Greenhouse gas emissions from burning US-sourced woody biomass in the EU and UK

Increased use of forest biomass for energy and associated impacts on the climate

Research paper Updated 17 October 2022 ISBN: 978 1 78413 493 8

Drax Power Station in Yorkshire, UK, illuminated at night

Duncan Brack

Associate Fellow, Environment and Society Programme

Dr Richard Birdsey

Senior Scientist, Woodwell Climate Research Center

Dr Wayne Walker

Carbon Program Director and Associate Scientist, Woodwell Climate Research Center

Existing national and intergovernmental policy frameworks treat forest biomass as zero-carbon at the point of combustion and grant it access to financial and regulatory support. Consumption of wood pellets is growing rapidly as a result. Yet these frameworks do not take full account of the contribution of biomass burning to increased carbon emissions.

This paper examines one particular source of woody biomass: pellets sourced from the US that are burnt for electricity and combined heat and power in the EU and UK. Accounting for emissions from their combustion, their supply chain, forgone removals of carbon dioxide from the atmosphere due to the harvesting of live trees and emissions from the decay of roots and unused logging residues left in the forest after harvest, it finds that US-sourced wood pellets burnt in the UK were responsible for 13 million–16 million tonnes of CO₂ emissions in 2019, equivalent to the emissions from between 6 million and 7 million passenger vehicles. This makes it less likely that the UK will achieve its climate targets.

A summary of the research findings is also available as a PDF download.

Annex: Emissions from wood pellet use for energy

A1. Emissions from pellet combustion

Calculations of emissions of CO₂ per weight of wood pellets in national reports to the UNFCCC vary. In its 2012 report, for example, the UK used the figure of 381.4 kg of carbon (1,400 kg CO₂) emitted per tonne of wood biomass (based on IPCC guidance from 1997),113 whereas in its 2014 report, it used figures of 1,055 kg CO₂ per tonne of wood for power stations, 1,416 kg for domestic combustion and 1,767 kg for other industrial combustion.114 The US Energy Information Administration uses the figure of 434 kg carbon (1,593 kg CO₂) per tonne of wood pellets.115 Other countries report emissions per unit of energy generated rather than per weight of wood. In reality, emissions from biomass will depend on the type of feedstock, the facility in which it is burnt and several other factors.

The second Chatham House paper on biomass energy, Woody Biomass for Power and Heat: Demand and Supply in Selected EU Member States (2018),116 which included this calculation for imported biomass, used the figure of 1.762 tonnes of CO₂ per tonne of wood pellets, based on reported emissions from Drax in 2013. For the purposes of this report, we have used the average of reported consumption and emissions from Drax for the years 2013–19, which gives a conversion factor of 1.80 tonnes CO₂ per tonne of wood pellets (see Table 7).

Table 7. Consumption of and emissions of CO₂ from wood pellets at Drax, 2013–19

	2013	2014	2015	2016	2017	2018	2019
Total pellets (kt)	1,596	3,900	5,893	6,591	6,756	7,171	7,051
Total emissions (ktCO₂eq)	2,799	7,150	10,372	11,836	12,212	13,019	12,795
Emissions / pellets (ktCO₂eq/kt)	1.75	1.83	1.76	1.80	1.81	1.82	1.81

Source: Drax Annual Reports.

Using this calculation of emissions from combustion, Table 8 sets out the CO₂ emissions attributable to wood pellets imported from the US and burnt for energy in the former EU28. (Figure 1 is based on these data, multiplied by 1.80).117

Table 8. Imports of US wood pellets to former EU28 and associated emissions of CO₂, 2014–19

	2014	2015	2016	2017	2018	2019
EU28 pellet imports from US (kt)	3,890	4,287	4,902	5,205	6,139	6,779
Emissions (ktCO₂)	7,002	7,716	8,823	9,370	11,049	12,202
Emissions as % of EU solid biomass emissions	1.61%	1.71%	1.92%	2.01%	2.37%	2.53%
Emissions as % of EU energy emissions (not included in emissions totals)	0.22%	0.24%	0.28%	0.29%	0.35%	0.41%

Source: Import figures generated from Eurostat data at http://epp.eurostat.ec.europa.eu/newxtweb/mainxtnet.do?noredirectnewsession=true.

