Energy use and carbon footprint of sawmilling

Why sawmill energy and carbon matter now

A sawmill turns logs into a product that stores carbon, and it does so with an energy bill that is small compared with the energy locked in the wood itself. That is why sawn timber has become a reference case in construction-sector carbon debates. For mill owners the practical point is different: the numbers customers now ask for (a global warming potential per cubic metre, an Environmental Product Declaration, a share of renewable energy) are built from a few energy flows that a mill can measure itself.

This article covers the main energy flows, what published studies report for Europe and North America, the role of kiln drying and residues, efficiency measures with documented results, life-cycle assessment (LCA) and EPD basics, biogenic carbon, and a practical starting point for a small mill.

A note on method. Every figure is attributed to a named source, with year, unit, region and system boundary where the source states them. Conversions and combinations are marked "own calculation". Where a source was internally inconsistent, or I could not verify a point, it is flagged here and in the closing sections. Board-foot figures need care: "thousand board feet" (MBF) is converted to cubic metres with different factors in different documents, so I convert only where I say so.

Where the energy goes in a sawmill

A sawmill has four broad energy consumers, using different carriers.

Electricity drives almost everything mechanical: the log yard (loaders, sorting, sprinklers), debarking, primary breakdown (headrig, band saw, circular saw line or chipper-canter), edger and resaw, sorting and stacking, chippers and hogs, conveyors, hydraulics, compressed air, dust extraction fans and lighting. The planer mill after drying adds further motors and extraction.

Thermal energy is used almost entirely for kiln drying, with smaller amounts for building heat and, in cold climates, log thawing. In mills with kilns it is usually the largest single energy flow.

Fuel for mobile equipment (mainly diesel) is small in energy terms but visible in the fossil footprint when most other energy is grid electricity and biomass heat. The 2025 California sawmill survey (Khatri and others, Forest Products Journal) found diesel, at 5.55 litres per MBF of lumber, to be the second hot spot after electricity.

Wood residues are both a by-product and a fuel. Bark, sawdust and shavings not sold to pulp or panel makers can fire a boiler for kiln heat, and in some mills a combined heat and power plant.

In the Swedish assessment by Johnsson, Andersson, Thollander and Karlsson (Energy, vol. 187, 2019), case studies of several Swedish wood-industry companies mapped energy end-use and greenhouse gas emissions by process and carrier. Drying was the largest energy-consuming and emitting process. The authors also recommend splitting energy performance indicators by wood type, end product and kiln technology, because one "kWh per m3" figure hides large differences.

Typical specific energy use: what studies report

"Specific energy use" means energy per unit of product, usually kWh or MJ per m3 of sawn timber. Two issues make comparison treacherous. The reference volume differs (logs, rough green lumber, kiln-dried or planed dry lumber), and the system boundary differs (electricity only, with or without kiln, with or without planer, or starting in the forest).

Electricity

Sources: the four documents named in the first column (see Sources used). No unit conversion is applied in the table because the papers use different MBF-to-m3 factors.

Own calculation, orientation only: the 2014 paper applies 1 nominal MBF = 1.623 m3, giving about 62 kWh/m3 (about 52 to 68). A factor of about 2.36 m3 per MBF applied to the California figure gives roughly 82 kWh/m3. These factors are not the same and I could not reconcile them, so the order of magnitude, roughly 50 to 85 kWh/m3 in North America and 53 to 67 kWh/m3 in the Norwegian data, is what to take away.

The US CORRIM life-cycle inventory of softwood lumber (Forest Products Journal 67(5-6), 2017; survey year 2012) breaks electricity down by process, a good template for self-benchmarking:

Source: Forest Products Journal 67(5-6), 2017; US Pacific Northwest and Southeast; cradle-to-gate for 1 m3 planed dry lumber, boiler energy surveyed separately. The rows have different reference units, so I did not add them.

Total energy and thermal energy

The same CORRIM study reports total cradle-to-gate energy for 1 m3 of planed dry softwood lumber of 3,434 MJ (Pacific Northwest) and 5,151 MJ (Southeast), with fossil energy of 878 and 1,160 MJ, and biomass energy of 2,449 MJ in the Pacific Northwest. The paper states that each region is more than 70% bio-based. Cubic recovery was 0.60 and 0.55.

