass wood pellet industry in the
and branches from logging operations). Using the model,
southeastern United States is expanding rapidly. Wood pellet
we estimated the total amount of carbon released over time
exports from the United States doubled from 1.6 million tons
(cumulative CO2 emissions) for each scenario and compared
in 2012 to 3.2 million tons in 2013, and they are expected to
those emissions with those from coal and natural gas.
reach 5.7 million tons in 2015.2 This growth is driven largely
by exports to Europe in response to flawed policy incentives
on renewable resources that regard all biomass as carbon
How we moDeleD Pellet CaRboN
neutral.3
emiSSioNS
Although recent science shows that many forms of forest
biomass are high-carbon sources of fuel, under the right
The SIG model first generates a working forest landscape—
circumstances, true wood waste could serve as a low-carbon
including timber harvest and forest regrowth—based on
option for producing pellets. For example, sawdust and chips
typical forest management in the sourcing ecoregion.
from sawmills that would otherwise quickly decompose and
The model then estimates the emissions from removing
release carbon anyway—can be a low-carbon source.
additional forest materials each year for wood pellet
4 On the
other hand, burning whole trees can produce higher carbon
production and by burning the pellets to produce electricity.
emissions than coal, and this elevated CO
(See the Technical Appendix for information on the SIG
2 level with respect
to coal can persist in the atmosphere for decades.
model and analytic methods.)
5 Therefore,
the composition of wood pellets matters greatly: the amount
Our modeling assumed that the biomass feedstocks used
of whole trees used in wood pellets can have a significant
to produce pellets were typical of the pellet industry.8 They
impact on the estimated carbon emissions of this fuel source.
are:
NRDC therefore modeled the carbon impacts of burning
n
forestry residues—tops and limbs from forestry
wood pellets of varying composition in power plants to
operations that are non-merchantable to other markets;
produce electricity. We used a carbon accounting model
n
whole trees—merchantable pulpwood, trees from
developed by the Spatial Informatics Group (SIG) to model
thinning operations, and non-merchantable trees; and
scenarios in which pellets sourced from bottomland
hardwood forests in Atlantic plain6 of North Carolina and
n
mill waste—by-products of sawmill operations such as
South Carolina supplied a typical power plant in the United
sawdust and chips.
Kingdom7. In our analysis, we modeled a range of scenarios
(See the Technical Appendix for more information on
in which pellets are made of varying amounts of whole trees,
feedstock definitions.)
mill waste, and non-merchantable forestry residues (tops
for more information,
www.nrdc.org
please contact:
www.facebook.com/nrdc.org
Sami yassa
www.twitter.com/NRDC
Senior Scientist
xxxxxx@xxxx.xxx
(415) 269-1355
We varied the amounts of each of these three feedstocks,
Figure 1: Cumulative emmissions (mgCo
running scenarios in which the proportion of whole trees in
2e/mw)
the wood pellets ranged from 20 percent to 70 percent.9 In
Pellets made of 70 percent whole trees
1,000,000
each of these scenarios, we modeled the emissions resulting
from burning the pellets in a typical power plant over a
800,000
100-year period. We compared these estimated biomass CO2
emissions with the CO2 that would have been emitted by
600,000
fossil fuels to produce the same amount of electricity.
400,000
200,000
ReSultS
0
In the figures below we show the results for three
representative scenarios: pellets made of 70 percent, 40
-200,000
Power (coal) old plants
percent, and 20 percent whole trees. The solid black line
Power (coal) new plants
-400,000
Power (natural gas) new plants
represents the estimated cumulative carbon emissions per
Biomass power
megawatt of power from a power plant burning wood pellets,
-600,000
accounting for the effects of forest regrowth. The dashed
lines represent cumulative emissions from fossil fuels per
2015
2025
2035
2045
2055
2065
2075
2085
2095
2105
2115
megawatt of power.