Table 9 contains the equivalent data for US wood pellets imported to the UK and for emissions associated with consumption of US wood pellets in Drax.118

Table 9. Imports of US wood pellets to UK, consumption of US wood pellets in Drax and associated emissions of CO₂ from combustion, 2014–19

	2014	2015	2016	2017	2018	2019
UK pellet imports from US (kt)	2,895	3,528	4,128	4,266	4,880	5,484
Emissions (ktCO₂)	5,211	6,350	7,430	7,678	8,784	9,870
Emissions as % of UK solid biomass emissions	32.94%	33.77%	40.68%	38.12%	39.52%	42.29%
Emissions as % of UK energy emissions (not included in emissions totals)	1.29%	1.63%	2.01%	2.14%	2.50%	2.91%
Drax consumption (kt)	2,300	3,135	3,892	3,996	4,464	4,613
Emissions (ktCO₂)	4,140	5,644	7,005	7,193	8,034	8,303

Source: Import figures generated from Eurostat data at http://epp.eurostat.ec.europa.eu/newxtweb/mainxtnet.do?noredirectnewsession=true; Drax consumption figures from Drax annual reports.

A2. Emissions from the supply chain

The use of biomass for energy also produces emissions from the supply chain: in this case, from the energy used in harvesting, processing and transporting wood pellets. Drax includes estimates for these in its annual reports. In 2019 50 per cent of these emissions derived from the pelletizing process, 20 per cent from shipping, 19 per cent from other transport, 8 per cent from drying the wood and about 4 per cent from cultivation, harvesting and chipping.119 Estimates are included for these supply chain emissions per unit of electricity generated, which enables the calculation of total supply chain emissions from the wood pellets burnt for energy in each year. As Table 10 shows, the average of the supply chain figures across the six years is 0.24 tonnes CO₂ per tonne of pellets. (In April 2021 Drax published its supply-chain figures for 2020, showing a fall to 109 kgCO₂eq/MWh, though it did not provide a breakdown or explain how that reduction had been achieved.120)

Table 10. Supply-chain emissions of CO₂ from the use of wood pellets in Drax, 2014–19

	2014	2015	2016	2017	2018	2019
Supply chain emissions (kgCO₂eq/MWh)	123	114	122	130	131	124
Generation from biomass (TWh)	7.9	11.5	12.7	13	13.8	13.7
Total emissions (ktCO₂eq)	970	1,311	1,549	1,690	1,808	1,699
Emissions/pellets (ktCO₂eq/kt)	0.25	0.22	0.24	0.25	0.25	0.24

Source: Drax Annual Reports.

Unlike the combustion figures, these emissions should be reported in the greenhouse gas reports of the countries in which the harvesting, processing and transport takes place. Countries are allocated shares of emissions from international transport.

A3. Impacts of pellet production on forest carbon stocks and sequestration in the southern United States

The use of woody biomass for energy affects carbon concentrations in the atmosphere from the impacts of harvesting live trees on carbon stocks, emissions from logging and mill residues, and forgone sequestration rates in the forests from which the wood is sourced. Calculating these effects is significantly more complex than calculating emissions from combustion and the supply chain. This is in large part because publicly available data from forest inventories does not allow for directly estimating the impact of harvesting of biomass for pellets to be distinguished from harvesting for other wood products. There is, however, sufficiently detailed data available to make reasoned assumptions about how additional harvesting of biomass for pellets – above what would have occurred if pellets were not being produced – impacts forests and net emissions of CO₂. This assertion is based on economic analyses that show how growing demand for wood products leads to increased harvesting.121

The approach taken here is a partial systems analysis, focused on two major parts of a full systems analysis: the forest ecosystem carbon cycle; and the harvested wood processing by mills for pellets and other wood products. The effects of substituting wood pellets for other energy sources that would typically be the target of comparative life cycle analyses are not estimated here, nor are indirect impacts of increasing harvest for pellets, such as the effects of induced land-use change on forest area, or effects on wood supply for the pulp and paper industry.

The approach aims to quantify how harvesting wood in the southern US for producing pellets for export to the UK affects the carbon stocks and stock change in the forests of the source regions for the pellet mills. To put these forest ecosystem and wood product estimates in context, the impacts of all sources used and exported were estimated, while their emissions upon combustion were also considered. The approach is focused on the cellulosic carbon and CO₂ emissions associated with the different sources of biomass used for pellets and the uncertainty of information about quantities of different sources used. The estimates are based on the production of pellets rather than consumption, so they may not be 100 per cent consistent with estimates from consuming countries. This is largely the result of separately compiled datasets that describe production and exports in the producing country and consumption and imports in the consuming countries.