For hardwood in the north-eastern United States, Bergman's gate-to-gate inventory (USDA Forest Service thesis, 2007) reports per m3 of planed dry hardwood lumber, counted from incoming logs: 608 MJ electricity and 5,800 MJ thermal, with 43.7% log-to-lumber volume conversion. Bergman and Bowe (10th IUFRO Wood Drying Conference, 2007) report the same averages, but their table shows 597 MJ and 5,400 MJ, so expect a few percent of uncertainty. Own calculation: about 169 kWh electricity and 1,611 kWh thermal per m3. The thermal figure is far above the softwood totals in CORRIM; species, schedules and boundaries differ, and I found no source isolating the reason.

Europe: a hardwood drying example and an audit

I found no verifiable open European softwood study giving both electricity and heat per m3. The closest data on heat is a Serbian beech model (Popadic, Furtula and Milic, BioResources 14(3), 2019): 90 beech logs, 4 m long, 30 to 49 cm diameter, boundary from sawmill residues to usable heat, with kiln drying in a biomass-fired water-heated kiln. Drying used about 435 kWh/m3 for 25 mm lumber and about 480 kWh/m3 for 50 mm. Residue energy was about 840 to 1,350 kWh per m3 of logs. Drying used under 10% of residue energy for small-diameter logs but about 30% for 40 to 49 cm logs. The paper reports no electricity.

A 2009 TU Wien diploma thesis analysed energy audit data from Austrian sawmills; the author notes the sample is small and not representative. Heat supply had the largest savings potential, about 25%, because drying chambers are the main heat consumers and non-adjustable heating and poor insulation cause the largest losses.

The share of drying

Sources are named in the table; all appear under Sources used.

Garrahan also gives a practical reference: a conventional heat-and-vent kiln needs about 1.6 kWh per kg of water removed (a theoretical figure including expected losses), and 60 to 70% of softwood drying energy goes to evaporating moisture. His quoted 1,514 MJ/m3 total (1,000 MJ/m3 for drying) is far below the CORRIM total, and its boundary is unclear, so I treat it as unverified.

An illustrative drying calculation (own calculation)

This is a worked example, not a measurement. Take the Swedish Wood EPD product (489 kg/m3 at 16% moisture), so oven-dry mass is 489 / 1.16, about 421 kg per m3. Assume entry at 60% moisture (dry basis; my assumption) and exit at 16%: water removed is about 421 x 0.44, roughly 185 kg. At 1.6 kWh/kg that is about 296 kWh of heat per m3 (about 1,070 MJ). At an assumed entry of 100%, water removed is about 354 kg and heat about 566 kWh/m3. This range brackets the 435 to 480 kWh/m3 measured for beech, though species and thickness differ. The lesson: heat demand is nearly proportional to the water you remove, so starting moisture and yard drying matter.

Residues as fuel

In CORRIM, cubic recovery was 0.60 (Pacific Northwest) and 0.55 (Southeast), so roughly 40 to 45% of log volume leaves as chips, sawdust, bark and shavings. Bergman's hardwood inventory gives 43.7%. CORRIM allocated by bone-dry mass, giving lumber 50.1% (Pacific Northwest) and 52.2% (Southeast) of the sawmill-step impacts. Bark, sawdust and hog fuel are used as boiler fuel or sold off site.

So many mills fire kilns with their own residues. In CORRIM, biomass supplies about 70% or more of cradle-to-gate energy; Bergman and Bowe report on-site wood fuel supplying about 74% of thermal energy in the hardwood mills surveyed. EPDs show a similar picture: non-renewable primary energy is about 740 to 1,140 MJ per m3 in the Swedish and British Columbian declarations, while renewable primary energy as an energy carrier is 252 to 342 MJ (Swedish Wood), 1,540 MJ (Fiskarheden), 3,382 MJ (British Columbia) and 7,110 MJ (SCA). That spread probably reflects how studies classify biomass.