Figure 2: Cumulative emmissions (mgCo
For the first several decades of the plant’s operations,
2e/mw)
the burning of pellets creates a pulse of emissions to the
Pellets made of 40 percent whole trees
1,000,000
atmosphere—namely, increased carbon emissions resulting
from combustion. This pulse occurs because wood is less
800,000
energy-dense than fossil fuels, so burning biomass generally
emits more carbon than fossil fuels to produce the same
600,000
amount of energy.10 Over time, however, forest regrowth
400,000
reduces this atmospheric carbon.11 And after many decades,
this regrowth can recapture enough carbon to reduce the
200,000
cumulative emissions below those of fossil fuels. These
results are similar to the “carbon debt” outcomes found in
0
recent biomass studies.12
-200,000
Power (coal) old plants
Figures 1 and 2 together show the modeled emissions
Power (coal) new plants
when the proportion of whole trees in pellets ranges from
-400,000
Power (natural gas) new plants
Biomass power
40 percent to 70 percent. The modeling shows that it will
-600,000
take approximately 55 years for forest regrowth to recapture
enough carbon from the atmosphere to reduce the plant’s
2015
2025
2035
2045
2055
2065
2075
2085
2095
2105
2115
cumulative emissions below those of coal. At levels greater
than 40 percent, pellets emit more carbon than coal for most
Figure 3: Cumulative emmissions (mgCo2e/mw)
of this period. In addition, as the percentage of whole trees
Pellets made of 20 percent whole trees
increases above 70 percent (not shown in the figures), the
1,000,000
level of carbon emissions continues to increase.13
800,000
When whole trees make up 20 percent of the wood in
pellets, emissions are slightly higher than natural gas and
600,000
slightly lower than coal for a period of approximately 55
years, as shown in Figure 3. Even when whole trees make
400,000
up as little as 12 percent of pellets, our modeling showed
200,000
that burning pellets still produces emissions comparable
to natural gas trend line for approximately 50 years. (See
0
Technical Appendix for information on additional scenarios.)
-200,000
Power (coal) old plants
Power (coal) new plants
-400,000
Power (natural gas) new plants
Biomass power
-600,000
2015
2025
2035
2045
2055
2065
2075
2085
2095
2105
2115
PaGe 2 |
bioenergy model
CoNCluSioNS
In sum, our modeling shows that wood pellets made of whole
trees from bottomland hardwoods in the Atlantic plain of
the U.S. Southeast—even in relatively small proportions—
will emit carbon pollution comparable to or in excess of
fossil fuels for approximately five decades. This 5-decade
time period is significant: climate policy imperatives
require dramatic short-term reductions in greenhouse gas
emissions, and emissions from these pellets will persist in the
atmosphere well past the time when significant reductions
are needed. Moreover, several studies have concluded that
logging residuals alone may be unable to meet bioenergy
demands in the region we modeled, and that pulpwood trees
may need to be used to meet the increasing demand.14
These results have significant implications for the wood
pellet industry and its expansion. Pellet manufacturer Enviva
LP and British utility Drax Power are at the head of this
industry. Drax operates the United Kingdom’s largest coal-
fired power plant and is converting half of its six generating
units to run solely on wood pellets.15 Enviva is the largest
producer and exporter of wood pellets in the United States
enviva facility in ahoskie, NC.
and a primary biomass supplier to Drax. The company owns
and operates five production plants in the Southeastern U.S.
that have a combined wood pellet production capacity of
as the calls to curb carbon pollution grow louder, power
approximately 1.7 million metric tons per year.
companies face increased pressure to find cleaner sources
16
Enviva has claimed that its wood pellets are a clean source
of energy. Many have turned to woody biomass for fuel,
of fuel for electricity production and that they do not increase
much of which comes from forests. The wood is chipped
carbon emissions.
or turned into pellets—small, compressed cylinders of
17 Yet the company has not publicly
disclosed the composition of its wood pellets. If Enviva’s
woody material. These pellets are burned in power plants
pellets were comprised of whole trees from the modeled
just like coal. Most suppliers are operating under the
Southeast bottomland hardwoods—even in relatively small
false assumption that, since trees can grow back and
proportions—they would emit carbon pollution comparable
resequester carbon, then they are a carbon-neutral fuel
to fossil fuels for decades. The company has the responsibility
when burned. However, mounting scientific evidence
to come clean with the public, investors, and regulators by
shows that it could take many decades for forest regrowth
disclosing the makeup of its fuel, and to take steps to ensure
to offset stack emissions from power plants.
that its pellets do not increase carbon emissions.
