PosterPDF Available

Establishment of the bamboo panel industry sector in East Africa: Development of a scrimber panel based on Ethiopian highland bamboo



Abstract Urban areas in Sub-Saharan Africa are the fastest growing economies worldwide. The increasing demand for sustainable construction materials and efficient value-chains is closely interlocked to social and economic benefits of the rural population. Ethiopia as the politically leading nation in Africa has a very high potential due to its vast raw material stock (67 % of Africa’s bamboo resources) and the relatively stable political setting. First approaches on the utilization of native bamboo were explored in the late 1990’s. The current wood-processing branch in Ethiopia is still a net importer and widely processes eucalyptus, which is a seriously underestimated environmental problem. However, lacking process knowledge and the withholding involvement of industrial players had limited the bamboo sector in East Africa to low-value products in the past. To overcome these constraints, the strategic alliance “Development of the industrial bamboo sector in Ethiopia” was initiated by Gesellschaft für Internationale Zusammenarbeit (GIZ). African Bamboo PLC was created as a spin-off out of one of those eucalypt-based furniture companies. Inspired by Chinese panel producers and facilitated by institutional support from Germany (develoPPP, GIZ), Netherlands (Ministry of Foreign Affairs, PSI Funding) and Ethiopia (Development Bank of Ethiopia), a pilot plant was installed in Addis Ababa in 2013. The network project was led by Fraunhofer WKI and combined expertise of glue suppliers (Dynea AS, Solenis, BASF), wood-based panel industry (Dieffenbacher GmbH, Raute), woodworking machinery supplier (Michael Weinig AG, BigonDry) as well as research institutions. Adapt state-of-art industrial process to African highland bamboo The industrial bamboo sector in Ethiopia de-facto does not exist. However, WKI showed that highland bamboo is suitable for MDF, PB, OSB and compressed bamboo panels (CBP). CBP are highly densified board materials (>1.2 g/cm³) which are primarily used in high wear areas as container flooring or outdoor decking. Therefore, bamboo is split and mechanically crushed to strand mats by profiled feed rollers. The strand mats are subsequently kiln-dried and optionally heat treated. Commonly, the mats are then compacted either in a cold press process within a steel mould (block type) or in a panel batch process. Inspired by the Australian scrimber wood process, Chinese producers used Phyllostachys spp. for this kind of bamboo panel in the late 1990’s. The private-public partnership project improved and adapted the scrimber process to the Ethiopian highland bamboo species Yushania alpina. The different raw material brings along a complex supply system, new challenges in pre-processing, thermal treatment, glue application and hot pressing. The R&D part of the project is structured in two phases. In phase 1, we first investigated the raw material properties. Subsequently we produced lab-scale panels (500x500 mm) with three adhesive types (PF-, soya- and acrylate-based), eight thermal modification parameters (160 °C, 180 °C 200 °C, 220 °C; 3 h, 5 h), two glue application methods (spray, powder) and a variation with an hydrophobic agent (Sasol ProA18). Phase 2 was then started in 2016, aiming to cover an extended range of testing procedures. Those shall on the one hand address the product requirements for outdoor utilization and on the other hand deepen the understanding on how process parameters would influence the commercial size panel. Overcome raw material challenges As known from other lignocellulosic non-wood raw materials (bagasse, straw, agricultural residues), high contents of silica, waxes, non-structural carbohydrates and water cause challenges for the panel industry. The silica content in the analysed samples was lower than commonly expected. The amount found in material from different plots ranged from 0.1 to 1.6 [%]. The silica concentrates in extracellular spaces, hence a simple raw material washing would eliminated most of the silica. Usually seen as a technological constraint of bamboo, the epidermal wax layer in mature Y. alpina was negligible. Bamboo further shows high fluctuations in moisture content (MC) and sugars with height, season and growing site. The culms as well as the rhizome store fluctuating contents of water, starch and metabolic sugars. On the one hand, they enable the emerging shoots for fast height growth; on the other hand they attract fungi, stains and insects. This serious risk to quality of intermediate and final products was analysed. In mature (4yr) highland bamboo culms, starch fluctuates from 4.68 in the dry season to 3.05 m% in the rainy season. MC ranges from 85.7 % and 129.43 %. Asian producers’ thumb-rule says that a cut culm shall be processed within three days to not suffer biological attack. Contrary to Chinese recommendations, tests showed that simple air-drying is sufficient