MARKET ACCESS

Life Cycle Inventory and Environmental Product Declarations for Australian Wood Products: Full Report

Project number: PNA384-1516 February 2019

Level 11, 10-16 Queen Street Melbourne VIC 3000, Australia T +61 (0)3 9927 3200 E [email protected] W www.fwpa.com.au Life Cycle Inventory and Environmental Product Declarations for Australian Wood Products: Full Report

Prepared for

Forest & Wood Products Australia

by

J. Vickers, B. Fisher, T. Betten and S. Mitchell Publication: Life Cycle Inventory and Environmental Product Declarations for Australian Wood Products: Full Report Project No: PNA384-1516

IMPORTANT NOTICE

This work is supported by funding provided to FWPA by the Department of Agriculture and Water Resources (DAWR).

© 2019 Forest & Wood Products Australia Limited. All rights reserved.

Whilst all care has been taken to ensure the accuracy of the information contained in this publication, Forest and Wood Products Australia Limited and all persons associated with them (FWPA) as well as any other contributors make no representations or give any warranty regarding the use, suitability, validity, accuracy, completeness, currency or reliability of the information, including any opinion or advice, contained in this publication. To the maximum extent permitted by law, FWPA disclaims all warranties of any kind, whether express or implied, including but not limited to any warranty that the information is up-to-date, complete, true, legally compliant, accurate, non-misleading or suitable.

To the maximum extent permitted by law, FWPA excludes all liability in contract, tort (including negligence), or otherwise for any injury, loss or damage whatsoever (whether direct, indirect, special or consequential) arising out of or in connection with use or reliance on this publication (and any information, opinions or advice therein) and whether caused by any errors, defects, omissions or misrepresentations in this publication. Individual requirements may vary from those discussed in this publication and you are advised to check with State authorities to ensure building compliance as well as make your own professional assessment of the relevant applicable laws and Standards.

The work is copyright and protected under the terms of the Copyright Act 1968 (Cwth). All material may be reproduced in whole or in part, provided that it is not sold or used for commercial benefit and its source (Forest & Wood Products Australia Limited) is acknowledged and the above disclaimer is included. Reproduction or copying for other purposes, which is strictly reserved only for the owner or licensee of copyright under the Copyright Act, is prohibited without the prior written consent of FWPA.

ISBN: 978-1-925213-90-4

Researchers:

J. Vickers, B. Fisher, T. Betten thinkstep Pty Ltd 25 Jubilee Street, South Perth WA 6151, Australia

S. Mitchell Stephen Mitchell Associates PO Box 309, Earlwood NSW 2206, Australia

Final report received by FWPA in March 2018 Forest & Wood Products Australia Limited Level 11, 10-16 Queen St, Melbourne, Victoria, 3000 T +61 3 9927 3200 F +61 3 9927 3288 E [email protected] W www.fwpa.com.au

Version history:

No. Description

2.0 Completely revised version, incorporating:

• New life cycle inventory data collected for the 2015/16 financial year • More detailed assessment of variation within data • Sixth EPD added for glulam

1.0 Original life cycle assessment report supporting FWPA’s environmental product declarations for sawn softwood, sawn hardwood, particleboard, MDF and plywood: Vickers, J., Loske, F., Slocinski, C. and Mitchell, S. (2015). EPDs for Australian Wood Products. Final report of 22 October 2015 following verification of all EPDs. Prepared for Forest and Wood Products Australia. This life cycle assessment was based on 2005/06 life cycle inventory data collected by the Commonwealth Scientific and Industrial Research Organisation: Tucker, S.N.; Tharumarajah, A.; May, B.; England, J.; Paul, K.; Hall, M.; Mitchell, P.; Rouwette, R.; Seo, S. and Syme, M. (2009). Life Cycle Inventory of Australian Forestry and Wood Products. Final report prepared for Forest and Wood Products Australia. CSIRO.

This report has been prepared by thinkstep with all reasonable skill and diligence within the terms and conditions of the contract between thinkstep and the client. thinkstep is not accountable to the client, or any others, with respect to any matters outside the scope agreed upon for this project.

Regardless of report confidentiality, thinkstep does not accept responsibility of whatsoever nature to any third parties to whom this report, or any part thereof, is made known. Any such party relies on the report at its own risk. Interpretations, analyses, or statements of any kind made by a third party and based on this report are beyond thinkstep’s responsibility.

If you have any suggestions, complaints, or any other feedback, please contact us at [email protected].

LCI and EPDs for Australian Wood Products 3 of 159

Table of Contents

List of Figures ...... 9 List of Tables ...... 10 List of Acronyms ...... 14 Glossary ...... 16 Executive Summary ...... 18 1. Background ...... 20 2. Goal of the Study ...... 21 3. Scope of the Study ...... 22 3.1. Product description and application area ...... 22

3.2. Declared unit ...... 23

3.3. System boundaries ...... 24

3.3.1. General system boundaries ...... 24 3.3.2. Time boundaries ...... 26 3.3.3. Geographical boundaries ...... 26 3.3.4. Boundaries for manufacturing of equipment and for employees ...... 26 3.3.5. Boundaries to other product life cycles...... 27 3.4. Reference service life ...... 27

3.5. Data quality requirements ...... 27

3.6. Cut-off criteria ...... 28

3.7. Selection of LCIA methodology and impact categories...... 28

3.7.1. Life cycle impact assessment (LCIA) indicators ...... 28 3.7.2. Life cycle inventory indicators ...... 29 3.8. Interpretation to be used ...... 30

3.9. Allocation and recycling ...... 31

3.9.1. Allocation of upstream data ...... 31 3.9.2. Allocation in foreground data ...... 31 3.10. Software and database ...... 31

3.11. Critical review and EPD verification ...... 32

4. Life Cycle Inventory Analysis ...... 33 4.1. Methods ...... 33

4.1.1. Primary data collection ...... 33

LCI and EPDs for Australian Wood Products 4 of 159

4.1.2. Calculation of net calorific value for wood ...... 34 4.1.3. Calculation of net calorific value for resins ...... 34 4.1.4. Calculation of carbon sequestration, density and moisture content ...... 35 4.2. Background data ...... 36

4.2.1. Fuels and energy ...... 36 4.2.2. Raw materials and processes ...... 38 4.2.3. Preservative treatments ...... 39 4.2.4. Transportation ...... 40 4.2.5. Elemental composition and calorific value of materials ...... 40 4.3. Softwood forestry ...... 41

4.3.1. Overview ...... 41 4.3.2. Production process ...... 42 4.3.3. Establishment ...... 43 4.3.4. Forest management...... 45 4.3.5. Harvesting ...... 46 4.3.6. Allocation ...... 47 4.3.7. Assumptions ...... 48 4.4. Hardwood forestry ...... 53

4.4.1. Overview ...... 53 4.4.2. Production process ...... 54 4.4.3. Stand establishment ...... 55 4.4.4. Forest management...... 56 4.4.5. Harvesting ...... 57 4.4.6. Allocation ...... 57 4.4.7. Assumptions ...... 58 4.5. Sawn softwood ...... 59

4.5.1. Overview ...... 59 4.5.2. Production process ...... 62 4.5.3. Storage, de-barking and milling ...... 62 4.5.4. Kiln-drying ...... 64 4.5.5. Planing ...... 65 4.5.6. Treatment ...... 66 4.5.7. Packaging ...... 67 4.5.8. Allocation ...... 67 4.5.9. Assumptions ...... 68 4.6. Sawn hardwood ...... 70

4.6.1. Overview ...... 70

LCI and EPDs for Australian Wood Products 5 of 159

4.6.2. Production process ...... 72 4.6.3. Storage, de-barking and milling ...... 73 4.6.4. Kiln-drying ...... 74 4.6.5. Planing ...... 76 4.6.6. Treatment ...... 77 4.6.7. Packaging ...... 77 4.6.8. Allocation ...... 77 4.6.9. Assumptions ...... 78 4.7. Particleboard ...... 80

4.7.1. Overview ...... 80 4.7.2. Production process ...... 82 4.7.3. Manufacturing ...... 84 4.7.4. Allocation ...... 86 4.7.5. Assumptions ...... 86 4.8. Medium Density Fibreboard (MDF) ...... 87

4.8.1. Overview ...... 87 4.8.2. Production process ...... 90 4.8.3. Manufacturing ...... 91 4.8.4. Packaging ...... 92 4.8.5. Allocation ...... 93 4.8.6. Assumptions ...... 93 4.9. Plywood ...... 94

4.9.1. Overview ...... 94 4.9.2. Production process ...... 97 4.9.3. Veneer production ...... 98 4.9.4. Plywood manufacture ...... 101 4.9.5. Packaging ...... 103 4.9.6. Allocation ...... 103 4.9.7. Assumptions ...... 104 4.10. Glued-laminated timber (Glulam) ...... 105

4.10.1. Overview...... 105 4.10.2. Production process ...... 107 4.10.3. Glulam production ...... 109 4.10.4. Packaging ...... 111 4.10.5. Allocation ...... 112 4.10.6. Assumptions ...... 112 4.11. Wood preservative treatment ...... 113

LCI and EPDs for Australian Wood Products 6 of 159

4.12. End-of-life ...... 116

4.12.1. Landfill ...... 118 4.12.2. Energy recovery ...... 121 4.12.3. Reuse ...... 122 4.12.4. Recycling ...... 122 5. LCIA Results ...... 123 5.1. Sawn softwood ...... 123

5.1.1. EN 15804 results (A1-A3) ...... 123 5.1.2. Detailed results ...... 125 5.2. Sawn hardwood ...... 126

5.2.1. EN 15804 results (A1-A3) ...... 127 5.2.2. Detailed results ...... 128 5.3. Particleboard ...... 130

5.3.1. EN 15804 results (A1-A3) ...... 130 5.3.2. Detailed results ...... 132 5.4. MDF ...... 134

5.4.1. EN 15804 results (A1-A3) ...... 134 5.4.2. Detailed results ...... 135 5.5. Plywood ...... 137

5.5.1. EN 15804 results (A1-A3) ...... 137 5.5.2. Detailed results ...... 139 5.6. Glulam ...... 141

5.6.1. EN 15804 results (A1-A3) ...... 141 5.6.2. Detailed results ...... 143 5.7. Wood preservative treatments ...... 146

5.8. Release of hazardous substances during the use stage ...... 146

6. Interpretation ...... 148 6.1. General ...... 148

6.2. Identification of relevant findings ...... 149

6.3. Assumptions and limitations ...... 149

6.3.1. Key assumptions ...... 149 6.3.2. Limitations of the study ...... 150 6.4. Data quality assessment ...... 151

6.4.1. Precision and completeness ...... 151 6.4.2. Representativeness ...... 151

LCI and EPDs for Australian Wood Products 7 of 159

6.4.3. Completeness ...... 152 6.4.4. Consistency ...... 152 6.4.5. Reliability ...... 152 6.5. Conclusions and recommendations ...... 152

References ...... 154

Annexes

Annex A: Product specific LCI ...... Annex page 1 Annex B: Product specific LCIA ...... Annex page 64 Annex C: Wood preservative treatments LCIA ...... Annex page 169 Annex D: Hotspot analysis ...... Annex page 174 Annex E: Electricity grid mixes (confidential) ...... Excluded Annex F: Resin properties ...... Annex page 176 Annex G: EPDs ...... Annex page 181 Annex H: Verification statement ...... Annex page 351

LCI and EPDs for Australian Wood Products 8 of 159

List of Figures

Figure 4-1 Shrinkage versus moisture content (Timber Queensland, 2014) ...... 36 Figure 4-2 Process flow diagram of softwood forestry ...... 42 Figure 4-3 System boundary of the plantation softwood forest production module (red line) (CSIRO, 2009a) ...... 49 Figure 4-4: Process flow diagram of native hardwood forestry ...... 54 Figure 4-5 Process flow diagram of sawn softwood production ...... 62 Figure 4-6 Process flow diagram of sawn hardwood production ...... 72 Figure 4-7 Process flow diagram of particleboard production ...... 83 Figure 4-8 Process flow diagram of MDF production ...... 90 Figure 4-9 Process flow diagram of plywood production ...... 98 Figure 4-10 Process flow diagram of glulam production ...... 108 Figure 4-11 Key assumptions for biogenic carbon in landfill (reproduced from Carre (2011)) ...... 121 Figure 5-1 Softwood, KD, dressed A1-A3 impacts by production stage ...... 125 Figure 5-2 Softwood, KD, dressed A1-A3 comparison to other softwood studies ...... 126 Figure 5-3 Hardwood, KD, dressed A1-A3 impacts by production stage...... 129 Figure 5-4 Hardwood, KD, dressed A1-A3 comparison to other hardwood studies ...... 130 Figure 5-5 PB, MR, E1, Melamine 18mm A1-A3 impacts by production stage ...... 132 Figure 5-6 PB, MR, E1, Melamine 18mm A1-A3 comparison to other particleboard studies ...... 133 Figure 5-7 MDF, MR, E1, Melamine, 18mm A1-A3 impacts by production stage ...... 136 Figure 5-8 MDF, MR, E1, Melamine, 18mm A1-A3 comparison to other MDF studies ...... 137 Figure 5-9 17mm, formply, A-Bond A1-A3 impacts by production stage ...... 139 Figure 5-10 17mm, formply, A-Bond A1-A3 comparison to other plywood studies ...... 141 Figure 5-11 Softwood glulam A1-A3 impacts by production stage ...... 143 Figure 5-12 Hardwood glulam A1-A3 impacts by production stage ...... 145 Figure 5-13 Softwood glulam and hardwood glulam comparison to CORRIM ...... 146 Figure 6-1 Urea formaldehyde repeat unit (Groover, 2010) ...... 170 Figure 6-2 Melamine formaldehyde repeat unit ...... 170 Figure 6-3 Melamine urea formaldehyde repeat unit ...... 171 Figure 6-4 Phenol formaldehyde repeat unit ...... 172 Figure 6-5 Resorcinol formaldehyde repeat unit ...... 172 Figure 6-6 Phenol resorcinol formaldehyde ...... 173

LCI and EPDs for Australian Wood Products 9 of 159

List of Tables

Table 3-1 Technical specifications applying to the products in this background report ...... 22 Table 3-2 Declared unit for each EPD ...... 23 Table 3-3 Modules of the production life cycle included in the EPD as per EN15804 (X = declared module; MND = module not declared) ...... 25 Table 3-4: Life cycle impact assessment indicators ...... 28 Table 3-5: Life cycle inventory indicators on use of resources ...... 30 Table 3-6: Life cycle inventory indicators on waste categories ...... 30 Table 3-7: Life cycle inventory indicators on output flows ...... 30 Table 4-1 Elemental composition and net calorific value of resins ...... 35 Table 4-2 Key energy datasets used in inventory analysis ...... 37 Table 4-3 Unit process data for “Diesel consumption in forest engine” ...... 37 Table 4-4 Fuel properties ...... 38 Table 4-5 Key material datasets used in inventory analysis ...... 38 Table 4-6 Processes and flows used for modelling preservative treatments ...... 39 Table 4-7 Elemental composition of materials ...... 40 Table 4-8 Heating values of materials ...... 41 Table 4-9 Softwood forestry products by type ...... 42 Table 4-10 Softwood forestry production area and output – per m3 of wood harvested ...... 43 Table 4-11 Plantation establishment inventory – per ha established...... 44 Table 4-12 Forest residue burning – per t of bone dry residues ...... 45 Table 4-13 Plantation management – per ha of managed land ...... 45 Table 4-14 Thinning – per m3 of thinning output ...... 47 Table 4-15 Clearfelling – per m3 of clearfelled wood ...... 47 Table 4-16 Forest product allocation factors ...... 48 Table 4-17: Estimated annual evapo-transpiration and water use per unit wood produced for each case study region (weighted for wood production for water use and area for rainfall and ET) (CSIRO, 2009a) ...... 51 Table 4-18 Hardwood forestry products by type ...... 53 Table 4-19 Hardwood forestry production area and output – per m3 of wood harvested ...... 55 Table 4-20 Native hardwood forest establishment inventory – per ha established ...... 55 Table 4-21 Native hardwood forest management – per ha of managed land ...... 56 Table 4-22 Harvesting – per m3 of harvested wood ...... 57 Table 4-23: Forest product allocation factors ...... 58 Table 4-24 Contributors to sawn softwood LCI ...... 60 Table 4-25 Sawn softwood material properties ...... 61 Table 4-26 Residue densities and expansion factors ...... 61 Table 4-27 Softwood storage, de-barking and milling inventory – per m3 of sawn wood ...... 63 Table 4-28 Softwood kiln-drying inventory – per m3 of kiln-dried wood ...... 64

LCI and EPDs for Australian Wood Products 10 of 159

Table 4-29 Softwood planing inventory – per m3 of planed wood ...... 66 Table 4-30 Softwood packaging – per m3 of wood packaged ...... 67 Table 4-31 Sawing co-product allocation factors ...... 67 Table 4-32 Planing co-product allocation factors ...... 68 Table 4-33 Contributors to sawn hardwood LCI ...... 70 Table 4-34 Sawn hardwood material properties per m3 ...... 71 Table 4-35 Hardwood storage, de-barking and milling inventory – per m3 of sawn wood ...... 73 Table 4-36 Hardwood kiln-drying inventory – per m3 of kiln-dried wood ...... 74 Table 4-37 Hardwood planing inventory – per m3 of planed wood ...... 76 Table 4-38 Hardwood packaging – per m3 of wood packaged ...... 77 Table 4-39 Sawing co-product allocation factors ...... 78 Table 4-40 Planing co-product allocation factors ...... 78 Table 4-41 Contributors to particleboard LCI ...... 81 Table 4-42 Material composition of the particleboard products [%] ...... 81 Table 4-43 Elemental composition including water ...... 82 Table 4-44 Elemental composition excluding water ...... 82 Table 4-45 Net calorific values of softwood, paper and particleboard products ...... 82 Table 4-46: Contributors to MDF LCI ...... 87 Table 4-47 Moisture resistant (MR), melamine coated, 18 mm – material properties ...... 88 Table 4-48 Material composition of the MDF products [%] ...... 88 Table 4-49 Elemental composition including water ...... 89 Table 4-50 Elemental composition excluding water ...... 89 Table 4-51 Net calorific values of softwood, paper and MDF products ...... 90 Table 4-52 Moisture resistant (MR), melamine coated, 18 mm – inventory per m2 of board ...... 91 Table 4-53 MDF packaging inventory – per m2 of 18 mm MR, melamine coated board ...... 92 Table 4-54 Contributors to plywood LCI ...... 95 Table 4-55 17 mm formply, A-Bond – material properties ...... 95 Table 4-56 Material composition of the plywood products [%] ...... 96 Table 4-57 Elemental composition including water ...... 96 Table 4-58 Elemental composition excluding water ...... 97 Table 4-59 Net calorific values of softwood, paper and plywood products ...... 97 Table 4-60 Veneer production for 17 mm formply, A-Bond – inventory per m3 of veneer ...... 100 Table 4-61 Plywood manufacture for 17 mm formply, A-Bond – inventory per m2 of plywood ...... 102 Table 4-62 Plywood packaging inventory per m3 of plywood...... 103 Table 4-63 Veneer production co-product allocation factors ...... 103 Table 4-64 Plywood production co-product allocation factors ...... 104 Table 4-65: Contributors to Glulam LCI ...... 106 Table 4-66 Softwood glulam – material properties and distances ...... 106 Table 4-67 Hardwood glulam – material properties and distances ...... 107 Table 4-68 Softwood glulam – inventory per m3 of glulam ...... 109 Table 4-69 Hardwood glulam – inventory per m3 of glulam ...... 110 Table 4-70 Softwood glulam packaging inventory per m3 of glulam ...... 111

LCI and EPDs for Australian Wood Products 11 of 159

Table 4-71 Hardwood glulam packaging inventory per m3 of glulam ...... 111 Table 4-72 Glulam co-product allocation factors ...... 112 Table 4-73: Covered treatment types and sources ...... 113 Table 4-74: Energy inputs for treatment per m3 of timber ...... 114 Table 4-75: Bill of materials for the different treatment types ...... 114 Table 4-76: Concentration of treatments according to the different hazard classes ...... 115 Table 4-77: End-of-life scenarios per product group ...... 116 Table 4-78: End-of-life scenarios from EN 16485:2014 ...... 117

Table 4-79: Degradable organic carbon fractions (DOCf) of timber and wood products from bioreactor research (Wang, et al., 2011; Ximenes, et al., 2013) ...... 119 Table 5-1 Environmental impacts for production stage (A1-A3) for 1 m3 kiln-dried, dressed softwood ...... 124 Table 5-2 Resource use for production stage (A1-A3) for 1 m3 kiln-dried, dressed softwood ...... 124 Table 5-3 Waste categories and output flows for production stage (A1-A3) for 1 m3 kiln-dried, dressed softwood ...... 124 Table 5-4 Environmental impacts for production stage (A1-A3) for 1 m3 kiln-dried, dressed hardwood ...... 127 Table 5-5 Resource use for production stage (A1-A3) for 1 m3 kiln-dried, dressed hardwood ...... 127 Table 5-6 Waste categories and output flows for production stage (A1-A3) for 1 m3 kiln-dried, dressed hardwood ...... 128 Table 5-7 Environmental impacts for production stage (A1-A3) for 1 m2 MR, E1, Melamine 18mm particleboard ...... 131 Table 5-8 Resource use for production stage (A1-A3) for 1 m2 MR, E1, Melamine 18mm particleboard ...... 131 Table 5-9 Waste categories and output flows for production stage (A1-A3) for 1 m2 MR, E1, Melamine 18mm particleboard ...... 131 Table 5-10 Environmental impacts for production stage (A1-A3) for 1 m2 MR, E1, Melamine 18mm MDF ...... 134 Table 5-11 Resource use for production stage (A1-A3) for 1 m2 MR, E1, Melamine 18mm MDF . 134 Table 5-12 Waste categories and output flows for production stage (A1-A3) for 1 m2 MR, E1, Melamine 18mm MDF ...... 135 Table 5-13 Environmental impacts for production stage (A1-A3) for 1 m2 17mm, formply, A-Bond138 Table 5-14 Resource use for production stage (A1-A3) for 1 m2 17mm, formply, A-Bond ...... 138 Table 5-15 Waste categories and output flows for production stage (A1-A3) for 1 m2 17mm, formply, A-Bond ...... 138 Table 5-16 Environmental impacts for production stage (A1-A3) for 1 m3 softwood glulam and hardwood glulam ...... 142 Table 5-17 Resource use for production stage (A1-A3) for 1 m3 softwood glulam and hardwood glulam ...... 142 Table 5-18 Waste categories and output flows for production stage (A1-A3) for 1 m3 softwood glulam and hardwood glulam ...... 142 Table 5-19: Formaldehyde emission classes for Australian engineered wood products ...... 147 Table 6-1: Coverage of Australian production ...... 148 Table 6-2: Impacts of wood preservative treatments for hardwood, per m3 of treated wood ...... 162

LCI and EPDs for Australian Wood Products 12 of 159

Table 6-3: Impacts of wood preservative treatments for softwood (non-copper treatments), per m3 of treated wood ...... 163 Table 6-4: Impacts of wood preservative treatments for softwood (copper treatments), per m3 of treated wood ...... 164 Table 6-5 Key softwood forestry impact contributions >1% by source (baseline 2006) ...... 167 Table 6-6 Key hardwood forestry impact contributions >1% by source (baseline 2006) ...... 168 Table 6-7 Electricity mix by product and state ...... 169

LCI and EPDs for Australian Wood Products 13 of 159

List of Acronyms

ABARES Australian Bureau of Agricultural and Resource Economics and Sciences ACT Australian Capital Territory ACQ Ammoniacal Copper Quaternary preservative ADP Abiotic Depletion Potential ADPE Abiotic Depletion Potential, Elements ADPF Abiotic Depletion Potential, Fossil AP Acidification Potential of soil and water APVMA Australian Pesticides and Veterinary Medicines Authority CCA Chromated Copper Arsenate preservative CML Institute of Environmental Sciences at Leiden University CORRIM Consortium for Research on Renewable Industrial Materials CRU Components for Re-Use CSIRO Commonwealth Scientific and Industrial Research Organisation DOC Degradable Organic Carbon

DOCf Fraction of Degradable Organic Carbon dissimilated DQI Data Quality Indicator EEE Exported Energy, Electrical EET Exported Energy, Thermal EoL End-of-Life EP Eutrophication Potential EPD Environmental Product Declaration ET Evapotranspiration FSP Fibre Saturation Point FW Net use of fresh water FWPA Forest and Wood Products Australia GaBi Ganzheitliche Bilanzierung (German for holistic balancing) GBCA Green Building Council of Australia GCV Gross Calorific Value GHG Greenhouse Gas GWP Global Warming Potential GWPT Global Warming Potential, Total GWPB Global Warming Potential, Biogenic GWPBC Global Warming Potential, Biogenic Carbon dioxide GWPEB Global Warming Potential, Excluding Biogenic Carbon dioxide GWPF Global Warming Potential, Fossil HWD Hazardous Waste Disposed ISCA Infrastructure Sustainability Council of Australia ISO International Organization for Standardization LCA Life Cycle Assessment LCI Life Cycle Inventory

LCI and EPDs for Australian Wood Products 14 of 159

LCIA Life Cycle Impact Assessment LOSP Light Organic Solvent Preservative LEED Leadership in Energy and Environmental Design rating tool MC Moisture Content MDF Medium Density Fibreboard MER Materials for Energy Recovery MFR Materials For Recycling MR Moisture Resistant MSDS Material Safety Datasheet MUF Melamine-Urea-Formaldehyde resin NCV Net calorific value NGA National Greenhouse Accounts NHWD Non-Hazardous Waste Disposed NMVOC Non-Methane Volatile Organic Compound NRSF Use of Non-Renewable Secondary Fuels NSW New South Wales ODP Ozone Depletion Potential PCR Product Category Rules PENRE Use of Primary Energy (Non-Renewable) as Energy PENRM Use of Primary Energy (Non-Renewable) as Material PENRT Use of Primary Energy (Non-Renewable) in Total PERE Use of Primary Energy (Renewable) as Energy PERM Use of Primary Energy (Renewable) as Material PERT Use of Primary Energy (Renewable) in Total PF Phenol Formaldehyde resin PRF Phenol Resorcinol Formaldehyde resin POCP Photochemical Ozone Creation Potential PU Polyurethane resin PubCRIS Public Chemical Registration Information System QA Quality Assurance QLD Queensland RSF Use of Renewable Secondary Fuels RWD Radioactive Waste Disposed SM Use of Secondary Material TAS Tasmania TDA Timber Development Association (New South Wales) UF Urea Formaldehyde resin USGBC United States Green Building Council VIC Victoria VOC Volatile Organic Compound WA Western Australia

LCI and EPDs for Australian Wood Products 15 of 159

Glossary

Life cycle A view of a product system as “consecutive and interlinked stages … from raw material acquisition or generation from natural resources to final disposal” (ISO 14040:2006, section 3.1). This includes all material and energy inputs as well as emissions to air, land and water. Life Cycle Assessment (LCA) “Compilation and evaluation of the inputs, outputs and the potential environmental impacts of a product system throughout its life cycle” (ISO 14040:2006, section 3.2) Life Cycle Inventory (LCI) “Phase of life cycle assessment involving the compilation and quantification of inputs and outputs for a product throughout its life cycle” (ISO 14040:2006, section 3.3) Life Cycle Impact Assessment (LCIA) “Phase of life cycle assessment aimed at understanding and evaluating the magnitude and significance of the potential environmental impacts for a product system throughout the life cycle of the product” (ISO 14040:2006, section 3.4) Life cycle interpretation “Phase of life cycle assessment in which the findings of either the inventory analysis or the impact assessment, or both, are evaluated in relation to the defined goal and scope in order to reach conclusions and recommendations” (ISO 14040:2006, section 3.5) Functional unit “Quantified performance of a product system for use as a reference unit” (ISO 14040:2006, section 3.20) Allocation “Partitioning the input or output flows of a process or a product system between the product system under study and one or more other product systems” (ISO 14040:2006, section 3.17) Closed-loop and open-loop allocation of recycled material “An open-loop allocation procedure applies to open-loop product systems where the material is recycled into other product systems and the material undergoes a change to its inherent properties.” “A closed-loop allocation procedure applies to closed-loop product systems. It also applies to open- loop product systems where no changes occur in the inherent properties of the recycled material. In such cases, the need for allocation is avoided since the use of secondary material displaces the use of virgin (primary) materials.” (ISO 14040:2006, section 4.3.4.3.3)

LCI and EPDs for Australian Wood Products 16 of 159

Foreground system “Those processes of the system that are specific to it … and/or directly affected by decisions analysed in the study.” (JRC, 2010, p. 97) This typically includes first-tier suppliers, the manufacturer itself and any downstream life cycle stages where the manufacturer can exert significant influence. As a general rule, specific (primary) data should be used for the foreground system. Background system “Those processes, where due to the averaging effect across the suppliers, a homogenous market with average (or equivalent, generic data) can be assumed to appropriately represent the respective process … and/or those processes that are operated as part of the system but that are not under direct control or decisive influence of the producer of the good….” (JRC, 2010, pp. 97-98) As a general rule, secondary data are appropriate for the background system, particularly where primary data are difficult to collect. Critical Review “Process intended to ensure consistency between a life cycle assessment and the principles and requirements of the International Standards on life cycle assessment” (ISO 14040:2006, section 4.45)

LCI and EPDs for Australian Wood Products 17 of 159

Executive Summary

The goals of this study are: 1. To update Forest and Wood Products Australia’s (FWPA’s) existing life cycle inventory (LCI) and environmental product declarations (EPDs) for sawn softwood, sawn hardwood, particleboard, medium density fibreboard (MDF) and plywood. 2. To add an additional LCI and EPD for glued-laminated timber (glulam). All EPDs are intended to represent average Australian production and each EPD covers a range of specific sub-products, e.g. the sawn hardwood EPD includes both green and kiln-dried rough-sawn timber, as well as kiln-dried dressed timber. The intended audience for the EPDs is primarily building professionals, such as architects and specifiers, to help Australian wood products qualify for points under green rating tools. The EPDs comply with the requirements for an industry-wide EPD under the Green Building Council of Australia’s (GBCA’s) Green Star rating system, the Infrastructure Sustainability (IS) rating scheme of the Infrastructure Sustainability Council of Australia (ISCA), and the LEED rating tool of the US Green Building Council (USGBC). This study uses the 2015/16 Australian financial year as its reference and serves as a 10-year update to earlier work done by CSIRO, which was based primarily on 2005/06 data. It therefore provides a valuable benchmark for understanding changes within the Australian timber industry over the last decade and provides insights into developments affecting the environmental credentials of timber products. Improvements in manufacturing processes have been observed for many of the product categories. The most notable include:

• An increase in the conversion rate from logs to green sawn timber for both sawn softwood and sawn hardwood. • Decreased energy use in kiln-drying per cubic metre of sawn softwood and sawn hardwood. • A reduction in resin loadings for both MDF and particleboard products. • The substitution of natural gas for biomass as a thermal energy source for particleboard production. • Lower energy use per cubic metre of production for veneer and plywood products. Notable opportunities for further improvement in the manufacturing process include:

• Reducing the amount of wood residues burnt within the forestry stage, where feasible. This will lead to significant improvements in summer smog, eutrophication and acidification potentials of all wood products. • Further improving recovery rates for sawn products, as demonstrated by the variation in recovery rates found between sites. Sites with lower recovery rates may be able to achieve higher rates by improving processes or technology upgrades, though the authors recognise that recovery rates are linked to the quality of the incoming logs. • A further shift toward biomass for thermal energy and away from fossil energy (mainly natural gas and LPG). This would decrease carbon emissions, though it may increase emissions leading to summer smog. (It should be noted that summer smog is not currently

LCI and EPDs for Australian Wood Products 18 of 159

a major issue in areas of Australia outside of major cities (Department of the Environment and Heritage, 2005).) • For sites that have their own dams: use dam covers or reduce the number of dams to reduce unnecessary water loss. New additions in this study include packaging for all products and the inclusion of preservative treatments for sawn timber and glulam. Packaging was found to be of minor significance over the life cycle of the product. Preservative treatments were also found to be of minor significance for the assessed environmental indicators, despite their benefits for increasing the lifespan of the timber products which they protect. The updated data behind the EPDs allow them to see out their validity to 2020. The information contained in these documents will enable building professionals to make informed decisions about the environmental credentials of construction materials, driving the design of modern structures which have a lower impact on their surroundings.

