Skip to content

Example data mode. The sample companies, offers and requests shown here are fictional (AI pictures) and are not counted in the figures below.

Sawing and recovery

Hardwood sawmilling in Europe: oak, beech, ash and the grading of boards

How European hardwood logs are sawn (live, cant, quarter, boules), why they split, how red heart, sapwood and ash dieback cut value, what EN 975-1 and EN 1316 actually cover, German and Polish customs, and what open data say about markets and prices.

Country / region of the study or data: Europe (Germany, Poland, France, Romania, Switzerland, UK focus); NHLA contrast from USA

This article was drafted with AI assistance and has not yet been reviewed by an expert. Check important figures against a standard or your contract before relying on them.

Why hardwood sawmilling is a different trade

A softwood mill sells volume at a standard specification. A hardwood mill sells appearance. The same oak log can become prime furniture boards, flooring lamellas or pallet timber, and the spread in value between those outcomes is far larger than between strength grades in softwood. That fact explains why valuable logs are still often sawn by a sawyer who reads the log, why unedged boards remain a normal commercial form, why the log end matters as much as the middle, and why a coloured heart in beech can halve the revenue of a stem.

This article collects what can be checked in open sources about hardwood sawing in Europe, concentrating on oak, European beech and ash. It covers sawing patterns, growth stress, colour and heart defects, steaming, grading standards and trade customs, end markets and economics. It is explicit about what could not be verified: the EN grading standards are paywalled, and where only a title, abstract or supplier summary could be read, the text says so. Figures carry region and year. "Own calculation" marks the author's arithmetic on a sourced number. General sawmill practice not tied to an opened source is flagged as such.

The resource: what is growing and what is dying

The scientific advisory board on forest policy at the German federal ministry commented on the fourth national forest inventory (BWI 2022, Germany; Bauhus et al., 2024). Oak forest area is 1.13 million ha, up 11 % on 2012; beech is 1.68 million ha, up 8 %. Own calculation: that implies roughly 1.02 million ha of oak and 1.56 million ha of beech in 2012. Ash forests lost 22 % of their area to ash dieback, and the ash growing stock fell, while stocks of all species except spruce and ash rose. Trees older than 160 years increased by 48 % for oak and 36 % for beech. Regeneration tilts the long-run mix: oak is only 12.6 % of regeneration under oak stands, beech 73 % under beech stands. The regional BWI 2022 summary for Baden-Wuerttemberg on waldwissen.net says the same in words only: oak is harvested cautiously so that increment clearly exceeds use, and ash has "drastically" lost area.

Ash dieback as a raw-material problem

Ash dieback is caused by the fungus Hymenoscyphus fraxineus. In Norway, where it has been present since 2006, a national inventory comparison (Diaz-Yanez et al., 2020) found ash mortality 74 % higher than expected and surviving ash growing 13 % slower than expected. Ash is under 1 % of Norwegian trees, so the figure shows how the disease behaves, not a trade volume.

Evidence on wood quality is thinner. A 2024 study at a site near Holzminden, Lower Saxony (Peters et al.), isolated fungi from stem collar necroses of ten ash trees: H. fraxineus was 40.5 % of isolates, Armillaria 22 % and Diplodia fraxini 15.2 %. The paper does not assess timber quality, but notes white rot in nearly all sampled trees and links D. fraxini to wood discoloration. The practical reading is that butt rot and discolouration can sit invisibly in the most valuable part of the stem.

The best quantified link between crown condition and value is from Croatia (Ursic, Zecic and Vusic, 2025): 351 narrow-leaved ash (Fraxinus angustifolia, not the common ash of northern Europe) on eight floodplain sites, grouped by defoliation. In the 62.5 cm class, trees with 61 to 80 % defoliation averaged 183.25 EUR/m3 and dead trees 59.76 EUR/m3 (own calculation: about one third). Sawmill yield was not measured, so this quantifies log assortments, not boards. Trees of 17.5 to 22.5 cm yielded only fuel wood.

How hardwood logs are converted

The patterns

The sawyer reads the log end, bumps indicating overgrown branches or wounds, pith position and any coloured core before choosing a pattern. Fixed scanner-driven patterns suit uniform, low-value runs; valuable logs are often still sawn decision by decision (general practice; no open source on the extent of scanner use).