PaGe 3 |
bioenergy model
teCHNiCal aPPeNDiX: How tHe SPatial
landscape over time (except for 19 percent of the acreage,
iNFoRmatiCS GRouP’S (SiG) CalCulatoR
which was put off-limits due to operational constraints and
woRKS
protected status). Within the woodsheds, they prescribed
more intensive silviculture to produce additional biomass
This analysis is based on a Greenhouse Gas Calculator
material to meet the existing facility’s demand—typically
developed for NRDC by the Spatial Informatics Group
an extra thinning cycle and removal of boles smaller than
(SIG). The Spatial Informatics Group’s (SIG) Greenhouse
4 inches dbh (hereinafter “biomass harvest”).20 This mix of
Gas Calculator estimates the carbon emissions from woody
default silviculture and biomass harvest across the 17-facility
biomass energy sourced from forests in the southeastern
landscape constitutes the
business-as-usual scenario.
United States. The user first specifies a power-generating
Colnes et al. then modeled additional timber harvesting
facility type, plant efficiency, mix of feedstock types, sourcing
above the
business-as-usual associated with new demand in
ecoregion, and forest type to generate a
biomass power
additional 50-mile-radius woodsheds, which had the more
scenario. For this analysis, NRDC set parameters consistent
intensive “biomass harvest” prescriptions described above.
with wood pellets sourced from a mix of mill waste, forestry
The modeling in Colnes et al. meets this increased demand
residues, and whole tree boles from bottomland hardwoods
first with residues and non-merchantable boles, then with
in the Atlantic Coastal Flatwoods ecoregion.
pulp sized boles and tree tops. The difference between
The calculator models the greenhouse gas emissions
emissions from new demand and emissions from the
and sequestration over time in the chosen region under
business-as-usual scenario produces net emissions factors on
the specified scenario. The calculator then compares these
a per-output-energy (per-MWh) basis.
against a
business-as-usual scenario—namely, the emissions
The SIG Calculator builds on Colnes et al.’s results in two
and sequestration that would occur in the absence of the
main ways. First, it allows the user to simulate the sourcing
specified biomass sourcing and combustion. The difference
location and feedstock type by disaggregating emission
between the two trajectories represents the net cumulative
factors associated with a single ecoregion and forest type
greenhouse gas emissions on a carbon equivalent per-
inside the Colnes et al. study area. Second, it scales the
output-energy (per-MWh) basis attributable to the user-
net, per-MWh emission factors by the size of the power-
specified facility’s sourcing and power generation operations.
generating facility specified in a
biomass power scenario. For
The calculator relies entirely on data and results from
the purposes of this analysis, NRDC assumed that the new
the study
Biomass Supply and Carbon Accounting for
demand was generated in the Atlantic Coastal Flatwoods
Southeastern Forests (Colnes et al.) to generate both the
(Forest Service Ecoregion 232) of North Carolina.
biomass power scenarios and business-as-usual scenarios.18
Besides the analysis of forest growth, the SIG Calculator
These researchers assumed: (1) typical and customary
folds in greenhouse gas emissions from feedstock
silvicultural systems implemented in each ecoregion and
processing, transportation, and energy conversion at the
forest type, and (2) typical markets and end uses for the
power plant. It does not include greenhouse gas emissions
varying size classes of wood products harvested (pulpwood
from plant construction or plantation management. SIG
4–10 inches in diameter at breast height [dbh] and sawlogs
added parameters regarding the additional emissions of
greater than 10 inches dbh). Materials less than 4 inches in
transporting the biomass fuel from the U.S. Southeast to
diameter were considered non-merchantable and left to
the United Kingdom, along with analogous mining and
decay in forests in the
business-as-usual scenario.19
transportation emissions for the displaced coal.