LCI and EPDs for Australian Wood Products 19 of 159

1. Background

This is the second major iteration of the Environmental Product Declarations (EPDs) for Australian wood products, produced for Forest and Wood Products Australia (FWPA). The primary aim of this version was to collect a new set of Life Cycle Inventory (LCI) data and to update all EPD results to ensure that the EPDs see out their current validity period to 2020. Following EN 15804 section 6.3.7, data “shall have been updated within the last 10 years for generic data and within the last 5 years for producer specific data”. The aim of the current report is to fulfil this requirement. The original EPDs (relating to v1.0 of this background report) were produced by thinkstep and the Timber Development Association of New South Wales. The Life Cycle Assessment (LCA) results included within these EPDs were based on an LCI study completed in 2009 by the Commonwealth Scientific and Industrial Research Organisation (CSIRO) that inventoried data from the 2005/06 Australian financial year (1st July 2005 to 30th June 2006) (CSIRO, 2009a). These datasets were checked and partially updated in 2015 before the EPDs were produced. However, to protect the confidentiality of individual sites, only the average LCI data were available to thinkstep so it was not possible to spot-check or update site-specific data. As a result, it was decided that the best course of action was to collect all data again, building on past work wherever possible. The LCI data in version 2.0 of this report were collected through industry surveys conducted from June 2016 to July 2017 by thinkstep and Stephen Mitchell Associates. The time period covered is the 2015/16 Australian financial year (1st July 2015 to 30th June 2016), making the present study a 10-year update of CSIRO’s original LCI. The authors of this report formally acknowledge the work done by CSIRO from 2006 to 2009 which laid the foundation for this study. CSIRO produced a public summary (CSIRO, 2009a) and then separate module reports which were not part of the public communication (CSIRO, 2009b). These reports served as an important input into both the previous and the current versions of the EPDs, even though the underlying LCA data are no longer used as of version 2.0. Five EPDs were produced in the first version of this study (sawn softwood, sawn hardwood, particleboard, MDF and plywood). This second version adds a sixth EPD for glued-laminated timber (glulam) to the series. It also adds treatments for both sawn softwood and sawn hardwood, which were excluded from the initial scope. Additionally, while there are no EPDs for forest products, the scope of work also includes an update to the LCI for plantation softwood forestry and native hardwood forestry, as these two forest types provide the inputs used in all downstream products.

LCI and EPDs for Australian Wood Products 20 of 159

2. Goal of the Study

The goals of this study are: 1. To update FWPA’s existing LCI and EPDs for sawn softwood, sawn hardwood, particleboard, MDF and plywood. 2. To add an additional LCI and EPD for glulam. All EPDs are intended to represent the average Australian production mix. Imports are deliberately excluded due to challenges in collecting data from many producers in multiple countries. Each EPD applies to a range of specific sub-products – e.g. green and kiln-dried hardwood – to suit different applications within the construction industry. The target audience for the EPDs is building professionals, such as architects and specifiers, in order to help Australian wood products qualify for points under green rating tools, such as:

• The Green Star rating tool of the Green Building Council of Australia (GBCA); • The IS rating tool of the Infrastructure Sustainability Council of Australia (ISCA); and • The LEED rating tool of the US Green Building Council (USGBC). The EPDs are also intended to provide a verified carbon footprint for wood products that can be used in marketing the environmental credentials of wood products (with reference made to the EPDs). To help enable the use of the EPD data in green building, a follow-up to this study will be to make the industry-average LCI data available in Life Cycle Assessment (LCA) databases for use by LCA practitioners. Each EPD and corresponding LCA study has been produced according to EN 15804 (EN 15804:2012+A1:2013), ISO 14025 (ISO 14025:2006) and PCR 2012-01 Construction products and construction services v2.2 (2017-05-30) of the International EPD® System (IEPDS, 2017). The EPDs are published under the Australasian EPD Programme.

LCI and EPDs for Australian Wood Products 21 of 159

3. Scope of the Study

The following section describes the general scope of the project to achieve the stated goals. This includes, amongst others, the identification of specific product systems to be assessed, the functional unit and reference flows, the system boundary, allocation procedures, and cut-off criteria of the study.

3.1. Product description and application area

This background report applies to six groups of Australian-produced wood products used in the building and construction industry: 1. Softwood 2. Hardwood 3. Particleboard 4. Medium Density Fibreboard (MDF) 5. Plywood 6. Glued-laminated timber (glulam) Imports are not within the scope of this study and, as such, all EPDs represent a domestic production mix rather than a consumption mix. These products are used for a range of purposes within the built environment (see Table 3-1). Softwood is commonly used for structural applications, particularly framing. Hardwood is commonly used for flooring, decking, joinery, cladding and veneers. Particleboard and MDF are used in cabinet-making, joinery and flooring. Plywood is used for bracing, joinery and flooring. Glulam is used for structural elements.

Table 3-1 Technical specifications applying to the products in this background report

Product group Relevant Australian Standard Applications Softwood AS/NZS 1748 Timber - Solid - Stress- Structural applications, e.g. framing graded for structural purposes (AS/NZS 1748.1:2011; AS/NZS 1748.2:2011) AS 2858 Timber – Softwood – Visually Structural applications, e.g. framing stress-graded for structural purposes (AS 2858-2008) AS 4785 Timber – Softwood – Sawn and Furniture, shelving, lining, flooring, milled products (AS 4785.1-2002; AS DIY and remanufacturing 4785.2-2002) Hardwood AS 2082 Timber – Hardwood – Visually Structural applications, e.g. beams, stress-graded for structural purposes (AS bearers, floor joists 2082:2007) AS 2796 Timber – Hardwood – Sawn Furniture, joinery, flooring, decking, and milled products (AS 2796.1:1999; cladding AS 2796.2:2006)

LCI and EPDs for Australian Wood Products 22 of 159

Product group Relevant Australian Standard Applications Particleboard AS/NZS 1859.1:2004 Reconstituted Cabinet-making and joinery wood-based panels – Specifications – Particleboard (AS/NZS 1859.1:2004) AS/NZS 1859.3:2005 Reconstituted Cabinet-making and joinery wood-based panels – Specifications – Decorative and overlaid wood panels (AS/NZS 1859.3:2005) AS/NZS 1860.1:2002 Particleboard Flooring flooring – Specifications (AS/NZS 1860.1:2002) MDF AS/NZS 1859.2:2004 Reconstituted Cabinet-making and joinery wood-based panels – Specifications – Dry-processed fibreboard (AS/NZS 1859.2:2004) AS/NZS 1859.3:2005 Reconstituted Cabinet-making and joinery wood-based panels – Specifications – Decorative overlaid wood panels (AS/NZS 1859.3:2005) Plywood AS/NZS 2269.0:2012 Plywood – Structural applications, e.g. bracing Structural – Part 0: Specifications (AS/NZS 2269:2004) AS/NZS 2270:2006 Plywood and Joinery, internal wall lining, internal blockboard for interior use (AS/NZS ceiling lining 2270:2006) AS/NZS 2271:2004 Plywood and Exterior cladding, roof lining, walls blockboard for exterior use (AS/NZS 2271:2004) Glulam AS/NZS 1328.1:1998 Glued laminated Structural applications, e.g. beams, structural timber - Performance joists, columns, portal frames requirements and minimum production requirements

3.2. Declared unit

The declared unit for each EPD is set out below in Table 3-2. All EPDs represent an average Australian-produced product across multiple forests and multiple manufacturers’ plants.

Table 3-2 Declared unit for each EPD

Product group Unit Product Softwood 1 m3 kiln-dried, rough sawn 12.2% moisture content (dry basis), average density of 551 kg/m3 1 m3 kiln-dried, dressed 12.2% moisture content (dry basis), average density of 551 kg/m3 Hardwood 1 m3 kiln-dried, rough sawn 10.4% moisture content (dry basis), density of 735 kg/m3

LCI and EPDs for Australian Wood Products 23 of 159

Product group Unit Product 1 m3 kiln-dried, dressed 10.4% moisture content (dry basis), density of 735 kg/m3 1 m3 green, rough sawn 26.4% moisture content (dry basis), density of 768 kg/m3 Particleboard 1 m2 E0 & E1 standard, melamine coated, 16 mm thick 1 m2 E0 & E1 standard, melamine coated, 18 mm think 1 m2 E0 & E1 moisture resistant, melamine coated, 16 mm thick 1 m2 E0 & E1 moisture resistant, melamine coated, 18 mm thick 1 m2 Flooring, 19 mm thick 1 m2 Flooring, 22 mm thick 1 m2 Flooring, 25 mm thick MDF 1 m2 E0 & E1 standard, melamine coated, 16 mm thick 1 m2 E0 & E1 standard, melamine coated, 18 mm thick 1 m2 E0 & E1 standard, melamine coated, 25 mm thick 1 m2 E0 & E1 moisture resistant, melamine coated, 16 mm thick 1 m2 E0 & E1 moisture resistant, melamine coated, 18 mm thick 1 m2 E0 & E1 moisture resistant, melamine coated, 25 mm thick Plywood 1 m2 exterior, A-Bond, 7 mm thick (bracing) 1 m2 exterior, A-Bond, 9 mm thick (bracing) 1 m2 formply, A-Bond, 17 mm thick (formwork) 1 m2 formply, B-Bond, 17 mm thick (formwork) 1 m2 flooring, tongue and groove, 15 mm thick 1 m2 flooring, tongue and groove, 25 mm thick Glulam 1 m3 glued-laminated softwood 1 m3 glued-laminated hardwood & cypress

3.3. System boundaries

3.3.1. General system boundaries

The scope of this EPD is ‘cradle-to-gate with options’ as per EN 15804. The declared modules are shown in Table 3-3 below. It includes the environmental impacts associated with raw material extraction and processing (A1), material transport to the manufacturer (A2), manufacturing processes (A3), waste processing (C3), waste disposal (C4), and reuse, recovery and recycling potential (D). Impacts and indicators related to waste are considered in the module in which the waste occurs in line with the polluter pays principle specified in EN 15804. Other life cycle stages concerning transport to the construction site (A4), the construction process (A5), the use stage (B1-B7), deconstruction and demolition (C1), and end-of-life transport (C2) are not included in this EPD. These life cycle stages vary by end use and are best considered at the building level.

LCI and EPDs for Australian Wood Products 24 of 159

Table 3-3 Modules of the production life cycle included in the EPD as per EN15804 (X = declared module; MND = module not declared)

Construction Recovery Product stage Use stage End-of-life stage

process stage stage

supply

potential

Use

Repair

Disposal

Transport Transport Transport

Installation

Maintenance

Replacement

Manufacturing

Refurbishment

recovery

Wasteprocessing

Raw material Raw

Operationaluse water

Operational energyOperationaluse

Deconstruction/ demolition Future reuse, recycling or energy orenergy Future recycling reuse,

A1 A2 A3 A4 A5 B1 B2 B3 B4 B5 B6 B7 C1 C2 C3 C4 D X X X MND MND MND MND MND MND MND MND MND MND MND X X X

Upstream processes

The following upstream processes/life cycle stages are included: A1) Raw material supply

• Production of seeds and seedlings, stand establishment, forest road construction and maintenance, fire prevention and control, forest management, thinning and clear felling of softwood and harvesting of hardwood. • Extraction and processing of raw materials (e.g. mining processes). • Generation of electricity and heat from primary energy resources such as natural gas, LPG and solid biomass, also including their extraction, refining and transport. This also includes energy needed for raw material supply and energy for manufacturing in core process. • Energy recovery and other recovery processes from secondary fuels such as hog fuel and forest residues. • Processing up to the end-of-waste state.

Core processes

The following core processes are included. A2) Transportation

• Haulage of the harvested softwood and hardwood logs.

A3) Manufacturing:

• Manufacturing of sawn softwood and hardwood consisting of log storing, debarking, milling, drying, optional planing and packaging. • Manufacturing of particleboards consisting of the production of fibres and particles, drying, blending, board manufacture, mat forming, pressing, finishing and packaging. • Manufacturing of MDF consisting of debarking, chipping, pulping, drying and adding of wax and resin in the blowline, mat forming, pressing, finishing and packaging. • Manufacturing of plywood consisting of veneer production, glue mixing, overlay sheet production, plywood manufacturing, finishing and packaging.

LCI and EPDs for Australian Wood Products 25 of 159

• Manufacturing of glulam consisting of sawing wood to size, gluing wood layers, trimming to size and packaging.

Downstream processes

The following end-of-life scenarios are included:

• Landfill (typical): Disposal in a construction and demolition waste landfill without

landfill gas utilisation, with a degradable organic carbon fraction (DOCf) of 0.0-1.6% (see Table 4-79) (C4). • Landfill (NGA): Disposal in a municipal waste landfill with methane capture with a

DOCf of 10% (Australian Government, 2016a). Landfill gas is used in a combined heat and power plant to generate electricity to be supplied to the national grid mix. (C4) A credit is given for replaced electricity (D). • Energy recovery: Combustion (C3) with recovered energy (D) offset against average electricity from the Australian grid and thermal energy from natural gas (EN 16485:2014). • Reuse: Carbon sequestered in wood leaves the system boundary (C3) and is exported to future product systems (D). A credit is provided for offsetting primary production of the respective product (D) (EN 16485:2014). • Recycling: Wood waste is chipped (C3) and assigned credits relative to the avoided production of woodchips from virgin wood (D). In line with the reuse scenario, carbon sequestered in wood leaves the system boundary at C3 (EN 16485:2014).

3.3.2. Time boundaries Primary data should cover the 2015/16 Australian financial year (1st July 2015 to 30th June 2016) for participating sites or be representative of this time period. All data in the background system is taken from the GaBi Database 2017 (thinkstep, 2017). Most datasets have a reference year between 2013 and 2016. The specific reference year for all important background datasets used in this study can be found in section 4.2. Long-term emissions (>100 years) are not taken into consideration in the impact estimate. Waste to landfill is modelled assuming a 100-year time horizon.

3.3.3. Geographical boundaries All EPDs shall represent an Australian average with foreground data collected from operations within Australia. For some materials, background data may represent European conditions where no matching Australian LCI dataset is available within the GaBi databases. The specific reference location for all important background datasets used in this study can be found in section 4.1.2.

3.3.4. Boundaries for manufacturing of equipment and for employees

As per the PCR (IEPDS, 2017, section 7.5.4 ), the following are excluded from the system boundary:

• Infrastructure, construction, production equipment and tools not consumed in the production process; and • Impacts due to employees, e.g. employees commuting to and from work.

LCI and EPDs for Australian Wood Products 26 of 159

3.3.5. Boundaries to other product life cycles

Allocation of recycled material is reported in the LCI as an input or output flow when such materials leave or enter the specific product system. The boundary between the current and the next product system is defined by the willingness to pay for the recycled material. This implies that from the moment the user of a secondary material pays for the material, this (secondary) product system will also be responsible for its environmental burdens from that point onward. This is referred to as the Polluter Pays Principle within EN 15804. Consequently, if there is an inflow of recycled material to the production system, the recycling process and transportation of the recycled material to site are both included. If there is an outflow of material to recycling, both dismantling and transportation of the material to a sorting/recycling facility are included. The material intended for recycling is then an outflow from the production system.

3.4. Reference service life

As all EPDs have a cradle-to-gate scope with options as per EN 15804, no reference service life is specified.

3.5. Data quality requirements

The data used to create the inventory model shall be as precise, complete, consistent, and representative as possible with regards to the goal and scope of the study under given time and budget constraints.

• Measured primary data are considered to be of the highest precision, followed by calculated data, literature data, and estimated data. The goal is to model all relevant foreground processes using measured or calculated primary data. An assessment of data precision is used for each primary data flow within the LCI using a data quality indicator (DQI). • Completeness is judged based on the completeness of the inputs and outputs per unit process and the completeness of the unit processes themselves. The goal is to capture all relevant data in this regard. • Consistency refers to modelling choices and data sources. The goal is to ensure that differences in results reflect actual differences between product systems and are not due to inconsistencies in modelling choices, data sources, emission factors, or other artefacts. • Reproducibility expresses the degree to which third parties would be able to reproduce the results of the study based on the information contained in this report. The goal is to provide enough transparency with this report so that third parties are able to approximate the reported results. This ability may be limited by the exclusion of confidential primary data and access to the same background data sources. • Representativeness expresses the degree to which the data matches the geographical, temporal, and technological requirements defined in the study’s goal and scope. The goal is to use the most representative primary data for all foreground processes and the most representative industry-average data for all background processes. Whenever such data were not available (e.g., no industry-average data available for a certain country), best- available proxy data were employed. An evaluation of the data quality with regard to these requirements is provided in section 6.4 of this report.

LCI and EPDs for Australian Wood Products 27 of 159

3.6. Cut-off criteria

Environmental impacts relating to personnel, infrastructure, and production equipment not directly consumed in the process are excluded from the system boundary as per the PCR IEPDS, section 6.5.4 (2017). All data collected from producers were included in the LCA model without the use of cut-off criteria. Any cut-off criteria applied for background data are described in the GaBi documentation (thinkstep, 2017). All GaBi datasets include at least 95% of mass and energy input and output flows, and 98% of input and output flows by environmental relevance (based on expert judgment).

3.7. Selection of LCIA methodology and impact categories

3.7.1. Life cycle impact assessment (LCIA) indicators

Table 3-4 outlines the impact assessment indicators declared within this report as per EN 15804:2012+A1:2013 Annex C. Information on environmental impacts is expressed with the impact category parameters of LCIA using characterisation factors. In addition to the impact categories required, global warming potential due to fossil emissions (GWPF) has also been declared within the EPD to aid the user in understanding the global warming potential of each product. GWPF is calculated as the difference between the total global warming potential (GWP) and global warming potential from biogenic removals and emissions (GWPB). The reported impact categories represent impact potentials, i.e., they are approximations of environmental impacts that could occur if the emissions would (a) follow the underlying impact pathway and (b) meet certain conditions in the receiving environment while doing so. In addition, the inventory only captures that fraction of the total environmental load that corresponds to the chosen functional unit (relative approach). LCIA results are therefore relative expressions only and do not predict actual impacts, the exceeding of thresholds, safety margins, or risks. Long-term emissions (>100 years) are not taken into consideration in the impact estimate.

Table 3-4: Life cycle impact assessment indicators

Impact Category Description Unit Reference

Global Warming A measure of greenhouse gas emissions, kg CO2 (IPCC, 2007)

Potential (GWPT, such as CO2 and methane. These emissions equivalent GWPF and are causing an increase in the absorption of GWPB) radiation emitted by the earth, increasing the natural greenhouse effect. This may in turn have adverse impacts on ecosystem health, human health and material welfare. Global Warming Potential is calculated with a time horizon of 100 years. Abiotic Resource The consumption of non-renewable resources kg Sb (van Oers, et Depletion, non- leads to a decrease in the future availability of equivalent al., 2002) fossil resources the functions supplied by these resources. (ADPE) Depletion of mineral resources and non-

LCI and EPDs for Australian Wood Products 28 of 159

Impact Category Description Unit Reference Abiotic Resource renewable energy resources are reported MJ (net Depletion, fossil separately. Depletion of mineral resources is calorific resources assessed based on ultimate reserves. value) (ADPF)

3- Eutrophication Eutrophication covers all potential impacts of kg PO4 (Guinée, et Potential (EP) excessively high levels of macronutrients, the equivalent al., 2002) most important of which nitrogen (N) and phosphorus (P). Nutrient enrichment may cause an undesirable shift in species composition and elevated biomass production in both aquatic and terrestrial ecosystems. In aquatic ecosystems increased biomass production may lead to depressed oxygen levels, because of the additional consumption of oxygen in biomass decomposition.

Acidification A measure of emissions that cause acidifying kg SO2 (Guinée, et Potential (AP) effects to the environment. The acidification equivalent al., 2002) potential is a measure of a molecule’s capacity to increase the hydrogen ion (H+) concentration in the presence of water, thus decreasing the pH value. Potential effects include fish mortality, forest decline and the deterioration of building materials.

Photochemical A measure of emissions of precursors that kg C2H4 (Guinée, et Ozone Creation contribute to ground level smog formation equivalent al., 2002)

Potential (POCP) (mainly ozone O3), produced by the reaction of VOC and carbon monoxide in the presence of nitrogen oxides under the influence of UV light. Ground level ozone may be injurious to human health and ecosystems and may also damage crops. Ozone Depletion A measure of air emissions that contribute to kg CFC-11 (Guinée, et Potential (ODP) the depletion of the stratospheric ozone layer. equivalent al., 2002) Depletion of the ozone leads to higher levels of UVB ultraviolet rays reaching the earth’s surface with detrimental effects on humans and plants.

3.7.2. Life cycle inventory indicators

Table 3-5 describes parameters that come directly from the LCI, namely the use of renewable and non-renewable material resources, renewable and non-renewable primary energy, and water.

LCI and EPDs for Australian Wood Products 29 of 159

Table 3-5: Life cycle inventory indicators on use of resources

Unit Indicator Use of renewable primary energy as energy carrier (PERE) MJ, net calorific value Use of renewable primary energy as raw materials (PERM) MJ, net calorific value Total use of renewable primary energy (PERT) MJ, net calorific value Use of non-renewable primary energy as energy carrier (PENRE) MJ, net calorific value Use of non-renewable primary energy as raw materials (PENRM) MJ, net calorific value Total use of non-renewable primary energy (PENRT) MJ, net calorific value Use of secondary material (SM) kg Use of renewable secondary fuels (RSF) MJ, net calorific value Use of non-renewable secondary fuels (NRSF) MJ, net calorific value Net use of fresh water (FW) m³

The inventories for the basic materials contain the information on the “Total use of renewable/non- renewable primary energy”. The indicators “Use of primary energy as raw materials” are assessed based on the net calorific value of the product. The “Use of primary energy as energy carrier” can be calculated as the “Total primary energy” minus the “Use of primary energy as raw materials”. Table 3-6 and Table 3-7 set out the indicators on waste materials and materials/energy that are exported from the product system respectively. The reader should also be aware that water consumption does not account for relative water stress with respect to geographical location, meaning that it provides no information about the potential impacts of any water consumption that does occur.

Table 3-6: Life cycle inventory indicators on waste categories

Unit Indicator Hazardous waste disposed (HWD) kg Non-hazardous waste disposed (NHWD) kg Radioactive waste disposed (RWD) kg

Table 3-7: Life cycle inventory indicators on output flows

Unit Indicator Components for re-use (CRU) kg Materials for recycling (MFR) kg Materials for energy recovery (MER) kg Exported electrical energy (EEE) MJ Exported thermal energy (EET) MJ

3.8. Interpretation to be used

The results of the LCI and LCIA have been interpreted according to the Goal and Scope. The interpretation addresses the following topics:

LCI and EPDs for Australian Wood Products 30 of 159

• Identification of significant findings, such as the main process step(s), material(s), and/or emission(s) contributing to the overall results • Evaluation of completeness, sensitivity, and consistency to justify the exclusion of data from the system boundaries as well as the use of proxy data. • Conclusions, limitations and recommendations

3.9. Allocation and recycling

Allocation rules for foreground processes as well as upstream data used within this project are conformant with the allocation principles outlined in EN 15804.

3.9.1. Allocation of upstream data

For all refinery products, allocation by mass and net calorific value has been applied. The specific manufacturing route of every refinery product is modelled and so the impacts associated with the production of these products are calculated individually. Two allocation rules are applied: 1. The raw material (crude oil) consumption of the respective stages, which is necessary for the production of a product or an intermediate product, is allocated by energy (mass of the product * calorific value of the product); 2. The energy consumption (thermal energy, steam, electricity) of a process, e.g. atmospheric distillation, being required by a product or an intermediate product, are charged to the product according to the share of the throughput of the stage (mass allocation). Materials and chemicals needed used in the manufacturing process are modelled using the allocation rule most suitable for the respective product. For further information on a specific product see the GaBi software documentation (thinkstep, 2017). In addition to the above-mentioned allocation methods for refinery products and materials, inventories for electricity and thermal energy generation also include allocation by economic value for some co-products (e.g. gypsum, boiler ash and fly ash). In case of plants for the co-generation of heat and power, allocation by energy is applied.

3.9.2. Allocation in foreground data

Impacts due to multifunctional processes within the foreground system (e.g. sawmilling) are allocated to co-products by economic value. This is done because the difference in the value of co- products is large (>25% in most cases, in line with IEPDS (2017), section 7.7), yet most co-products still have some economic value and therefore cannot be considered as wastes. Regardless of the use of economic allocation, inherent properties of the co-products – such as carbon content and energy content – are always allocated based on physical flows such as mass.

3.10. Software and database

The LCA model was created using the GaBi software system for life cycle engineering, developed by thinkstep AG. The GaBi Database 2017 provides the life cycle inventory data for the background system.

LCI and EPDs for Australian Wood Products 31 of 159

3.11. Critical review and EPD verification

Review of the LCI data and LCIA results was carried out by:

• Andrew Carre, Lecturer, RMIT University • Brad Ridoutt, Principal Research Scientist, CSIRO (focusing solely on water modelling) • Fabiano Ximenes, Research Scientist, New South Wales Department of Primary Industries • Kimberly Robertson, Consultant, Catalyst Limited All reviews were carried out independently, rather than as a panel. Verification of the EPDs was carried out by Kimberly Robertson, Consultant at Catalyst Limited. Kimberly is an independent verifier registered under the Australasian EPD Programme Limited. A copy of her verification statement can be found in Annex H.