  • Live sawing (through-and-through, boule). Parallel cuts across the log give flat-sawn boards in the centre and rift or quartered boards towards the sides. The boards are edged later or sold unedged in log sequence, the traditional form for oak.
  • Cant sawing. The log is opened into a cant that is resawn into square-edged boards, which suit automated trimming and grading.
  • Grade sawing and quadrant sawing. In grade sawing the log is rotated so the best face meets the saw each time. Quadrant sawing quarters the log first and cuts each quarter parallel to a face, giving many rift and quarter-sawn boards.
  • Quarter sawing. Rings meet the board face at a steep angle. In oak this shows the medullary rays as the "mirror" figure. Quarter-sawn boards are generally more stable across the width, but lower yield and need larger, straighter logs (general wood knowledge, not from an opened source).

What the research says

The best open evidence is old. At the USDA Forest Products Laboratory (United States, 1980, paper FPL-356), Richards, Adkins, Hallock and Bulgrin simulated the sawing of mathematical hardwood log models of 10 to 28 inches. All four-sided methods gave similar lumber values. Live sawing beat them with good-quality logs and lost for 10 and 12 inch logs with large defective cores. Live sawing followed by ripping gave the highest values in almost all cases. The abstract gives no dollar figures. It is a simulation on idealised logs under the American NHLA grading rules, so the ranking may not carry over to European appearance grading. The logic still explains why mills often sell clean oak as a boule and cant-saw logs with large hearts.

Boules and unedged trading

Selling a log as a boule lets the buyer see the whole pattern, but moves the grading problem into the contract. A UK supplier's summary of EN 975-1 (a secondary source, see the grading section) states that in a boule at least 65 % must be the specified grade, under 25 % may be one grade lower and under 10 % two grades lower, and that boule width is measured at the midpoint of the central board (limit 350 mm for QBA, 250 mm for QB1 to QB4). Check these against the standard before writing them into a contract.

The German trade customs (Tegernseer Gebraeuche, edition of 4 July 2023) say unedged hardwood is measured piece by piece at the middle, using the narrow and broad sides, the narrow side with the bark edge, or block-lying, with length rounded down to a full 10 cm. Sound, firm oak sapwood is measured; damaged or loose sapwood is not measured and is compensated.

Stresses, end checks and wounds

Growth stress

Standing trees carry maturation stress. When a log is cross-cut it releases as end splits, and when sawn the boards can spring. A study of 440 mature European beech in nine stands in five countries (Jullien et al., 2013) found a mean maturation stress of about 9.64 MPa (assuming a green-wood modulus of 12 GPa). In 45 % of trees the stress range was 15 MPa or more, and only 15 % were below 8 MPa; the 5th and 95th percentile trees differed more than fourfold. Slenderness was the strongest predictor of high stress and a large crown the strongest predictor of low stress. Trunk straightness and inclination were weak indicators, so they should not be relied on alone to predict splitting. End splitting itself was not measured.

For the sawyer, no external sign reliably identifies the log that will split, and the spread between trees is large. Quick cross-cutting, end sealing and records by stand are the usual responses (general practice, not a finding of the cited study).

Bark wounds

A forest research study summarised on waldwissen.net (no authors, region or year on the page; the address points to the FVA institute in Baden-Wuerttemberg) examined 40 each of beech, spruce and fir two years after injury. In beech, skidding wounds at the base caused minor discolouration (average penetration 0.7 cm), while felling wounds in the upper stem were worse (3.2 cm, more frequent intensive rot, most often Hypoxylon fragiforme). The study concludes that discolouration and rot reduce timber value and felling damage in beech should be avoided.

Drying stress

Oak and beech dry slowly and crack easily when stressed or when sapwood and heartwood differ strongly. That is the reason for the very long seasoning of stave wood described below. No kiln schedules were found in the open sources read, so none are quoted.

Colour and heart defects

Beech red heart

False heartwood in beech (red heart, Rotkern) is a coloured core that develops with age and after wounds. It is not decay in itself but often accompanies it, and it breaks the colour uniformity that furniture, stair and flooring buyers pay for.