Silvicultural scenarios were modeled by Colnes et al.
The SIG Calculator estimated emissions from harvesting
using 2010 U.S. Forest Service Forest Inventory and Analysis
equipment using a factor of 0.015 ton of CO2 per bone dry ton
(FIA) data for each ecoregion. The U.S. Forest Service Forest
(BDT) of harvested material. Truck transportation emissions
Vegetation Simulator Southern Variant was used to account
were estimated using a factor of 0.000134 ton of CO2 per BDT-
for changes in on-site forest carbon pools. The forest carbon
mile, which assumed 12.5 tons per truck, 6 miles per gallon,
accounting included above- and belowground live trees,
and 22.2 lbs. CO2 per gallon of diesel fuel (see sources cited in
standing deadwood, belowground deadwood, and down
Colnes et al., pg. 84).
deadwood. Carbon stored in wood products in use and in
The SIG Calculator assumes that mill residues are carbon
landfill pools was simulated separately in the SIG Calculator.
neutral—meaning neither biogenic nor fossil fuel (e.g.,
To establish the
business-as-usual scenario, Colnes et
transport and processing-related) carbon emissions are
al. modeled a landscape including woodsheds around 17
associated with their use.
existing biomass facilities in the region as of 2010. Each
The SIG Calculator further assumes that logging residues
woodshed was specified by a 50 mile-radius around each
constitute 32 percent of harvest volume for bottomland
known existing biomass facility. The researchers first
hardwoods in the Atlantic Coastal Flatwoods. In order to
prescribed default silvicultural practices across the entire
generate representative ratios of logging residues versus
PaGe 4 |
bioenergy model
roundwood for fuelsheds in the Atlantic Coastal Flatwoods
Products Output Database.21 In 2009, the most recent year for
(Forest Service Ecoregion 232) of North Carolina, we retrieved which reports were available, logging residues constituted an
reported volumes of logging residue and roundwood by
average of 32 percent of the harvest volume in hardwoods by
weight for six counties in the ecoregion from the Timber
weight.
endnotes
1 Colnes, a., et al.,
Biomass Supply and Carbon Accounting for Southeastern Forests, The biomass energy resource Center, forest Guild, and spatial
informatics Group, f
ebruary 2012, www.biomasscenter.org/images/stories/se_Carbon_study_fiNaL_2-6-12.pdf. Harmon, M., Impacts of Thinning on
Carbon Stores in the PNW: A Plot Level Analysis, Oregon state university, May 2011. Mitchell, s., M. Harmon, and K. O’Connell, “Carbon Debt and
Carbon sequestration Parity in forest bioenergy Production,”
GCB Bioenergy 4, no. 6 (November 2012): 818-827. repo, a., et al., “sustainability of
forest bioenergy in europe: Land-use-related Carbon Dioxide emissions of forest Harvest residues,”
GCB Bioenergy, published online March 2014.
stephenson, a. L., and D. MacKay,
Life Cycle Impacts of Biomass Electricity in 2020: Scenarios for Assessing the Greenhouse Gas Impacts and Energy
Input Requirements of Using North American Woody Biomass for Electricity Generation in the UK, u.K. Department of energy and Climate Change,
July 2014, www.gov.uk/government/uploads/system/uploads/attachment_data/file/349024/beaC_report_290814.pdf. Ter-Mikaelian, M., et al.,
“Carbon Debt repayment or Carbon sequestration Parity? Lessons from a forest bioenergy Case study in Ontario, Canada,”
GCB Bioenergy,
published online May 2014. Walker, T., et al.,
Biomass Sustainability and Carbon Policy Study, Manomet Center for Conservation sciences, June 2010,
www.mass.gov/eea/docs/doer/renewables/biomass/manomet-biomass-report-full-hirez.pdf.
