Sweet chestnut

Overview and range in Europe

Sweet chestnut (Castanea sativa, also Spanish or European chestnut) is a warm-temperate broadleaf grown both for nuts and for timber. The European Atlas of Forest Tree Species text (ILVO copy) says its native range is uncertain because of long human management, that it now spans north-western Africa, Europe and western Asia, and that Europe holds over 2.5 million ha, 89% of it in a few countries led by France and Italy, followed by Spain, Portugal and Switzerland. It grows from 200 to 1,800 m, prefers well-drained, acidic to neutral, nutrient-poor soils, does not thrive on limestone and is highly sensitive to summer drought.

For the sawmill, the key fact is the management type. The Atlas states that coppice covers about 80% of chestnut forests, yielding firewood, charcoal, poles for fences and pit props, and small items such as barrels and shingles, while pure high forest is rare, about 10%. Regional detail supports this: a Calabrian study reports 101,600 ha of chestnut, about 21.1% of the region's woodland, and an English Nature report (Buckley and Howell, 2004) gives 18,788 ha for Britain in the 1995-99 census, with coppice rotations of 12 to 16 years. In the Pfälzerwald of Germany, all current stands at one forest office originated from coppice (Wambsganß and co-authors, AFZ-DerWald 2013).

The Atlas notes that susceptibility to ring shake limits high-quality timber use; it is the central issue for this species.

Appearance and identification

The Wood Database describes light to medium brown heartwood darkening to reddish brown, with narrow pale sapwood. The LWF paper (Richter and Ehmcke) puts sapwood at about 1 to 1.5 cm, which is removed when round stock is processed. Grain is straight to spiral or interlocked, texture coarse and uneven. The wood is ring-porous, with 2 to 4 rows of large solitary earlywood pores, many small latewood pores, common tyloses and distinct rings; rays are not visible without a lens (Wood Database).

Ring-shaped cracks are visible on the log end where shake is present (LWF). I found no published key for separating chestnut from oak in sawn wood.

Properties that matter

Sweet chestnut is a medium-weight, relatively soft and weak hardwood. The LWF paper gives a mean density of 0.59 g/cm3 (0.57 to 0.66) against 0.71 for oak and beech, a stiffness of 9,000 N/mm2 against 13,000 for oak, and a Brinell hardness of 15 to 23 N/mm2 against 23 to 42 for oak.

Property	Value	Source	Region
Density, mean	0.59 g/cm3 (0.57-0.66)	LWF Wissen 81	Germany/Europe; moisture and origin not stated
Average dried weight	590 kg/m3 at 12%	Wood Database	specimen origin not stated
Apparent density	590 kg/m3 at 12%	Gabarro encyclopedia	Spain (trade datasheet)
Density, dried	about 560 kg/m3	Vincent Timber datasheet	UK (supplier; undated)
Bending strength (MOR)	80 N/mm2 (LWF); 71.4 MPa (Wood Database)	as listed	Europe; origin unstated
Modulus of elasticity	9,000 N/mm2 (LWF); 8.61 GPa (Wood Database)	as listed	Europe; origin unstated
Compression parallel	49 N/mm2 (LWF); 43.8 MPa (Wood Database)	as listed	Europe; origin unstated
Hardness	Brinell 15-23 N/mm2 (LWF); Janka 3,010 N at 12% (Wood Database)	as listed	Europe; origin unstated
Shrinkage R / T / vol.	4.3 / 6.4 / 11.3-11.6% (LWF); 4.2 / 6.9 / 12.6% green to oven-dry (Wood Database)	as listed	Europe; origin unstated
Fibre saturation point	below 22%	LWF Wissen 81	Europe
Heartwood pH	below 4	LWF Wissen 81	Europe
Durability class	2 (EN 350, 1995 edition)	LWF Wissen 81	Europe

The LWF paper also gives tensile strength parallel to grain of 135 N/mm2, impact bending work of 55 to 59 kJ/m2 along the grain, and a tangential-to-radial ratio of about 1.7 (Wood Database: 1.6). The LWF source does not state moisture conditions for its strength values. The LWF paper states that chestnut has no building timber grade and no EN 338 strength class, which matters for structural sales.