LCI and EPDs for Australian Wood Products 32 of 159

4. Life Cycle Inventory Analysis

4.1. Methods

4.1.1. Primary data collection

Primary data were collected from a sample of Australian forest and wood products manufacturers for each product category. Quota sampling was used with the aim of covering at least 50% of total Australian domestic production by volume for all product categories except sawn hardwood where 25% coverage was specified due to the large number of small producers in the sawn hardwood industry. Quota sampling (a non-probabilistic technique) was selected over stratified random sampling (its probabilistic equivalent) due to the difficulty in securing commitment for the study. Each site was expected to contribute approximately 16 hours of staff time – a commitment that no company made lightly. All FWPA members were given the opportunity to participate. A non-random sampling technique was chosen to allow the practitioners to focus effort on large sites and those smaller sites that were considered most likely to participate. While the sampling technique used was non- random, the only two criteria used for site selection were site size and perceived willingness to cooperate. The perceived environmental performance of the site was not a factor in site selection. The project ran from April 2016 to November 2017, with most of the focus on collecting high-quality data from the participating facilities. Surveys were conducted by thinkstep and Stephen Mitchell Associates between June 2016 and September 2017. Sites were requested to supply data from the Australian financial year starting 1st July 2015 and ending 30th June 2016 wherever possible. Where the site used a different reporting period (which was uncommon), they were asked to provide data from their most recent reporting year instead. Data were collected through either: 1. A web-based questionnaire in thinkstep's SoFi software; or 2. A desktop-based questionnaire in Microsoft Excel. Both questionnaires had the same set of questions. Initially only the web-based questionnaires were provided to all data collectors. The Excel questionnaires were developed during the process when some sites found the web-based questionnaires difficult to complete. In the end, roughly half of all the data were collected through the web-based questionnaires and half were collected through the Excel questionnaires. To help sites complete the questionnaires, support was provided by phone, email, and, in rare cases, through in-person visits. Upon receipt of a completed questionnaire, the project team ran several checks to determine the validity of the data. The most important of these were:

• Wood balance: does dry wood in = dry wood product out + dry wood residues out? Measured in kilograms of bone dry wood (0% water content). • Water balance: does direct fresh water input + water incorporated in incoming wood = waste water out + evaporation + water incorporated in outgoing products and wastes?

LCI and EPDs for Australian Wood Products 33 of 159

• Energy check: does stated total thermal energy input = expected thermal energy input to dry wood from its incoming moisture content to its final moisture content within a minimum and maximum efficiency range? • Deviation between sites: can deviation between sites be explained? • Deviation from 2006 baseline: can deviation from older data be explained? • Deviation from CORRIM: can deviation from CORRIM data be explained? (CORRIM in the USA has produced many detailed LCI reports for wood products in the past 10 years, making them a helpful public benchmark against which to assess the FWPA LCI.)

4.1.2. Calculation of net calorific value for wood

The net calorific value (NCV) of softwood is derived from its gross calorific value (GCV) using the following equation (van Loo & Koppejan, 2002): 푤 푤 ℎ 푤 푁퐶푉 = 퐺퐶푉 ∗ (1 − ) − 2.447 ∗ − 2.447 ∗ ∗ 9.01 ∗ (1 − ) 100 100 100 100 Where: NCV = net calorific value, wet basis (MJ/kg) GCV = gross calorific value, dry basis (MJ/kg) w = water content of the fuel (mass water over green mass) (weight-%) h = hydrogen content of the fuel, dry basis (weight-%) The GCV of wood is sourced from the Phyllis 2 database from Energy Research Centre of the Netherlands (2017), by evaluating the average of "wood, pine". The GCV is derived from the dry and ash free GCV of 20.68 MJ/kg with a dry ash content of 0.7%. The GCV is thus determined as 20.53 MJ/kg. The hydrogen content is taken from EN 16449:2014, and is 6%. The water content is calculated from the given moisture using the following formula: 푀퐶 푤 = 푀퐶 1 + 100 Where: w = water content of the fuel (mass water over green mass) (weight-%) MC = moisture content of the fuel (mass water over oven dry mass) (weight-%)

4.1.3. Calculation of net calorific value for resins

The net calorific value of the resins was calculated based on the elemental composition by the following Boie formula (Filho, et al., 2013): 푀퐽 푁퐶푉 [ ] = 35푐 + 94.3ℎ + 6.3푛 − 10.8표 − 2.44푤 푘푔 with: c = Mass of carbon per kg of resin h = Mass of hydrogen per kg of resin n = Mass of nitrogen per kg of resin o = Mass of oxygen per kg of resin w = Mass of water per kg of resin Table 4-1 shows the elemental composition and the net calorific values of the resins.

LCI and EPDs for Australian Wood Products 34 of 159

Table 4-1 Elemental composition and net calorific value of resins

Resin C [g/kg] H [g/kg] N [g/kg] O [g/kg] Net calorific value [MJ/kg] Urea formaldehyde 333 55.9 389 222 17.0 Melamine formaldehyde 381 48.0 444 127 19.3 Melamine urea formaldehyde 367 50.2 428 154 18.6 Phenol formaldehyde 803 53.9 0.0 143 31.7 Resorcinol formaldehyde 703 47.2 0.0 250 26.4 Phenol resorcinol formaldehyde 750 50.3 0.0 200 28.8

The determination of the molecular formula for each resin type shown in Table 4-1 is documented in Annex F.

4.1.4. Calculation of carbon sequestration, density and moisture content

During tree growth, carbon dioxide from the air is sequestered as biogenic carbon within the tree. Various forestry steps (thinning, felling, etc.) and also natural processes (e.g. fire) release this sequestered carbon back to the air, while some biogenic carbon remains sequestered in the wood products leaving the forest. In line with the PCR, embodied carbon is treated as an inherent property of the wood (IEPDS, 2017, section 7.7) and the removal of carbon dioxide included within the wood must equal the carbon contained in the finished product in line with ISO/TS 14067:2013 (ISO/TS 14067:2013, section 6.4.9.6) and the calculation specified in EN 16449:2014 (described later in this section). Including sequestered carbon is appropriate, provided that it is reported separately in the EPD (ISO/TS 14067:2013, section 6.4.9.6) and that forests are sustainably managed (EN 16485:2014, section 6.3.4.2). There is further discussion of sustainable management of Australian forests in sections 4.3.7 for softwood forestry. This section describes the procedure for calculation of the biogenic carbon sequestered in the wood, a parameter which is affected by the density and the moisture content of the wood. The moisture content (MC) of wood can be stated relative to either its oven dry mass (i.e. the mass of wood only, after all water is evaporated) or the total mass of the wood including water. The former is known as the oven-dry basis (dry basis) and the latter is known as the wet or original basis (wet basis) (Briggs, 1994). These moisture contents are calculated as follows (Briggs, 1994):

• % MCOD = 100 * mass of water / oven-dry mass

• % MCW = 100 * mass of water / original mass including water The dry basis is more common in the solid wood industry, while the wet basis is more common in the wood fuels industry. Therefore, wood moisture contents used throughout this study refer to dry basis. Foresters commonly use volume rather than mass when dealing with cut logs as, although mass can change dramatically as logs lose moisture, volume remains relatively stable until the logs are sawn and mechanically dried. This is due to the fact that until the moisture content falls below the wood’s fibre saturation point (FSP; typically around 30% moisture content relative to oven-dry weight), free water is found in cell cavities rather than in cell walls. Below the FSP, water is lost from the cell walls, causing the fibres to contract and the wood to shrink. This is illustrated below in Figure 4-1: shrinkage approaches zero after the moisture content crosses the FSP.

LCI and EPDs for Australian Wood Products 35 of 159

Figure 4-1 Shrinkage versus moisture content (Timber Queensland, 2014)

The CO2 sequestered per cubic metre of wood was calculated using the formula specified in European standard EN 16449 (EN 16449:2014): 44 휌 × 푉 푃 = × 푐푓 × 휔 휔 퐶푂2 휔 12 1 + 100 Where:

푃퐶푂2 is the biogenic carbon sequestered in the wood that can be oxidised to a carbon dioxide emission to air 44 is the molecular weight of carbon dioxide divided by the atomic weight of carbon 12 푐푓 is the carbon fraction of oven dry mass of woody biomass (0.5 is the default value) 휔 is the moisture content of the product on a dry basis, e.g. 12 (%) 3 휌휔 is the density of woody biomass at that moisture content (kg/m ) 3 푉휔 is the volume of the solid wood product at that moisture content (m ) In Australia, 50±2% of the dry weight of the wood of native species and non-native Pinus radiata is carbon (Gifford, 2000). As a result, the default value of 0.5 is applied for cf.

The basic density (휌휔, the mass of wood at 0% MC contained in a cubic metre of green wood) was calculated as a volume-weighted average of participating sites.

4.2. Background data

4.2.1. Fuels and energy

Data for fuel inputs and electricity grid mixes were obtained from the GaBi Database 2017. National and regional Australian average data sets were used as far as possible, global or European average data were used where no suitable Australian datasets were available. Electricity grid mixes were modelled as a production-weighted average of the individual state grid mixes. Production data were based on ABARES production volumes by state for sawn softwood and sawn hardwood, where data were available on a state level. For particleboard, MDF, plywood and glulam, where ABARES production data were not disaggregated by state, the electricity mix

LCI and EPDs for Australian Wood Products 36 of 159

was based on electricity consumption of the sites involved in this study for each product category and the state where each site is located. The following tables (Table 4-2 to Table 4-5) show the most relevant LCI datasets used in modelling the product systems. Documentation for all non-project-specific datasets except diesel combustion in forest engines can be found in the GaBi software documentation (thinkstep, 2017). Unit process data for diesel consumption in forest engines is included in Table 4-3 below and is used for all flows written as “Diesel consumption in forest engine” later in this section.

Table 4-2 Key energy datasets used in inventory analysis

Material/process Dataset name Source Year Region Electricity Electricity grid mix thinkstep 2013 AU Electricity Electricity grid mix (NSW) thinkstep 2016 AU Electricity Electricity grid mix (QLD) thinkstep 2016 AU Electricity Electricity grid mix (SA) thinkstep 2016 AU Electricity Electricity grid mix (TAS) thinkstep 2016 AU Electricity Electricity grid mix (VIC) thinkstep 2016 AU Electricity Electricity grid mix (WA) thinkstep 2016 AU Thermal energy Thermal energy from natural gas thinkstep 2013 AU Thermal energy Thermal energy from biomass (solid) thinkstep 2013 AU Thermal energy Thermal energy from LPG thinkstep 2013 EU-28 Diesel combustion Diesel consumption in forest engine thinkstep 2013 GLO Diesel fuel Diesel mix at filling station thinkstep 2013 AU LPG (transport) Distance travelled by liquified petroleum thinkstep 2012 EU-27 gas (LPG) car <1,4 l (direct)

Table 4-3 Unit process data for “Diesel consumption in forest engine”

Type Flow Unit Amount Inputs Diesel [Refinery products] kg 1 Outputs Ammonia [Inorganic emissions to air] kg 2.02E-05 Benzene [Group NMVOC to air] kg 3.80E-05 Benzo{a}pyrene [Group PAH to air] kg 7.69E-09 Carbon dioxide [Inorganic emissions to air] kg 3.175 Carbon monoxide [Inorganic emissions to air] kg 0.00804 Dust (PM2.5) [Particles to air] kg 0.002964 Formaldehyde (methanal) [Group NMVOC to air] kg 0.000161 Hydrogen chloride [Inorganic emissions to air] kg 9.88E-07 Methane [Organic emissions to air (group VOC)] kg 4.67E-05 Nitrogen oxides [Inorganic emissions to air] kg 0.039769 Nitrous oxide (laughing gas) [Inorganic emissions to air] kg 0.000331 NMVOC (unspecified) [Group NMVOC to air] kg 0.001899 Polychlorinated dibenzo-p-dioxins (2,3,7,8 - TCDD) kg 6.01E-14 [Halogenated organic emissions to air] Sulphur dioxide [Inorganic emissions to air] kg 0.000902

LCI and EPDs for Australian Wood Products 37 of 159

Table 4-4 provides an overview of fuel properties used to convert fuel inputs across all product categories.

Table 4-4 Fuel properties

Property Value Unit Source Diesel density 0.838 kg/l thinkstep Petrol density 0.735 kg/l thinkstep LPG density 0.529 kg/l thinkstep Diesel NCV 36.0 MJ/l thinkstep Petrol NCV 32.3 MJ/l thinkstep LPG NCV 24.4 MJ/l thinkstep

4.2.2. Raw materials and processes

Data for up- and downstream raw materials and unit processes were obtained from the GaBi Database 2017. Australian data were used in preference to European data; however, appropriate Australian data were often not available in GaBi. Table 4-5 shows the most relevant LCI datasets used in modelling the product systems.

Table 4-5 Key material datasets used in inventory analysis

Material/Process Dataset name Source Year Region Municipal water Tap water thinkstep 2016 EU-28 Lubricating oil Lubricants at refinery thinkstep 2013 AU Steel part Steel hot rolled coil (EN 15804 thinkstep 2016 DE A1-A3) Phenol Formaldehyde (PF) Phenol formaldehyde resin thinkstep 2016 DE resin Urea Formaldehyde (UF) resin Adhesive system Urea thinkstep 2016 DE

Formaldehyde (UF) (35% H2O) Melamine-Urea-Formaldehyde Adhesive system MUF (35%M thinkstep 2016 DE

(MUF) resin i.T., 35% H2O) Phenolic overlay for plywood Phenolic resin (45% thinkstep 2016 DE concentration) Acrylic putty Acrylate resin (solvent- thinkstep 2016 DE systems) Wheat (for glue mixes) Wheat flour thinkstep 2016 DE Paper (primary) Kraftliner (2015) thinkstep 2016 EU-28 /FEFCO Sodium hydroxide (100 % Sodium hydroxide from thinkstep 2016 AU caustic soda) chlorine-alkali-electrolysis (diaphragm) Gravel for forestry roads Gravel (Granulation 2/32) thinkstep 2016 DE

The materials/processes in Table 4-5 with a non-Australian geographical relevance were used to approximate the respective materials/processes used/applied in Australia. This was necessary

LCI and EPDs for Australian Wood Products 38 of 159

since no equivalent Australian datasets were available in either GaBi or the Australian LCI database.

4.2.3. Preservative treatments

Background data used for modelling preservative treatments are provided in Table 4-6.

Table 4-6 Processes and flows used for modelling preservative treatments

Material/Process Dataset name Source Year Region Ammonia Ammonia (liquid, agriculture) thinkstep 2016 DE Aromatic hydrocarbon (proxy) Benzene mix thinkstep 2016 EU-28 Bifenthrin (proxy) Insecticide unspecific thinkstep 2016 DE (Pyrethroid) Boric acid Boric acid (estimation) thinkstep 2016 EU-28 Chromated copper arsenate Chromated copper arsenate thinkstep 2016 US (CCA) Copper acetate (proxy) Acetic acid from methanol thinkstep 2016 EU-28 Copper (II) carbonate Cupric carbonate basic by- thinkstep 2016 EU-28 hydroxide product sodium sulphate (estimation) Copper ammonium carbonate Cupric carbonate basic by- thinkstep 2016 EU-28 (proxy) product sodium sulphate (estimation) DDAC (Didecyl dimethyl Ammonium chloride (Salmiac, thinkstep 2016 DE ammonium chloride) (proxy) Solvay-process) Didecyl tertiary amine (proxy) Dimethylamine (Production thinkstep 2016 EU-28 from Alcohols) Dispersant Dispersing agent (unspecific) thinkstep 2016 GLO Dye Dyes thinkstep 2016 DE Methanol Methanol mix thinkstep 2016 EU-28 Monoethanolamine Monoethanolamine (MEA) thinkstep 2016 DE N,N-Didecyl-N,N- Ammonium hydrogen thinkstep 2016 EU-28 Dimethylammonium carbonate carbonate (proxy) N-Methyle-2-pyrolidone n-Methylpyrolidone mix (NMP) thinkstep 2016 DE Permethrin (proxy) Insecticide unspecific thinkstep 2016 DE (Pyrethroid) Propiconazole (proxy) Fungicide unspecific thinkstep 2016 DE Propylene glycol (Ppropane- Propylene glycol thinkstep 2016 EU-28 1,2-Diol) Sodium nitrite Sodium nitrate thinkstep 2016 EU-28 Sodium tetraborate Borax pentahydrate thinkstep 2016 EU-28 pentahydrate Tebuconazole (proxy) Fungicide unspecific thinkstep 2016 DE Water Tap water thinkstep 2016 EU-28 Emissions to air VOC to air Organic emissions to air - thinkstep - Benzene

LCI and EPDs for Australian Wood Products 39 of 159

Material/Process Dataset name Source Year Region Water vapour Inorganic emissions to air - thinkstep - water vapour Energy Electricity Electricity grid mix thinkstep 2013 AU Diesel Diesel at refinery thinkstep 2013 AU

4.2.4. Transportation

Average transportation distances and modes of transport are included for the transport of the raw materials to production and assembly facilities. The GaBi database for transportation vehicles and fuels was used to model transportation. These are representative datasets for a wide range of transport options representing different vehicle types, sizes and technologies (e.g. different Euro-rated engines for trucks). These datasets are parameterised and have been adjusted to fit the specific vehicle loading efficiencies, carrying capacities, transport distances, etc. wherever transport processes are required. The most environmentally relevant truck dataset used within this study is for transport from forest to sawmill. For this, the GaBi dataset for a 34-40 tonne gross weight (27 tonne payload) truck and trailer unit with average performance across the Euro 0 to Euro 5 emissions bands has been used for both softwood and hardwood.

4.2.5. Elemental composition and calorific value of materials

The elemental compositions of the different materials are listed in Table 4-7. For the elemental composition of the resins see Table 4-1.

Table 4-7 Elemental composition of materials

Component C H N S O Cl Ash Source Softwood (dry) 50 6 0 0 44 0 0 (EN 16449:2014) Hardwood (dry) 50 6 0 0 44 0 0 (EN 16449:2014) Paraffin wax 85 15 0 0 0 0 0 (Energy Information Administration, 2002) Lamination paper 46 6 0 0 39 0 8 Mean of category "paper" from Phyllis2 (dry) database (ECN, 2017).

Tongue 86 14 0 0 0 0 0 Based on monomer C3H6 (polypropylene)

Ammonium 0 6 21 24 48 0 0 Based on chemical formula (NH4)2SO4 sulphate

Water 0 11 0 0 89 0 0 Based on chemical formula H2O

The heating values of the different materials are shown in Table 4-8. For the heating values of the resins see Table 4-1.

LCI and EPDs for Australian Wood Products 40 of 159

Table 4-8 Heating values of materials

Component GCV [MJ/kg] NCV [MJ/kg] Source (elemental comp.)

Softwood (dry) 20.53 Mean of category "wood, pine" from Phyllis2 database (ECN, 2017). Dry value, calculated from value 20.67 MJ/kg and an ash content of 0.7%. Hardwood (dry) 19.99 Mean of category "wood, eucalyptus" from Phyllis2 database (ECN, 2017). Dry value, calculated from value 20.31 and an ash content of 1.57% Paraffin wax 42.32 40.20 (IPCC, 2016)

Lamination 19.32 Mean of category "paper" from paper (dry) Phyllis2 database (ECN, 2017). Dry value, calculated from value 20.96 MJ/kg and an ash content of 7.81%. Tongue 45.8 42.66 (Walters, n.d.) (polypropylene)

4.3. Softwood forestry

4.3.1. Overview

This section covers softwood plantation forestry in Australia. The dominant softwood species used to produce timber products in Australia is Pinus radiata (radiata pine). Other softwood species used are Araucaria cunninghami (hoop pine), Pinus pinaster (maritime pine) and the southern pines: Pinus elliottii (slash pine), Pinus caribaea (Caribbean pine) and hybrids thereof. The plantation managers included within the survey were:

• Forest Products Commission Western Australia • Forestry Corporation of New South Wales • Hancock Queensland Plantations (HQP) • Hancock Victorian Plantations (HVP) • OneFortyOne Plantations (South Australia) • Timberlands Pacific, Penola Plantations (South Australia and Victoria) • Timberlands Pacific, Taswood Estate (Tasmania) Six of these seven plantation managers provided enough data to be included within the LCI. These six plantations collectively produced 9,966,729 m3 of softwood logs in 2015/16 (reference year for the study), equating to 61% of total Australian production of approximately 16,346,000 m3 (based on the 2015/16 total from ABARES (2017). All averaged data displayed in the following sections are calculated as production-weighted averages of the six included plantations. Outputs of the softwood forestry process have been categorised to match the Australian Pine Log Price Index (KPMG, 2016). However, “intermediate” and “medium” sawlogs have both been considered within a single “medium log” category for the purpose of this study. The average proportions of forest products from plantations included in this study is displayed in Table 4-9.

LCI and EPDs for Australian Wood Products 41 of 159

Baseline (2006) refers to data collected by the Commonwealth Scientific and Industrial Research Organisation (CSIRO) that inventoried data from the 2005/06 Australian financial year (1st July 2005 to 30th June 2006) (CSIRO, 2009a), as updated in v1 of this background report (2015).

Table 4-9 Softwood forestry products by type

Type Flow Volume Volume Baseline DQI* (1000 m3) (%) (2006) Outputs Large sawlogs (>43.9 cm) 604 6.06% 5.93% M Medium sawlogs (24-43.9 cm) 3660 36.74% 42.20% M Small sawlogs (<23.9 cm) 1270 12.71% 18.70% M Preservation logs 204 2.05% - M Pulplogs 4070 40.87% 28.44% M Salvage logs/other 157 1.58% 4.72% M * measured (M) / calculated (C) / estimated (E) / literature (L)

The range of log products produced by plantations participating in this study generally shows a similar distribution to that observed in the baseline study. However, the current study shows an increase in pulplogs harvested relative to sawlogs. Good agreement between surveyed plantations is seen with regard to production of pulplogs and medium sawlogs as shown in Annex A. Some variation exists between sites for large sawlogs, small sawlogs and preservation logs. High variation exists between sites for production of salvage logs as this is a lower value product and recovery varies greatly between sites, depending on cost and sale price.

4.3.2. Production process

The production process of softwood sawlogs and other softwood plantation products is illustrated in Figure 4-2.

CO2 sequestration in wood

System boundary Burning of residues Seeds, energy, water, Stand establishment fertiliser, herbicides Seedling production and planting Fuel reduction burning

Liquid fuels, water & Forest management Monitoring, fire prevention and control, road Primary gravel building energy, water, Emissions raw Thinning material Liquid fuels Diesel consumption of thinning vehicles and appliances

Clearfelling Liquid fuels Diesel consumption of harvesting vehicles and appliances

Saw logs Pulp logs

Salvage logs & Preservation logs Other Figure 4-2 Process flow diagram of softwood forestry

LCI and EPDs for Australian Wood Products 42 of 159

The softwood forestry process has been broken down into four unit processes: establishment, management, thinning and clearfelling. Establishment covers the clearing and preparation of post-harvest land by methods such as burning of harvest residues and mechanical land preparation. Establishment also includes the production of seeds within a nursery including: water use, fertiliser and chemical application for seedling raising and activities associated with the planting of seedlings on cleared land. Forest management covers monitoring of the plantation area, evapotranspiration, construction and maintenance of forest infrastructure such as roads, chemical application such as fertilisers, pesticides and herbicides and fire prevention and control activities including undertaking fuel reduction burns. Harvesting is the final stage in the cycle. This stage has been divided into thinning and clearfelling. The thinning process includes machine and fuel usage for one or more thinning operations within the managed area, where trees are de-limbed, or selectively harvested from the plantation. Clearfelling includes all machine and fuel used during end of rotation harvest. The area of plantation managed and established per cubic metre of wood product harvested is displayed in Table 4-10. The proportion of products harvested during thinning and clearfelling is also displayed.

Table 4-10 Softwood forestry production area and output – per m3 of wood harvested

Type Flow Unit Total Average Baseline DQI* (2006) Inputs Total area established ha 19,800 0.00199 0.00162 M Total area managed ha 599,000 0.0601 0.0545 M Volume clearfelled m3 7,600,000 0.763 0.577 M Volume thinned m3 2,370,000 0.237 0.423 M Outputs Total wood harvested m3 9,970,000 1 1 M * measured (M) / calculated (C) / estimated (E) / literature (L)

The current study shows that fewer wood products are harvested from softwood plantations per ha of land compared to the baseline study. More product output is also reported to come from clearfelling operations than thinning operations. Some variability in data does exist for the surveyed sites as shown in Annex A.

4.3.3. Establishment

Establishment operations include slash burning, site preparation and planting. Following harvesting, large woody debris is pushed into heaps with bulldozers and burnt. Remaining smaller material is then broken down using chopper rollers which are pulled behind bulldozers or skidders. The site may be ripped and mounded prior to planting seedlings or cuttings by hand. Establishment operations vary between different regions and forest owners depending on soil and terrain, forest silviculture and markets for products. In particular, the amount of harvest residue burnt is highly variable and depends on the market for low value wood products, stand silviculture and terrain. Inventory data for plantation establishment are given in Table 4-11.

LCI and EPDs for Australian Wood Products 43 of 159

Table 4-11 Plantation establishment inventory – per ha established

Type Flow Unit Average Baseline DQI* (2006)

Inputs Seedlings established pcs. 1.26E+03 1.27E+03 M Diesel kg 1.15E+02 9.10E+01 C Residues burnt (bone dry content) kg 5.85E+03 7.68E+03 E Petrol vehicle travel km 2.45E+00 8.89E+01 C

Seed Production Seedlings raised/produced pcs. 1.26E+03 1.27E+03 M Electricity kWh 1.71E+01 2.24E+01 C Diesel kg 2.89E+00 4.32E+00 C Water kl 7.47E+01 7.16E+01 C Petrol vehicle travel km 9.03E-01 1.19E+00 C Magnesium sulphate kg 1.90E-01 1.90E-01 C Potassium chloride (agrarian, 60% K2O) kg 9.89E-04 9.92E-04 C Triple superphosphate (agrarian, 45% 6.84E-01 6.86E-01 C kg P2O5) Urea (agrarian) kg 1.98E-01 1.98E-01 C Outputs Land established ha 1.00E+00 1.00E+00 M * measured (M) / calculated (C) / estimated (E) / literature (L)

The number of seeds established per ha of land has remained similar in comparison to the baseline study. However, the data from the current study shows an increase in diesel use and a decrease in residues burnt during establishment activities. Both of which have significant impacts on forestry emissions as shown by the hotspot analysis in Annex D. Large variability exists between sites for the amount of forest residues burnt as shown in Annex A. This is dependent on to the amount of residues produced, available markets and different approaches to the treatment of post-harvest residues by different plantations. Some plantations burn a higher proportion of residues, whilst others may prepare the land mechanically depending on the size and type of residues produced.

The production of seedlings in a nursery remains very similar to the baseline study. Electricity and diesel consumed per seedling has decreased relative to the baseline study. Gaps in seedling production data have been filled with data from the baseline study, as not all sites could provide sufficient data to construct a reliable average. The hotspot analysis showed seedling production to be of minor significance. Most other inputs therefore remain very similar to the baseline study.

Forest residue burning

Burning of forest residues releases substantial amounts of non-CO2 greenhouse gases (CH4, N2O and NOx) in addition to biogenic CO2. Since the forests and the carbon they contain are assumed to be in a steady state (i.e. no net change in above- or below-ground carbon over time) only the non-

CO2 GHGs need be accounted for. The literature source for emissions associated with the burning of forest residues has been updated in this study to a more recent source (Commonwealth of Australia, 2016). Emissions are taken for the burning of coarse residues in temperate forests. Notable differences to the emissions data associated with burning of forest residues used in the baseline study (IPCC, 2006) are: an increase in methane emissions and the addition of non-

LCI and EPDs for Australian Wood Products 44 of 159

methane volatile organic compounds (NMVOC). A decrease in nitrous oxide emissions is also observed as shown in Table 4-12. No specific information on the amount of harvest residues remaining after burning for softwood plantations was available during the time of this study. The amount of burnt material has been estimated as 80% by mass, based on the rate from the original CSIRO module B report (CSIRO, 2009b). The moisture content of wood burnt in the forest has been taken to be 112% on a dry basis, equal to the average log moisture content received by the surveyed sawmills. In addition to air emissions, burning wood also produces small quantities of ash. Evaluating the average of "wood, pine" the Phyllis2 database (ECN, 2017) delivers an ash content of 0.7% (relative to dry matter). Given its small amount, it is assumed to remain on site and be assimilated over time and has not been considered further when modelling this process. Emissions from burning of forest residues are given in Table 4-12.