2 Wood resources international LLC, “Global Timber and Wood Products Market update,” news brief, October 11, 2012.
3 2009 renewable energy Directive, which sets an overall eu
target of a 20 percent share of renewable energy by 2020. ec.europa.eu/clima/policies/
package/index_en.htm.
4 stephenson, a. L., and D. MacKay,
Life Cycle Impacts of Biomass Electricity.
5 Pingoud, K., T. ekholm, and i. savolainen, “Global Warming Potential factors and Warming Payback Time as Climate indicators of forest biomass
use,”
Mitigation and Adaptation Strategies for Global Change 17, no. 4 (January 2012): 369-386. schulze, e. D., et al., “Large-scale bioenergy from
additional Harvest of forest biomass is Neither sustainable Nor Greenhouse Gas Neutral,”
GCB Bioenergy 4, no. 6 (November 2012): 611-616. Walker,
T., et al.,
Biomass Sustainability.
6 The sourcing region is the atlantic Coastal flatwoods (forest service ecoregion 232)
7 We assumed a capacity factor of 85 percent.
8 suz-anne Kinney, “Dispelling the Whole Tree Myth: How a Harvested Tree is used,” forest2Market (f2M), December 20, 2013,
www.forest2market.com/blog/dispelling-the-whole-tree-myth-how-a-harvested-tree-is-used. The author states: “While whole trees are certainly being
harvested, any whole tree that ends up in the wood yard of a pellet facility is either defective in some way (unmerchantable), a pulpwood-sized tree that
took 10–15 years to grow or the upper section of a larger tree…. it is impossible to tell which of the latter two categories any single pulpwood-sized log
falls into.”
9 for each scenario, we held the amount of mill waste constant at 20 percent.
10 Walker, T., et al.,
Biomass Sustainability.
11 in the model, this sequestration also includes avoided decay (decay that would have otherwise occurred).
12 Ter-Mikaelian, et al., “Carbon Debt repayment.” Colnes, a., et al., biomass supply and Carbon accounting . Walker, T., et al., biomass sustainability.
13 This is the case all the way to 100 percent whole trees— not illustrated in figures.
14
http://naldc.nal.usda.gov/download/46157/PDf, https://nicholasinstitute.duke.edu/sites/default/files/publications/forest-biomass-supply-in-the-
southeastern-united-states-implications-for-industrial-roundwood-and-bioenergy-production-paper.pdf.
15
Livermore, M., “The Contradictions of biomass,” scientific alliance, March 10, 2014, www.cambridgenetwork.co.uk/news/the-contradictions-of-
biomass.
16
http://247wallst.com/energy-business/2015/04/29/enviva-ipo-generating-heat-and-light.
17 enviva LP,
Inherent Sustainability & Carbon Benefits of the US Wood Pellet Industry, white paper
, 2012, www.envivabiomass.com/wp-content/
uploads/inherent-sustainability-carbon-benefits-20121005.pdf.
18 Colnes, a., et al.,
Biomass Supply and Carbon Accounting for Southeastern Forests, southern environmental Law Center, 2012,
www.biomasscenter.org/images/stories/se_Carbon_study_fiNaL_2-6-12.pdf.
19 Materials less than 4 inches in diameter were collected in certain biomass power scenarios only.
20 The “biomass harvest” generates additional timber harvest volume in the forms of (1) non-merchantable tops and limbs, (2) non-merchantable boles
less than 4 inches dbh, and (3) pulpwood boles 4–10 inches dbh.
21 u.s. Department of agriculture forest service, southern research station, Timber Product Output (TPO) reports, Table C10: “Volume of timber
removals by state/County
, species group, removals class and source,” srsfia2.fs.fed.us/php/tpo_2009/tpo_rpa_int1.php, accessed November 24, 2014.
The model partitions logging residues from new harvest to logging residues from existing harvest at a ratio of approximately 4:1.
PaGe 5 |
bioenergy model