In 42-year-old Calabrian standards, shaken trees had lower shrinkage and healthy trees higher density (Springer tomography paper), from one study.

Sawing and log handling

Ring shake decides yield. The LWF paper describes it as a common feature in middle-aged and mature trees that devalues large parts of the stem, especially the valuable butt section. It forms along the earlywood vessels when stress is released, and abrupt growth fluctuations, especially in youth, are a factor.

The numbers on frequency come from several places and regions, and they are not directly comparable:

• Rhineland-Palatinate, Germany: about 30% of trees above 45 cm diameter at breast height are affected (Husmann, Saborowski and Hapla, Forstarchiv 2013, from 55 trees in six stands).
• Piedmont, Italy: in one coppice, 37.9% of shoots and 96.2% of standards were affected, and ring shake was found in 51% of stools (Macchioni and Pividori, 1996).
• Calabria, Italy: 61 of 100 standing 42-year-old standards had ring shake, detected by sonic tomography with eight sensors (Springer paper).
• In the Pfälzerwald, ring shake typically starts around age 30 or 25 cm diameter, and only 30 to 50% of stronger timber could be sawn into sawlogs, the rest going to firewood (Wambsganß and co-authors).

Causes: the Husmann paper links ring shake to growth fluctuation, which is strongest in the first 25 to 30 years and is increased by variable weather and stronger thinning, and recommends restricting growth in young stands during the first 20 to 25 years. Macchioni and Pividori mention trauma from felling, fire and animals and say the predisposing genetic, soil and nutritional factors are unknown. Mutabaruka and co-authors (iForest 2015) also call the causes unknown. The English Nature report says shake losses can be reduced with rotations of no more than 70 years.

A sawing trial in the Pfälzerwald (10 trees about 100 years old, 20 log sections, 7.704 m3) estimated 20% loss from sawing and 19.5% from ring shake or cracks, giving 60.5% average sawn yield. Of 139 boards of 40 mm, 39 were defect-free, 89 had only end cracks and 11 were rejected, mostly heart boards with pith. This is a single German trial, and the page does not say how the trees were selected. I would not extrapolate it to other regions.

Practical measures from the sources: inspect log ends for ring cracks, sort logs and sawn boards consistently (Wambsganß and co-authors), and remove the thin sapwood from round stock.

Drying

Sweet chestnut dries slowly and is prone to cracking. The LWF paper and waldwissen.net say moisture should be removed slowly to avoid cracking and warping, and that excessively high drying gradients can cause cell collapse; no drying times are given. Gabarro warns of collapse, case hardening and internal and external cracks; the Vincent datasheet says drying is difficult, with moisture retained in patches, a tendency to collapse and honeycomb, and poor response to reconditioning.

Unsal and Kantay (Turkish workshop paper, Acta Horticulturae 815, 2009) wrote that Anatolian C. sativa is denser than American C. dentata and so needs milder conditions: ambient temperature should not exceed 43 degrees C until fibre saturation to guard against collapse and internal checks, whereas C. dentata tolerates up to 54 degrees C. The full schedule tables were not available, and I offer no schedule.

For round wood, Bérard, Laurent and Dumonceaud (EJWWP 2005, France) report that small-diameter coppice chestnut is very sensitive to cracking while drying; their hot oil bath (130 degrees C for one hour, then drying to 12%) was designed to reduce surface checking.

Machining, gluing and finishing

Dried chestnut is easy to work by hand and machine (Wood Database), sanding, polishing, turning and drilling are possible (LWF), and it finishes and stains well. It splits easily, so pre-drill for nails and screws (Wood Database, Gabarro); the coarse texture makes turning mediocre. The LWF paper reports that OSB strands from chestnut produced many fines because the earlywood rings act as predetermined breaking points.

Gluing is reported as satisfactory to good for face and finger joints even with high extractives (LWF), so large elements can be made from small stock. Gabarro warns of reaction with alkaline glues because the wood is acidic, and the Vincent datasheet says it takes glue well. Sample before production.