Table 4-12 Forest residue burning – per t of bone dry residues

Type Flow Unit Average Baseline DQI* (2006) Inputs Forest residues t 1 1 L Outputs Carbon monoxide kg 106 107 L Methane (biotic) kg 8.4 4.7 L NMVOC kg 12.8 - L Nitrogen oxides kg 2.71 3 L Nitrous oxide kg 0.058 0.26 L * measured (M) / calculated (C) / estimated (E) / literature (L)

4.3.4. Forest management

Forest management includes monitoring, fire control and prevention activities, road construction and maintenance, and fertiliser additions. Inputs for each activity in softwood plantations are expressed in terms of total plantation area. Water use for forests is described further in section 4.3.7. The process inventory for forest management is given in Table 4-13.

Table 4-13 Plantation management – per ha of managed land

Type Flow Unit Average Baseline DQI* (2006) Inputs Roads constructed/maintained m 1.35E+01 1.15E+01 M Diesel kg 2.21E-02 2.27E-02 C Petrol vehicle travel km 4.61E+00 7.19E+00 C

Roading Bitumen kg 1.26E-01 1.38E+00 E Gravel kg 3.92E+02 4.30E+03 M Diesel kg 2.07E+00 2.33E+00 C

Chemical application Aircraft use h 1.11E-03 1.11E-03 E Helicopter use h 1.24E-03 1.24E-03 E

LCI and EPDs for Australian Wood Products 45 of 159

Type Flow Unit Average Baseline DQI* (2006) Diesel kg 2.14E-01 1.27E+00 C Magnesium sulphate kg 2.08E-03 2.90E-03 C Potassium chloride (agrarian, 60% K2O) kg 2.75E+00 3.85E+00 C Triple superphosphate (agrarian, 45% 2.34E+01 3.27E+01 C kg P2O5) Urea (agrarian) kg 7.13E+00 9.63E+00 C Diammonium phosphate (DAP) kg 1.38E+00 - C NPK (15N-15P2O5-15K2O) kg 4.65E-01 - C NPK (60% phosphate-water-solution) kg 5.36E-01 - C Monoammonium phosphate (MAP) kg 1.36E+00 - C

Fire prevention Aircraft use h 2.59E-04 2.59E-04 E Helicopter use h 6.44E-04 6.44E-04 E Diesel kg 5.57E-01 5.62E-01 C Debris burnt kg 2.52E+02 2.03E+02 E Petrol vehicle travel km 1.40E+00 1.40E+00 E Outputs Land managed ha 1.00E+00 1.00E+00 M * measured (M) / calculated (C) / estimated (E) / literature (L)

A significant difference relative to the baseline study comes from an increase in the average amount of debris burnt for fire prevention. This has been identified as contributing significantly across a range of impact categories for softwood plantation forestry in the hotspot analysis (see Annex D). Large variability in the amount of debris burnt for fire prevention activities is observed between sites (see Annex A), as some sites have indicated that they do not undertake any burning for fire prevention activities, whilst the amount of debris burnt by those who do undertake burning activities varies greatly. The number of roads constructed or maintained has increased relative to the baseline study, however the amount of gravel and bitumen used in road construction has decreased significantly. The amount of gravel used also varies over a wide range (see Annex A) as this is dependent on the type of new roads built and length needing maintenance due to weather and usage patterns. Data for chemical application and aircraft use were not provided by many of the participants, therefore the baseline data were used to fill gaps for many inputs.

4.3.5. Harvesting

Table 4-14 shows fuel inputs required for thinning – the partial harvest of stands in which a percentage of trees are removed. Normal thinning operations (known as thinning from below) aim to remove smaller and poorer formed trees and leave larger, higher quality trees to grow faster and produce higher value or more uniform products. In some cases, larger trees may be removed (known as thinning from above) in order to increase product uniformity and meet the requirements of processing facilities. Although different products are produced from different harvest operations the chief aim of thinning is not to produce low value co-products but to increase the amount and value of the final product at clearfelling. Thus, rather than attempting to allocate inputs and emissions from separate thinning operations to the products produced during each operation, it is more realistic to allocate inputs

LCI and EPDs for Australian Wood Products 46 of 159

from the thinning and clearfelling across the combined products from those operations. The fuel requirements associated with clearfelling are given in Table 4-15. Diesel consumption from thinning and clearfelling was only provided by half of the participating plantations due to harvesting operations often being carried out by contractors. In cases where the diesel usage was not able to be provided, a volume-weighted average of the sites which were able to provide data was applied.

Table 4-14 Thinning – per m3 of thinning output

Type Flow Unit Average Baseline DQI* (2006) Inputs Diesel kg 1.07 2.17 C Outputs Thinned wood m3 1 1 M * measured (M) / calculated (C) / estimated (E) / literature (L)

Table 4-15 Clearfelling – per m3 of clearfelled wood

Type Flow Unit Average Baseline DQI* (2006) Inputs Diesel kg 1.12 1.36 C Outputs Clearfelled wood m3 1 1 M * measured (M) / calculated (C) / estimated (E) / literature (L)

Diesel consumption during thinning has decreased significantly relative to the baseline study as seen in Table 4-14. Diesel consumption per cubic metre of wood harvested has also decreased for clearfelling relative to the baseline study as seen in Table 4-15. These changes result in reductions in impacts from forestry. There is some variability between sites as shown in Annex A. Thinning diesel consumption reported by all sites was greater than the diesel consumption for clearfelling per cubic metre of wood harvested. However, since the average was calculated as a weighted average by volume of thinned or clearfelled respectively, the thinning diesel consumption per cubic metre thinned is lower than for clearfelling because the site contributing the largest share of thinning uses the least diesel per cubic metre of thinned wood.

4.3.6. Allocation

Impacts were allocated to co-products by economic value (Table 4-16). Economic allocation was applied because the difference in the value of co-products is typically large (>25% in many cases, in line with IEPDS 2017, section 7.7). Since sequestered carbon and primary energy are physical properties of wood, they have been allocated on a mass basis to the different co-products. As inputs and emissions were allocated to products on an economic basis, it was assumed that all forest management activities (including thinning) were aimed at producing a final wood product (sawlog or pulplog etc.). It is assumed that the only product of the plantation process is wood, which is harvested and sold. Thus, it did not make sense to try to allocate inputs or emissions from different processes during the rotation to products harvested at different times in the rotation, because they also contributed to product harvest at the end of the rotation. Instead, all inputs and emissions up to and including those from felling were allocated to the total wood products harvested throughout the rotation. Processes that occurred after final harvest (e.g. haulage) were allocated to the particular product to which they applied.

LCI and EPDs for Australian Wood Products 47 of 159

All prices are taken as a five year volume-weighted average from the financial year 1 January 2012 to 31 December 2016 based on weighted average log prices published by KPMG in its Australian Pine Log Price Index (KPMG, 2016). The category “medium sawlog” is taken as a volume-weighted average of “intermediate sawlog” and “medium sawlog” categories given in the KPMG report. These prices are based on stumpage values (i.e. wood value prior to harvest and haulage). Thus, they are lower than the value at mill door. However, it was considered reasonable to assume that the relative prices of the various products would remain roughly the same.

Table 4-16 Forest product allocation factors

Current study Price ($/m3) Volume Value Allocation (1000 m3) (M$) (%) Large sawlogs (>43.9 cm) 82.32 604 49.7 12.6% Medium sawlogs (24-43.9 cm) 61.43 3660 225 56.9% Small sawlogs (<23.9 cm) 36.63 1270 46.4 11.7% Preservation log 28.12 204 5.74 1.45% Pulplog 15.70 4070 64.0 16.2% Salvage log 28.43 157 4.48 1.13% Total 39.65 9970 395 100%

4.3.7. Assumptions

Assumptions made during the construction of the softwood forestry LCI are listed within the following section.

Steady state forestry

A key assumption is that all forests included in the study were in a steady state with regard to management inputs and wood products produced as well as for above- and below-ground carbon stocks. Softwood plantations were assumed to be sustainably managed with no reduction in carbon stocks in soil, litter or live biomass over time. To meet the requirements of a steady state approach in softwood plantations, all newly planted land not previously managed for wood production, and all activities relating to the establishment operations on this land, are excluded from the module boundary (Figure 4-3). Newly planted (1st rotation) land not previously managed for wood production was excluded from the LCI. However, 1st rotation forest that had previously been established was included.

LCI and EPDs for Australian Wood Products 48 of 159

Figure 4-3 System boundary of the plantation softwood forest production module (red line) (CSIRO, 2009a)

The assumption of steady-state forestry is important for the global warming potential impact category due to the effect of land use change on stored carbon. Furthermore, some standards (e.g. EN 16485, which complements EN 15804 for wood products) only allow sequestered carbon to be counted within the LCA system boundary if wood is sourced from sustainably managed forests. The approach used in this study is that all forestry operations (tree growth, thinning, clear felling, etc.) are considered carbon neutral (i.e. CO2 uptake = CO2 released) except where this is known not to be the case. The two cases where there is considered to be a change in the CO2 balance are: 1. Carbon leaving the forest embodied in a product (see section 4.1.3); and 2. Forest residue burning operations that result in carbon-containing emissions to air that are not carbon dioxide (see section 4.3.3). Currently, all major Australian publicly owned native forest managed for timber production, as well as almost all of the softwood and hardwood plantations managed for timber to produce sawn hardwood, sawn softwood, particleboard, MDF, plywood and glulam are independently certified to one or both of the internationally accepted forest management standards. These are the Australian Standard for Sustainable Forest Management (AS 4708), which is recognised under the Programme for the Endorsement of Forest Certification (PEFC), and/or one of the Forest Stewardship Council’s (FSC) interim forest management standards. It is therefore appropriate to include biogenic CO2 sequestration in this EPD in line with EN 16485 (section 6.3.4.2). For more information on certification by forest owner or manager please see http://www.forestrystandard.org.au/find-certified/certified-forest-managers and http://info.fsc.org/certificate.php.

Biodiversity

Like other land uses, forestry operations for timber and wood production can have both positive and negative effects on biodiversity. However, as biodiversity varies considerably by region and as data are often limited, assessing potential biodiversity impacts within LCA is challenging. An Australian study (Turner, et al., 2014) demonstrated a new method – BioImpact – to discern the biodiversity impacts of different land uses. A trial of this method was conducted using case studies in three different regions and four production systems in New South Wales: native hardwood forestry, plantation softwood forestry, mixed cropping and rangeland grazing. The results showed

LCI and EPDs for Australian Wood Products 49 of 159

that the biodiversity impacts of native hardwood production in the region studied were significantly lower than the land uses in the other regions. The management of planted softwood forests resulted in similar biodiversity impacts to those of cropping/grazing systems.

Water use in forests

PCR 2012:01 (Section 16.1) states that all water loss from a drainage basin is considered consumption, including any net loss of rain water. According to the PCR, net loss should be interpreted as any additional water loss beyond what would occur in the original, natural system. For plantation softwood forestry, the natural system might be a native forest or a grassland (Quinteiro, et al., 2015). The initial versions of the EPD (v1.0 and v1.1) included estimated losses of rain water in the main results tables, labelled as green water consumption. These values have not been updated and are now reported separately rather than in the main results tables to reflect their uncertainty. At the time of writing, there is no internationally agreed method for calculating green water consumption due to evapotranspiration relative to a hypothetical natural state (Manzardo, et al., 2016). As such, different calculation methods may yield significantly different results, introducing a high level of uncertainty. Water consumption values also have no direct correlation to water stress or a water footprint. The methodology and assumptions made for water use in the original study are as follows: Total water use by both softwood plantations and hardwood native forests was estimated from a relationship between total rainfall and evapotranspiration for forests (Zhang, et al., 2001), and allocated to the harvested wood products. These estimates were compared with actual measurements of rainfall and catchment water yield for forested catchments (Benyon, et al., 2006; Bubb & Croton, 2002; Brown, et al., 2005; Cornish & Vertessy, 2001; Vertessy, et al., 2001). Even for bare soil, some moisture is lost through evaporation. Therefore, water use for wood production from plantations was expressed relative to water use by a base case land use (pasture). This base case water use was estimated using the Zhang relationship for pasture and average rainfall figures for the forest region from the Bureau of Meteorology. In contrast, water use for native forests available for wood production was expressed relative to that for forests reserved for conservation purposes, using results from previous studies showing the change in evapotranspiration and water yields for stands of different ages. These data indicated that average evapotranspiration for a forest harvested once every 60-120 years was likely to be 10-20% greater than that for a stand burnt and regenerated once every 240 years (the nominal frequency of wildfire in ash-type forests prior to European settlement) (Moran & O'Shaughnessy, 1984; Vertessy, et al., 2001). Thus, the additional water use by a stand harvested for wood production was assumed to be 15% of that for a stand reserved for conservation purposes (or 13% of the total water use by the harvested stand estimated from the Zhang relationship). The approach taken by CSIRO appears to be in line with PCR 2012:01 (IEPDS, 2017, Annex D): Net use of fresh water shall be calculated and reported. The term ‘net’ (as opposed to gross) in relation to fresh water use, is used to show that:

• Use of water that it is not consumed should not be considered within the indicator (e.g. water used for river transport, used to power hydroelectric turbines or used as coolant and returned to the original source). • Water that would have been lost from the original, natural system is not considered within the losses from the studied technical system (e.g. from evaporation of rainwater or from a body or water).

LCI and EPDs for Australian Wood Products 50 of 159

• Evaporated fresh water is considered consumption unless it is demonstrated otherwise. • For each process, the water flows should be identified in terms of volumes extracted, volumes discharged and the source or destination, e.g. surface water, ground water, seawater. CSIRO’s water use calculations are presented in Table 4-17. However, it is important to note that CSIRO itself stated that, “We stress that these calculations are very uncertain, given the major assumptions involved.” Given both the uncertainty of these calculations and the potential for them to be misinterpreted (e.g. by a reader taking total water consumption and comparing it to that of another product, without considering local water stress), these results are reported in their own “Water consumption” section of each EPD (see Annex G) rather than in the main results table.

Table 4-17: Estimated annual evapo-transpiration and water use per unit wood produced for each case study region (weighted for wood production for water use and area for rainfall and ET) (CSIRO, 2009a)

Case study Forest Rainfall Evapo-transpiration Water use

Annual wood Avg Forest1 Reference2 Difference Absolute3 Relative4 Area production

000 ha 000 m3 m3 ha-1 mm mm mm mm ML m-3 ML m-3 Softwood plantation Green Triangle 176 3,534 20.1 738 665 534 131 0.33 0.07 Tumut NSW 90 1,489 16.5 1125 911 666 245 0.55 0.15 SE Qld 87 1,124 12.9 1314 1012 716 296 0.79 0.23 WA 85 857 10.1 866 752 582 170 0.74 0.17 Wtd. Average 16.6 957 802 607 195 0.49 0.12 Hardwood native forest Vic CH 84 758 9.0 1421 1053 916 137 1.2 0.15 Tasmania 182 867 4.8 1207 939 817 123 2.0 0.26 NE NSW 224 229 1.0 1263 978 851 128 9.6 1.25 Wtd. Average 3.7 1270 977 849 127 2.7 0.35

1Estimated from average annual rainfall and a relationship between rainfall and ET for forests in equilibrium (Zhang 2001).

2Reference land use assumed to be pasture for plantations and unlogged native forest for native forest.

3Total evapo-transpiration from forests divided by total wood extracted.

4Difference between ET for forests relative to that for reference land-use divided by wood production.

Seedling production

Electricity supply has been based on the Australian average supply mix because production represents national average production (i.e. softwood plantations are located in all states excluding NT). All input water to seedling production is assumed to evaporate, since it is not possible to determine the fraction of water sequestered within the seedlings.

LCI and EPDs for Australian Wood Products 51 of 159

Chemical application

Gaps in chemical application data have been filled using data taken from the baseline study. Such emissions to air and water due to fertiliser assumes that the inputs of fertiliser were correct. Only N- P-K fertilisers have been included. Emissions of fertilisers are based on published emissions to air and water.

- + 3- + • Emissions of fertiliser to water can include products of N (NO3 , NH4 ), P (PO4 ), K (K ) and 2- S (SO4 ) fertilisers. The following emissions of fertiliser were assumed: - N as NO3 : (fertiliser N)*0,3*62/14 = NO3 (IPCC, 2006)

• Emissions to air from fertilisers include NH3 and N2O from N fertiliser:

N2O-N 1% of applied N (IPCC 2006).

NH3 Emissions: (fertiliser N*(0,008+(0,0001*9,1))) 17/14 (Rösemann et al., 2011) Emissions from herbicide and fungicide were assumed to be zero based on published data from Tasmania (Lee-Archer, 2007) and unpublished data from forest growers.

Production of CuSO4, ZnSO4 and MnSO4 are modelled as magnesium sulphate.

Residue burning

Amount of forest residues burnt are based on estimates by forest managers which are assumed to be accurate. Emissions are assumed to be applicable to radiata pine residues from harvesting operations in Australia.

Wildfire

Wildfires are relatively rare in plantations. For example, ABARES (2013, Table 5.1) reports carbon emissions from wildfires in plantations as 0 Mt for the period 2001-10. Due to the rarity of these events and the lack of historical data to create an accurate estimate of occurrence rates, emissions from wildfire have not been included for softwood forestry.

Fire prevention

Data for aircraft use are originally based on a single aerial fire contractor. These data are taken to be representative of aircraft use for other plantations across Australia. Aircraft fuel use data are based on published estimates from NSW Fire Service. These are assumed to be accurate and applicable to all of Australia. Forest residue mass burnt is based on estimates from forest managers. These are assumed to be representative of actual residue burning.

Roading

Bitumen is assumed to be applied in the same proportion to gravel as in the baseline study. This assumes the proportion of sealed roads to gravel roads across the managed area remains the same and the same mix of gravel to bitumen is used for sealed roads.

Harvesting

The forest is in a steady state (i.e. inputs into forest establishment, management and harvest and the volume harvested do not change over time). Forest manager estimates of fuel consumption for harvesting operations are assumed to be accurate for work done by a third party.

LCI and EPDs for Australian Wood Products 52 of 159

Electricity grid mix

The electricity grid mix for softwood forestry has been based on the downstream product grid mix. For example, when modelling sawn softwood, the electricity grid mix for sawn softwood has also been applied to the forestry stage. See the assumptions section of each downstream process for further details. The electricity grid mixes used for each product are given in Annex E.

4.4. Hardwood forestry

4.4.1. Overview

This section covers native hardwood forestry in Australia. Dozens of different species are harvested across Australia and the dominant species vary significantly by region. Important species in the (north)east include blackbutt (Eucalyptus pilularis), spotted gum (Corymbia/Eucalyptus maculata), flooded gum (Eucalyptus grandis) and grey ironbark (Eucalyptus drepanophylla); important species in the south include mountain ash (Eucalyptus regnans), alpine ash (Eucalyptus delegatensis) and messmate (Eucalyptus obliqua); important species in the west include jarrah (Eucalyptus marginata) and karri (Eucalyptus diversicolor). The native forest managers included within the survey were:

• Department of Agriculture and Fisheries (Queensland) • Forestry Corporation of New South Wales • Forestry Tasmania • VicForests • Western Australia Forest Products Commission Four of these five native forest managers provided enough data to be included within the LCI. These four plantations collectively produced 3,490,431 m3 of native hardwood logs in 2015/16 (reference year for the study), equating to 88% of total Australian production of approximately 3,958,000 m3 (based on the 2015/16 total from ABARES (2017). All averaged data displayed in the following sections are calculated as production-weighted averages of the four included plantations. The average proportions of forest products from native forests included in this study is displayed in Table 4-18 below. Baseline (2006) refers to data collected by the Commonwealth Scientific and Industrial Research Organisation (CSIRO) that inventoried data from the 2005/06 Australian financial year (1st July 2005 to 30th June 2006) (CSIRO, 2009a), as updated in v1 of this background report (2015).

Table 4-18 Hardwood forestry products by type

Type Flow Volume Volume Baseline DQI* (1000 m3) (%) (2006) Outputs Sawlogs 1,368 39.2% 34.8% M Pulplogs 1,714 49.1% 64.0% M Veneer logs 203 5.80% 0.54% M Preservation logs 143 4.10% - M Other 63.1 1.81% 0.59% M Total 3,490 100% 100%

LCI and EPDs for Australian Wood Products 53 of 159

* measured (M) / calculated (C) / estimated (E) / literature (L) Poles, piles and girders have been grouped as ‘preservation logs’ for the purposes of this study. ‘Other’ represents firewood and other low value forestry products. The range of log products produced by native forests participating in this study generally shows a similar distribution to that observed in the baseline study. However, the current study shows a decrease in the proportion of pulplogs harvested relative to the baseline study, with an increase in all other forestry products. The product outputs vary by forest region as seen in Annex A, as the southern regions tend to produce a higher proportion of pulplogs compared to the northern regions.

4.4.2. Production process

The activities in a typical hardwood production system and associated inputs and emissions are illustrated in Figure 4-4.

CO2 sequestration in wood

System boundary Burning of residues

Seeds & energy Stand establishment Seed collection, preparation and planting Primary Fuel reduction burning energy, Forest management water, Liquid fuels, gravel Monitoring, fire prevention and control, road Emissions raw building material

Harvesting Liquid fuels Diesel consumption of harvesting vehicles and appliances

Veneer logs Pulplogs Other Preservation logs Sawlogs Figure 4-4: Process flow diagram of native hardwood forestry

The native hardwood process is similar to the plantation softwood process with a few major differences. Native forests are generally re-established after harvest by aerial seeding rather than hand planting so there is no need for nurseries and seedling production. The exception to this is selection harvesting systems where seed for regeneration is from standing trees. For clearfell systems, seed production involves the collection, processing and drying of seed. Seed capsules are collected from trees in the region either following felling operations or from standing trees (by climbing and cutting branches). Seed is extracted, cleaned and dried with electricity or gas being used in the drying process. In some cases, where hand seeding is used, seed is coated (pelletised) to enable more effective delivery. Site preparation is typically in the form of broadcast burning of slash rather than heaping and there is little in the way of mechanical site preparation. These activities are usually carried out by the forest manager. Management activities across the estate are similar to those in softwood plantations with the exception that chemical and fertiliser application is rarely carried out. Most activities are untaken by the forest manager, although different activities (e.g. fire control) may be undertaken by separate Government agencies. Accounting for management activities of native forests in an LCI is complicated by the fact that they also involve management for non-wood products such as recreation, biodiversity and water quality. As a result, it is difficult to separate inputs directly related to wood production from those for other land uses.

LCI and EPDs for Australian Wood Products 54 of 159

Thinning is currently a low proportion (<5%) of the total harvesting in native forests, and most harvesting activity is from clearfelling or selection logging methods. Other methods used are regrowth retention harvesting, single-tree selection harvesting and salvage harvesting. Felling is usually mechanised, as for softwood plantations, but occasionally, may be by hand. In steep terrain, cable logging may be used instead of skidding to winch logs or trees to a landing on the top of a hill ready for loading.

Table 4-19 Hardwood forestry production area and output – per m3 of wood harvested

Type Flow Unit Total Average Baseline DQI* (2006) Inputs Total area established ha 1.67E+05 6.14E-03 1.69E-03 M Total area managed ha 2.05E+07 5.88E+00 2.03E-01 M Outputs Total wood harvested m3 3.49E+06 1.00E+00 1.00E+00 M * measured (M) / calculated (C) / estimated (E) / literature (L)

The total area managed and established with respect to the volume of logs harvested has increased relative to the baseline study as shown in Table 4-19. This observed decrease in forestry intensity is partly due to different management areas and states included in the current study compared with the case study areas included in the baseline study.

4.4.3. Stand establishment

Establishment consists of slash burning and/or mechanical disturbance, seed production/collection and seed planting. Native hardwood forests are established after the previous stand has been harvested. Following harvest, woody residues (stems, branches and leaves) are generally left on the ground. Prior to seeding/regeneration, these residues are pushed into heaps using a bulldozer, or left spread out on the ground. They may then be burnt (although some areas are left mechanically disturbed without burning). In clearfell systems, seed is generally applied by aerial seeding but in some cases, by hand. Selective systems rely on seedfall from surrounding standing trees. Site preparation activities are generally undertaken by the forest management agency. Establishment operations vary between different regions depending on soil, terrain, forest silviculture and markets for products. Seed is collected by contractors from native forest in the region. Generally, electricity or gas is used to heat kilns which dry the seed. Electricity is also used in seed storage (generally under constant temperature conditions). Seed processing may be operated by the forest owner or externally. LCI data for native hardwood forestry are given in Table 4-20.

Table 4-20 Native hardwood forest establishment inventory – per ha established

Type Flow Unit Average Baseline DQI* (2006)

Inputs Seeds planted kg 2.41E-01 5.10E-01 M Diesel consumed kg 3.81E+00 1.21E+01 E Kerosene kg 2.00E+00 4.00E+00 E Diesel travel distance Km 3.10E+01 9.87E+01 E Residues burnt t 2.15E+01 6.65E+01 C

LCI and EPDs for Australian Wood Products 55 of 159

Type Flow Unit Average Baseline DQI* (2006)

Seed production Seeds collected / produced kg 1.00E+00 1.00E+00 M Diesel travel distance km 1.21E+01 5.03E+01 E Thermal energy from natural gas MJ 2.26E+02 2.29E+02 E Electricity kWh 2.95E+01 3.75E+01 M Outputs Land established ha 1.00E+00 1.00E+00 M * measured (M) / calculated (C) / estimated (E) / literature (L)

Forest residue burning

Emissions resulting from the burning of forest residues is described in section 4.3.3 “Forest residue burning”. The moisture content of residues burnt in the forest has been taken to be 80% (dry basis), equal to the average log moisture content received by the surveyed sawmills. It has been assumed that 68% of residues burnt are consumed (based on dry mass) in native hardwood systems based on IPCC guidelines for post-logging slash burns in eucalyptus forests (IPCC, 2006). This proportion is lower than in softwood plantation systems due to less heaping of residues before burning.

4.4.4. Forest management

Forest management includes monitoring, fire control and prevention activities and road construction and maintenance. Inputs for each activity in native forests are expressed in terms of managed area. Water use for forests is as described in section 4.3.7. Road construction and maintenance consists of construction of new roads and maintenance of existing roads. Fire prevention activities consist of fuel reduction burning, fire surveillance and fire control. As with slash burning, CO2 emissions from fuel reduction burning are not included in the forest system as these are assumed to derive from biomass which sequestered the carbon; however, non-CO2 GHG emissions are included. More details are provided in section 4.4.3 “Forest residue burning”. The process inventory for forest management is given in Table 4-21.

Table 4-21 Native hardwood forest management – per ha of managed land

Type Flow Unit Average Baseline DQI* (2006) Inputs Kerosene kg 3.05E-02 8.85E-01 E Diesel kg 1.38E-01 7.20E-01 C Road length constructed/maintained m 3.22E+00 4.70E+00 C

Road construction Gravel kg 1.29E+01 4.49E+00 C Diesel use kg 1.32E-01 9.26E-03 C Petrol travel km 4.95E-06 1.44E-04 E LCV travel km 4.88E-02 1.42E+00 E

Fire prevention

LCI and EPDs for Australian Wood Products 56 of 159

Type Flow Unit Average Baseline DQI* (2006) Diesel use kg 6.55E-03 1.90E-01 E Kerosene kg 6.55E-03 1.90E-01 E LCV travel km 2.04E-02 5.90E-01 E Residues burnt kg 5.85E+00 5.29E+01 M Outputs Land managed ha 1.00E+00 1.00E+00 M * measured (M) / calculated (C) / estimated (E) / literature (L) Gaps in the forest management data existed for some sites inputs, these have been filled based on sites that were able to provide data. Where this has been done, inputs are assumed to be equal to the weighted average of other participating sites and prorated based on total harvest volume. The collected data showed that forest management activities followed the harvested wood volume more closely than the managed forest area. This is due to very large differences in production intensity and the share of managed forest area harvested between states.

4.4.5. Harvesting

Harvesting of native hardwood forests involves either felling or selective harvesting systems. Theoretically rotation lengths range from 70 to 110 years for clearfelling systems in regrowth forest. Clearfelling consists of complete removal of the stand, while selective harvesting removes single trees or groups of trees leaving small gaps. Trees may be felled with a mechanical harvester which also delimbs the stems, cuts them into lengths and sorts them into different products. These are then collected by a forwarder. Sometimes a feller buncher, processor, skidder combination is used instead.

Table 4-22 Harvesting – per m3 of harvested wood

Type Flow Unit Average Baseline DQI* (2006) Inputs Diesel use kg 3.35E+00 3.61E+00 M LCV travel km 1.16E+00 1.16E+00 E Outputs Wood harvested m3 1.00E+00 1.00E+00 M * measured (M) / calculated (C) / estimated (E) / literature (L)

A minor decrease has been seen in harvesting diesel use in the current study compared to the baseline study. Where data were not able to be provided by a specific forest manager, gaps in diesel use have been filled using data from the baseline study.