to make the intermediate products (strips, strand mats) resistant against biological degradation. Save process time, energy and resources Kiln-drying and thermal treatment of bamboo were commonly done in a two-step batch process. The strand mats are either treated with a saturated steam atmosphere (SST) or hot-dry air (HDAT). Both procedures need a time and energy-intense moisture correction afterwards. We integrated both steps into one batch chamber (BigonDry). First, the air-dry strand mats (12 – 15 % MC) are kiln-dried at low temperatures (80 °C). Then, a high-temperature phase modifies the mats during 3 to 5 h). The combination of both steps delivers “press-ready” strand mats (4 % MC) and therefore saves handling time and process heat. In common productions, the strand mats are bundled and dipped into resin tanks for some minutes. The bundles are then taken out and the excess glue simply drips off the mats. The PF amount PF in the final panel is influenced only by the leaking time. The results are high PF contents (> 25 %), heterogeneous distribution and high process MC. The mats necessarily need to be dried before hotpressing. As a consequence, uncontrolled pre-curing of the PF occurs. As an alternate solution we applied four different PF resins (Dynea), Acrodur (BASF) and Soyad (Solenis) with two application systems. Powder application showed to be no technical viable method. However, the spray-application brought up promising results and is currently under development at Raute, Finland. Panels containing only 12 % PF achieved satisfactory results in IB, TS and bending. The process bottleneck is the hot press. The vapour pressure caused by the released process heat bursts the panel when opening the press. Hence, producers operate cycles up to 2 h to cool down the panel core temperature below 95 °C before press opening. Contrarily, the core temperature must be hold above 105 °C at least during 4 min due to curing of the phenolic resin. To find a suitable compromise, we pressed panels with varying “turn-off” temperature levels. We found a way to reduce maximum temperatures and shorten the cooling phase by successively opening the press in decile mm steps. Nevertheless, the high specific pressure of up to 15 N/mm² (lab-scale) challenges the press engineering. In an industrial scale 16-daylight press, a sophisticated cooling system will be needed to reach appropriate short cycle times. Preliminary testing phase Testing in phase 1 focused basic physical and mechanical properties in order to show that the process is technical feasible. Using characteristic benchmark values (beech LVL and OSB 3), we defined threshold values for thickness swelling after 24 h (TS < 6%), internal bond (IB >0,6 N/mm²), flexural strength (MOR > 90 N/mm²) and stiffness (MOE > 16 kN/mm²) as minimum requirement for the labsize CBP. Based on these results we identified the “best” variants for later investigation in phase 2. The MOR of untreated, spray-applied panels ranged from 126.1 to 195.5 N/mm², while MOE reached 19.23 to 26.49 kN/mm². The results of for heat-treated materials were in parts lower than the threshold limit. They will be cross-checked in phase 2. The internal bond of untreated panels easily fulfilled the requirements (0.8 to 1.0 N/mm²). For TS satisfactory results (2.3 to 3.1 %) were found when wax (concentration 2 %) was added to the mix. Product testing phase In phase 2 a comprehensive range of tests according on WPC decking standard EN 15534-1 and EN 15534-4 is conducted. Furthermore, biodegradation effects are analysed. Bamboo possesses protective strategies against biological agents only in a standing culm. Once the culm is split or crushed, fungi easily attack the material, since there are no extractives in the tissue. For finished outdoor products, biodegradation is a well-known risk. Bacteria and mould affect aesthetic aspects, while termites, white/brown rot fungi (Basidiomycetes) and soft rot fungi (Ascomycetes) are the technological threats to decking. Literature shows that phenolic compounds (PF), thermal modification and a high density decrease the overall attack by biological agents. Nevertheless, few is known on how each factor influences the final panel resistance to biodegradation. In the moment lab-scale tests are started in Hamburg to close this knowledge gap. Outlook The bamboo panel industry in Africa still is in its infancy. Products made of alternate lignocellulosic raw materials like bamboo offer not only market potentials but development perspectives for other Sub-Saharan economies. Besides the ecological advantages, a developing rural value-chain will benefit thousands of Ethiopian farmer families. The unique character of this project is a result of a complex structure of researchers, stakeholders, and industry experts. Hence, the actual innovation in this project is the combination of state-of-art technologies. The result is a highly efficient bamboobased panel with properties which promise to be interesting not only in outdoor decking applications.