Approach and methods

The analysis approach and methods involved the following elements:

A set of pellet mills in the southern US that export pellets to the UK was identified, along with their assumed wood procurement areas.
The type of source material used by each mill was also identified from available databases.
Forest inventory data and harvest records from the USDA Forest Service were used to characterize the impact of harvesting roundwood (live trees of merchantable size) and smaller trees for pellets on the carbon stocks and stock changes in the procurement areas.
The emissions from changes in carbon stocks were then compared with estimates of the emissions from combustion of biomass from different sources (roundwood, logging debris and mill waste).
The proportion of the total emissions was estimated for each source category.

Combined, these estimates represent the total amount of forest-based emissions plus lost forest growth and decay of unused post-harvest debris associated with burning pellets in the UK that were sourced from the southern US.

Figure 8 below shows the flow of biomass and the associated CO₂ emissions and removals pathways. A forest goes through a typical management cycle beginning with an existing forest, then a thinning or stocking control treatment, then a final harvest, and finally regeneration and regrowth. When thinned or clearcut, there is a temporary reduction in growth rate (referred to here as ‘lost forest growth’) and reduced removal of CO₂ from the atmosphere. Harvested tree boles, logging residues and mill residues are all sources of products. Here, we separate the accounting of each so that only the proportion of these sources associated with producing pellets is counted as emissions attributed to burning pellets for energy.

Figure 8. Flow of biomass and associated CO₂ emissions and removals pathways

Pellet mills in the southern US that were in operation for all of 2019 were identified from purchased data compiled by Forisk Consulting.122 Pellet mills that were likely to export pellets to the UK were identified based on export statistics from the US International Trade Commission123 and Eurostat124 trade data. Based on this analysis, 10 mills were selected for further analysis. These included three operated by Drax, six operated by Enviva and one operated by Georgia Biomass (Figure 9).

Figure 9. Locations of mills analysed in this study

— Source: Adapted from Southern Environmental Law Center.

Similarly to previous studies,125 and as per common practice when assessing the procurement area for a mill, a circle with a 50-mile radius centred on each mill was used to approximate this area. The analysis assumed no overlap in the procurement areas between mills. One exception was the three Enviva mills situated close together around the North Carolina-Virginia border, near the company’s Port of Chesapeake facility. In this specific example, a procurement area with a 75-mile radius was established, centred at the mid-point between the three mills.

All of the pellets produced by the three Drax-owned mills were assumed to be exported to the UK. In order to match the total quantity of pellets imported to the UK (according to UK import data) and used by Drax from all southern US mills (as shown in Table 9 above), the quantity of pellets exported from each of the remaining seven selected mills was reduced by 29.8 per cent; the adjusted figures are shown in Table 11. This accounting adjustment was necessary because the data on production of pellets by the different mills in the US is not directly associated with exports from US ports to specific countries, meaning that not all of the production of pellets by most mills (except those operated by Drax) would be exported to the UK.

Table 11. Assumed exports to Drax Power Station in 2019

Mill	Dry tonnes of pellets
Enviva Cottondale	565,492
Enviva Greenwood	323,138
Enviva Northampton Enviva Southampton Enviva Ahoskie	1,168,214
Enviva Sampson	429,914
Drax Amite	578,000
Drax La Salle	496,000
Drax Morehouse	578,000
Georgia Biomass	474,047
Total	4,612,805

Source: Compiled by the authors, based on mill production data and export/import data.

Note: All production by the Drax-owned mills was assumed shipped to Drax Power Station. Assumed exports from the remaining mills were reduced by 29.8 per cent so that the total exported from all mills was equal to the total imported from the southern US to the Drax facility in the UK.

The main categories of source material used for wood pellets in the US are waste wood (logging residues and mill wastes) and roundwood. Logging residues are defined by the USDA Forest Service as tree limbs, tops and stumps that are typically left behind in the forest after logging operations remove the merchantable boles of live trees. Also, by definition logging residues may include trees damaged or killed in the course of logging and left behind, unmerchantable live trees such as small-diameter trees or undesirable species and live trees from pre-commercial thinning (including other intermediate harvest operations in naturally regenerated forests). There is scant evidence that pre-commercial thinning is a common practice in the southern US.126, 127 Mill wastes include sawdust and trimmings such as slabs and edgings from tree boles that are processed for various products in lumber mills. Much of this material is burnt for power in the lumber mills that produce the waste, but some of it is now being used to produce wood pellets and other composite wood products. Roundwood consists of merchantable-sized live trees that are harvested for various wood products such as pulp and paper, lumber and plywood. In the southern US, roundwood is typically harvested by commercial thinning and clearcut. The wood harvested from any specific logging operation may be used for multiple products. Often the poorest quality and lowest value wood is used for pellets.