Residue use is a business decision: a mill that cuts kiln heat frees residues for sale, so heat savings have a revenue side. Carbon accounting treats residue combustion differently by method (see biogenic carbon).

Energy efficiency measures with documented results

The evidence is mixed: some figures come from trade guides and old demonstration projects, so source and year are given.

Electricity side

• Variable speed drives on fans and pumps. The Energy Trust of Oregon wood-products guide (printed January 2024) cites fan affinity laws: at 80% flow a fan uses 51% of full power, at 50% flow about 13%. Typical simple paybacks: kiln fans 2 to 6 years, dust collection fans as short as 2 years, motors and hydraulic pumps 1 to 5 years. Garrahan (2011) says fan upgrades and variable-speed drives can cut fan electricity by 30 to 40%; fan electricity is 5 to 10% of kiln energy but up to 20% of its energy cost. The Austrian audit thesis (2009) estimated about 7% savings for drives and 4% for ventilators.
• Compressed air. In the Austrian audits, pneumatic systems had the highest electrical saving potential, about 28%. The Energy Trust guide says each 2 psi pressure reduction saves about 1%, and leak repair, zero-loss drains and master controls pay back quickly. A US Department of Energy case at a Weyerhaeuser sawmill in Coburg, Oregon (2000) reports about 1.3 million kWh per year saved for a $55,000 project with one-year payback. The Norwegian programme also reported compressed air measures with payback under a year.
• Scheduling and idle running. Interlocking auxiliaries to the saw line costs little; I found no sawmill-specific study, so no number is given.

Heat side

• Kiln controls, venting and heat recovery. Garrahan (2011): heat exchangers on the vent stream can preheat incoming air to within 10 to 20 degrees Fahrenheit of kiln temperature. The Energy Trust guide lists fan controller tuning (payback under one year) and sealing and insulating the kiln, without percentages.
• Pre-drying. Garrahan states each 1% moisture reduction before the kiln cuts kiln energy by about 40,000 BTU (not verified), consistent in direction with the calculation above.
• Kiln technology. Garrahan reports dehumidification kilns use up to 50% less energy than heat-and-vent kilns but run on electricity, and one southern pine study showed 13% energy reduction with high-temperature drying. Bergman and Bowe (2007) found hardwood mills with dehumidification kilns used 45.3% more electricity but under 5% of average thermal energy.
• Boiler and steam. Leaky steam traps (payback under a year) and excess-air optimisation (as short as 2 weeks), per the Energy Trust guide.
• Energy management. Olsson (2020): a sawmill energy management system (policy, review, team, targets, action plan, weekly review) went with a fall in specific electricity from 66.8 to 53.1 kWh/m3, about 20%, worth roughly 90 thousand euros a year, between the 2015 and 2020 data points. Awareness alone gave 2 to 5% in the first year. Energy is only about 5% of mill costs, which weakens the incentive.

Payback figures are typical estimates from US and Norwegian sources, not guarantees; the Energy Trust guide does not tie its ranges to a year.

Life-cycle assessment basics

LCA (ISO 14040 and 14044) adds up environmental flows over a defined life cycle. Four choices define a sawmill result.

1. Declared unit. Typically 1 m3 of sawn, dried, planed or rough timber at a stated moisture content and density.
2. System boundary. Cradle-to-gate covers EN 15804 modules A1 (raw material supply, including forestry), A2 (transport to the factory) and A3 (manufacturing: sawing, drying, planing, packaging). Further modules: A4 and A5 (transport to site, construction), B1 to B7 (use), C1 to C4 (end of life), D (benefits beyond the system).
3. Allocation. A sawmill makes several products at once; impacts are shared between timber and co-products by mass, economic value or system expansion. CORRIM allocated by bone-dry mass. The method can change the result.
4. Data quality. One mill, a group, or an industry average, and from which year? The Swedish Wood EPD uses 2019 data from 16 planing mills, about 25% of Swedish planed wood output.

An LCA has goal and scope, inventory, impact assessment and interpretation.