1) PhD student, Center for Wood Sciences, University Hamburg
2) Project leader, Dr., Fraunhofer WKI Institute for Wood Research
3) Wood scientist, Fraunhofer WKI Institute for Wood Research
Establishment of the bamboo panel industry
sector in East Africa
Development of a scrimber panel based on Ethiopian highland bamboo
Goran Schmidt1), Dirk Berthold2), Mathias Belda3)
Set up a cooperation network
Ethiopia is one of the rising economies in Sub-Saharan
Africa with a large gap between available raw
materials and strong industrial growth.
Ethiopia owns ca. 1 million ha bamboo lands (2/3 Africa)
Lack of process knowledge and industrial involvement
limited the bamboo sector to low-value products
The strategic alliance framework “Development of the
industrial bamboo sector in Ethiopia” funded by GIZ
Incubator company, African Bamboo established a pilot
plant in Addis Abeba in 2013
Combined expertise of glue suppliers (Dynea, Solenis,
BASF), machinery developers (Dieffenbacher, Raute,
BigonDry) and research (WKI)
Bamboo scrimber panel in brief
Common bamboo scrimber are highly densified board
materials with densities >1.2 g/cm³ for high wear areas.
Chinese producers refer to it as woven strand board
Decentral supply chain involving about 2.000 bamboo
farmers and microenterprises
Culms are split and roller-crushed to strand mats
Strand mats are kiln-dried and “carbonized” in hot air
or saturated steam chambers
Glue applied by dipping bundles into tanks, controlling
glue concentration (>20 %) by soaking and drip-off time
Hot-pressing with >15 N/mm² and cooling down
Goals to achieve
We aimed to develop a lighter and more resource
efficient scrimber bamboo panel based on the African
highland bamboo Yushania alpina.
Analyse raw material properties and fluctuations
Adopt Asian WSB process to existing machinery
Find appropriate glue and application method
Identify process steps for enhancing product quality
Conduct testing programme for market entrance
The most promising variant was selected for further
investigation in commercial scale.
Spraying application of 10 % PF-System 2 (50/50)
Mix of 2 % Pro A18 Sasol wax
Heat treatment was not necessary to fulfill thresholds
Resistance against biodeterioration remains unclear
Market entrance facilitation
A test programme according to European quality
requirements shall help market entrance.
Commercial scale panels were produced and undergo
comprehensive standard testing
Absence of full standard for outdoor products made
from alternative lignocellulosic raw materials
First oriented to WPC standard EN 15534
Recently drafted EN 17009 will regulate requirements
Ongoing investigations
Does raw material assorting enhance product quality?
The fibre content increases with culm height
Density profiles and physical properties will be
analyzed for panels made from three different raw
material groups (top, middle, bottom)
Does thermal treatment effect biological performance?
The influence of thermal treatment on the resistance
against different biodegrading agents (soft rot, brown
rot, white rot, termites) is currently investigated
Overcome raw material challenge
Bamboo contains high amounts of silica, waxes, water
and sugars.
Silica 0,1 to 1,6 [m%] was found in extracellular spaces
Only negligible wax was found in the epidermal layer
Significant moisture fluctuations (86 < MC < 130 [%])
Metabolic sugars fluctuate less in mature culms
Starch changes amounts to 4,68 in dry and 3,05 m% in
the rainy season
Investigate production parameters
A screening test of a production parameters was
conducted for lab-scale panels.
Untreated vs. treated (160, 180, 200, 220 [°C]; 3, 5 [h])
Three different types of glue (four PF-based, one
acrylate and one soy-based)
Different glue applications (spray, powder, curtain)
With or without hydrophobic agents
The results were compared to LVL and OSB3 references
threshold values.