The quantities of the different sources of biomass used by each selected pellet mill in 2019 were taken from the Forisk Wood Bioenergy US Database,128 which reports green tons of pulpwood (referred to here as roundwood), logging residues and mill residues. The Forisk data allowed for simulating individual mill contributions to exports of pellets to the UK and to Drax specifically, as well as the impacts of pellet production on the mills’ supply regions. Results from individual mills were aggregated and then the exports sourced from non-Drax mills were adjusted as described earlier to match the actual imports of pellets for use by Drax in the UK.

Two adjustments to the reported amounts of source material in the Forisk database were made. First, Drax reports for 2019 suggest that about 55 per cent of all source material was from roundwood. Therefore the amount reported by Forisk for the mills that export to the UK (61 per cent) was reduced accordingly (Table 11) and considered part of a ‘scenario 1’ estimate of roundwood used for pellets. Second, after this first adjustment, a ‘scenario 2’ estimate was made for roundwood. This was because estimates reported by Forisk and by Drax are likely to be biased downward, because of a reluctance by pellet mills to acknowledge their harvesting of live trees.129 To account for this possibility in scenario 2, 50 per cent of the total reported logging residues used for pellets was reallocated to the roundwood category (Table 12). This approach is based on the broad legal definition of logging residues, which as described earlier can include pre-commercial thinning of live trees and other non-commercial live trees that may be harvested along with merchantable trees.130 Based on these adjustments, the estimate of roundwood used to produce pellets in the US for export to the EU and UK amounts to  55–67 per cent of the total supply of biomass from all sources.

Table 12. Sources of biomass used for pellets in dry tonnes of biomass

Source	Scenario 1	Scenario 2
Roundwood	2,537,043	3,083,764
Logging residue	1,093,442	546,721
Mill residue	982,320	982,320
Total	4,612,805	4,612,805

Source: Compiled by the authors: as reported in mill production records compiled in the US.

Note: The reported total was adjusted downward to match estimated imports from the US by Drax. The amount of roundwood reported by Forisk was also adjusted downward to match the percentage reported by Drax (55 per cent) for scenario 1. In scenario 2, one-half of the reported logging residues was shifted to the roundwood category.

Since there is no publicly available quantitative data on the amount of roundwood that comes from commercial thinning131 vs final harvest, and since the type of harvest determines the impact on growth, both scenarios were further defined to help establish bounds for the impact of using roundwood on forest carbon. Scenario 1 assumed that all harvest was from commercial thinning, while scenario 2 assumed that all harvest was from final clearcut. This approach provided for a wide range of estimated impacts on forest carbon stocks and sequestration following harvest, with the actual values expected to be within those extremes.

The USDA Forest Service forest inventory database132 was queried to estimate the characteristics of the assumed procurement areas based on the most recent forest inventories for the southern states. Inventory dates based on the time of field measurements ranged from 2012 to 2019, and the time between successive measurements varied from five to seven years among states depending on implementation schedules. Variables used in this analysis were estimated by forest type and stand origin – i.e. whether planted or artificially regenerated. Variables derived from the inventory data included: forest area; the main non-soil (biomass, dead wood, litter, and understory) carbon stocks and stock changes; harvest areas and above-ground biomass harvested; and data to calculate the proportion of total tree biomass allocated to bole, stump plus roots, and branches plus tops. The total number of inventory sample plots falling within the procurement areas associated with the 10 mills was 4,640.

The previous estimates from publicly available forest inventory data and mill surveys were used to infer the impact of pellet production on biomass stocks, carbon accumulation in the source areas and emissions from burning pellets by type of source material. To estimate the direct emissions from burning pellets sourced from roundwood, logging residues or mill residues, the average emissions figure reported by Drax for 2013–19 – 1.80 tonnes CO₂ per tonne of wood pellets – was used (see Section A1).

(In fact, the same figure results from converting the dry weight of the pellets to CO₂, assumed to be 100 per cent emitted when burnt, using the following unit conversion factors:

One US ton = 0.907 metric tonne
One green unit of roundwood or residue = 0.45 dry unit
One dry metric tonne biomass = 0.49 metric tonnes carbon
One metric tonne carbon = 3.67 metric tonnes CO₂
One dry metric tonne of wood pellets = 1.8 metric tonnes of CO₂.)