EPDs for sawn timber

An EPD is a third-party-verified document reporting LCA results in a standard format. In Europe EN 15804 is the core rule set for construction products. EN 15804+A2 (2019) splits climate impact into GWP-fossil, GWP-biogenic and GWP-luluc (land use and land use change), plus GWP-total. The older +A1 reported a single indicator.

Declarations are published in programme libraries: EPD International (environdec.com), EPD Norge (via epd-global.com) and, in the German-speaking area, IBU. In North America EPDs follow ISO 21930 and are verified under programmes such as UL Environment or ASTM International. The German ÖKOBAUDAT database holds Thünen Institute's "Nadelschnittholz - getrocknet (Durchschnitt DE)" (version 2023-I, under EN 15804+A2); the metadata record I could check has no numbers, so I quote none. I did not retrieve an IBU sawn-timber EPD.

EPDs have a validity period, usually five years; several below have passed or are close to it.

Reported GWP values for A1-A3 per m3

All values are cradle-to-gate (A1 to A3), per 1 m3 of the product shown. Biogenic values are negative because they count carbon taken up by the tree.

Unit: kg CO2 eq per m3. Sources: the documents named in the second column (environdec.com and epd-global.com EPD files, the AWC/CWC and ASTM declarations, Forest Products Journal papers), all under Sources used. North American values use TRACI 2.1 and different product definitions from the EN 15804+A2 Swedish values; do not compare line by line.

Reading the table:

• Fossil GWP tracks mill energy. Swedish values are about 29 to 49 kg CO2 eq per m3; North American cradle-to-gate values about 36 to 81. Grid mix, forestry diesel and transport, and the fossil share of kiln heat can explain much of the spread.
• A3 dominates. In the AWC/CWC declaration A3 is 42.56 of 63.12 (about two thirds, own calculation); in the BC EPD 34.28 of 45.86 (about three quarters, own calculation). Manufacturing, meaning electricity and drying, is what a mill controls.
• Primary energy. The Swedish Wood EPD reports non-renewable primary energy (excluding material) of 900 MJ per m3 for a combination plant and 1,140 MJ for a standalone planing mill, about 250 kWh for the former (own calculation). Fiskarheden reports 737 MJ non-renewable and 9,920 MJ renewable (8,380 MJ as material); SCA 755 and 15,000 MJ (7,920 MJ as material); BC 760.87 and 3,381.89 MJ as energy. These are primary energy, not the final energy a meter reads.
• Electricity and heat are not split in the Swedish EPDs I read, and the drying fuel is not specified in the Fiskarheden or SCA documents. Your own records must answer "how is your kiln heated?".

Biogenic carbon accounting

Wood is about half carbon by dry mass; Cardinal and others (Ecological Indicators, 2024) use 50% of oven-dry weight. The EPDs agree: 218 kg C per m3 (Fiskarheden), 211 (Swedish Wood), 206 (SCA) and 214.3 kg C, or 785.7 kg CO2 eq (BC). CO2 is carbon times 44/12, about 3.67. Check (own calculation): 218 x 3.667 = 799 kg CO2, matching the declared -800; 211 gives 774, matching -773.

How it appears depends on the standard:

• EN 15804+A2. Uptake is counted as negative in A1 and release at end of life (module C) as positive. In the Stora Enso EPD guide (a CLT product, used only to illustrate the method) the A1 biogenic value is -762 kg CO2 eq per m3 and +762 appears in C3; they offset. GWP-total is the sum of fossil, biogenic and luluc; GWP-GHG is the same with biogenic CO2 set to zero. The Swedish Wood EPD declares A1-A3 as one total, with C1-C4 and D declared separately.
• North American ISO 21930 style. The WoodWorks explainer describes GWPTRACI as extraction, transport and manufacturing emissions, and GWPBIO as GWPTRACI plus biogenic values. In the AWC/CWC declaration, biogenic carbon removed in the product is -2,052.87 kg CO2 per m3; emitted from the product 1,868.67 (1,025.02 in A3 and 843.66 in C3/C4); emitted from combustion of renewable waste 184.80. Net biogenic emissions across A1-A3 are zero, and all stored carbon is conservatively assumed released at end of life, so no net storage is claimed.
• A mill example. The BC EPD shows biogenic flows of -1,520.04 in A1 and +1,561.91 in A3 beside a 45.86 kg CO2e total. It says the stored carbon (214.3 kg C per m3) is included for later cradle-to-grave work but not counted as long-term storage.