4.4.6. Allocation

Impacts were allocated to co-products by economic value (Table 4-23). Economic allocation was applied because the difference in the value of co-products is typically large (>25% in many cases, in line with IEPDS 2017, section 7.7). Since sequestered carbon and primary energy are physical properties of wood, they have been allocated on a mass basis to the different co-products.

LCI and EPDs for Australian Wood Products 57 of 159

Table 4-23: Forest product allocation factors

Product Price ($/m3) Volume Value Allocation (%) (1000 m3) (M$)

Sawlogs $62 1,370 85.1 60% Pulplogs $11 1,710 18.7 13% Veneer logs $69 203 14.0 10% Preservation logs $158 143 22.6 16% Other $4 63.1 0.26 0.2%

Total 3,490 141 100%

Poles, piles and girders have been grouped as ‘preservation logs’ for the purposes of this study. ‘Other’ represents firewood and other low value forestry products. Unlike plantation softwood forestry, where there is a comprehensive source of data on historical prices provided by KPMG, no such reference exists for native hardwood forests. The national ABARES statistics only provide highly aggregated data. Economic stumpage data were obtained directly from three forest managers. It has been assumed that the log prices provided by forest managers are representative of similar products from other sites which provided no stumpage value data and that the product values provided by forest managers are representative of the relative values of the different products at mill door. Each forest manager uses a different sawlog grading system. In order to construct an average log price, sawlogs have been aggregated into a single category based on production-weighted averages. Weighted average stumpage prices were provided by one forest manager for the period 2010/11 to 2014/15. The second provided weighted average stumpage prices for the period 2009 to 2014. The remaining forest manager provided stumpage prices for the 2015 period. A weighted average has been constructed using production volumes provided by these forest managers for the 2015/16 year.

4.4.7. Assumptions

Assumptions made during the construction of the hardwood forestry LCI are listed within the following section.

General assumptions

Inputs of fuel, oil and consumables (e.g. bars) resulting from chainsaw use during seed collection (cutting down branches) were excluded because they were considered very minor and no data were available. The only product considered from native hardwood forests is wood which is harvested and sold.

Steady state forestry

A key assumption is that all forests included were in a steady state in regard to management inputs and wood products produced as well as for above and below ground carbon stocks. Thus, both plantations and native forests were assumed to be sustainably managed with no reduction in carbon stocks in soil, litter or live biomass over time. The majority of native forests currently being harvested are regrowth forests. The assumption is that for these forests, CO2 sequestration and carbon emissions are balanced over a forest rotation.

LCI and EPDs for Australian Wood Products 58 of 159

For information on forest certification and inclusion of sequestered biogenic CO2, please refer to section 4.3.7, “Steady state forestry”.

Wildfire

Carbon emissions from wildfires have been excluded from this study as there is no accepted method for the treatment of wildfire in LCA of temperate forests within Australia, due to the high variability (and hence high uncertainty) of emissions from wildfires, and difficulties in attributing the relative contribution to emissions from different forest management actions (S. H. Roxburgh, personal communication, 9 October 2017). While wildfires do occur in native forests, controlled/ prescribed burning and other forest management actions may lead to either a decrease or an increase in net carbon emissions depending on the specific forest ecosystem, the amount of fuel available for combustion at any given time, and the level of intervention (Bradstock, et al., 2012). To provide some context, 148 Mt of carbon stored in biomass was lost through wildfire from Australian native forests over the period 2001-10 (ABARES, 2013, Table 5.1). This is roughly equivalent to the amount of carbon removed through prescribed fire (90 Mt) and selective logging (66 Mt) over the same time period (ABARES, 2013, Table 5.1). The equivalent emissions if the forest was unmanaged are unknown, and may have been either higher or lower than 238 Mt of carbon released through both planned and unplanned fire over the 2001-10 period.

Land use and biodiversity

Please refer to section 4.3.7, “Biodiversity”.

Water use in forests

Please refer to section 4.3.7, “Water use in forests”

Electricity grid mix

The electricity grid mix for hardwood forestry has been based on the downstream product grid mix. For example, when modelling sawn softwood, the electricity grid mix for sawn hardwood has also been applied to the forestry stage. See the assumptions section of each downstream process for further details. The electricity grid mixes used for each product are given in Annex E.

4.5. Sawn softwood

4.5.1. Overview

The following sawn softwood products are included within the LCI:

• 1 m3 of sawn kiln-dried softwood • 1 m3 of dressed kiln-dried softwood LCI data were collected using detailed surveys of six of the 29 softwood mills in Australia that are FWPA members, as shown in Table 4-24 (note that one company contributed data for two different mills). All companies listed have contributed financially to this project through their FWPA levies. These six mills, of an estimated 61 mills throughout Australia (Gavran, et al., 2014), collectively produced 1,229,419 m3 of sawn softwood in 2015/16 (reference year for the study), equating to 33- 35% of total Australian production of approximately 3,510,000-3,784,000 m3 (based on the 2015/16 total from ABARES (2017), which has been adjusted to saleable volume following Houghton (2017)).

LCI and EPDs for Australian Wood Products 59 of 159

Table 4-24 Contributors to sawn softwood LCI

Producer company Total Financial Data mills contributor contributor Allied Timber Products Pty Ltd 2 2 Associated Kiln Driers Pty Ltd trading as A.K.D. Softwoods 2 2 Australian United Timbers Pty Ltd 1 1 Auswest Timbers Pty Ltd 1 1 Boral Timber 1 1 Carter Holt Harvey Woodproducts Australia 6 6 1 D&R Henderson Pty Ltd 1 1 Highland Pine Products Pty Ltd 1 1 Hyne Timber 2 2 1 KSI Sawmills Pty Ltd 1 1 LM Hayter & Sons Pty Ltd 1 1 Lorimer Timber Pty Ltd trading as Davids Timber 1 1 McDonnell Industries Pty Ltd trading as NF McDonnell & 1 1 Sons Penrose Pine Products Pty Ltd 1 1 Rodpak Pty Ltd 1 1 SA Sawmilling Pty Ltd 1 1 Tarmac Sawmilling Pty Ltd 1 1 TASCO trading as Dongwha Timbers Pty Ltd 1 1 1 Timberlink Australia 2 2 2 Wespine Industries Pty Ltd 1 1 1 Whiteheads Timber Sales Pty Ltd 1 1 Total 29 29 6

Treatment is covered separately in section 4.11 rather than as part of the main inventory. This has been done because of the large number of treatment types and hazard classes available. In the final EPD, the impacts associated with treatment can be simply added to the impacts associated with the relevant untreated product. The properties of wood products included in this study are given in Table 4-25. All moisture contents are given as dry basis. A comparison is given to the baseline study and also to two CORRIM studies of softwood timber in the United States: one covering the Northeast and North Central region (Puettmann, et al., 2012), the other covering the Inland Northwest region (Puettmann & Oneil, 2012). Baseline (2006) refers to data collected by the Commonwealth Scientific and Industrial Research Organisation (CSIRO) that inventoried data from the 2005/06 Australian financial year (1st July 2005 to 30th June 2006) (CSIRO, 2009a), as updated by EPD Registration No. S-P-00560, Version 1.1, Revised 13 July 2015, Valid until 22 June 2020.

LCI and EPDs for Australian Wood Products 60 of 159

Table 4-25 Sawn softwood material properties

Softwood properties Unit Average Baseline CORRIM NE CORRIM DQI* (2006) & NC USA INW USA

3 Sequestered carbon (per m ) kg CO2e 9.00E+02 9.00E+02 6.80E+02 7.16E+02 L Green log density kg/m3 9.71E+02 9.93E+02 7.07E+02 6.07E+02 M Green, sawn density kg/m3 9.42E+02 9.93E+02 7.07E+02 6.98E+02 M Kiln-dry density kg/m3 5.51E+02 5.50E+02 4.18E+02 4.36E+02 M Shrinkage % 6.68E+00 8.50E+00 - - C Moisture content green logs % 1.12E+02 1.17E+02 9.65E+01 6.00E+01 C Moisture content green, sawn % 1.06E+02 1.17E+02 9.65E+01 6.00E+01 C Moisture content kiln-dry % 1.22E+01 1.20E+01 1.60E+01 1.50E+01 M Moisture content of residues burnt % 7.73E+01 - - 5.00E+01 C NCV of kiln-dried product MJ/kg 1.69E+01 1.69E+01 - - C NCV of residues burnt MJ/kg 9.83E+00 - - - C Distance travelled by logs km 5.39E+01 6.62E+01 - - M * measured (M) / calculated (C) / estimated (E) / literature (L)

The material properties of outputs from a softwood sawmill remain very similar to the previous study as there has been little change in the softwood species grown across Australia. The reduction in densities of logs and sawn wood is linked to the reduced moisture content of logs received by participants of the study, however the basic density (density of wood excluding water) remains almost identical. The distance travelled by the logs from the forest to sawmill has also decreased relative to the baseline study. Variability between sites included in this study was low for most of the properties listed in Table 4-25. Some variability does exist between the moisture content of the residues burnt by some sites as shown in Annex A. This is due to differences in the mix of green and dry residues used as a fuel source. The distance between the forest and sawmill also shows large variability for the participating sites. Table 4-26 shows expansion factors from Briggs (1994, Table 7-1), for residues from sawing and planing. These expansion factors are used to provide average bulk densities based on the average density of green sawn wood and kiln-dried wood.

Table 4-26 Residue densities and expansion factors

Residue Green Dens.1 Dry Dens.2 Exp. Factor (kg/m3) (kg/m3) Wood chips 329 193 2.86 Sawdust 377 220 2.50 Bark 401 - 2.35 Planer shavings - 138 4.00 1 Based on sawn wood density (Table 4-25) and expansion factor 2 Based on dry density (Table 4-25) and expansion factor

LCI and EPDs for Australian Wood Products 61 of 159

4.5.2. Production process

The production process of sawn softwood and other sawmill co-products is illustrated in Figure 4-5.

Sawlogs from softwood forestry

System boundary

Liquid fuels, water Log haulage Transport Emissions Waste (recycling Liquid fuels, water Log storage & landfill)

Softwood timber, green Electricity, thermal energy, water, consumed materials Debarking and sawmilling Wood chips & other residues

Primary Electricity, water, thermal energy, Kiln-drying Softwood timber, kiln-dried, rough sawn water, energy raw Burning of residues material Residues Liquid fuels, electricity, water Planing Softwood timber, kiln-dried, dressed

Softwood timber, kiln-dried, dressed, treated Chemicals, electricity, liquid fuels, water Treatment Softwood timber, kiln-dried, rough sawn, treated

Packaging material, transport packaging Packaging

Figure 4-5 Process flow diagram of sawn softwood production

Softwood sawmills produce a range of different products from green logs depending on the number of processes applied. The sawmill data have been split into four unit processes to differentiate the impacts of products output from each process. These are described in sections 4.5.3 to 4.5.6 below. Indicators assessing the variability in LCI data are given in Annex A. The products within the scope of this study include:

• Kiln-dried, rough-sawn, untreated timber • Kiln-dried, rough-sawn, treated timber • Kiln-dried, dressed, untreated timber • Kiln-dried, dressed, treated timber Green timber is produced and sold by some mills, however this is a lower value product and is not produced in as high volumes as kiln-dried wood, hence it has not been included within the scope of the study. All products are assumed to receive an equal share and type of packaging material.

4.5.3. Storage, de-barking and milling

The first unit process covers log storage, debarking and milling. Logs are delivered to the mill by truck from a nearby forest, and are stored in a log yard and moved using mobile plant. Sprinklers may be used in the yard to keep the logs wet. From the log yard, logs are debarked, then sawn to dimension, producing green, rough-sawn timber. The sawing stage also produces lower value wood residues as co-products, which may be either sold or used internally as fuel. The process inventory is given in Table 4-27. Due to the large difference in price between the co-products, inputs and emissions are allocated on an economic basis. This methodology is described in section 4.5.8.

LCI and EPDs for Australian Wood Products 62 of 159

Table 4-27 Softwood storage, de-barking and milling inventory – per m3 of sawn wood

Flow Unit Average Baseline CORRIM CORRIM DQI* (2006) NE & NC INW USA USA

Inputs Softwood Logs m3 1.85E+00 2.21E+00 2.13E+00 1.92E+00 M Large logs >43.9 cm m3 1.35E-01 - - - M Medium logs 24-43.9 cm m3 1.29E+00 - - - M Small logs <24 cm m3 4.25E-01 - - - M Electricity kWh 2.49E+01 6.31E+01 4.37E+01 2.25E+01 M Diesel kg 7.34E-01 2.12E+00 1.40E+00 1.12E+00 M Thermal energy MJ - 6.00E+00 - - M Municipal water kg 2.26E+01 5.10E+01 - - M Ground water kg 1.52E+01 - 5.90E+02 - C Lake water kg 1.14E+00 - - - C Rain water kg 4.39E-03 - - - M Steel parts Kg 5.00E-01 5.00E-01 - - E Lubricating oil kg 3.26E-02 3.26E-02 - - E Petrol kg 5.47E-04 5.70E-02 8.05E-01 6.53E+00 M Outputs Green, sawn softwood m3 1.00E+00 1.00E+00 1.00E+00 1.00E+00 M Bark** m3 9.86E-02 1.60E-01 1.30E-01 4.04E-02 M Wood Chips** m3 5.34E-01 7.10E-01 3.57E-01 3.10E-01 M Sawdust (green)** m3 3.23E-02 3.40E-01 8.92E-02 1.66E-01 M Steel scrap kg 5.00E-01 5.00E-01 - - E Sludge (solids) kg 6.68E-04 3.30E+00 - - M Used oil kg 3.26E-02 3.10E-02 - - E Municipal waste kg 5.20E-01 3.90E-01 3.82E-01 - M Waste water kg 3.24E-01 5.10E+01 - - C Water to land kg 2.66E+01 - - - C Water vapour kg 6.49E+01 - - - C Wood residues used in m3 1.86E-01 - - - C kiln** * measured (M) / calculated (C) / estimated (E) / literature (L) **Given at density and MC of green sawn wood

Electricity for storage, de-barking and milling showed a decrease when compared to the baseline study. Electricity is also significantly lower than the CORRIM NE & NC USA study but is comparable to the CORRIM INW USA study. The difference in electricity compared to the baseline study may be due to a combination of increased efficiency, improved technology and different mills participating in the two studies. Good agreement exists between sites for electricity use (see Annex A).

LCI and EPDs for Australian Wood Products 63 of 159

Diesel use in the log yard is highly dependent on the distance travelled between production lines and equipment used. It is therefore expected that differences in diesel consumption compared to the baseline study are due to different site layouts, equipment and participating sites across the two studies. Thermal energy was reportedly used during log storage, de-barking and milling during the baseline study. This was not found to be the case for sites participating in the current study or either of the two CORRIM USA comparison studies. Another important finding is that the recovery rate (yield) is higher than for all three comparison studies. This results in lower upstream impacts per cubic metre of sawn wood produced. Good agreement exists between sites on recovery rates with low variability. Diesel during storage, de- barking and milling has decreased relative to the baseline study. The water input has been classified according to source in the current study, however the water input is lower for the current study than the three comparison studies. Water usage does vary between sites as does the source of water used. Variability in water usage is likely due to climate differences, cost and availability of on-site recycling facilities.

4.5.4. Kiln-drying

Green sawn timber is the input to the second unit process: kiln-drying. Green timber is dried within a kiln to a moisture content close to 12% (dry basis) (depending on end-product). Thermal energy for the kiln is provided by the combustion of either natural gas and/or wood residues produced from other processes by the site. The output of the kiln-drying process is kiln-dried, rough-sawn timber. The process inventory for kiln-drying is given in Table 4-28.

Table 4-28 Softwood kiln-drying inventory – per m3 of kiln-dried wood

Flow Unit Average Baseline CORRIM CORRIM DQI (2006) NE & NC INW USA USA

Inputs Green, sawn softwood m3 1.07E+00 1.09E+00 1.09E+00 1.00E+00 C Diesel kg 2.84E-01 9.30E-04 1.15E+01 4.68E+00 M LPG (transport) MJ 3.08E+00 - - - M Electricity kWh 2.48E+01 3.60E+01 1.99E+01 1.67E+01 M Thermal energy from MJ 2.39E+03 2.50E+03 2.15E+03 1.02E+03 C biomass Wood residues burnt kg 2.45E+02 - - - C (77% MC) Thermal energy from natural MJ 5.38E+02 5.80E+02 0.00E+00 1.27E+03 M gas Municipal water kg 7.99E+01 3.50E+02 - - M Ground water kg 9.95E+01 - - - C Lake water kg 7.99E+00 - - - C Rain water kg 3.93E-02 - - - M

Outputs Kiln-dried, sawn softwood m3 1.00E+00 1.00E+00 1.00E+00 1.00E+00 M VOCs kg 4.10E-01 4.10E-01 - 1.68E-01 L

LCI and EPDs for Australian Wood Products 64 of 159

Flow Unit Average Baseline CORRIM CORRIM DQI (2006) NE & NC INW USA USA Municipal waste kg 7.36E-02 - - - M Waste water kg 2.25E+00 3.50E+02 - - C Water vapour kg 7.50E+02 0.00E+00 - - C Butyraldehyde kg 2.60E-04 2.60E-04 - - L Crotonaldehyde kg 2.60E-04 2.60E-04 - - L Ethanol kg 7.50E-02 7.50E-02 - - L Formaldehyde (methanal) kg 1.10E-03 1.10E-03 - - L Formic acid (methane acid) kg 7.70E-03 7.70E-03 - - L Hexane (isomers) kg 9.60E-04 9.60E-04 - - L Methanol kg 2.90E-02 2.90E-02 - - L Propionaldehyde kg 9.60E-04 9.60E-04 - - L * measured (M) / calculated (C) / estimated (E) / literature (L)

A comparison of average kiln-drying data to the baseline study shows a decrease in electricity consumption, wood residues burnt and water consumption. However, all three are higher than both CORRIM USA studies. A significant difference in the reported diesel consumption for kiln-drying exists when compared to the baseline study. This is likely due to an error in the baseline study data as reported consumption was very low. Diesel consumption reported for kiln-drying in the current study is more reasonable when compared to the other comparison studies and the expected use based on other sawn softwood production processes. Some of the participant sites use different fuel types for thermal energy such as natural gas or biomass, the total energy usage however is comparable to the baseline study. Water used for kiln-drying has decreased in the current study relative to the baseline study. This is likely to be due to different participating sites in the two studies, as the amount of water used as well as the source type varies significantly between sites. The variability of kiln-drying data is displayed further in Annex A. Note that water outputs from the kiln-drying process appear to be greater than water inputs. A water balance however, is maintained due to moisture being removed from green wood and from any wood residues that are burnt during the drying process and leaving the system as water vapour. The corresponding mass of water vapour can be calculated from the difference in density and moisture content between the sawn wood and dry wood given in Table 4-25. All water originally contained in the wood residues (77% MC) is considered to be lost as water vapour during the kiln-drying process.

4.5.5. Planing

Most sawn, kiln-dried wood enters a third unit process: planing. During this stage wood is sawn and finely planed to produce a relatively smooth surface finish and final dimensions. This process also produces valuable (relative to green) co-products that may be either sold or used to provide thermal energy to the kiln-drying process. Due to the large difference in price between the co-products, inputs and emissions are allocated on an economic basis. This methodology is described in section 4.5.8. The main output from this process is kiln-dried, dressed timber. The process inventory for planing is given in Table 4-29.

LCI and EPDs for Australian Wood Products 65 of 159

Table 4-29 Softwood planing inventory – per m3 of planed wood

Flow Unit Average Baseline CORRIM NE CORRIM DQI (2006) & NC USA INW USA

Inputs

Kiln-dried, sawn softwood m3 1.30E+00 1.17E+00 1.27E+00 1.09E+00 C Diesel kg 2.98E-01 3.49E-01 3.42E-01 3.65E-01 M Electricity kWh 2.61E+01 2.10E+01 1.96E+01 3.41E+01 M Gasoline kg 2.17E-04 1.30E-02 5.59E-03 4.61E-02 M LPG (transport) MJ 2.75E+00 1.22E-03 - - M Municipal water kg 2.05E+01 - - - M Ground water kg 7.40E+00 - - - C Lake water kg 7.13E-01 C

Outputs Kiln-dried, dressed softwood m3 1.00E+00 1.00E+00 1.00E+00 1.00E+00 M Chips** kg 3.15E+00 9.65E+00 2.29E+01 3.97E+00 M Hogfuel** kg 0.00E+00 6.80E+00 - - M Sawdust** kg 0.00E+00 2.90E+01 - - M Shavings** kg 9.18E+01 4.58E+01 8.06E+01 3.65E+01 M Waste water kg 8.76E-04 - - - C Water to land kg 8.30E+00 - - - C Water vapour kg 2.03E+01 - - - C Municipal waste kg 1.92E+00 - - - M

Wood residues used in kiln** kg 6.82E+01 - - - C * measured (M) / calculated (C) / estimated (E) / literature (L) **Wood residues given at MC of kiln-dried wood (12%)

A key comparison to note from the current study is the planing recovery rate, which is lower than all three comparison studies. This was reported by all sites with very low variability (see Annex A). The densities of kiln-dried, sawn timber, kiln-dried dressed timber and planing residues are given in Table 4-25.

4.5.6. Treatment

The final unit process is timber treatment. A range of different preservative treatment types may be used to treat the wood depending on the suitability and hazard class requirement. Treatment is applied to the surface of the wood or within pressurised chambers depending on the requirements of the treatment type used. Preservative treatment may be applied to either kiln-dried, rough sawn timber or kiln-dried, dressed timber. Alternatively, timber may be sold without preservative treatment as untreated timber. Detailed information about wood preservative treatment can be found in section 4.9.

LCI and EPDs for Australian Wood Products 66 of 159

4.5.7. Packaging

Average amounts of packaging materials used by each site per cubic metre of wood is given in Table 4-30.

Table 4-30 Softwood packaging – per m3 of wood packaged

Flow Unit Average DQI*

Wood packaged m3 1.00E+00 M Softwood gluts m3 8.60E-03 M LDPE wrap kg 1.43E-01 M Plastic strapping kg 1.40E-01 M Paper labels kg 2.61E-03 M Inks kg 5.45E-02 M Paints kg 1.27E-03 M * measured / calculated / estimated / literature

Packaging data was not included within the scope of the baseline study; therefore, no comparison is available. Packaging materials are assumed to be used evenly on all products produced per site per cubic metre of output.

4.5.8. Allocation

Allocation to the co-products of sawmilling is based on economic value as the difference in the value of co-products is large (>25%, in line with IEPDS (2017, section 7.7)). Most co-products have some economic value and therefore cannot be considered as wastes. Economic values have been provided by participants and are displayed as volume-weighted averages in Table 4-31 for sawing and Table 4-32 for planing.

Table 4-31 Sawing co-product allocation factors

Sawing Units Amount Average Economic Allocation (Units) price value ($) (%) ($/Unit)

Green, sawn softwood m3 1.00 235 235 82.7% Woodchips (green)* m3 0.53 70.6 37.7 13.3% Bark* m3 0.10 35.5 3.51 1.24% Sawdust (green)* m3 0.03 39.7 1.28 0.45% Sawing residues burnt (bone dry) kg 85.4 0.08 6.61 2.33%

Total 284 100% *Given at density and MC of green sawn wood

Although no economic value was actually specified for the ‘sawing residues burnt’, which are used internally. Instead this value has been calculated based on the volume-weighted average of bark and green sawdust. These are the fuel types typically burnt onsite during the kiln-drying process.

LCI and EPDs for Australian Wood Products 67 of 159

Table 4-32 Planing co-product allocation factors

Planing Units Amount Average Economic Allocation (Units) price value ($) (%) ($/Unit)

Kiln-dried, dressed softwood m3 1 391 391 94.3% Shavings (kiln-dry) kg 91.8 0.18 16.8 4.05% Woodchips (kiln-dry) kg 3.15 0.08 0.26 0.06% Planing residues burnt (bone dry) kg 60.8 0.11 6.58 1.59%

Total 414 100%

No economic value was specified for the ‘planing residues burnt’, which are used internally. Instead this value has been calculated based the average price of shavings weighted by the volume of residues burnt by each site, as shavings are typically the main dry fuel burnt onsite during the kiln- drying process.

4.5.9. Assumptions

Assumptions made during the construction of the sawn softwood LCI are listed below.

General assumptions

All data received from manufacturers have been accepted as accurate (after cross-checks for mass balance, plausibility, etc.). Energy used is reported by energy source (e.g. coal, diesel fuel and electricity), but the upstream profiles of energy sources are modelled based on Australian averages. Upstream profiles for intermediate products that are consumed in small quantities, such as lubricants, are modelled based on available LCI data. Where onsite dams are used, water use has been calculated as follows: annual, site specific rainfall on dam area is reported as lake water use and is taken from the nearest Bureau of Meteorology measurement station (BoM, 2017). Dam evaporation has been calculated using the tool Readyreckoner (NCEA, 2017). This has been used to calculate the net evaporation relative to a similar area of pasture land with the same location and annual rainfall using the Zhang relationship (Zhang, et al., 2001). Net use of ground water has been calculated to balance water according to the following: ground water in + rainfall (lake water) = water use from dam + net evaporation.

Computational assumptions

Input and output materials and processes are a production-weighted average based on the data provided from Australian manufacturers.

Storage, de-barking and milling

Mass of product and residues sold by each site are assumed to be accurate. Any difference in mass between log input and sold output has been assumed to be combusted on site during the kiln-drying process. Sawn green timber output from sawing process is based on the volume of purchased logs and site sawing recovery rates (yield).

LCI and EPDs for Australian Wood Products 68 of 159

Where a water input exceeded water discharged to a drain or treatment facility (water imbalance), remaining water is assumed to be discharged to land. The relationship stated by Zhang et al. (2001) has been used to estimate the water returning to groundwater and water evaporated using the annual precipitation in the nearest major city (ABS, 2017). Where residues from sawing were reported by participating sites on a bulk (expanded) volumetric basis, the measurement was assumed accurate. Densities provided by the participants had a high level of uncertainty, therefore expansion factors were used for each residue type as given in Table 4-26 based on the site’s sawn wood density to convert to mass.

Kiln-drying

All residues produced but not sold from either the sawing or planing processes have been assumed to be burnt for thermal energy used in the kiln-drying process. Shrinkage during kiln-drying is calculated from the density and moisture content of wood entering the kiln relative to the density and moisture content of the wood exiting the kiln. Where a water input exceeded water discharged to a drain or treatment facility (water imbalance), remaining water is assumed to be evaporated, including moisture leaving the wood during drying.

Planing

Masses of products and residues sold by each site are assumed to be accurate. Any difference in mass between rough-sawn input and sold output has been assumed to be combusted on site during the kiln-drying process. Rough sawn input into the planing process is based on the volume of dressed product output and site planing recovery rates (yield). Where a water input exceeded water discharged to a drain or treatment facility (water imbalance), remaining water is assumed to be discharged to land. The relationship stated by Zhang et al. (2001) has been used to estimate the water returning to groundwater and water evaporated using the annual precipitation in the nearest major city (ABS, 2017). Where residues from planing were reported by participating sites on a bulk (expanded) volumetric basis, the measurement was assumed accurate. Densities provided by the participants had a high level of uncertainty, therefore expansion factors were used as given in Table 4-26 for each residue type based on the site’s kiln-dried wood density to convert to mass.

Packaging

Packaging materials such as wooden gluts are assumed to be carbon neutral with respect to sequestered carbon, as the installation phase (module A5), where the sequestered carbon would be released, is not declared within this study. When modelling the production of paper, any scrap paper used has been assumed to be burden free. Inherent properties: biogenic carbon and primary energy have been accounted for within the dataset. Similarly, no credit has been applied for any waste paper that is recycled in the product life cycle. This methodology of cutting-off impacts has been applied throughout the model as it was not possible to credit recycled paper as no consistent, high-quality data were available for 100% primary paper production and 100% secondary paper production. The carbon balance has been corrected via CO2 emissions (biotic), assuming decomposition or incineration in the time frame of 100 years.

LCI and EPDs for Australian Wood Products 69 of 159

Electricity grid mix

The electricity grid mix for sawn softwood production (A1-A3) has been modelled based on the state-share of Australian sawn softwood production as a percentage of volume based on ABARES 2012/2013 data (ABARES, 2017). The same electricity grid mix for sawn softwood has also been applied to the softwood forestry stage. Electricity at end-of-life (module C) has been modelled using an average Australian electricity mix as the location where the product reaches end-of-life is unknown. The electricity grid mixes used for each product are given in Annex E.