Estimating the impact of harvesting roundwood for pellets on forest carbon accumulation required estimates of the area and biomass attributed specifically to pellet production. The supply of wood for pellets is typically integrated with the supply of wood from harvesting for other products like lumber and paper. Therefore, it was necessary to assume that wood for pellets has a proportional impact on forests, according to the mass of pellets produced compared with the mass of wood harvested for all purposes. The proportion of area and biomass represented by harvesting roundwood for pellets was estimated by dividing the dry biomass of pellets from roundwood by the total biomass harvested in the procurement areas for all products, for scenarios 1 and 2 of estimated roundwood used. As described above, scenario 1 reflected reported estimates of roundwood used plus the assumption that all harvest of live trees for pellets was by thinning. Scenario 2 was defined by transferring some of the source material from the logging residue category to the roundwood category, plus assuming all harvest for pellets was done by clearcut.

The average annual forest carbon increment lost from harvesting for pellets was estimated by multiplying the area-equivalent harvested by the difference between the reduced growth from harvesting live trees and the growth that would have occurred if the trees had not been harvested, accumulated over the number of years it would take to reach the pre-harvest rate of carbon increment. These values were estimated from growth rates by age class from the forest inventory data. It assumed ages of 17.5 years for commercial thinning and 27.5 for final harvest of naturally regenerated forests, and 12.5 years for commercial thinning and 22.5 years for final harvest of planted forests.133, 134 Assumed ages of harvest were confirmed by examining forest inventory data showing harvest areas by forest age class, though the average ages of harvest may not reflect the fact that some older forests are also harvested. Calculations were carried out separately for planted and naturally regenerated forests, owing to the significant difference in average growth rates and harvest ages between the two (Table 13).

Table 13. Estimated number of years for the growth of a harvested forest to reach pre-harvest growth rates following thinning or clearcut, by forest type and stand origin

	Forest type
Harvest type	All types	Loblolly-shortleaf	Oak-pine	Oak-hickory
	Natural regeneration
Thinning	2½	3	2	2
Clearcut	7½	6	6	15
	Plantation
Thinning	1½	1	2½	2
Clearcut	6	6	8	15

Source: Compiled by the authors.

Note: These estimates were calculated from analysis of harvest and growth rates by age class from forest inventory data, for the respective forest types and stand origins. Thinning is sometimes associated with faster volume growth, but available evidence indicates that there the removal of live trees causes reduced CO₂ uptake for a period of time (see discussion in Chapter Two).

The loss of carbon from the decay of roots and from the decay of unused logging debris in the harvested forests were also estimated for harvested roundwood attributed to increasing pellet production, using average ratios of roots and logging debris to the biomass of the harvested tree bole of 1.26 and 1.172, respectively. These ratios are based on estimates of the carbon stored in different tree parts as reported in the forest inventory database for the assumed procurement areas, using component ratio biomass equations. Logging debris used for other products besides pellets was excluded from the calculations.

All of the estimates were summarized for all mills combined using common units of carbon and CO₂, accounting for the proportion of all pellets produced by the mills that were shipped to the UK and used at Drax.

Results – procurement areas and net emissions by source of material

The procurement areas around the 10 pellet mills contain 11 million hectares of forest (Table 14). About 4 million hectares are plantation forests and 7 million hectares are naturally regenerated. Loblolly-shortleaf pine is the most common forest type, with more than 4 million hectares. Of this, 61 per cent or 2.6 million hectares are planted.

Table 14. Area of forestland within procurement areas by forest type and stand origin (hectares)

Stand origin	All types	Loblolly-shortleaf pine	Oak-pine	Oak-hickory	Other types
Natural regeneration	7,119,802	1,662,147	948,479	1,874,467	2,634,709
Plantation	3,976,266	2,621,233	246,656	142,866	965,512
All origins	11,096,068	4,283,381	1,195,135	2,017,332	3,600,220

Source: Compiled by the authors.

Note: The most common forest types that supply timber are specified. Areas are unadjusted for small amounts of overlap between procurement areas.

The average annual net change in carbon stocks within the procurement areas is highest for loblolly-pine plantations at 4.72 tonnes C/ha/yr, followed closely by plantations of ‘other types’, mainly composed of longleaf-slash pine and elm-ash-cottonwood (Table 15). Naturally regenerated forests have lower carbon growth rates than plantations.

Table 15. Average annual net change in carbon stocks (carbon growth) of forests within procurement areas by forest type and stand origin (tonnes C/ha/year)

Stand origin	All types	Loblolly-shortleaf pine	Oak-pine	Oak-hickory	Other types
Natural regeneration	2.29	3.43	2.59	2.04	1.63
Plantation	4.45	4.72	2.38	1.50	4.67
All origins	3.06	4.22	2.55	2.00	2.45

Source: Compiled by the authors.