Thickness swelling threshold TS < 6 %
Internal bond threshold IB > 0,5 N/mm²
Flexural strength threshold MOR > 90 N/mm²
Flexural strength threshold MOE > 16 kN/mm²
1) Loading a combined kiln-drying and thermal treatment chamber with a stack of strand
mats 2) Spray application of PF resin with industrial scale equipment 3) Mat formation
showing narrow strands in the core and broad strands in the face layers
1) Average content of sugars in 4 yr. old culms of the dry and rainy season based on culm
dry mass, 2) Manual culm harvest in the dry season in Ethiopia‘s South
2 3 1
1 2
Promising test results for spray application and identification of „best variants“
1. Plantation 2. Culm harvest 3. Post-harvest 4. Mechanical
8. Condition, Sand,
Finish 7. Hot press 6. Glue application 5. Thermal
Simplified process flow scheme showing the main technological challenges
1) Fungal mycel appears in voids in between compressed fibres 2) Phenolic glue did not
hinder basidiomycetes 3) Intact parenchyma attracts fungal attack (Coniophora puteana)
2 3
1) Planks cut from final product produced at WKI 2) Top view of untreated panel
1 2
04812 16 20
Density kg/m³
Panel thickness [mm]
Density Average sample density
Density profile of a scrimber panel using untreated and unsorted bamboo strands
Glue type
TS (%)
0 % Wax
TS (%)
2 % Wax
IB (N/mm²) MOR
160°C / 3h PF-System 1 9,8 4,1 0,6 109,3 24400,2
160°C / 3h PF-System 2 4,3 2,8 0,3 87,2 18361,0
160°C / 5h PF-System 1 4,6 4,9 1,0 171,7 26048,8
160°C / 5h PF-System 2 2,1 1,8 1,3 136,0 21906,1
180°C / 3h PF-System 1 3,2 3,1 1,0 146,5 23595,3
180°C / 3h PF-System 2 6,4 2,6 0,2 119,7 26571,0
180°C / 5h PF-System 1 2,8 3,4 0,6 104,7 23098,7
180°C / 5h PF-System 2 2,0 1,3 0,7 130,6 24322,6
200°C / 3h PF-System 1 6,0 2,4 0,4 87,3 26245,7
200°C / 3h PF-System 2 3,6 3,0 0,8 115,0 26464,6
200°C / 5h PF-System 1 2,9 1,8 0,6 97,3 22052,2
200°C / 5h PF-System 2 4,3 1,9 1,0 131,0 25580,2
220°C / 3h PF-System 1 3,2 0,5 72,0 21807,8
220°C / 3h PF-System 2 3,2 0,4 73,4 21453,6
Untreated PF-System 1 1,8 2,3 0,8 126,1 19282,0
Untreated PF-System 2 5,8 3,1 1,0 195,5 26490,4
Full-text available
A novel engineered scrimber was manufactured from heat treated African highland bamboo Yushania alpina (K. Schum.) W.C. Lin on an industrial scale. Scrimber are a group of engineered wood products that consist of long, slender particles resulting from a non-cutting defibration technique. Those are heat treated, impregnated with phenolic resin, highly densified and the product is believed to be resistant to biodegradation by wooddestroying fungi. The highland bamboo scrimber was tested for resistance against basidiomycete monocultures and soft rot in a soil bed test. The soft rot soil bed test caused nearly 20 % mass loss and 61 % stiffness loss for the scrimber made from untreated bamboo. Heat treatment reduced the ML to 5% for the 200 °C variant, whereas stiffness loss was only slightly reduced for heat treated variants. All treatment temperatures led to 50 % fungal stiffness loss. The notably high abiotic mass loss in sterile samples reduced with treatment temperature. The heat treatment did not affect the durability class resulting from the test against basidiomycete monocultures. Although the variability in heat treated samples was slightly lower, all variants achieved durability class 1. The soil bed test against soft rot resulted in durability classes 1, 2, 3 and 4. When based upon stiffness as parameter, the durability class was 4, whereas mass loss led to DC 1 for treatment at 200 °C, DC 2 for 180 and 160 °C and DC 3 for the untreated control, respectively. The development of substrate moisture content over time indicated that fungal growth can possibly be delayed via heat treatment, but not stopped. The minimum target treatment temperature to achieve a notable improvement in mass loss was 180 °C.
ResearchGate has not been able to resolve any references for this publication.