A 2010 Romanian study (Campu, Transilvania University of Brasov; nine trees of 120, 140 and 160 years in the Tesla felling area) found red heartwood in about 34 % of the 65.638 m3 of round wood, varying from 5 % to 48 % between logs, and reaching 50 to 70 % of the diameter in the 160-year class. About 9 % of logs had inner decay (roughly 2 % of volume). Under the old Romanian industrial sorting, about 47 % of the volume was usable only as timber (class C) and about 48 % only as structural veneer. Under the qualitative classification of SR EN 1316-1:2001, no wood reached class A, only about 5 % reached class B, and the rest fell to C and D. The author estimates the downgrading cuts beech round wood revenue by 40 to 50 %. This is a small sample from one felling area: it shows the mechanism, not a European average.

A 2017 study in Bosnia and Herzegovina (Gurda et al., Sarajevo), in two compartments, found that red heart increases with diameter and age and occurs more often at a given size on poorer sites. The abstract reports no percentages.

The age and diameter that make a beech attractive as a sawlog also raise the chance of red heart. Mills therefore buy specifically for white beech, accept lower top-grade yield, or steam.

Oak sapwood

European oak sapwood is paler and not durable, while the heartwood is (general wood knowledge, not from an opened source), so grading treats sapwood explicitly. The UK supplier summary of EN 975-1 shows sapwood in the grade code with X (one face) or XX (both faces), and says sound sapwood is allowed on two arrises of top-grade beams if under 15 % of the face width. The German customs measure firm sapwood and not loose sapwood.

Ash colour

Healthy ash is pale and prized for uniform colour, so discolouration from dieback or Armillaria damages value. The UK national grading guide (Forest Research, 2005) lists "coloured ash, beech or sycamore" as a special feature, so coloured heart is traded as a specialist product. No opened source gave numbers on brown heart in ash.

Steaming of beech

Steaming before drying is the standard treatment of beech lumber. It homogenises colour between white sapwood and reddish false heartwood; relaxation of growth stress and easier drying are commonly claimed but no opened source quantified them.

The best open data are from Dzurenda and Dudiak (Forests 15(6), 1009, 2024), who steamed beech blanks of 32 x 50 x 800 mm from Polana, Slovakia, and measured CIE Lab* colour.

Mode Conditions Result
I Saturated moist air, about 95 C, up to 24 h Sapwood darkened (L* 78.5 to 71.0), heartwood lightened slightly (63.8 to 65.5); contrast fell from 14.8 to 5.5, zones still unequal
II Saturated steam, about 105 C, about 18 h Zones unified at L* 61.3, brown tone; sapwood darkened substantially
III Saturated steam, about 120 C, about 9 h Strongest homogenisation, dark brown-grey at L* 55.9; colour shift 22.7 (sapwood) and 11.8 (heartwood)

Source: Dzurenda and Dudiak 2024, laboratory blanks, Slovakia.

Higher temperature reached uniform colour faster but darker, and sapwood did most of the change in every mode. A mill therefore chooses a point on a trade-off between uniformity and darkness, and sells steamed beech as its own product beside white beech. Small blanks are not industrial boards, so schedules for 25 to 50 mm stock must be found empirically.

Grading: the standards and the customs

Which standard covers what

Standard Product and species Status as verified
EN 975-1 Sawn timber, appearance grading, oak and beech AFNOR: 1995 edition cancelled, replaced by April 2009 edition (current); visual grading, designations, lot composition
EN 975-2 Sawn timber, appearance grading, poplar Named in the commissioning brief; not confirmed in any opened source
EN 1316-1 Hardwood round timber, oak and beech, classes A to D 2012 edition (approved 18 August 2012), supersedes 1997
EN 1316-2 Round timber, poplar Listed in the foreword of EN 1316-1:2012; edition not checked
EN 1316-3 Round timber, ash, maples and sycamore BS EN 1316-3:1998; no scope text read

Sources: AFNOR, iTeh catalogue, BSI Knowledge.

Two points. First, EN 975-1 is titled and scoped for oak and beech. No opened source shows an EN appearance standard for sawn ash, maple or sycamore; EN 1316-3 covers them only as round timber. Ash and sycamore boards are therefore typically traded by agreement, under oak-style rules or merchant grading, and the UK guide says its grades can be applied to minor timbers such as alder, birch and walnut. Second, EN 1316 classes logs for unknown end use, while EN 975 grades boards. A mill buys on one and sells on the other, and no formula links them.