A large negative GWP-biogenic does not mean the mill removed carbon from the atmosphere. It means the product contains carbon the tree absorbed. Whether that is a climate benefit depends on forest regrowth and on the product's fate, outside a cradle-to-gate declaration. Always check whether a footprint is fossil only, includes biogenic flows, or includes land-use change.

Carbon storage in wood products

National inventories count storage in the harvested wood products (HWP) pool. The Swiss inventory documentation (FOEN, National Inventory Document 2026, covering 1990 to 2024) uses default half-lives of 35 years for sawnwood, 25 for wood panels and 2 for paper, from IPCC (2014) Tier 2 defaults, with the production approach (carbon in products from domestic harvest, wherever the product ends up). I could not open the paper on IPCC 2019 Refinement defaults, so I do not state how those half-lives differ.

Two caveats. Stored carbon per m3 (roughly 700 to 800 kg CO2) is large compared with manufacturing fossil emissions (roughly 30 to 80 kg CO2 eq); this is an own comparison across different sources and boundaries. And the benefit is conditional on the product staying in use, the forest regrowing and what the wood replaces. Cardinal and others (2024) report average substitution factors of about 0.80 tC/tC for sawnwood in construction against steel and reinforced concrete alternatives, excluding biogenic carbon, and advise that factors should fall as those materials decarbonise. A mill's declared GWP should not claim substitution benefits unless an EPD includes module D.

System boundaries and caveats

• Reference unit. Per m3 of logs, green, dried or planed: the same mill gives different numbers. Check moisture content (8%, 15% and 16% appear above).
• Board feet. Factors of 1.623, 1.59 and about 2.36 m3 per MBF appear in the sources. The AWC/CWC declaration is inconsistent: one statement gives about 39.7 kg CO2 eq per MBF, another would give about 100.
• Primary versus final energy. EPDs report primary energy; meters report final energy.
• Representativeness. The West Virginia survey ranges from 31 to 588 kWh/MBF, a factor of nearly 19.
• Validity. Swedish Wood EPD S-P-02657 ended 2026-06-14 per the document, before this article's date (2026-10-26); renewal not checked.
• Grids. The SCA EPD models Swedish production-mix electricity at 35.3 g CO2 eq/kWh; a coal-heavy grid gives a very different electricity footprint.

How a small sawmill can start measuring

This follows the Norwegian programme structure (Olsson, 2020) and the indicator advice of Johnsson and others (2019).

1. Fix one reference unit. Cubic metre of sawn output at a stated moisture; record how volume is measured. Also keep log intake volume to compute recovery.
2. Collect twelve months of bills and production counts. Electricity, diesel, propane or boiler fuel, residues sold and burned. Divide by production for a first kWh per m3, and compare with the ranges above, boundary caveats in mind.
3. Add meters on the big loads. Main incoming supply, saw line, kiln fans and compressor. Compare weekly readings with weekly production. Metering the compressor on a stopped weekend shows leak load directly.
4. Log every kiln batch. Volume, species, thickness, initial and final moisture, schedule, run time and any heat reading. Compute kWh per kg of water and compare with Garrahan's 1.6 kWh/kg reference.
5. Separate carriers. Electricity, fossil fuel and biomass (tonnes or loose m3, with moisture), with stated conversion factors (3.6 MJ per kWh).
6. Pick two or three quick wins. Compressed air leaks and pressure, idle fans and conveyors, kiln sealing and fan control, steam traps. Payback depends on your equipment.
7. Build a routine. A named person, a target, a short weekly review. Olsson reports 2 to 5% from awareness in year one.
8. Then estimate carbon. Own method: electricity kWh x grid factor plus fuel litres x fuel factor, divided by m3 produced, with official emission factors stated. A1 and A2 need log supplier data; start by asking for diesel use and haul distance. This is not an EPD and must not be marketed as one, but it is a defensible internal figure for talking to a verifier or consultant.
9. If customers demand an EPD, ask your trade association whether a sector EPD exists and whether a mill may take part (participation rules not verified).