Tree harvests for all products within the procurement areas total 373,000 hectares annually, affecting 3.4 per cent of the total forest area (Table 16). Nearly two-thirds of the harvest area is plantation forest. Of the 18.5 million tonnes of biomass harvested annually for all types of wood products, about 12 per cent is harvest residues that are either left in the forests or used for timber products that can accept low-grade wood, such as pellets and particleboard. Of the total biomass harvested for all products within the procurement area, 10 per cent was attributed to pellet production in scenario 1 and 12 per cent in scenario 2.

Table 16. Summary of biomass harvest data within procurement areas by stand origin

Harvest data	All forests	Natural regeneration	Plantations
Area harvested (ha/yr)	372,626	141,905	230,721
Biomass harvested (tonnes C/yr)	18,493,140	8,127,292	10,254,466
Harvest residues (tonnes C/yr)	2,568,822	1,172,184	1,225,676

Source: Compiled by the authors.

CO₂ emissions from different sources of biomass used for pellets burnt in Drax

As a result of this analysis, annual net emissions of CO₂ from pellets sourced from the southern US and burnt in Drax were estimated at between 11 million and 13 million tonnes in 2019 (Table 17). This estimate includes:

Emissions to the atmosphere from burning the pellets (calculated from emissions reported by Drax – Section A1): 8.3 million tonnes.
Emissions from the supply chain (calculated from emissions reported by Drax – Section A2): 1.1 million tonnes.
Emissions from the decay of roots and unused logging residues left in the forest after harvest and forgone removals of CO₂ from the atmosphere due to the harvest of live trees (calculated as explained above): between 1.5 million tonnes (scenario 1) and 3.6 million tonnes (scenario 2).

Table 17. Net emissions in 2019 (tonnes CO₂) by source, from biomass used for pellets exported from the southern US to the UK for Drax, regardless of where emitted

Source	Scenario 1	Scenario 2
Combustion emissions
Roundwood used for pellets	4,566,677	5,550,775
Logging residue used for pellets	1,968,196	984,098
Mill residues used for pellets	1,768,176	1,768,176
Supply chain emissions
Supply chain	1,114,334	1,114,334
Impact on forest carbon
Decay of logging residues	1,432,293	1,740,945
Lost forest growth	115,280	1,882,909
Total	10,964,956	13,041,237
Additional percentage of combustion from lost forest growth and decay of residues	19%	44%

Source: Compiled by authors.

Note: ‘Lost forest growth’ and ‘decay of logging residues’ represents the removal of CO₂ from the atmosphere that would not have occurred if the roundwood had not been harvested.

The largest difference in estimates among the source categories for scenarios 1 and 2 is for lost forest growth. This is due to the dual effects of using less roundwood and only wood from thinnings (which has a faster growth recovery) in scenario 1, compared with using more roundwood and only using wood from clearcut harvest in scenario 2. All sources considered, the actual emissions associated with burning the pellets are 19 to 44 per cent higher than combustion emissions alone when accounting for changes in the net carbon balance of the forests (decay of residues plus lost forest growth) in the procurement areas.

Of the different emission sources, the midpoint of the two scenarios suggests that the combustion of roundwood accounted for about 42 per cent of total emissions (Figure 10) and 61 per cent of the combustion emissions (Table 17) from burning pellets. Based on the amount of biomass harvested by stand origin for all products including pellets (not shown in table), 56 per cent of the roundwood used for pellets comes from forest plantations and 44 per cent from stands established by natural regeneration. The combustion of logging and mill residues, which are often cited as a main source of wood for pellets, accounted for 12–15 per cent of the total net emissions, respectively, for the midpoint of the two scenarios.

Figure 10. CO₂ emissions associated with US-sourced wood pellets burnt at Drax, 2019

The estimate of forgone removals from lost forest growth ranges from 0.1 million to 1.9 million tonnes per year, with the midpoint of the two scenarios estimated at 9 per cent of total emissions. This large range represents the uncertainty associated with the amount of roundwood included in the reported use of forest residues, as well as the specific forest types, ages, type of harvesting (thinning or clearcut), and locations of roundwood harvested specifically for pellets. Decay of unused logging residue accounted for 1.4 million–1.7 million tonnes per year or 13 per cent of all emissions, according to the average of scenarios. Carbon accumulation rates (Table 15) were consistently estimated across the southern US and therefore do not contribute significantly to the uncertainty estimation.