The standards' texts are paywalled and were not read. Grade criteria below come from secondary sources.

The grade codes

A UK supplier guide (EcoChoice, undated) describes the EN 975-1 code: Q for Quercus, then a product letter (S, B, F or P), then a number from A (exceptional) through 1 (good) to 4 (poor), with X or XX for sapwood on one or both faces. Listed grades include QBA, QB1 to QB4, QSA, QS1 to QS3, QF1a, QF1b, QF2 to QF4, QPA, QP1 and QP2.

Grade Knot criterion (supplier page) Wane
QF1a Sound knots under 10 mm only; 20 % of boards may have one extra not stated
QF2 Sound knots under 5 mm disregarded; up to 3 knots no larger than 25 mm not stated
QF3 Sound knots under 10 mm disregarded; up to 3 knots no larger than 40 mm; one dead knot under 20 mm under 3 mm
QF4 Sound knots under 70 mm; two dead or unsound knots (under 20 mm and under 35 mm) under 10 % of width, 20 % of length

Source: EcoChoice guide, UK. A commercial summary of a paywalled standard; some values on the page were illegible and shakes are not covered. For orientation only. The page also states that a selected lot may contain no more than 10 % of boards below the specified grade.

The UK commentary "Making the Grade" (Davies and Watt, Forest Research, 2005) follows EN 975-1 closely but simplifies it: grade 1 (uniform, "clear" or "prime"), grade 2 (some knots or splits, usable areas clear) and grade 3 (any knots, splits and colour variation, subject to suitability), with optional A and B sub-grades. It covers all UK hardwoods and one system for square- and waney-edged timber, and lists special features such as pippy, burr, quarter-sawn, spalted beech and tiger oak.

German and Polish customs

Germany. The Tegernseer Gebraeuche (2023) have a hardwood chapter but no hardwood grade classes; their quality annex covers softwood only. Hardwood must meet the agreed grades, otherwise normal growth; coarse defects are compensated in measurement but straight cracks are not, and spoiled, heavily twisted or rotten wood may be rejected. Without an agreement the wood must be dimensionally stable at a reference moisture of 18 %. Standard lengths are 3 to 6 m (up to 15 % of 2.5 to 2.9 m), in 10 cm steps. Hardwood is usually bought on inspection and collected by the buyer. Ash is listed among the "Bunthoelzer" with relaxed length rules; beech is not mentioned at all. A German buyer therefore relies on a contractual grade description, an EN 975-1 designation, merchants' own grade names or inspection of the stack.

Poland. Polish merchant grading of sawn hardwood could not be verified in open sources, so it is not described. What was read is the 2004 Polish act on classification of unprocessed round timber: a voluntary national system with quality classes A/EEC (healthy, highest quality), B/EEC (average, limited defects) and C/EEC (too defective for A or B but industrially usable) and dimension classes. It does not mention EN 1316, it lists tension wood ("drewno ciagliwe") as a hardwood quality defect, and its current legal status was not confirmed.

Switzerland as a structural reference. A 2009 study for the Swiss environment office (Luethi) counted 64 sawmills cutting some hardwood, five large ones accounting for 54 % of volume. Intake was about 130,000 m3 of roundwood (about 60 % beech, 18 % oak, 9 % ash), output about 85,000 m3 of sawn timber and veneer. Most mills use a band saw as main machine; only two of those visited use a gang saw. About a quarter of output is exported, mainly to Italy, and wood is 50 to 65 % of costs. The data are 15 years old and for a small country.

The American contrast: NHLA

NHLA grades are yield-based: a board is assessed by the share of clear cuttings of minimum size, and the grade follows from that percentage and the board dimensions. According to the American Hardwood Export Council (undated), FAS needs boards at least 6 in. wide and 8 ft long with 83 1/3 % to 100 % clear yield; No. 1 Common needs 3 in. by 4 ft with 66 2/3 % up to 83 1/3 %; No. 2A Common 50 % up to 66 2/3 %; Selects and F1F need one FAS face. The USDA FPL Wood Handbook chapter on commercial lumber confirms that grade is set from the poorer face except for F1F and Selects, and that the official NHLA rules are the only full description (its table extracted garbled and was not otherwise used).