Sources used

Open journal papers, USDA Forest Service documents, EPD files from the International EPD System, EPD Norge, AWC/CWC and ASTM, and Nordic, Austrian, Serbian and Swiss technical documents. The full list follows and is attached in structured form.

• Appalachian hardwood sawmill electricity use, Forest Products Journal 64(1-2), 2014: https://fpj.kglmeridian.com/view/journals/fpro/64/1-2/article-p33.xml
• Appalachian hardwood sawmill energy survey and audits, Forest Products Journal 62(1), 2012: https://fpj.kglmeridian.com/view/journals/fpro/62/1/article-p32.xml
• CORRIM life-cycle inventory of softwood lumber (Pacific Northwest and Southeast), Forest Products Journal 67(5-6), 2017: https://fpj.kglmeridian.com/view/journals/fpro/67/5-6/article-p331.xml
• Bergman, Gate-to-Gate Life-Cycle Inventory on Hardwood Sawmills in the Northeastern Region of the United States, USDA Forest Service Treesearch, 2007: https://research.fs.usda.gov/treesearch/46320
• Bergman and Bowe, Environmental Impact of Producing Hardwood Lumber Using a Life-Cycle Inventory, IUFRO Wood Drying Conference 2007 (Treesearch): https://research.fs.usda.gov/download/treesearch/33313.pdf
• Khatri and others, On-Site Energy Consumption and Life-Cycle Environmental Impacts of Sawmills in California, Forest Products Journal 75(2), 2025: https://research.fs.usda.gov/treesearch/69017
• Swedish Wood, EPD: Swedish sawn and planed wood product of spruce or pine (S-P-02657), 2021: https://api.environdec.com/api/v1/EPDLibrary/Files/8ef05152-0174-4b7f-4141-08d92a4c0fd0/Data
• International EPD System library: Swedish sawn timber of spruce or pine (S-P-01325): https://www.environdec.com/library/epd1325
• International EPD System library: Swedish sawn dried timber of spruce or pine, Svenskt Trä (S-P-02537): https://environdec.com/library/epd2537
• EPD Norge NEPD-4821-4043, Sawn timber of spruce or pine u 16%, Fiskarhedens Trävaru AB, 2023: https://epd-global.com/getfile.php/1343783-1693993697/EPDer/Byggevarer/Heltreprodukter/NEPD-4821-4043_Sawn-timber-made-of-spruce-or-pine--u-16-.pdf
• EPD Norge NEPD-3442-2053, Sawn dried timber of spruce and pine, SCA Wood AB, 2022: https://epd-global.com/getfile.php/1324058-1653305090/EPDer/Byggevarer/Heltreprodukter/NEPD-3442-2053_Sawn-dried-timber-of-Spruce-and-Pine-with-an--average-moisture-content-of-16-.pdf
• AWC and CWC, EPD North American Softwood Lumber (UL Environment 4788424634.102.1), 2020 (copy hosted by Iowa DNR): https://www.iowadnr.gov/media/3251/download
• Athena / ASTM EPD 398, Softwood Lumber BC Average Product, 2023: https://www.astm.org/v3/assets/blt5eb0a2cb04534832/bltbe61f32d92123225/697b20b02cca71df239cb563/903.Softwood_Lumber_BC_Average_Product_EPD.pdf
• American Wood Council launches Northern U.S. softwood lumber EPD, Wood & Panel, 2025: https://www.woodandpanel.us/news/article/american-wood-council-launches-northern-u-s-softwood-lumber-epd-to-complete-regional-sustainability-data-suite/
• WoodWorks, Understanding the carbon numbers in a wood EPD: https://www.woodworks.org/resources/understanding-the-carbon-numbers-in-a-wood-epd/
• Stora Enso EPD guide (CLT example of EN 15804+A2 indicators): https://storaenso.com/-/media/documents/download-center/documents/sustainability/stora-enso-epd-guide.pdf
• Thünen Institute, Nadelschnittholz getrocknet (Durchschnitt DE), ÖKOBAUDAT 2023-I metadata (openagrar): https://www.openagrar.de/servlets/MCRFileNodeServlet/openagrar_derivate_00058087/Metadaten_openagrar_mods_00094595.pdf
• Johnsson, Andersson, Thollander, Karlsson, Energy savings and greenhouse gas mitigation potential in the Swedish wood industry, Energy 187, 2019 (record): https://agris.fao.org/search/en/records/65dfd9924c5aef494fe5e7f5
• Olsson, Energy management in the Norwegian sawmill industry, ECEEE Industrial Efficiency 2020: https://www.eceee.org/library/conference_proceedings/eceee_Industrial_Summer_Study/2020/3-energy-management/energy-management-in-the-norwegian-sawmill-industry/2020/3-157-20_Olsson_pre.pdf/
• Reisenbichler, Energieeffizienzsteigerung durch standardisierte Energieaudits, TU Wien diploma thesis, 2009: https://repositum.tuwien.at/handle/20.500.12708/11740
• Popadic, Furtula, Milic, Influence of beech log diameter and quality on energy in sawmill residue, BioResources 14(3), 2019: https://bioresources.cnr.ncsu.edu/wp-content/uploads/2019/06/BioRes_14_3_6331_Popadic_FM_Infl_Diameter_Quality_Beech_Logs_Energy_Sawmill_Residue_15658.pdf
• Garrahan (FPInnovations), How Energy Efficient is Your Drying Operation?, Wood Business, 2011: https://www.woodbusiness.ca/how-energy-efficient-is-your-drying-operation-352/
• Energy Trust of Oregon, Wood products energy savings guide, 2024: https://www.energytrust.org/fact-sheets/wood-products-energy-savings-guide/
• US DOE / OSTI, Weyerhaeuser compressed air system improvement at a sawmill, 2000: https://www.osti.gov/biblio/1216079
• FOEN (Switzerland), Harvested wood products in the Swiss GHG inventory, NID 2026: https://www.bafu.admin.ch/dam/en/sd-web/zKzAQbxiCPJd/FOEN_2026e_HWP_GHG_inventory.pdf
• Cardinal and others, Climate change substitution factors for Canadian forest-based products and bioenergy, Ecological Indicators 160, 2024: https://espace2.etsmtl.ca/28547/1/Levasseur-A-2024-28547.pdf