The average time needed to regrow the biomass harvested for pellets varied by stand origin (plantation or natural) and whether the harvest was a thinning or final clearcut. Plantations had the shortest average regrowth periods of 5–10 years for thinning and 25–30 years for final harvest. The average of 10 years indicates that more of the supply of wood for pellets came from thinning. Because of slower growth rates, stands of natural origin would take 10–15 years to recover harvested biomass from thinning, and 40–60 years after final harvest, with an average of 25 years for thinning and final harvest combined (estimates from this analysis, not shown in table). Note that these estimates are not the same as the commonly used carbon payback period, as they do not include all elements of a full system accounting. The range of estimates is inferred from forest inventory data indicating the age and biomass stock when harvests are most common. The averages represent how long it would take the area-equivalent harvested for pellets to recover the live biomass present before the harvest based on average biomass growth rates.

Uncertainties

As mentioned in the introduction to the Annex, numerous assumptions had to be made owing to a lack of data specific to the harvest areas that supplied the different sources of biomass for pellets and the inability to track the specific flow of biomass from the harvesting sites to the mills and through US ports that shipped to the UK. A scenario-based approach was used to illustrate some of the primary sources of uncertainty and how these affected the final results.

What follows is a summary of the main assumptions:

Procurement areas of 50-mile radius represent actual source areas.
Mills selected are likely to export to the UK, based on export data from US ports and import data from the UK.
Assumed proportion of pellet production from each mill that is actually exported to the UK.
Some of the source material labelled as ‘logging residue’ actually comprises harvested live trees.
Harvesting practices range from 100 per cent thinning and partial harvest to 100 per cent clearcut.
Forest growth rates are relatively uniform across the whole procurement region (mainly the southern US coastal plain and Piedmont region).
Carbon emissions from lost forest growth and decomposition of logging debris are consistent with averages within the procurement areas, based on forest inventory data.
Forest ages at time of harvest are based on averages within the source areas, according to forest inventory data and literature review.
Harvested wood in the assumed procurement areas is proportionally allocated to pellets.
Estimated forest emissions from pellet production are based on proportion of the total wood harvest that is used for pellets.

Department of Energy and Climate Change (DECC) (2014), UK Greenhouse Gas Inventory, 1990 to 2012, Annex A3, Table A 3.2.5, http://unfccc.int/files/national_reports/annex_i_ghg_inventories/national_inventories_submissions/application/zip/gbr-2014-nir-15apr.zip.

UK Parliament (2016), ‘Baroness Neville-Rolfe answer to Parliamentary Question HL2685’, 2 November 2016, https://questions-statements.parliament.uk/written-questions/detail/2016-10-26/HL2685.

See EPA (2016), Inventory of US Greenhouse Gas Emissions and Sinks: 1990–2014, pp. 3–92,  https://www.epa.gov/ghgemissions/inventory-us-greenhouse-gas-emissions-and-sinks-1990-2014.

Brack, D., Hewitt, J. and Marchand, T. M. (2018), Woody Biomass for Power and Heat: Demand and  Supply in Selected EU Member States, Research Paper, London: Royal Institute of International Affairs,  https://www.chathamhouse.org/2018/06/woody-biomass-power-and-heat.

The figures for imports are not precisely comparable with those for consumption, as pellets can be imported in one year and burnt in the next.

As above.

Drax Group PLC (2020), Enabling a zero carbon, lower cost energy future: Annual report and accounts 2019, p. 43.

Drax Group PLC (2021), Driven by our purpose: Annual report and accounts 2020, p. 53,  https://www.drax.com/wp-content/uploads/2021/03/Drax_AR2020.pdf.

USDA Forest Service (2016), Future of America’s Forests and Rangelands; Favero, Daigneault and Sohngen (2020), ‘Forests: Carbon sequestration, biomass energy, or both?’; Buchholz, Gunn and Sharma (2021), ‘When Biomass Electricity Demand Prompts Thinnings in Southern US Pine Plantations’.

Forisk Consulting, LLC (2020). ‘Wood Bioenergy US Database’, https://forisk.com (data as at 17 April 2020).

US International Trade Commission (USITC) (2020). ‘USITC DataWeb’, https://dataweb.usitc.gov.

Eurostat (2020), ‘European Union Trade Statistics’, https://ec.europa.eu/eurostat/web/international-trade-in-goods/data/database.