4.6. Sawn hardwood

4.6.1. Overview

The following sawn hardwood products are included within the LCI:

• 1 m3 of sawn green hardwood • 1 m3 of sawn kiln-dried hardwood • 1 m3 of dressed kiln-dried hardwood LCI data were collected using detailed surveys of 11 of the 43 hardwood sawmills in Australia, who are FWPA members, as shown in Table 4-33 (note that two companies contributed data for two different mills each). Pole mills and other hardwood mills not producing sawn products have not been included. All companies listed have contributed financially to this project through their FWPA levies. These 11 mills, of an estimated 200 mills throughout Australia (including non-members of FWPA), collectively produced 145,151 m3 of sawn hardwood in 2015/16 (reference year for the study), equating to 25-27% of total Australian production of approximately 547,000-587,000 m3 (based on the 2015/16 total from ABARES (2017), which has been adjusted to saleable volume following Houghton (2017), allowing for nominal volume conversions, stock losses and packaging).

Table 4-33 Contributors to sawn hardwood LCI

Producer company Total Financial Data mills contributor contributor A E Girle and Sons 1 1 A G Brown Pty Ltd 1 1 Australian Solar Timbers 1 1 1 Australian Sustainable Hardwoods Pty Ltd 1 1 1 Auswest Timbers Pty Ltd 5 5 1 Blueleaf Corporation Pty Ltd trading as Whittakers Timber 1 1 Products Boral Timber Division 4 4 1 Britton Bros Pty Ltd 1 1 1 Dale & Meyers Operations Pty Ltd trading as DTM Timber 1 1 Endeavour Foundation trading as Nangarin Timbers 1 1 Fenning Investments Pty Ltd trading as Fenning Bairnsdale 1 1 Hallmark Oaks Pty Ltd 1 1 Hexan Holdings Pty Ltd trading as Whiteland Milling 1 1 Hurford Sawmilling Pty Ltd 2 2 1

LCI and EPDs for Australian Wood Products 70 of 159

Producer company Total Financial Data mills contributor contributor Intech Operations Pty Ltd trading as Muckerts Sawmill 1 1 Ironwood Taree Pty Ltd 1 1 J Notaras & Sons Pty Ltd 1 1 Jarrahwood Australia Pty Ltd 1 1 Machin’s Sawmill Pty Ltd 1 1 McCormack Demby Timbers Pty Ltd 1 1 McKay Timber 1 1 Millmerran Timbers Pty Ltd 1 1 Nannup Timber Processing (NTP) 1 1 1 Neville Smith Forest Products 2 2 2 Parkside Building Supplies Pty Ltd 2 2 2 Porta Mouldings Pty Ltd 1 1 Radial Timber Australia 1 1 Ravenshoe Timbers Pty Ltd 1 1 Ryan & McNulty Pty Ltd 1 1 Saunders Sawmill 1 1 Schiffke Sawmill Pty Ltd 1 1 Urgenty Pty Ltd trading as Mary Valley Timbers 1 1 Wade Sawmill 1 1 Total 43 43 11

Treatment is covered separately in section 4.11 rather than as part of the main inventory. This has been done because of the large number of treatment types and hazard classes available. In the final EPD, the impacts associated with treatment can be simply added to the impacts associated with the relevant untreated product. The properties of wood products included in this study are given in Table 4-34. All moisture contents are given as dry basis. A comparison is given to the baseline study and also to a CORRIM study of hardwood timber in the Northeast and North Central United States (Bergman & Bowe, 2007). Baseline (2006) refers to data collected by the Commonwealth Scientific and Industrial Research Organisation (CSIRO) that inventoried data from the 2005/06 Australian financial year (1st July 2005 to 30th June 2006) (CSIRO, 2009a), as updated by EPD Registration No. S-P-00561, Version 1, Issued 13 August 2015, Valid until 13 August 2020.

Table 4-34 Sawn hardwood material properties per m3

Hardwood properties Unit Average Baseline CORRIM DQI* (2006) NE & NC USA

3 Sequestered carbon (per m ) (green) kg CO2e 1114 1168 - L

3 Sequestered carbon (per m ) (kiln-dry) kg CO2e 1221 1283 - L Green log density kg/m3 1092 1100 854 M Green, sawn density kg/m3 768 828 854 M

LCI and EPDs for Australian Wood Products 71 of 159

Hardwood properties Unit Average Baseline CORRIM DQI* (2006) NE & NC USA Kiln-dry density kg/m3 735 784 623 M Shrinkage % 9.3 9.3 6.6 C Moisture content green logs % 79.6 72.6 86.9 C Moisture content green, sawn % 26.4 30 86.9 C Moisture content kiln-dry % 10.4 12 10.6 M Moisture content of residues burnt % 10.4 - 10.6 M NCV of green product MJ/kg 14.7 14.2 - C NCV of kiln-dried product MJ/kg 17.2 16.9 - C NCV of residues burnt MJ/kg 17.2 - - C Distance travelled by logs km 98 116 125 M * measured (M) / calculated (C) / estimated (E) / literature (L)

The material properties of outputs from a hardwood sawmill remain very similar to the previous study as there has been little change in the hardwood species grown across Australia. The average distance travelled by the logs from the forest to sawmill has decreased a small amount relative to the baseline study.

Variability between the sites included in this study was low for most of the properties listed in Table 4-34. An assessment of variability is shown in Annex A.

4.6.2. Production process

The production process of sawn hardwood and other sawmill co-products is illustrated in Figure 4-6.

Sawlogs from hardwood forestry

System boundary

Liquid fuels, water Log haulage Transport Emissions Waste (recycling Liquid fuels, water Log storage & landfill)

Hardwood timber, green Electricity, thermal energy, water, consumed materials Debarking and sawmilling Wood chips & other residues

Primary Electricity, water, thermal energy, Kiln-drying Hardwood timber, kiln-dried, rough sawn water, energy raw Burning of residues material Residues Liquid fuels, electricity, water Planing Hardwood timber, kiln-dried, dressed

Hardwood timber, kiln-dried, dressed, treated Chemicals, electricity, liquid fuels, water Treatment Hardwood timber, kiln-dried, rough sawn, treated

Packaging material, transport packaging Packaging

Figure 4-6 Process flow diagram of sawn hardwood production

LCI and EPDs for Australian Wood Products 72 of 159

Hardwood sawmills produce a range of different products from green logs depending on the number of processes applied. The sawmill data has been split into four unit processes in order to differentiate the impacts of products output from each process. These are described in sections 4.6.3 to 4.6.6 below. Indicators assessing the variability in LCI data are given in Annex A. The products within the scope of this study include:

• Green, rough-sawn, untreated timber • Kiln-dried, rough-sawn, untreated timber • Kiln-dried, rough-sawn, treated timber • Kiln-dried, dressed, untreated timber • Kiln-dried, dressed, treated timber All products are assumed to receive an equal share and type of packaging material per m3.

4.6.3. Storage, de-barking and milling

The first unit process covers log storage, debarking and milling. Logs are delivered to the mill by truck from a forest, and are stored in a log yard and moved using mobile plant. Sprinklers may be used in the log yard – particularly at hotter times of the year – to prevent the logs from drying out. From the log yard, logs are debarked, then sawn to dimension, producing green, rough-sawn timber. The sawing stage also produces lower value wood residues as co-products, which may be either sold or potentially used internally as fuel. None of the 11 participating sites was found to burn green residues on-site, electing to use residues from the dry mill for thermal energy generation instead. The process inventory has been constructed as a volume-weighted average and is given in Table 4-35. Due to the large difference in price between the co-products, inputs and emissions are allocated on an economic basis. This methodology is described in section 4.6.8.

Table 4-35 Hardwood storage, de-barking and milling inventory – per m3 of sawn wood

Flow Unit Average Baseline CORRIM NE DQI* (2006) & NC USA

Inputs Average hardwood logs m3 2.32E+00 2.80E+00 1.83E+00 M Electricity kWh 5.20E+01 1.36E+02 5.78E+01 M Thermal energy MJ - 6.00E+00 - M Diesel kg 2.85E+00 2.16E+00 2.26E+00 M LPG transport kg 5.19E-03 - - M Municipal water kg 3.03E+02 4.20E+02 - M Ground water kg 2.22E+02 - - E Lake water kg 6.67E+01 - - E Steel kg 5.00E-01 5.00E-01 - E Lubricating oils kg 3.50E-02 3.50E-02 - E Outputs Green, sawn hardwood m3 1.00E+00 1.00E+00 1.00E+00 M Bark** m3 3.52E-02 - 1.95E-01 M Sawdust** m3 4.05E-01 1.20E+00 2.65E-01 M

LCI and EPDs for Australian Wood Products 73 of 159

Flow Unit Average Baseline CORRIM NE DQI* (2006) & NC USA

Wood chips** m3 8.45E-01 6.00E-01 3.82E-01 M Waste kg 2.20E+01 5.00E-01 - M Steel scrap waste kg 5.00E-01 5.00E-01 - E Waste water municipal kg 3.59E+01 4.20E+02 - M Water discharged to land kg 3.45E+02 - - C Water vapour kg 9.67E+02 - - C * measured (M) / calculated (C) / estimated (E) / literature (L) **Given at density and MC of green sawn wood A key observation from the hardwood sawmilling data is that a lower recovery rate was observed relative to the baseline study and CORRIM study (Bergman & Bowe, 2007). The lower recovery rate results in more upstream impacts from forestry relative to a product with a higher recovery rate. Sawing recovery rates were found to range from 29-54%. Recovery rates varied depending on the size and efficiency of the facility as well as the grade and species of logs delivered and the product they were cutting for. Electricity consumption for sawing was found to be lower than in the baseline study, however, it is comparable to the electricity consumption in the CORRIM study. The types of residues created by the sawmills in this study also differ from the baseline study. A larger proportion of the residues produced in this study are woodchips, with a lower proportion of sawdust. This is likely linked to the markets available and the lower recovery rate observed for co- products, as larger sized offcuts are produced, chipped and sold as wood chip. Once cut into slabs the hardwood slabs are carefully stacked with gluts and stored onsite for a period of time to naturally dry to fibre saturation moisture content (~25-30% MC). This drying time is a function of the end dimensions size (larger slabs take longer to dry) and species.

4.6.4. Kiln-drying

Green sawn timber is the input to the second unit process: kiln-drying. Green timber is placed within a kiln, or a series of kilns, that control air-flow, relative humidity and temperature to dry the hardwood to a moisture content close to 10% (dry basis) – depending on the product being produced. Thermal energy and steam for the kiln is provided by the combustion of natural gas, LPG and/or wood residues which are usually produced from other processes by the site. Electricity is used to drive fans to control air flow. The output of the kiln-drying process is kiln-dried, rough-sawn timber. The process inventory for kiln-drying is given in Table 4-36.

Table 4-36 Hardwood kiln-drying inventory – per m3 of kiln-dried wood

Flow Unit Average Baseline CORRIM NE DQI* (2006) & NC USA

Inputs Green, sawn hardwood m3 1.10E+00 1.10E+00 1.07E+00 M Electricity kWh 2.06E+01 1.50E+01 3.08E+01 M

LCI and EPDs for Australian Wood Products 74 of 159

Flow Unit Average Baseline CORRIM NE DQI* (2006) & NC USA

Thermal energy from natural MJ 6.68E+01 2.47E+01 - M gas Thermal energy from LPG MJ 1.55E+02 - - M Thermal energy from biomass MJ 8.99E+02 2.20E+03 3.85E+03 C Wood residues burnt kg 4.75E+01 C (10.4% MC) Diesel kg 6.03E-01 6.67E-03 4.04E-01 M LPG transport kg 3.87E-02 - - M Municipal water kg 1.89E+02 2.80E+02 - M Ground water kg 1.15E+02 - - E Lake water kg 8.14E+00 - - E

Outputs Kiln-dried, sawn hardwood m3 1.00E+00 1.00E+00 1.00E+00 M Waste water kg - 2.80E+02 - E Water vapour kg 4.18E+02 - - C Acetic acid kg 1.30E-02 1.30E-02 L Acrolein kg 4.90E-03 4.90E-03 L Aldehyde (unspecified) kg 8.70E-03 8.70E-03 L Benzaldehyde kg 1.50E-04 1.50E-04 L Butyraldehyde kg 2.60E-04 2.60E-04 L Crotonaldehyde kg 2.60E-04 2.60E-04 L Ethanol kg 7.50E-02 7.50E-02 L Formaldehyde (methanal) kg 1.10E-03 1.10E-03 L Formic acid (methane acid) kg 7.70E-03 7.70E-03 L Methanol kg 2.90E-02 2.90E-02 L Propanil kg 9.30E-04 9.30E-04 L VOC (unspecified) kg 3.00E-01 3.00E-01 L * measured (M) / calculated (C) / estimated (E) / literature (L)

A key difference from kiln-drying compared to the baseline and CORRIM study (Bergman & Bowe, 2007) is a decrease in the total amount of thermal energy used during the kiln-drying process. The dry mills included within the study were also found to use less thermal energy for kiln-drying than those that operated as combined green- and dry-mills. Two of the three included dry-mills used natural gas or LPG as a thermal energy source for the kilns which had a lower thermal energy use per cubic metre of wood dried. The study also included a site which used solar kilns for drying most of its wood. This contributes to the lower observed energy consumption overall. All other sites included in the study used biomass residues created during the planing stage as the thermal energy source for the kilns. Diesel usage for transport in the kiln-drying stage previously reported in the baseline study is expected to be an error, as it is significantly lower than the value reported in both the current study and the CORRIM study.

LCI and EPDs for Australian Wood Products 75 of 159

4.6.5. Planing

Sawn, kiln-dried hardwood may enter a third unit process: planing. During this stage wood is planed (or moulded) and sawn to produce a smooth surface finish and final dimension. This process also produces valuable co-products that may be either sold or used to provide thermal energy to the kiln- drying process. Due to the large difference in price between the co-products, inputs and emissions are allocated on an economic basis. This methodology is described in section 4.6.8. The main output from this process is kiln-dried, dressed timber. The process inventory for planing is given in Table 4-37.

Table 4-37 Hardwood planing inventory – per m3 of planed wood

Flow Unit Average Baseline CORRIM NE DQI* (2006) & NC USA

Inputs Kiln-dried, sawn hardwood m3 1.26E+00 1.29E+00 1.33E+00 M Diesel kg 2.10E+00 3.56E-01 1.10E+00 M Electricity kWh 5.79E+01 4.31E+01 3.38E+01 M Gasoline kg - 3.31E-02 - M LPG transport kg 1.79E-02 - - M Natural Gas MJ - 9.18E-02 - M Outputs Kiln-dried, dressed hardwood m3 1.00E+00 1.00E+00 1.00E+00 M Wood chips** kg 2.18E+01 - - M Hogfuel kg - 1.21E+02 8.22E-02 E Sawdust** kg 1.86E+01 2.19E+01 8.60E-02 M Shavings** kg 6.43E+01 8.57E+01 1.61E-01 M Wood residues to kiln** kg 7.57E+01 - - C Waste to landfill kg 1.12E+01 - - M Water discharged to land kg 3.06E-01 - - C Water vapour kg 8.58E-01 - - C * measured (M) / calculated (C) / estimated (E) / literature (L) **Wood residues given at MC of kiln-dried wood (10.4%) The planing recovery rate observed was higher than the recovery rates reported in the baseline and CORRIM study (Bergman & Bowe, 2007). This varied by site from 66-86%. Planing residues were treated in different ways depending on the site. Residues were burnt for thermal energy by most of the included sites with any remainder sold. Some of the sites burnt all of their planing residues. One site was found to be send all of their planing residues to a landfill but were in the process of buying equipment to utilise this residue in saleable co-products within 12 months of the survey. Energy consumption was found to be higher for the current study than in the baseline study and CORRIM study. A number of sites were examining their electricity usage as a result of recent price increases.

LCI and EPDs for Australian Wood Products 76 of 159

4.6.6. Treatment

The final unit process is timber treatment. A range of different preservative treatment types may be used to treat the wood depending on the suitability and hazard class requirement. Treatment is applied to the surface of the wood or within pressurised chambers depending on the requirements of the treatment type used. Preservative treatment may be applied to either kiln-dried, rough sawn timber or kiln-dried, dressed timber. Alternatively, timber may be sold without preservative treatment as untreated timber. Treatment of sapwood of susceptible species (to H1 and H3 hazard levels) occurs in mills in NSW and Queensland. Preservative treatment was not occurring at mills in Southern States or WA. Detailed information about wood preservative treatment can be found in section 4.9.

4.6.7. Packaging

Average amounts of packaging materials used by each site per cubic metre of wood is given in Table 4-38.

Table 4-38 Hardwood packaging – per m3 of wood packaged

Flow Unit Average DQI

Wood packaged m3 1.00E+00 M Softwood gluts m3 1.64E-03 M Hardwood gluts m3 1.40E-03 M LDPE wrap kg 2.29E-01 M Plastic strapping kg 1.57E-01 M Steel strapping kg 1.41E-01 M Paper labels kg 9.94E-01 M Inks kg 1.67E-03 M Paints kg 9.33E-02 M Pallets No. 3.42E-04 M * measured / calculated / estimated / literature

Packaging data were not included within the scope of the baseline study; therefore, no comparison is available. Packaging materials are assumed to be used evenly on all products produced per site per cubic metre of output.

4.6.8. Allocation

Allocation to the co-products of sawmilling is based on economic value as the difference in the value of co-products is large (>25%, in line with IEPDS (2017, section 7.7)). Most co-products have some economic value and therefore cannot be considered as wastes. Economic values have been provided by participants and are displayed as volume-weighted averages in Table 4-39 for sawing and Table 4-40 for planing.

LCI and EPDs for Australian Wood Products 77 of 159

Table 4-39 Sawing co-product allocation factors

Sawing Units Amount Average Economic Allocation (Units) price value ($) (%) ($/Unit)

Green, sawn hardwood m3 1 596 596 90.98% Bark* m3 0.035 58.8 2.07 0.32% Woodchips (green)* m3 0.845 46.8 39.5 6.03% Sawdust (green)* m3 0.405 43.4 17.6 2.68%

Total 655 100% *Given at density and MC of green sawn wood

Volumes displayed are given at the density of the sawn wood. Prices have been converted to match the densities using the expansion factors given in Table 4-26.

Table 4-40 Planing co-product allocation factors

Planing Units Amount Average Economic Allocation (Units) price value ($) (%) ($/Unit)

Kiln-dried, dressed hardwood m3 1 1,533 1533 99.4% Woodchips (kiln-dry)* kg 22 0.066 1.43 0.09% Sawdust (kiln-dry)* kg 19 0.056 1.04 0.07% Shavings (kiln-dry)* kg 64 0.037 2.37 0.15% Wood residues to kiln (kiln-dry)* kg 76 0.062 4.77 0.31%

Total 1543 100% *Wood residues given at MC of kiln-dried wood (10.4%)

As the majority of the sites surveyed burnt most of the residues produced on-site, they were not able to provide sales prices for these residues. Where dry residues were sold and burnt on-site, the value of the burnt residues has been taken to be the same as the lower value residue sold. For combined green and dry mills, the price of the residues has been taken as equal to the green residue price per dry mass. Where dry mills were not able to provide a price for residues, a price has been calculated as a mass-weighted average of the remaining sites.

4.6.9. Assumptions

Assumptions made during the construction of the sawn hardwood LCI are listed below.

General assumptions

All data received from manufacturers have been accepted as accurate (after cross-checks for mass balance, plausibility, etc.). Energy used is reported by energy source (e.g. coal, diesel fuel and electricity), but the upstream profiles of energy sources are modelled based on Australian averages. Upstream profiles for intermediate products that are consumed in small quantities, such as lubricants, are modelled based on available LCI data.

LCI and EPDs for Australian Wood Products 78 of 159

Where onsite dams are used, water use has been calculated as follows: annual, site specific rainfall on dam area is reported as lake water use and is taken from the nearest Bureau of Meteorology measurement station (BoM, 2017). Dam evaporation has been calculated using the tool Readyreckoner (NCEA, 2017). This has been used to calculate the net evaporation relative to a similar area of pasture land with the same location and annual rainfall using the Zhang relationship (Zhang, et al., 2001). Net use of ground water has been calculated to balance water according to the following: ground water in + rainfall (lake water) = water use from dam + net evaporation.

Computational assumptions

Input and output materials and processes are a production-weighted average based on the data provided from Australian manufacturers.

Storage, de-barking and milling

For the distance a mass-weighted average distance travelled by logs from forest to sawmill has been estimated by each sawmill. Where a sawmill was not able to provide an estimate, a distance of 100 km has been used. Mass of product and residues sold by each site are assumed to be accurate. Any difference in mass between log input and sold output has been assumed to be combusted on site during the kiln-drying process. Sawn green timber output from sawing process is based on the volume of purchased logs and site sawing recovery rates (yield). Where a water input exceeded water discharged to a drain or treatment facility (water imbalance), remaining water is assumed to be discharged to land. The relationship stated by Zhang et al. (2001) has been used to estimate the water returning to groundwater and water evaporated using the annual precipitation in the nearest major city (ABS, 2017). Where residues from sawing were reported by participating sites on a bulk (expanded) volumetric basis, the measurement was assumed accurate. Densities provided by the participants had a high level of uncertainty, therefore expansion factors were used for each residue type as given in Table 4-26 based on the site’s sawn wood density to convert to mass.

Kiln-drying

All residues produced but not sold from either the sawing or planing processes have been assumed to be burnt for thermal energy used in the kiln-drying process. Shrinkage during kiln-drying is calculated from the density and moisture content of wood entering the kiln relative to the density and moisture content of the wood exiting the kiln. Where a water input exceeded water discharged to a drain or treatment facility (water imbalance), remaining water is assumed to be evaporated, including moisture leaving the wood during drying. Some of the surveyed sites reported a small amount of waste being sent to landfill originating from the kiln-drying phase. This has not been included in the study due to a lack of sufficient data on the type of waste and the very small volume of waste involved.

Planing

Mass of product and residues sold by each site are assumed to be accurate. Any difference in mass between rough-sawn input and sold output has been assumed to be combusted on site during the kiln-drying process.

LCI and EPDs for Australian Wood Products 79 of 159

Rough sawn input into the planing process is based on the volume of dressed product output and site planing recovery rates (yield). Where a water input exceeded water discharged to a drain or treatment facility (water imbalance), remaining water is assumed to be discharged to land. The relationship stated by Zhang et al. (2001) has been used to estimate the water returning to groundwater and water evaporated using the annual precipitation in the nearest major city (ABS, 2017). Where residues from planing were reported by participating sites on a bulk (expanded) volumetric basis, the measurement was assumed accurate. Densities provided by the participants had a high level of uncertainty, therefore expansion factors were used as given in Table 4-26 for each residue type based on the site’s kiln-dried wood density to convert to mass.

Packaging

Packaging materials such as wooden gluts are assumed to be carbon neutral with respect to sequestered carbon, as the installation phase (module A5), where the sequestered carbon would be released, is not declared within this study. When modelling the production of paper, any scrap paper used has been assumed to be burden free. Inherent properties: biogenic carbon and primary energy have been accounted for within the dataset. Similarly, no credit has been applied for any waste paper that is recycled in the product life cycle. This methodology of cutting-off impacts has been applied throughout the model as it was not possible to credit recycled paper as no consistent, high-quality data were available for 100% primary paper production and 100% secondary paper production. The carbon balance has been corrected via CO2 emissions (biotic), assuming decomposition or incineration in the time frame of 100 years.

Electricity grid mix

The electricity grid mix for sawn hardwood production (A1-A3) has been modelled based on the state-share of Australian sawn hardwood production as a percentage of volume based on ABARES 2012/2013 data (ABARES, 2017). The same electricity grid mix for sawn hardwood has also been applied to the hardwood forestry stage. Electricity at end-of-life (module C) has been modelled using an average Australian electricity mix as the location where the product reaches end-of-life is unknown. The electricity grid mixes used for each product are given in Annex E.

4.7. Particleboard

4.7.1. Overview

The following particleboard products are included within the LCI:

• 1 m2 of particleboard, 16 mm E0 & E1 standard melamine coated • 1 m2 of particleboard, 18 mm E0 & E1 standard melamine coated • 1 m2 of particleboard, 16 mm E0 & E1 moisture resistant (MR) melamine coated • 1 m2 of particleboard, 18 mm E0 & E1 moisture resistant (MR) melamine coated • 1 m2 of particleboard, 19 mm flooring (tongue & groove) • 1 m2 of particleboard, 22 mm flooring (tongue & groove) • 1 m2 of particleboard, 25 mm flooring (tongue & groove) LCI data were collected using detailed surveys of three of the seven particleboard plants in Australia. These three plants collectively produced 514,000 m3 of particleboard in 2015/16, equating

LCI and EPDs for Australian Wood Products 80 of 159

to 54% of total Australian production of approximately 955,000 m3 (based on the 2015/16 total from ABARES (2017)). Table 4-41 indicates the three companies that own these seven plants. All companies listed have contributed financially to this project through their FWPA levies. However, the specific companies that contributed data cannot be identified as there are only three producers.

Table 4-41 Contributors to particleboard LCI

Company Total Financial Data plants contributor contributor Carter Holt Harvey Woodproducts Australia 4 4 Confidential D&R Henderson Pty Ltd 1 1 Confidential The Laminex Group 2 2 Confidential Total 7 7 3

Table 4-42 shows the material composition of the different particleboard products according to the product-specific bills of material.

Table 4-42 Material composition of the particleboard products [%]

PB Floor PB Floor PB Floor PB MR PB MR PB Std. PB Std.

19mm 22mm 25mm 16mm 18mm 16mm 18mm Softwood (dry) 77.6 78.6 78.6 83.5 83.4 84.1 84.2 Urea formaldehyde 0.0 0.0 0.0 0.4 0.2 5.9 7.3 Melamine 0.0 0.0 0.0 0.8 0.6 0.8 0.6 formaldehyde Melamine urea 10.9 10.6 10.7 6.0 7.6 0.0 0.0 formaldehyde Paraffin wax 0.8 0.9 0.9 0.7 0.6 0.5 0.4 Lamination paper 0.0 0.0 0.0 1.3 1.1 1.3 1.1 (dry) Tongue 0.5 0.5 0.5 0.0 0.0 0.0 0.0 (polypropylene) Ammonium 0.3 0.0 0.3 0.2 0.2 0.1 0.2 sulphate Tannin 0.9 0.8 0.7 0.0 0.0 0.0 0.0 Preservative 0.0 0.0 0.0 0.0 0.0 0.0 0.0 Dyes 0.0 0.0 0.0 0.0 0.0 0.0 0.0 Water 8.9 8.6 8.4 7.2 6.3 7.3 6.2

Tannin, preservatives and dyes were neglected for the calculation of the elemental composition as data were unavailable and the shares are insignificant. The shares of the other materials were scaled up to account for the neglected elements in a worst-case approach. The elemental compositions of the different wood products were calculated according to the sum of the elemental composition of its elements. Table 4-43 shows the elemental composition including the incorporated water and Table 4-44 shows the elemental composition of the bone-dry product.

LCI and EPDs for Australian Wood Products 81 of 159

Table 4-43 Elemental composition including water

PB Floor PB Floor PB Floor PB MR PB MR PB Std. PB Std.

19mm 22mm 25mm 16mm 18mm 16mm 18mm C (biogenic) 39.16 39.63 39.58 42.33 42.18 42.65 42.58 C (fossil) 6.01 5.87 5.90 3.66 4.17 2.67 3.04 H 6.50 6.49 6.47 6.39 6.32 6.40 6.33 N 2.40 2.26 2.34 1.79 2.03 2.65 3.17 S 0.07 0.00 0.06 0.04 0.05 0.03 0.04 O 45.85 45.74 45.63 45.69 45.18 45.51 44.76 Cl 0.00 0.00 0.00 0.00 0.00 0.00 0.00 Ash 0.00 0.00 0.00 0.10 0.08 0.10 0.08

Table 4-44 Elemental composition excluding water

PB Floor PB Floor PB Floor PB MR PB MR PB Std. PB Std.

19mm 22mm 25mm 16mm 18mm 16mm 18mm C (biogenic) 42.99 43.37 43.20 45.64 45.02 46.03 45.39 C (fossil) 6.60 6.42 6.44 3.94 4.45 2.88 3.24 H 6.03 6.03 6.03 6.01 5.99 6.02 6.01 N 2.64 2.48 2.56 1.93 2.16 2.86 3.38 S 0.08 0.00 0.07 0.05 0.05 0.03 0.05 O 41.56 41.62 41.63 42.32 42.23 42.07 41.85 Cl 0.00 0.00 0.00 0.00 0.00 0.00 0.00 Ash 0.00 0.00 0.00 0.11 0.09 0.11 0.09

Ammonium sulphate, tannin, preservatives and dyes were neglected for the calculation of the elemental composition as data were unavailable and the shares are insignificant. No additional heating value was added for these materials to apply a worst-case approach. The net calorific values of softwood and paper were calculated from the gross calorific values and the corresponding water and hydrogen content. The net calorific value of the total wood products is calculated as the sum of the values for its components. The results are shown in Table 4-45.

Table 4-45 Net calorific values of softwood, paper and particleboard products

PB Floor PB Floor PB Floor PB MR PB MR PB Std. PB Std.