Limits of this article

• Extraction method. Source pages were read through a fetch tool that summarises documents; I did not read full PDFs side by side, and some tables were garbled. Components were cross-checked where they could be summed, but check each figure against the original before using it in a contract, bid or compliance document.
• Inconsistencies found in sources: SCA EPD biogenic GWP (-741 and -756); CORRIM GWP (57.8 and 81.5 versus 58.7 and 81.4); Bergman text versus table; AWC/CWC board-foot conversion; Norwegian slides' mill count.
• Coverage gaps. No verified open European softwood study with electricity and heat per m3; European coverage is Norwegian weekly electricity, a Serbian beech model, a 2009 Austrian audit thesis and Swedish EPD primary energy. North American surveys date from 2010 to 2012 apart from the 2025 California study.
• Not retrieved: numeric ÖKOBAUDAT values, IBU sawn-timber EPDs, the 2024 to 2025 AWC regional softwood EPDs and IPCC 2019 Refinement HWP defaults.
• EPD validity. The International EPD System page for S-P-01325 shows it as expired and not EN 15804 compliant; S-P-02657 reached its stated end in June 2026. Successors were not checked.
• Comparability. Standards, units, moisture contents and years differ; no ranking is implied.
• Own calculations are labelled and are arithmetic on cited numbers or illustrations with stated assumptions, not measurements.
• Not advice. General information, not an engineering audit or verified carbon claim.