The contrast: NHLA promises a statistical share of clear wood, which suits cutting-to-size industries. EN 975-1 and its national derivatives describe what the board face may show, which suits users who take whole boards. Neither describes strength. Mixing the two vocabularies in one order is a classic source of disputes, and no open conversion table is offered.

End markets

Flooring

Flooring is the largest quantified outlet for European oak. The European Federation of the Parquet Industry (FEP, June 2024, data for 2023) reports European parquet production of 64.74 million m2 in 2023, down 29.52 % from 92.31 million m2 in 2022; FEP countries produced 54.4 million m2. FEP blames weak construction, high interest rates, low consumer confidence and unfair competition, and sees 2023 as the bottom as destocking ended. A trade-press report on the 2025 FEP data (timber-online.net) gives 2024: FEP production 51.5 million m2 (down 5.3 %), Europe 62.2 million m2 (down 3.9 %), about 36 % below the 2021 peak of about 97.2 million m2 (as stated in the source).

Species share of wood used in parquet 2022 2023 2024
Oak 82.1 % 83.0 % 83.8 %
Ash not given 4.5 % 4.2 %
Beech not given 2.7 % 2.5 %
Tropical not given 2.0 % 1.9 %

Sources: FEP press release June 2024 (2022, 2023); timber-online.net 2025 (2024). Europe.

Multilayer was 84 % of the 2023 product mix (83 % in 2024), solid 14 to 15 %. By production share Poland led in 2023 (18.18 %) and 2024 (19.7 %), followed by Austria and Sweden; Germany was the largest consumer (17.44 % in 2023, 22.1 % in 2024). Oak lamella demand is thus concentrated in a few floor producers and moves sharply with construction.

Furniture and joinery

No quantified open source on the European share of furniture and joinery in sawn hardwood use could be read, so none is quoted. The buyers' needs are visible in the rules above: furniture wants uniform colour (white or steamed beech, colour-sorted ash), joinery wants durable heart (sapwood rules for oak), and both want dry, stable stock (the 18 % reference moisture).

Barrels and cooperage

Staves give oak its highest value per log at its lowest yield. A French spirits merchant's article on stave mills (undated) says around 10 to 20 % of a log becomes stave wood, with yields up to 40 % for oak from the Troncais forest; it admits its own figures do not add up, so treat them as indicative. Logs are split along the grain and then sawn; staves season outdoors for 24 months (moisture from about 80 % to roughly 20 to 25 %), are re-moistened and kiln-dried to 14 to 16 %, and 10 to 11 barrels of 225 litres come from 1 m3 of stave wood on average. A 2024 paper from Kosovo (Sejdiu, Marku and Sejdiu) on sawn Quercus petraea for barrels notes few firms make oak barrels and the exploitation coefficient is low; the numeric result was not in the page read.

Production, trade and yields

Trade-press reports on the European Organisation of the Sawmill Industry (EOS) give the order of magnitude. EUWID reported 5.73 million m3 for the EOS group in 2020, 4.8 % below 2019, against 6.14, 6.04 and 6.12 million m3 in 2016 to 2018. In 2020 Romania produced 1.62 million m3, France 1.24 million and Germany 991,000, below one million for the first time in many years. Croatia (1.39 million m3 in 2019) and Bosnia and Herzegovina (about 1.30 million in 2019) were not counted in the EOS totals.

For 2023 timber-online.net (November 2024) reported almost 5.9 million m3 for the member countries (France 1.35 million, Romania 1.1 million, Croatia 1.05 million), a forecast fall of about 3 % to just over 5.7 million in 2024 and recovery to just under 5.8 million in 2025. Demand was about 4 million m3 in 2023 (France 1 million, Germany 630,000, Romania 600,000), forecast down about 9 % to 3.67 million in 2024. Own calculation: production exceeds consumption by roughly 1.9 million m3 a year, which is the export dependence of the sector. The country sets differ between reports and the source is trade press with no species split.

Yield. The open sources read give no species-specific sawing yield for European oak, beech or ash. The quantified material is limited to the Swiss intake and output above (not a sawing yield, since veneer is included), the stave wood range of 10 to 20 % (up to 40 %), the 40 to 50 % revenue loss from beech red heart, and the Croatian ash value gap. Any quoted board-per-log percentage would be invented here. The number mills should measure themselves is top-grade yield by log class and supplier.