The tannins and acidity (heartwood pH below 4) affect finishing and fixings: damp wood in contact with iron-containing metals gives corrosion stains (LWF), the Vincent datasheet describes blue-black discolouration and recommends stainless fixings and protecting decorative surfaces below because tannins can exude, and it advises consulting coating manufacturers. The Wood Database lists chestnut as a possible skin sensitiser.

Durability and treatment

The LWF paper rates the heartwood class 2 under EN 350 (1995 edition), noting that a 2003 study reported high variability, while a 2014 study did not confirm it and found good resistance in historic timber. The Wood Database calls it durable to very durable but susceptible to insect attack. The European Atlas ascribes the durability to a high natural tannin content, and Wambsganß and co-authors state tannin up to 13% of dry weight and tyloses in the vessels.

A 2026 open study (Kiliç and co-authors, Drewno) buried 60 chestnut samples for three years under EN 252 at four sites in Türkiye. Mass loss ranged from 4.71% at Elazığ to 35.96% at Çanakkale, with visual rating from 1.0 to 3.6, and bending strength after exposure from 71.73 to 38.28 N/mm2. The page does not explain the site differences.

Piles and palisades are used in ground and water contact without chemical treatment once the thin sapwood is milled off, and untreated avalanche tripods are expected to last at least 30 years (waldwissen.net, LWF; these are service expectations, not measured results). Gabarro says sapwood can be impregnated but heartwood cannot. I found no standards class for use classes under EN 335 and state none.

Typical uses and market notes

Higher grades (logs above diameter class D4 in the German system) go to decorative knife-cut veneers, windows and doors, cask staves and interior and exterior construction timber (LWF, waldwissen.net). Lower grades go to garden and landscape work, decking, garden furniture and wet-area flooring. Poles and palisades come from young stems (classes D0 and D1), which do not yet tend to develop ring shake. House construction in the Mediterranean and ceiling construction in central Italy are also reported, and customers for large high-quality logs are mainly in southern Europe.

Fence products are the mainstay of British coppice. English Nature reports that coppice aged 11 to 15 years yields high-grade paling poles, 25 to 40 years post-and-rail crops, and 45+ years planking. Cleft pales were mostly 0.91 to 1.06 m and posts 1.37 to 1.68 m, and fencing partly replaced falling hop-pole demand. I do not reproduce the report's dated stumpage prices.

Price data from the Pfälzerwald (2004 to 2011, about 12 years old): sawlogs were 13% of sales at an average of 99 euros per cubic metre, versus 41% slope-protection round timber, 27% firewood and 19% particleboard wood; stems over 30 cm free of ring shake reached roughly 150 to 300 euros per cubic metre. A survey of 20 carpentry firms found constant but low demand, with high prices, poor sawn quality and ring shake as barriers (Wambsganß and co-authors). I found no current figures.

What is uncertain or varies

• The Cirad Tropix datasheet was not opened. The Wood Database page does not state specimen origin, ring shake or any price.
• No sourced drying times or full schedules were found; temperature limits come from an abstract of a Turkish paper on Anatolian chestnut.
• The EN 350 class 2 is from a 1995 edition cited in the LWF paper; I did not read a current class. Reported variability is high between studies and sites.
• Ring shake frequency varies from about 30% to over 90% depending on stand, age and tree type; causes are partly unknown.
• Sawn-yield and price figures come from one German forest office and one English report and are dated.
• Two trade datasheets (Gabarro, Vincent) disagree slightly on density and do not say how it was measured.

Key points

• Density is about 0.59 g/cm3, with low hardness and stiffness compared with oak; the LWF paper reports no EN 338 class.
• Ring shake, often starting around age 30 or 25 cm diameter, is the main yield loss; only 30 to 50% of stronger timber was sawable in one German district.
• Heartwood durability is class 2 under EN 350 (1995) and varies by site and sample; sapwood is thin and must be removed.
• Dry slowly and keep it away from iron; collapse and honeycomb are risks and a low temperature limit is advised.
• Coppice products such as paling, posts and poles dominate supply, with larger high-quality logs scarce.