19mm 22mm 25mm 16mm 18mm 16mm 18mm Softwood (dry) 17.3 17.3 17.4 17.6 17.8 17.6 17.9 Lamination paper (dry) 16.1 16.2 16.2 16.4 16.6 16.4 16.7 Total 16.9 17.0 17.0 17.5 17.6 17.5 17.7

4.7.2. Production process

The production process of particleboard products is illustrated in Figure 4-7. Material inputs such as wood, resins & waxes etc. were collected as a Bill of Materials (BOM) for each product, along with total production volumes of each product. Energy, water and waste data were collected as totals per

LCI and EPDs for Australian Wood Products 82 of 159

site, as reliable data was not available per individual product. Total material inputs were also requested for each site along with the site’s total particleboard output. Allocation of energy, water and waste to each product is described in section 4.7.4.

Pulp logs from softwood forestry

Post-consumer recycled wood Wood chips from sawmill

System boundary Electricity, water, Debarking, chipping, flaking and thermal energy screening Combustion of residues

Thermal energy Drying

Primary Resins, wax, electricity Blending energy, water, raw Chemicals, electricity, material liquid fuels, water Treatment (flooring)

Emissions Electricity Mat forming Waste (recycling & Chemicals, electricity, landfill) thermal energy Pressing and curing

Electricity Finishing Particle board, flooring products Cooling, trimming and sanding Combustion of residues Paper, resin, thermal Particle board energy Lamination

Packaging material, transport packaging Packaging

Figure 4-7 Process flow diagram of particleboard production

Particleboard manufacturing consists of a number of activities (depicted in Figure 4-7):

• Production of particles and fibres • Drying (reducing moisture content to 3-5 %) • Classifying • Blending fibres and resin/wax • Addition of treatment • Mat-forming • Pressing into boards and curing of resin • Finishing, trimming and sanding of both surfaces • Laminating • Packaging

Particleboard plants also have a boiler that produces heat for various processes in the manufacturing operation. The boiler is usually fuelled by wood waste from the production processes, natural gas or purchased biomass. Lamination of particleboards can be done onsite or offsite. It is an additional process to the particleboard manufacturing of some products, consisting of:

• Paper treatment (impregnating paper with resins) • Short cycle pressing (bonding paper with the boards) • Trimming and packaging

The particleboard plants providing data for this project did not have sufficient sub-metering to split the manufacturing process into a large number of unit processes. Therefore, the particleboard

LCI and EPDs for Australian Wood Products 83 of 159

manufacturing process has been treated as a black box, where only inputs and outputs crossing the system boundary have been defined.

4.7.3. Manufacturing

Due to the small number of producers and data contributors, and the possibility of disaggregation of confidential information, detailed production information has been omitted from this report. The variability in the properties of particleboard between sites is very low with the exception of the moisture content of some inputs. Key differences to the baseline study include a reduction in resin and wax loadings. Resin loadings data were known to include a high level of uncertainty in the baseline study and were likely overestimated based on a conservative assumption. Energy consumption overall has seen a small reduction relative to the baseline study, with the greatest reduction being from natural gas. Inputs and outputs with high variability between participating sites are often due to alternative material types or energy sources being used by each site e.g. more pulplog is used by some sites whereas others use more low-moisture from sawmills or post-consumer recycled wood.

Production of particles and fibres

Particleboard raw material may be round wood, such as forest thinnings or peeler cores from plywood mills and dockings, or sawmill residues ranging from green woodchips and dockings to planer shavings and sawdust. Post-consumer recycled wood such as end-of-life pallets as well as offcuts from frame and truss manufactures may also be used as an input in some cases. A wide range of chippers, flakers and size reduction mills are used to convert the different raw materials to the required particles. Screening is used to ensure tight control of final particle size, with oversize being returned for further breakdown while material that is too fine is usually used as an energy source for drying. For the purposes of this study, inputs into the particleboard have been categorised into three grouped categories: wood from a forest, wood residues from a sawmill and post-consumer recycled wood. Wood directly from a forest has been treated as softwood pulplogs, with the associated economic allocation between forestry products as described in section 4.3.6. Wood from a sawmill has been treated as green woodchip, a co-product from the sawmilling stage of sawn softwood production. The allocation production and upstream impacts has been assigned based on economic allocation at the sawmill as described in section 4.5.8. Post-consumer recycled wood input has been considered to be free of upstream impacts and to have reached the end of waste state at the point of collection.

Drying

In particleboard plants, particles are passed through driers that reduce their moisture content to 3- 5%. Most modern driers are direct-fired in that the particles are dried by direct contact with hot gas from the burners. Particles are then passed through screens and wind-sifters to sort the furnish into various size fractions, to remove heavy or thick particles or excessive dust and to grade the particles into sizes if layered boards are produced.

LCI and EPDs for Australian Wood Products 84 of 159

Blending (glue addition)

Resin, in the form of a liquid, is forced through nozzles and sprayed onto the particles in a separate blender. Prior to entering the nozzles, the other glue additives (including preservative treatment for flooring products) are correctly proportioned and added to the glue flow and mixed thoroughly. Wood furnish passes through weighing devices that automatically control the resin flow rate. Continual checks on flow rates and particle moisture contents ensure consistent blending with moisture contents increased to 10-16 % after blending.

Mat forming

Board quality depends on the quality of mat forming. For particleboard, glued particles pass from a silo into a weighing device that meters the correct weight of flakes onto a moving belt. Various rotating rakes distribute these evenly across the belt. This controlled wall of flakes falls from the belt cascading onto the mat-forming device that is also moving at controlled speed. The mat is built up to weight either by several passes under these spreaders or by one pass under several spreaders and is layered upon the mat transfer belt or caul which carries the mats to the hot press. There are various types of mat (single-layer, three-layer and multi-layer). Three-layer or sandwich construction of the mat is the most common industry practice giving an opportunity to tailor the board characteristics, e.g. fine surface for pre-finishing, flakes for strength. The two outside layers of particles are fine and contain more glue and moisture than the core layer of coarser particles. When pressed, the surfaces of the boards are of higher density than the core.

Pressing

The preformed mats of glued particles and fibres are transferred to the hot press for pressing and curing. This operation is critical and requires carefully controlled heat, pressure and timing. Prior to hot pressing, the mats may be pre-pressed cold to reduce their thickness and to make them easier to transport. Pressing can be by batch i.e. mats formed then pressed using single daylight presses or in groups in a multi daylight press. Single daylight presses take large single boards at each pressing cycle, while the multi daylight operation presses many boards at once. The thick mats are compressed in the press with thickness controlled either by thickness bars (stops) or other electronic thickness measuring devices. As soon as heat is applied, the glue curing process begins and full pressure is quickly applied to reach the desired thickness before cure. Full pressure at stops is held for the prescribed time then pressure is slowly reduced until the press is opened. Airing cycles are important to allow the steam generated to escape thus preventing “blown boards”. Typical pressures are 2-3 MPa, temperatures 140-220 °C and press-time 6-15 seconds per mm of board thickness plus the opening/closing times of the press.

Finishing

The hot boards are removed from the press (or sawn across on continuous presses) and further conditioned to equilibrate board moisture content and to stabilise and fully cure the resin. This conditioning usually follows cooling in star coolers for boards with urea formaldehyde resins. Panels are then trimmed and sanded on both faces to tight thickness tolerances. Sanded sheets are sawn to stock sizes or to suit special orders.

LCI and EPDs for Australian Wood Products 85 of 159

4.7.4. Allocation

Energy, water and waste inputs to each product are allocated based on the total dry mass of wood, dry mass of resin (including resin in any laminate) and wax in each product. Minor additives such as dyes, preservatives, tannin, paper and tongue were not a factor in the allocation due to their small volume, high variation between product types and are expected to have a relatively minor influence over the energy and water requirements of each product compared to wood, resins and wax.

4.7.5. Assumptions

Assumptions made during the construction of the particleboard LCI are listed below.

General assumptions

All data received from manufacturers have been accepted as accurate (after cross-checks for mass balance, plausibility, etc.) Energy used is reported by energy source (e.g. coal, diesel fuel and electricity), but the upstream profiles of energy sources are modelled based on Australian averages. Insufficient data were obtained for inbound transport of raw material inputs. A travel distance of 100 km has therefore been assumed for all inputs.

Manufacturing

Where a water input exceeded water discharged to a drain or treatment facility (water imbalance), remaining water is assumed to be evaporated, including water contained in wood and residue inputs Any difference between input mass and output mass has been assumed to be scrap or offcuts burnt on-site for thermal energy purposes. The total use of thermal energy for the entire site and associated emissions has been allocated based on the input mass of key inputs as described in section 4.7.4. Upstream resin manufacture has been modelled based on available GaBi datasets. The composition of resins within the product has been calculated according to Annex F. The combustion of wood residues and offcuts are assumed to contain resin; therefore, emissions are also based on the compositions described in Annex F. When modelling the production of paper, any scrap paper used has been assumed to be burden free. Inherent properties: biogenic carbon and primary energy have been accounted for within the dataset. Similarly, no credit has been applied for any waste paper that is recycled in the product life cycle. This methodology of cutting-off impacts has been applied throughout the model as it was not possible to credit recycled paper as no consistent, high-quality data were available for 100% primary paper production and 100% secondary paper production. The carbon balance has been corrected via CO2 emissions (biotic), assuming decomposition or incineration in the time frame of 100 years.

Packaging

Packaging materials such as wooden gluts are assumed to be carbon neutral with respect to sequestered carbon, as the installation phase (module A5), where the sequestered carbon would be released, is not declared within this study.

LCI and EPDs for Australian Wood Products 86 of 159

Computational assumptions

Input and output materials and processes are a production-weighted average based on the data provided from Australian manufacturers.

Electricity grid mix

The electricity grid mix for particleboard manufacture (A1-A3) has been modelled based on the percentage of production occurring in each state by volume. Production estimates by states are based primarily on data collected as part of this study and published values for the remaining sites which did not take part in the study. The same electricity grid mix particleboard has also been applied to upstream processes such as softwood forestry stage and sawmills providing woodchip. Electricity at end-of-life (module C) has been modelled using an average Australian electricity mix as the location where the product reaches end-of-life is unknown. The electricity grid mixes used for each product are given in Annex E.

4.8. Medium Density Fibreboard (MDF)

4.8.1. Overview

The following medium density fibreboard (MDF) products are included within the LCI:

• 1 m2 of MDF, 16 mm E0 & E1 standard melamine coated • 1 m2 of MDF, 18 mm E0 & E1 standard melamine coated • 1 m2 of MDF, 25 mm E0 & E1 standard melamine coated • 1 m2 of MDF, 16 mm E0 & E1 moisture resistant (MR) melamine coated • 1 m2 of MDF, 18 mm E0 & E1 moisture resistant (MR) melamine coated • 1 m2 of MDF, 25 mm E0 & E1 moisture resistant (MR) melamine coated LCI data were collected using detailed surveys of all three MDF plants operating in Australia, as shown in Table 4-46. These plants collectively produced 615,708 m3 MDF in 2015/16, which equates to 100% of Australian domestic production.

Table 4-46: Contributors to MDF LCI

Producer company Total Financial Data plants contributor contributor Alpine MDF Industries Pty Ltd 1 1 1 Borg Panels 1 1 1 The Laminex Group 1 1 1 Total 3 3 3

The scope of this study covers a range of different MDF product types. The full list of product inventories is displayed in Annex A. All LCI tables and comparisons within this section refer to a sample MDF product: moisture resistant (MR), melamine coated, 18 mm. This product has been chosen as it is the product produced in greatest volume that has a comparison available from the baseline study. The material properties of moisture resistant (MR), melamine coated, 18 mm MDF are shown in Table 4-47. Baseline (2006) refers to data collected by the Commonwealth Scientific and Industrial Research Organisation (CSIRO) that inventoried data from the 2005/06 Australian financial year (1st July 2005

LCI and EPDs for Australian Wood Products 87 of 159

to 30th June 2006) (CSIRO, 2009a), as updated by EPD Registration No. S-P-00563, Version 1, Issued 21 October 2015, Valid until 21 October 2020.

Table 4-47 Moisture resistant (MR), melamine coated, 18 mm – material properties

Property Unit Average Baseline DQI* (2006)

Board density kg/m3 7.21E+02 7.42E+02 M Board thickness mm 1.80E+01 1.80E+01 M Product moisture content (MC) % 7.11E+00 - M MC of pulplogs % 1.11E+02 - M MC of wood chips % 9.42E+01 - M MF resin water content % 0.00E+00 - M MUF resin water content % 2.91E+01 - M UF resin water content % 0.00E+00 - M Yield % 8.40E+01 - C Distance travelled by inputs km 1.00E+02 - E * measured (M) / calculated (C) / estimated (E) / literature (L)

The variability in the properties of moisture resistant, melamine coated, 18mm MDF is very low between sites as shown in Annex A with the exception of the moisture content of some inputs. Table 4-48 shows the material composition of the different MDF products according to the product- specific bills of material.

Table 4-48 Material composition of the MDF products [%]

MDF MR MDF MR MDF MR MDF Std. MDF Std. MDF Std.

16mm 18mm 25mm 16mm 18mm 25mm Softwood (dry) 80.8 81.0 80.3 81.8 82.4 80.7 Urea formaldehyde 0.0 0.0 0.0 4.7 7.2 3.3 Melamine 0.7 0.6 0.6 0.8 0.7 0.6 formaldehyde Melamine urea 10.3 10.5 11.5 4.7 1.9 7.7 formaldehyde Paraffin wax 0.7 0.6 0.6 0.6 0.5 0.6 Lamination paper 0.6 0.6 0.5 0.7 0.6 0.5 (dry) Tongue 0.0 0.0 0.0 0.0 0.0 0.0 (polypropylene) Ammonium sulphate 0.0 0.0 0.0 0.0 0.0 0.0 Tannin 0.0 0.0 0.0 0.0 0.0 0.0 Preservative 0.0 0.0 0.0 0.0 0.0 0.0 Dyes 0.0 0.0 0.0 0.0 0.0 0.0

LCI and EPDs for Australian Wood Products 88 of 159

MDF MR MDF MR MDF MR MDF Std. MDF Std. MDF Std.

16mm 18mm 25mm 16mm 18mm 25mm Water 6.8 6.6 6.6 6.7 6.6 6.6

Tannin, preservatives and dyes were neglected for the calculation of the elemental composition as data were unavailable and the shares are insignificant. The shares of the other materials were scaled up to account for the neglected elements in a worst-case approach. The elemental compositions of the different wood products were calculated according to the sum of the elemental composition of its elements. Table 4-49 shows the elemental composition including the incorporated water and Table 4-50 shows the elemental composition of the bone-dry product.

Table 4-49 Elemental composition including water

MDF MR MDF MR MDF MR MDF Std. MDF Std. MDF Std.

16mm 18mm 25mm 16mm 18mm 25mm C (biogenic) 40.71 40.78 40.36 41.20 41.49 40.61 C (fossil) 5.40 5.39 5.79 4.44 3.92 5.20 H 6.33 6.31 6.30 6.34 6.33 6.32 N 2.50 2.52 2.69 3.19 3.54 3.18 S 0.00 0.00 0.00 0.00 0.00 0.00 O 45.01 44.96 44.82 44.78 44.66 44.65 Cl 0.00 0.00 0.00 0.00 0.00 0.00 Ash 0.05 0.04 0.04 0.05 0.05 0.04

Table 4-50 Elemental composition excluding water

MDF MR MDF MR MDF MR MDF Std. MDF Std. MDF Std.

16mm 18mm 25mm 16mm 18mm 25mm C (biogenic) 43.69 43.68 43.20 44.18 44.43 43.47 C (fossil) 5.80 5.77 6.20 4.76 4.19 5.57 H 5.97 5.96 5.96 5.99 5.99 5.97 N 2.68 2.70 2.88 3.42 3.80 3.41 S 0.00 0.00 0.00 0.00 0.00 0.00 O 41.80 41.84 41.72 41.60 41.53 41.54 Cl 0.00 0.00 0.00 0.00 0.00 0.00 Ash 0.05 0.05 0.04 0.06 0.05 0.04

Ammonium sulphate, tannin, preservatives and dyes were neglected for the calculation of the elemental composition as data were unavailable and the shares are insignificant. No additional heating value was added for these materials to apply a worst-case approach. The net calorific values of softwood and paper were calculated from the gross calorific values and the corresponding water and hydrogen content. The net calorific value of the total wood products is calculated as the sum of the values for its components. The results are shown in Table 4-51.

LCI and EPDs for Australian Wood Products 89 of 159

Table 4-51 Net calorific values of softwood, paper and MDF products

MDF MR MDF MR MDF MR MDF Std. MDF Std. MDF Std.

16mm 18mm 25mm 16mm 18mm 25mm Softwood (dry) 17.7 17.8 17.8 17.7 17.8 17.8 Lamination paper (dry) 16.5 16.6 16.6 16.5 16.6 16.6 Total 17.5 17.5 17.5 17.5 17.5 17.5

4.8.2. Production process

The production process of MDF products is illustrated in Figure 4-8. Material inputs such as wood, resins & waxes etc. were collected as a BOM for each product, along with total production volumes of each product. Energy, water and waste data were collected as totals per site, as reliable data were not available per individual product. Total material inputs were also requested for each site along with the site’s total MDF output. Allocation of energy, water and waste to each product is described in section 4.8.5. Indicators assessing the variability in LCI data are given in Annex A

Pulp logs from softwood forestry Wood chips from sawmill

System boundary

Electricity, water, thermal energy Debarking, chipping and screening Combustion of residues

Primary Thermal energy Drying energy, water, raw material Resins, wax, electricity Refining and blending

Emissions Electricity Mat forming Waste (recycling Chemicals, electricity, & landfill) thermal energy Pressing and curing

Electricity Finishing Cooling, trimming and sanding Combustion of residues Paper, resin, thermal Medium Density Fibreboard energy Lamination

Packaging material, transport packaging Packaging

Figure 4-8 Process flow diagram of MDF production

MDF manufacturing consists of a number of activities, depicted in Figure 4-8:

• Debarking and chipping of logs • Screening and washing • Refining • Blending • Drying • Blending fibres and resin/wax • Mat forming • Pre-pressing • Hot-pressing into boards and curing of resin • Trimming, sanding of both surfaces • Laminating • Packaging

LCI and EPDs for Australian Wood Products 90 of 159

MDF plants also have a boiler that produces heat and steam for various processes in the manufacturing operation. The boiler is usually fuelled by wood waste from the production processes, natural gas or purchased biomass. Lamination of MDF can be done onsite or offsite. It is an additional process to the particleboard manufacturing, consisting of:

• Paper treatment (impregnating paper with resins) • Short cycle pressing (bonding paper with the boards) • Trimming and packaging.

The MDF plants providing data for this project did not have sub-metering to split the manufacturing process into a large number of unit processes. Therefore, the manufacturing process has been treated as a black box, where only inputs and outputs crossing the system boundary have been defined.

4.8.3. Manufacturing

Table 4-52 shows inputs and outputs for the production of 1 m2 of moisture resistant (MR), melamine coated, 18 mm MDF. The manufacturing process is treated as a “black box” with inputs and outputs crossing the system boundary included. A comparison is given to the baseline study. Indicators assessing the variability in LCI data are given in Annex A.

Table 4-52 Moisture resistant (MR), melamine coated, 18 mm – inventory per m2 of board

Type Flow Unit Average Baseline DQI* (2006) Inputs Wood chips (bone dry) kg 4.52E+00 1.10E+01 M Moisture contained in wood kg 4.26E+00 5.40E+01 M (94.2% MC) Pulp logs (bone dry) kg 8.01E+00 3.02E+00 M Moisture contained in wood kg 8.85E+00 5.40E+01 M (111% MC) UF mass (wet) kg 0.00E+00 0.00E+00 M MUF mass (wet) kg 2.30E+00 3.35E+00 M MF mass (wet) kg 1.00E-01 4.36E-01 M Mass of wax kg 9.09E-02 3.26E-01 M Electricity kWh 6.84E+00 6.56E+00 M Thermal energy from biomass MJ 4.55E+01 4.64E+01 C Thermal energy from LPG MJ 1.02E-02 0.00E+00 M Thermal energy from natural gas MJ 1.16E+01 2.39E+01 M Diesel kg 2.30E-02 3.65E-03 M Petrol kg 9.84E-05 - M LPG (transport) kg 9.15E-04 3.23E-03 M Municipal water kg 1.83E+00 1.48E+01 M Ground water kg 1.27E+00 - C Lake water kg 1.42E-01 - C

LCI and EPDs for Australian Wood Products 91 of 159

Type Flow Unit Average Baseline DQI* (2006) Rain water kg 2.23E+00 - M Paper kg 8.70E-02 1.77E-01 C Outputs MDF produced m2 1.00E+00 1.00E+00 M Waste to landfill kg 3.85E-02 - M Waste to recycling kg 1.23E-03 - M Sludge to land kg 6.03E-02 - M Water to groundwater kg 7.02E-02 - C Waste water kg 6.67E+00 1.48E+01 C Formaldehyde emissions kg 1.76E-02 1.76E-03 L VOCs kg 2.42E-02 2.42E-02 L Water vapour kg 1.10E+01 - C * measured (M) / calculated (C) / estimated (E) / literature (L)

Key differences to the baseline study include a reduction in resin and wax loadings. Resin loadings data were known to include a high level of uncertainty in the baseline study and were likely overestimated based on a conservative assumption. Energy consumption overall has seen a small reduction relative to the baseline study, with the greatest reduction being from natural gas. The variability of data between sites is given in Annex A. Inputs and outputs with high variability between participating sites are often due to alternative material types or energy sources being used by each site e.g. more pulplog is used by some sites whereas others use more woodchip from sawmills.

4.8.4. Packaging

Packaging materials used for MDF are presented in Table 4-53. Packaging data were not included within the scope of the baseline study; therefore, no comparison is available. Packaging materials are assumed to be used evenly on all products types produced per site per cubic metre of board output.

Table 4-53 MDF packaging inventory – per m2 of 18 mm MR, melamine coated board

Flow Unit Average DQI* MDF packaged m2 1.00E+00 M Softwood gluts m3 4.26E-04 M LDPE wrap kg 8.59E-05 M Plastic strapping kg 1.99E-03 M Paper labels kg 3.07E-03 C Inks kg 2.02E-06 C Paints kg 5.87E-05 C * measured (M) / calculated (C) / estimated (E) / literature (L)

LCI and EPDs for Australian Wood Products 92 of 159

4.8.5. Allocation

Energy, water and waste inputs to each product are allocated based on the total dry mass of wood, dry mass of resin (including resin in any laminate) and wax in each product. Minor additives such as dyes, preservatives, tannin and paper were not a factor in the allocation due to their small proportions and high variation between product types. They are expected to have a relatively minor influence over the energy and water requirements of each product compared to wood, resins and wax.

4.8.6. Assumptions

Assumptions made during the construction of the MDF LCI are listed within the following section.

General assumptions

All data received from manufacturers have been accepted as accurate (after cross-checks for mass balance, plausibility, etc.) Energy used is reported by energy source (e.g. coal, diesel fuel and electricity), but the upstream profiles of energy sources are modelled based on Australian averages. Any difference between input mass and output mass has been assumed to be scrap or offcuts burnt on-site for thermal energy purposes. The total use of thermal energy for the entire site and associated emissions has been allocated based on the input mass of key inputs as described in section 4.8.5. Insufficient data were obtained for inbound transport of raw material inputs. A travel distance of 100km has therefore been assumed for all inputs. Upstream resin manufacture has been modelled based on available GaBi datasets. The composition of resins within the product has been calculated according to Annex F. The combustion of wood residues and offcuts are assumed to contain resin; therefore, emissions are also based on the compositions described in Annex F.

Manufacturing

Where a water input exceeded water discharged to a drain or treatment facility (water imbalance), remaining water is assumed to be evaporated, including water contained in wood and residue inputs Where onsite dams are used, water use has been calculated as follows: Annual, site specific rainfall on dam area is reported as lake water use and is taken from the nearest Bureau of meteorology measurement station (BoM, 2017). Dam evaporation has been calculated using the tool Readyreckoner (NCEA, 2017). This has been used to calculate the net evaporation relative to a similar area of pasture land with the same location and annual rainfall using the Zhang relationship (Zhang, et al., 2001). Net use of ground water has been calculated to balance water according to the following: ground water in + rainfall (lake water) = water use from dam + net evaporation. Offcuts/residues from MDF manufacture that are used as fuel on-site are assumed to contain residue in the same proportions as the product of that type. The combustion of resin has been modelled separately to wood residue combustion. The water content of wet sludge spread on land has been assumed to partially evaporate and partially return to groundwater. The relationship stated by Zhang et al. has been used to estimate the water returning to groundwater and water evaporated using the annual precipitation in the nearest major city (ABS, 2017).

LCI and EPDs for Australian Wood Products 93 of 159

When modelling the production of paper, any scrap paper used has been assumed to be burden free. Inherent properties: biogenic carbon and primary energy have been accounted for within the dataset. Similarly, no credit has been applied for any waste paper that is recycled in the product life cycle. This methodology of cutting-off impacts has been applied throughout the model as it was not possible to credit recycled paper as no consistent, high-quality data were available for 100% primary paper production and 100% secondary paper production. The carbon balance has been corrected via CO2 emissions (biotic), assuming decomposition or incineration in the time frame of 100 years.

Packaging

Packaging materials such as wooden gluts are assumed to be carbon neutral with respect to sequestered carbon, as the installation phase (module A5), where the sequestered carbon would be released, is not declared within this study.

Computational assumptions

Input and output materials and processes are a production-weighted average based on the data provided from Australian manufacturers.

Electricity grid mix

The electricity grid mix for MDF manufacturing (A1-A3) has been modelled based on actual electricity consumption by each site included in this study and the state in which they are located. The same electricity grid mix MDF has also been applied to upstream processes such as softwood forestry stage and sawmills providing woodchip. Electricity at end-of-life (module C) has been modelled using an average Australian electricity mix as the location where the product reaches end- of-life is unknown. The electricity grid mixes used for each product are given in Annex E.

4.9. Plywood

4.9.1. Overview

The following plywood products are included within the LCI. All products included within the LCI are untreated.

• 1 m2 of plywood, 7 mm exterior, A-Bond • 1 m2 of plywood, 9 mm exterior, A-Bond • 1 m2 of plywood, 15 mm flooring (tongue & groove), A-Bond • 1 m2 of plywood, 25 mm flooring (tongue & groove), A-Bond • 1 m2 of plywood, 17 mm formply, A-Bond • 1 m2 of plywood, 17 mm formply, B-Bond LCI data were collected using detailed surveys of all five plywood plants operating in Australia, as shown in Table 4-54. These plants collectively produced 142,391 m3 plywood in 2015/16, which equates to 100% of Australian domestic production. The included products represent 66,312 m3 or 47% of all plywood products produced in Australia by volume.

LCI and EPDs for Australian Wood Products 94 of 159

Table 4-54 Contributors to plywood LCI

Producer company Total Financial Data plants contributor contributor Ausply Pty Ltd 1 1 1 Austral Plywoods 1 1 1 Big River Timbers 1 1 1 Carter Holt Harvey Woodproducts Australia 1 1 1 Ta Ann 1 1 1 Total 5 5 5

Four sites provided data for the production of veneer. All five of the surveyed sites provided data for the production of plywood. Therefore, the weighted average veneer production data reflects the production-weighted average of these four sites, whereas the plywood manufacturing data reflects the production-weighted average of all five sites. This is seen to have little influence on the results for veneer production. The scope of this study covers six different plywood product types. Data was provided by two sites for five different product types each, one site provided data for two product types and the two sites provided data for one product type each. Due to the small sample size and the possibility of the disaggregation of confidential data, the number of sites providing data for each product type has not been disclosed. The full list of product inventories is displayed in Annex A along with an assessment of inter-site variability. All LCI tables and comparisons within this section refer to a sample plywood product: 17 mm formply, A-Bond. This product has been chosen as it is the product produced in greatest volume that has a comparison available from the baseline study. The material properties of 17 mm formply, A-Bond plywood are shown in Table 4-55. The data quality indicator relates to the recent data. If the information was only used in the baseline scenario it is marked with n/a as it is not relevant for recent data. Baseline (2006) refers to data collected by the Commonwealth Scientific and Industrial Research Organisation (CSIRO) that inventoried data from the 2005/06 Australian financial year (1st July 2005 to 30th June 2006) (CSIRO, 2009a), as updated by EPD Registration No. S-P-00564, Version 1, Issued 14 October 2015, Valid until 14 October 2020.