Price drivers

  • The top of the log. The Lower Saxony State Forests oak valuable-timber submission of 5 March 2025 (between Ems and Elbe) sold 1,594 m3 of oak of 1,672 m3 offered (own calculation: 95 %), plus 64 m3 of ash, to 18 bidders (15 the year before). The average oak price was 701 EUR/m3, a slight decline attributed to quality but in line with the 2022 to 2023 peak, and the top log of 3.19 m3 reached 2,889 EUR/m3 (9,215 EUR). These are selected-log prices, not the price of an ordinary sawlog.
  • Defects. Red heart cut beech round wood revenue by 40 to 50 % in the Romanian sample; defoliated or dead ash was worth far less per m3 in the Croatian study.
  • Downstream demand. With oak at 83.8 % of parquet species and parquet output down from about 97.2 to 62.2 million m2 between 2021 and 2024, lamella-grade oak demand fell with it. EOS-group demand was forecast down 9 % for 2024.
  • Resource. Ageing oak and beech stands and a 22 % loss of ash area (Germany, 2012 to 2022) mean ash sawlogs are scarcer and more variable, which rewards accurate log grading.
  • Raw material share. Wood is 50 to 65 % of a hardwood mill's cost in the Swiss study (2009), so purchasing skill often decides margin.
  • Contract terms. Measurement method, sapwood and wane treatment, reference moisture and sale on inspection move money between seller and buyer; the German customs show they are matters of agreement.

Sources used

Limits of this article

  • The texts of EN 975-1, EN 975-2, EN 1316-1, -2 and -3 are paywalled and were not read. Status and scope come from catalogue pages. Oak grade criteria come from one UK supplier's summary with illegible values, and from a UK national commentary that is not the standard.
  • EN 975-2 (poplar) is named only in the brief and is unverified. EN 1316-2 is verified only as a foreword listing.
  • No opened source shows a European sawn appearance standard for ash, maple or sycamore.
  • Polish merchant grading for sawn hardwood, Polish customs for oak and beech, and the current status of the 2004 Polish act were not verified.
  • UNECE market documents could not be fetched, so no UNECE figures are used. The search allowance ran out before sources on quarter sawing, kiln schedules, end-check incidence and ash brown heart could be searched; those points are flagged as general practice or omitted.
  • EOS and ETTF volumes come from trade press, not the organisations' own publications, and the country sets differ between years.
  • No species-specific European sawing yield, beech or ash sawlog price series, or furniture and joinery share was found.
  • The FPL sawing simulation is a 1980 U.S. study of idealised logs under NHLA grading.
  • Several studies are small or local (nine beech in Romania, two compartments in Bosnia, ten ash in Lower Saxony) and show mechanisms, not European averages. The Croatian study concerns narrow-leaved ash and defoliation, not dieback diagnosis.
  • NHLA figures come from a trade page; the FPL table was garbled and used only for confirmation.
How this text was checked · checked 2026-10-26

Each source page opened one at a time via WebFetch; numbers quoted with region and year; own calculations labelled; paywalled standards read only as catalogue entries; web search budget exhausted before final topics

Could not be verified

  • Text of EN 975-1, EN 975-2, EN 1316-1/-2/-3 not read (paywalled); grade criteria from a UK supplier summary with illegible values
  • EN 975-2 (poplar) and EN 1316-2 edition not confirmed beyond a foreword listing / the brief
  • No EN sawn appearance-grading standard for ash, maple or sycamore found; brief's framing corrected in text
  • Polish merchant grading for sawn hardwood and current status of the 2004 Polish act not verified
  • UNECE documents could not be fetched; no UNECE figures used
  • EOS/ETTF volumes from trade press only, country sets vary between years
  • No species-specific European sawing yields, beech/ash sawlog price series or furniture/joinery shares found
  • Quarter sawing, scanner use, kiln schedules, oak sapwood durability, steaming effects on stress and end-check practice stated as general knowledge only
  • Bark-wound study on waldwissen.net: region and year not stated
  • NHLA figures from a trade page; FPL table garbled and used only to confirm system

Sources

Find companies offering this service

Last updated 2026-10-26

More in this category