Buchholz, T. and Gunn, J. (2015), Carbon Emission Estimates for Drax Biomass Powerplants in the UK Sourcing from Enviva Pellet Mills in US Southeastern Hardwoods Using the BEAC Model, Northampton, CT: Spatial Informatics Group for Southern Environmental Law Center, https://www.southernenvironment.org/uploads/audio/2015-05-27_BEAC_calculations_SE_hardwoods.pdf; Buchholz, T., Gunn, J. and Kittler, B. (2019), The Carbon Impacts of UK Electricity Produced by Burning Wood Pellets from Drax’s Three US Mills, Northampton, CT: Spatial Informatics Group for Southern Environmental Law Center, https://www.biomassmurder.org/docs/2019-05-27-sig-gis-the-carbon-impacts-of-uk-electricity-produced-by-burning-wood-pellets-from-drax-s-three-us-mills-english.pdf.

Gonzales-Benecke et al. (2011), ‘A Flexible Hybrid Model of Life Cycle Carbon Balance for Loblolly Pine (Pinus taeda L.) Management Systems’; Buchholz, Gunn and Kittler (2018), ‘UK wood pellet derived electricity: Carbon emission estimates from trees, thinnings and residues sourced in mixed pine-hardwood forests and pine plantations in the southeastern US’.

Ricardo Energy and Environment for DECC (2016), Use of North American woody biomass in UK electricity generation: Assessment of high carbon biomass fuel sourcing scenarios, London: DECC, https://assets.publishing.service.gov.uk/government/uploads/system/uploads/attachment_data/file/594046/Summary_-_Assessment_of_high_carbon_biomass_fuel_sourcing_scenarios.pdf.

Forisk Consulting, LLC (2020), Wood Bioenergy US Database.

Buchholz, Gunn and Kittler (2019), The Carbon Impacts of UK Electricity Produced by Burning Wood Pellets from Drax’s Three US Mills.

See also Buchholz, Gunn and Sharma (2021), ‘When Biomass Electricity Demand Prompts Thinnings in Southern US Pine Plantations’.

‘Commercial thinning’ is broadly defined to include partial harvest of naturally regenerated stands in addition to thinning of plantations.

Burrill, E. A. et al. (2020), The forest inventory and analysis database: database description and user guide version 8.0 for Phase 2, Washington, DC: USDA Forest Service, https://www.fia.fs.fed.us/library/database-documentation/current/ver80/FIADB User Guide P2_8-0.pdf.

Hoover et al. (2014), ‘Quantifying Greenhouse Gas Sources and Sinks in Managed Forest Systems’.

Previous chapter Next chapter

Greenhouse gas emissions from burning US-sourced woody biomass in the EU and UK

Show authors

Annex: Emissions from wood pellet use for energy

A1. Emissions from pellet combustion

Table 7. Consumption of and emissions of CO₂ from wood pellets at Drax, 2013–19

Table 8. Imports of US wood pellets to former EU28 and associated emissions of CO₂, 2014–19

Table 9. Imports of US wood pellets to UK, consumption of US wood pellets in Drax and associated emissions of CO₂ from combustion, 2014–19

A2. Emissions from the supply chain

Table 10. Supply-chain emissions of CO₂ from the use of wood pellets in Drax, 2014–19

A3. Impacts of pellet production on forest carbon stocks and sequestration in the southern United States

Approach and methods

Figure 8. Flow of biomass and associated CO₂ emissions and removals pathways

Figure 8. Flow of biomass and associated CO₂ emissions and removals pathways

Figure 9. Locations of mills analysed in this study

Figure 9. Locations of mills analysed in this study

Table 11. Assumed exports to Drax Power Station in 2019

Table 12. Sources of biomass used for pellets in dry tonnes of biomass

Table 13. Estimated number of years for the growth of a harvested forest to reach pre-harvest growth rates following thinning or clearcut, by forest type and stand origin

Results – procurement areas and net emissions by source of material

Table 14. Area of forestland within procurement areas by forest type and stand origin (hectares)

Table 15. Average annual net change in carbon stocks (carbon growth) of forests within procurement areas by forest type and stand origin (tonnes C/ha/year)

Table 16. Summary of biomass harvest data within procurement areas by stand origin

CO₂ emissions from different sources of biomass used for pellets burnt in Drax

Table 17. Net emissions in 2019 (tonnes CO₂) by source, from biomass used for pellets exported from the southern US to the UK for Drax, regardless of where emitted

Figure 10. CO₂ emissions associated with US-sourced wood pellets burnt at Drax, 2019

Figure 10. CO₂ emissions associated with US-sourced wood pellets burnt at Drax, 2019

Uncertainties