Table 4-55 17 mm formply, A-Bond – material properties

Property Unit Average Baseline DQI* (2006)

Density of finished plywood kg/m3 514 599 M Moisture content of finished plywood % 7.8 12 M Density of veneer kg/m3 509 551 M Moisture content of veneer % 6.6 8 M Density of softwood logs kg/m3 779 993 M Moisture content of softwood logs % 71 117 M Density of hardwood logs kg/m3 1257 1100 M Moisture content of hardwood logs % 97 73 M

LCI and EPDs for Australian Wood Products 95 of 159

Property Unit Average Baseline DQI* (2006)

Solids content of PF resin % 47 100 M Solids content of MUF resin % - - M Solids content of UF resin % - 100 - Solids content of acrylic putty % - - M * measured (M) / calculated (C) / estimated (E) / literature (L) / not applicable (-)

Table 4-56 shows the material composition of the different MDF products according to the product- specific bills of material.

Table 4-56 Material composition of the plywood products [%]

Ext, A- Ext, A- Floor, Floor, Formply, Int, B- Bond, Bond, T&G, A- T&G, A- A-bond, bond, 7mm 9mm bond, bond, 17mm 17mm

15mm 25mm Softwood (dry) 88.6 89.0 88.8 88.5 72.4 55.6 Hardwood (dry) 0.0 0.0 0.0 0.0 12.9 28.7 Phenol formaldehyde 3.5 3.2 3.2 3.9 5.2 1.1 Melamine urea 0.0 0.0 0.0 0.0 0.0 5.6 formaldehyde Lamination paper (dry) 0.0 0.0 0.0 0.0 2.3 1.7 Tongue 0.0 0.0 0.3 0.2 0.0 0.0 (polypropylene) Acrylic 0.6 0.5 0.3 0.1 0.0 0.0 Water 7.4 7.3 7.3 7.4 7.2 7.4

The elemental compositions of the different wood products were calculated according to the sum of the elemental composition of its elements. Table 4-57 shows the elemental composition including the incorporated water and Table 4-58 shows the elemental composition of the bone-dry product.

Table 4-57 Elemental composition including water

Ext, A- Ext, A- Floor, T&G, Floor, T&G, Formply, A- Int, B-bond, Bond, 7mm Bond, 9mm A-bond, A-bond, bond, 17mm 17mm

15mm 25mm C (biogenic) 44.3 44.5 44.4 44.2 43.7 42.9 C (fossil) 3.2 2.8 3.0 3.3 4.2 2.9 H 6.4 6.4 6.4 6.4 6.4 6.3 N 0.0 0.0 0.0 0.0 0.0 2.4 S 0.0 0.0 0.0 0.0 0.0 0.0 O 46.2 46.3 46.2 46.1 45.6 45.3 Cl 0.0 0.0 0.0 0.0 0.0 0.0

LCI and EPDs for Australian Wood Products 96 of 159

Ext, A- Ext, A- Floor, T&G, Floor, T&G, Formply, A- Int, B-bond, Bond, 7mm Bond, 9mm A-bond, A-bond, bond, 17mm 17mm

15mm 25mm Ash 0.0 0.0 0.0 0.0 0.2 0.1

Table 4-58 Elemental composition excluding water

Ext, A- Ext, A- Floor, T&G, Floor, T&G, Formply, A- Int, B-bond, Bond, 7mm Bond, 9mm A-bond, A-bond, bond, 17mm 17mm

15mm 25mm C (biogenic) 47.8 48.0 47.9 47.7 47.1 46.3 C (fossil) 3.4 3.1 3.3 3.6 4.5 3.2 H 6.0 6.0 6.0 6.0 6.0 5.9 N 0.0 0.0 0.0 0.0 0.0 2.6 S 0.0 0.0 0.0 0.0 0.0 0.0 O 42.8 42.9 42.8 42.7 42.2 41.8 Cl 0.0 0.0 0.0 0.0 0.0 0.0 Ash 0.0 0.0 0.0 0.0 0.2 0.1

The net calorific values of softwood and paper were calculated from the gross calorific values and the corresponding water and hydrogen content. The net calorific value of the total wood products is calculated as the sum of the values for its components. The results are shown in Table 4-59.

Table 4-59 Net calorific values of softwood, paper and plywood products

Ext, A- Ext, A- Floor, Floor, Formply, Int, B- Bond, Bond, T&G, A- T&G, A- A-bond, bond, 7mm 9mm bond, bond, 17mm 17mm

15mm 25mm Softwood (dry) 19.2 19.2 19.2 19.2 19.2 19.2 Hardwood (dry) 18.7 18.7 18.7 18.7 18.7 18.7 Lamination paper (dry) 17.9 17.9 17.9 17.9 17.9 17.9 Total 18.1 18.0 18.1 18.1 18.2 17.5

4.9.2. Production process

Plywood is a panel product comprising several thin layers of softwood, hardwood or a combination of veneers that are typically 1.5 to 3mm in thickness. The veneers are resin-bonded with the grain direction of each layer perpendicular to the previous, and then hot pressed. Plywood products are either engineered wood panels (such as structural plywood and formwork plywood) or non- structural panels (such as interior and exterior plywood). The difference between the engineered products and the non-structural products is that engineered wood products have defined and standardised structural properties, such as strength, stiffness, dimensional stability and reliability characteristics. The typical manufacturing operations for plywood are shown in Figure 4-9 which depicts the production including inputs and outputs from the process.

LCI and EPDs for Australian Wood Products 97 of 159

Large logs from softwood forestry Veneer logs from hardwood forestry

System boundary Electricity, water, thermal energy Log preparation & Conditioning

Electricity Green veneer production

Combustion Primary of residues Emissions energy, water, Electricity, thermal Drying & finishing Waste raw energy (recycling material & landfill) Resins & electricity Glue spreading

Electricity, thermal energy Pressing

Water, electricity Water spraying

Finishing Electricity Plywood Trimming and sanding

Packaging material, transport packaging Packaging

Figure 4-9 Process flow diagram of plywood production

Material inputs such as wood, resins & waxes etc. were collected as a BOM for each product, along with total production volumes and properties of each product. Energy, water and waste data were collected as totals per site, as reliable data were not available per individual product. Total material inputs were also requested for each site along with the site’s total plywood output. Allocation of energy, water and waste to each product is described in section 4.9.6. The plywood production process consists of many stages as depicted in Figure 4-9. As data were not able to be collected with sufficient accuracy at a sub-process level, the plywood manufacturing process has been aggregated based on two unit processes:

• Veneer production, including all inputs and outputs associated with the production and finishing of individual veneer layers • Plywood production, including all inputs and outputs associated with the production and finishing of plywood products from veneer inputs These unit processes have been chosen as some of the participating facilities produce a small quantity of veneer on-site and purchase dry veneer to manufacture plywood. Other sites also sell a portion of the produced veneer. Indicators assessing the variability in LCI data are given in Annex A

4.9.3. Veneer production

Veneers are thin sheets of solid wood rotary peeled or sliced from a log. Veneer used for plywood in Australia is produced mainly by rotary peeling debarked logs. Veneers can also be sliced in parallel sheets (i.e. longitudinal slicing, crosscut slicing or stay-log lathe slicing) allowing the grain patterns to be retained, for aesthetic purposes. These decorative veneers account for a very small percentage and are used mainly in high value furniture applications, so are not within the scope of this report.

LCI and EPDs for Australian Wood Products 98 of 159

Veneer production processes include:

• Log preparation (debarking) • Conditioning (steaming/moistening log) • Green veneer production (cutting to billets, peeling, clipping and chipping) • Drying and finishing (drying veneers & grading)

Log preparation

Large (generally greater than 250 mm diameter), high quality logs, are debarked for most hardwood species. This occurs before delivery from the plantation. When debarked on site, a mechanical scraping machine is utilized.

Conditioning

At some mills, depending on the log species, the log is conditioned by steaming, submersing in water or water spraying. Conditioning means the logs have a high, stable moisture content, which allows improved peeling operations for these specific logs. In addition, heating of the log softens the timber fibres further facilitating improved veneer quality and yield.

Green veneer production

Veneer log green veneer production consists of four principal activities: cutting to billets, peeling, clipping, and chipping. The logs are firstly cut to size. These cut logs are known as ‘billets’ and are cut longer than the size of the plywood panel. To obtain the veneer from the log, the log is rotary peeled by placing a log on a lathe and spinning at speed against a long knife (or lathe blade). The log is first loaded and centred on the lathe. Scanners are often used to ensure the log is correctly positioned for maximum yield. The log is then rotated on the lathe against the blade to peel the veneer off in one long continuous sheet with consistent thickness. As the continuous veneer ‘ribbon’ is cut it feeds along a conveyor belt and through an automated clipping machine which ‘clips’ the veneer to the desired length. Scanners allow yield from the veneer sheets to be maximized and off-cuts are diverted to the boiler. The wet veneer proceeds to the drier.

Drying and finishing

Once the veneers have been cut they are dried to approximately 5-8% moisture content. The optimum moisture content required for gluing depends on both the species and density of the veneer, as well as the adhesives and gluing process. The moisture content of the green veneer ranges from 40% to over 100%. The sheets require high drying temperatures of up to 300 °C. Veneers are generally dried using mechanical driers fitted with convection and/ or radiant type heaters. Compartment driers use convection heating, similar to timber kilns, while mechanical driers, which are prevalent, have both convection and/ or radiant type heaters. The veneers travel on conveyor belts through the drying chambers at a regulated speed. Drying can present several problems, such as buckling, cracking, splitting, shrinking and distortion. These defects can be repaired, by plugging or filling, or jointed, by edge gluing, stitching or taping more than one veneer. The sheets can also be cut into two sheets to form the ‘cross-band’ sheets, the panels which are at right angles to the face veneers. Dried veneers are generally graded before use in the plywood process and after grading can be stored for use in plywood manufacture or sold.

LCI and EPDs for Australian Wood Products 99 of 159

Veneer production inventory

The production-weighted average of veneer manufacturing data has been taken from all sites. All veneer inputs and yields from veneer manufacture are assumed to be equal regardless of type of veneer. However, as the type and density of veneer varies depending on the end plywood product, the log type and properties have been varied depending on the plywood products. This results in a different ratio of softwood and hardwood log inputs and a different volume of water vapour emitted during drying from each product due to differences in the moisture content of log types. The inventory for veneer production as an input into the manufacture of 17 mm formply, A-Bond plywood is given in Table 4-60. The full list of product inventories is displayed in Annex A along with an assessment of inter-site variability. As the baseline output data were grouped as totals and not split up between veneer and plywood manufacture, some outputs were allocated to only one process step. The wood for thermal energy was allocated to the veneer production as the recent data shows that the main output occurs at that stage. The water output is also allocated to the veneer production as current data show that it mainly occurs at this stage. All other outputs were allocated to the process where they occur in current data.

Table 4-60 Veneer production for 17 mm formply, A-Bond – inventory per m3 of veneer

Type Flow Unit Average Baseline DQI* (2006) Inputs Logs m3 2.36E+00 2.30E+00 M Softwood large log volume m3 2.09E+00 2.19E+00 C Hardwood veneer log volume m3 2.67E-01 1.15E-01 C Electricity kWh 4.66E+01 1.82E+02 M Thermal energy MJ 5.69E+03 7.12E+03 C Thermal energy from natural gas MJ 9.61E+02 4.83E+02 M Thermal energy from wood MJ 4.00E+03 6.60E+03 C Thermal energy from wood MJ 6.65E+02 - C (containing resin) Thermal energy from LPG MJ 6.24E+01 3.82E+01 M Municipal water kl 5.86E-01 1.95E-01 M Ground water kl 2.83E-01 - M Diesel kg 1.79E+00 - M LPG kg 2.75E-01 - M Outputs Dry Veneer m3 1.00E+00 1.00E+00 M Log core m3 4.20E-01 1.40E-01 M Bark m3 1.16E-01 2.33E-02 M Veneer clips m3 2.85E-01 1.86E-01 M Wood for thermal energy** kg 2.76E+02 4.95E+02 C Waste to landfill kg 3.33E-01 - M Waste to recycling kg 2.99E+00 - M Water total kl 1.20E+00 6.81E-01 C Waste water treatment kl 1.66E-01 - M

LCI and EPDs for Australian Wood Products 100 of 159

Type Flow Unit Average Baseline DQI* (2006) Water to site kl 6.74E-02 - C Water vapour kl 9.63E-01 - C * measured (M) / calculated (C) / estimated (E) / literature (L) / not applicable (-) **Given at moisture content of veneer

A comparison of the data from the current study to that of the baseline study shows similarities for inputs such as thermal energy and recovery rate. A higher percentage of hardwood was used for A- Bond formply than in the baseline study. Electricity consumption for veneer production was lower in the current study compared to the baseline study per cubic metre of veneer produced. Water consumption was also found to be significantly higher in the current study.

4.9.4. Plywood manufacture

Plywood manufacture consists of three principal activities:

• Glue spreading • Pre-pressing and hot-pressing • Water spraying

Glue spreading

The veneer is first assembled in layers in the correct order to form a specific plywood sheet which may be of differing thicknesses. The front-face veneer, core and cross-band veneers, and the back- face veneer are arranged to optimize the performance of the product in use. Adhesive is either applied to each veneer layer, or just to the cross-bands and cores from which the glue is transferred to the adjacent layers during pressing. The veneer sheets are fed through the glue spreader and stacked in sheet order.

Pre-pressing and hot pressing

There are two stages of pressing in plywood manufacture beginning with pre-pressing, which aids the glue spreading process. The pre-press is undertaken on a cold press into which a stack of glue- spread veneers (which will form many sheets of plywood) are fed. These veneers are fed into the cold-press and placed under pressure in normal atmospheric conditions. The glue is transferred to all veneer surfaces giving a better glue bond and developing strength in the individual panels which assists with panel handling and subsequent hot pressing which is the second pressing process.

Water spraying

The hot-press relies on high temperature and pressure to cure the glue. The presses are typically multi-opening, hydraulic hot presses with temperatures reaching 140 °C and pressures of 1 MPa. Each press can be loaded manually or automatically depending on mill technology. In both cases the plywood sheets need to be watered after pressing to return the moisture content of the plywood sheet to a normal range and to improve the panel’s stability and flatness.

Plywood manufacture inventory

Data for the manufacturing of plywood products from veneer was collected in the form of a bill-of- materials (BoM). This was provided by all sites producing each of the product types within the scope of the study. Site electricity, water and energy inputs were allocated to plywood products

LCI and EPDs for Australian Wood Products 101 of 159

based on production volumes. The production-weighted average inventory of sites which produce 17mm formply, A-Bond plywood is given in Table 4-61. The full list of product inventories is displayed in Annex A along with an assessment of inter-site variability.

Table 4-61 Plywood manufacture for 17 mm formply, A-Bond – inventory per m2 of plywood

Type Flow Unit Average Baseline DQI* (2006) Inputs Dry Veneer m3 2.19E-02 1.56E-02 M Electricity kWh 1.09E+00 4.98E-01 M Thermal energy MJ 2.78E+00 1.70E+01 C Thermal energy from natural gas MJ 1.38E+00 1.05E+00 M Thermal energy from wood MJ 1.22E+00 1.60E+01 C Thermal energy from wood MJ 1.80E-01 - C (containing resin) Diesel kg 2.75E-02 - M LPG kg 1.21E-02 1.47E-02 M Mass of PF resin kg 9.57E-01 7.24E-01 M Mass of MUF resin kg - - M Mass of acrylic putty kg - 9.86E-03 M Mass of phenol formaldehyde putty kg - 3.40E-03 - Phenolic overlay kg 4.72E-01 1.82E-01 M Tongue kg - - M Flour kg - 9.52E-02 - Filler kg - 6.75E-02 - Preservatives kg - 1.00E-03 - Municipal water kl 4.14E-03 2.53E-03 M Ground water kl 4.55E-03 - M Outputs Plywood m2 1.00E+00 1.00E+00 M Sawdust & trimmings m3 4.45E-03 2.09E-01 M Wood for thermal energy kg 1.15E+00 Allocated to C veneer production. Waste to landfill kg 3.02E-03 - M Waste water treatment kl 1.56E-03 Allocated to M veneer production. Water vapour kl 7.44E-03 Allocated to C veneer production. * measured (M) / calculated (C) / estimated (E) / literature (L) / not applicable (-)

A comparison of the current data to the baseline data shows an increase in the amount of electricity consumed per square metre of 17 mm formply, A-Bond. However, a decrease in the total amount of thermal energy is also observed. A similar mass of PF resin was found to be used per square metre of plywood in the current study. Data covering the amount of preservatives used in each product

LCI and EPDs for Australian Wood Products 102 of 159

were not provided by any of the facilities. The products in the current study therefore represent untreated plywood manufacture. Water use for plywood manufacture was found to be more than three times higher for the current study than in the baseline study. The recovery rate for plywood production was also found to be lower for the current study. However, the recovery rate calculated from the baseline data appears to be too high for the plywood manufacture process.

4.9.5. Packaging

Packaging materials used for plywood are presented in Table 4-62. Packaging was included within the scope of the baseline study, however it is difficult to make a meaningful comparison, as many of the packaging materials from the current study were not reported in the baseline study. Packaging materials are calculated as an average across all products types produced per site per cubic metre of plywood output.

Table 4-62 Plywood packaging inventory per m3 of plywood

Flow Unit Average Baseline DQI* Plywood packaged m3 1.00E+00 1.00E+00 M Softwood gluts m3 6.52E-03 - C LDPE wrap kg 4.00E-04 1.70E-02 C Plastic strapping kg 1.07E-01 - C Paper labels kg 1.68E-01 - C Steel strapping kg 1.10E-02 - C Inks kg 1.45E-02 1.13E-01 C Paints kg 3.35E-01 3.75E-01 C * measured (M) / calculated (C) / estimated (E) / literature (L)

4.9.6. Allocation

Allocation to the co-products of sawmilling is based on economic value, as the difference in the value of co-products is large (>25%, in line with IEPDS (2017, section 7.7)). Most co-products have some economic value and therefore cannot be considered as wastes. Economic values have been provided by participants and are displayed as volume-weighted averages in Table 4-63 for veneer production and Table 4-64 for plywood production.

Although no economic value was actually specified for the ‘residues burnt’, which are used internally. Instead this value has been calculated based on the volume-weighted average of other residues from the processes they are generated from. These are the fuel types typically burnt onsite during the kiln-drying process.

Table 4-63 Veneer production co-product allocation factors

Veneer production Units Amount Average Economic Allocation (Units) price value ($) (%) ($/Unit)

Dry veneer m3 1.00 409 409 96.35% Log core m3 0.330 20.7 6.83 1.61%

LCI and EPDs for Australian Wood Products 103 of 159

Veneer production Units Amount Average Economic Allocation (Units) price value ($) (%) ($/Unit)

Bark m3 0.112 15.7 1.75 0.41% Veneer clips m3 0.283 13.6 3.85 0.91% Residues burnt (bone dry) kg 246 0.01 3.04 0.72%

Total 424 100%

Table 4-64 Plywood production co-product allocation factors

Plywood production Units Amount Average Economic Allocation (Units) price value ($) (%) ($/Unit)

Plywood m3 1.00 1042 1042 99.4% Sawdust m3 0.05 37.5 1.92 0.18% Residues burnt (bone dry) kg 53.5 0.07 3.87 0.37%

Total 1048 100%

4.9.7. Assumptions

Assumptions made during the construction of the plywood LCI are listed below.

General assumptions

All data received from manufacturers have been accepted as accurate (after cross-checks for mass balance, plausibility, etc.) Energy used is reported by energy source (e.g. coal, diesel fuel and electricity), but the upstream profiles of energy sources are modelled based on Australian averages. Where a water input exceeded water discharged to a drain or treatment facility (water imbalance), remaining water is assumed to be evaporated, including water contained in wood and residue inputs Any difference between input mass and output mass has been assumed to be scrap or offcuts burnt on-site for thermal energy purposes. The total use of thermal energy for the entire site and associated emissions has been allocated based on the input mass of key inputs as described in section 4.9.6. Insufficient data were obtained for inbound transport of raw material inputs. A travel distance of 100 km has therefore been assumed for all inputs.

Computational assumptions

Input and output materials and processes are a production-weighted average based on the data provided from Australian manufacturers.

Veneer production

Veneer production data has been modelled using the volume-weighted average of inputs from each site. Softwood and hardwood veneer have been treated equally with respect to all other inputs such

LCI and EPDs for Australian Wood Products 104 of 159

as energy and water and outputs such as scrap rate. The weighted average ratio of hardwood and softwood used within plywood manufacturing has been varied based for each product.

Plywood manufacture

Offcuts/residues from plywood manufacture that are used as fuel on-site are assumed to contain residue in the same proportions as the product of that type. The combustion of resin has been modelled separately to wood residue combustion. Upstream resin manufacture has been modelled based on available GaBi datasets. The composition of resins within the product has been calculated according to Annex F. The combustion of wood residues and offcuts are assumed to contain resin; therefore, emissions are also based on the compositions described in section Annex F. Thermal energy used for the plywood manufacture process was not able to be measured by any of the participating sites. This has therefore been estimated based on personal communication with one of the sites as 5% of the thermal energy used in veneer production based on an equivalent volume of inputs. This assumption has been applied to all sites and prorated based on the relative amount of veneer and plywood produced per site. When modelling the production of paper, any scrap paper used has been assumed to be burden free. Inherent properties: biogenic carbon and primary energy have been accounted for within the dataset. Similarly, no credit has been applied for any waste paper that is recycled in the product life cycle. This methodology of cutting-off impacts has been applied throughout the model as it was not possible to credit recycled paper as no consistent, high-quality data were available for 100% primary paper production and 100% secondary paper production. The carbon balance has been corrected via CO2 emissions (biotic), assuming decomposition or incineration in the time frame of 100 years.

Packaging

Packaging materials such as wooden gluts are assumed to be carbon neutral with respect to sequestered carbon, as the installation phase (module A5), where the sequestered carbon would be released, is not declared within this study.

Electricity grid mix

The electricity grid mix for plywood production (A1-A3) has been modelled based on actual electricity consumption by each site included in this study and the state in which they are located. The same electricity grid mix for plywood has also been applied to upstream processes such as softwood forestry stage and hardwood forestry stage. Electricity at end-of-life (module C) has been modelled using an average Australian electricity mix as the location where the product reaches end- of-life is unknown. The electricity grid mixes used for each product are given in Annex E.

4.10. Glued-laminated timber (Glulam)

4.10.1. Overview

The following glued-laminated timber (glulam) products are included within the LCI:

• 1 m3 of softwood glulam, untreated • 1 m3 of hardwood glulam, untreated

LCI and EPDs for Australian Wood Products 105 of 159

LCI data were collected using detailed surveys of four glulam plants operating in Australia, as shown in Table 4-65. These plants collectively produced 19,052 m3 glulam in 2015/16, covering the majority of large-scale, structural glulam manufacturing in Australia. The total market size for structural glulam is currently unknown, however total glulam consumption in Australia has been estimated as 30,000 m3 (GLTAA, 2017), giving a coverage of approximately 64% if all consumed glulam was to be manufactured in Australia.

Table 4-65: Contributors to Glulam LCI

Producer company Softwood Hardwood Financial Data glulam glulam contributor contributor Australian Sustainable Hardwoods 1 1 1 Hyne Timber 1 1 1 1 Vicbeam 1 1 1 Warrnambool Timber Industries Pty. 1 1 1 Ltd. Total 3 4 2 4

The material properties and transport distances for untreated glulam are given in in Table 4-66 for softwood glulam and Table 4-67 for hardwood glulam. As glulam was not included within the scope of the previous study, no comparison was available. LCI data taken from the CORRIM SE USA study (Bowers, et al., 2017) have been used as a comparison.

Table 4-66 Softwood glulam – material properties and distances

Property Unit Average CORRIM SE USA DQI* (softwood)

Glulam density kg/m3 621 622 M Glulam MC % 12 12 M Glulam NCV MJ/kg 16 - M Softwood KD untreated dens kg/m3 619 622 M Softwood KD untreated MC % 12 12 M Weighted average wood distance km 84 234 M from supplier Phenol resorcinol formaldehyde (PRF) % 45 - M water content PRF truck distance km 1,845 - M Polyurethane (PU) water content % - - M PU truck distance km 737 - M Distance to landfill km 14 - M * measured (M) / calculated (C) / estimated (E) / literature (L)

LCI and EPDs for Australian Wood Products 106 of 159

Table 4-67 Hardwood glulam – material properties and distances

Property Unit Average DQI*

Glulam density kg/m3 674 M Glulam MC % 10 M Glulam NCV MJ/kg 17 M Hardwood KD untreated dens kg/m3 681 M Hardwood KD untreated MC % 10 M Cyprus Pine KD untreated dens kg/m3 527 M Cyprus Pine KD untreated MC % 12 M Weighted average wood distance km 242 M from supplier % 45 M PRF water content km 1,845 M PRF truck distance % - M PU water content km 183 M PU truck distance km 15 M * measured (M) / calculated (C) / estimated (E) / literature (L)

4.10.2. Production process

Glulam is an engineered wood product, made to standardised structural properties, such as strength, stiffness, dimensional stability and reliability characteristics. It consists of lengths of softwood, hardwood or cypress pine timber, stacked and laminated using water-resistant resins. Glulam may be produced as standard straight dimensions, or custom made in various shapes and profiles. The typical manufacturing operations for plywood are shown in Figure 4-10 which depicts the production including inputs and outputs from the process.

LCI and EPDs for Australian Wood Products 107 of 159

Kiln-dried, dressed softwood from Kiln-dried, dressed hardwood from sawmill sawmill

System boundary

Electricity Trimming Residues

Electricity, resins Finger-jointing Residues

Primary Electricity, resins Face bonding Emissions energy, water, Waste raw Electricity, thermal (recycling material energy, water Pressing & curing & landfill)

Electricity Finishing Residues

Chemicals, electricity, Treatment water

Packaging material, Packaging transport packaging

Untreated glulam Treated glulam

Figure 4-10 Process flow diagram of glulam production

The four surveyed sites produced glulam products from softwood, hardwood and cypress pine. For the purposes of this study cypress pine has been treated as a lower density hardwood, as the production of cypress pine timber is more similar to hardwood production from native forestry than plantation softwood and has therefore been included in the hardwood glulam life cycle inventory (LCI). Data for glulam production has been collected and treated as a single unit process including:

• Trimming • Finger-jointing • Face bonding • Pressing & curing, and • Finishing Additional processes included in glulam production include:

• Packaging, and • Treatment (optional)

Trimming

Kiln-dried, dressed timber is received by sawmills. This timber is then cut to length, defects removed, graded and sorted according to the desired product composition.

Finger-jointing

The sorted lengths are joined end-to-end by using finger-joints to create uniform lengths of timber. The finger-joints are cut and a resin is applied to the interlocking surfaces before pressing together and curing creating full length laminations.

LCI and EPDs for Australian Wood Products 108 of 159

Face bonding

Individual laminations have resin applied to the joining faces and are then layered with the number of laminations required to create the required depth section. Sorted laminations are ordered to optimise the strength characteristics of the glulam with the higher strength laminations placed toward the top and bottom of the member. The resins typically used for glulam manufacture in Australia are phenol resorcinol formaldehyde (PRF) or polyurethane (PU).

Pressing & curing

The glued laminations are placed in a jig to maintain alignment of the layers and are pressed together by clamps kept under high pressure until the resin is cured.

Finishing

Once the resin is fully cured, the glulam members are finished by planing and cutting to the final dimensions.

4.10.3. Glulam production

Inputs and outputs for the production of 1 m3 of softwood and hardwood glulam are given in Table 4-68 for softwood glulam and Table 4-69 for hardwood glulam. The manufacturing process is treated as a “black box” with inputs and outputs crossing the system boundary included. A comparison is given to a study completed by CORRIM (Bowers, et al., 2017), as glulam was not originally within the scope of the baseline study. Indicators assessing the variability in LCI data are given in Annex A.

Softwood glulam

Inputs and outputs of the production process for softwood glulam are given in Table 4-68. Polyurethane resin is used as the adhesive by three of the four manufacturers in this study, with phenol resorcinol formaldehyde (PRF) used by one manufacturer. As a result, industry-average glulam includes a mix of both polyurethane and PRF, even though only one resin type is used in any specific product.

Table 4-68 Softwood glulam – inventory per m3 of glulam

Type Flow Unit Average CORRIM SE DQI* USA (softwood) Inputs Softwood KD untreated m3 1.17E+00 1.14E+00 M PRF amount kg 7.01E+00 - M PU amount kg 1.96E+00 - M MUF amount kg - 4.31E+00 M Electricity kWh 9.51E+01 8.45E+01 M Natural gas MJ 2.54E+02 1.60E+02 M Diesel transport kg 6.27E-02 - M LPG transport kg 1.24E+00 - M Water municipal kg 2.60E+01 - M Outputs Glulam m3 1.00E+00 1.00E+00 M

LCI and EPDs for Australian Wood Products 109 of 159

Type Flow Unit Average CORRIM SE DQI* USA (softwood) Shavings and trimmings m3 1.98E-01 1.40E-01 M Water vapour kg 2.59E+01 - C Waste to landfill kg 9.90E+00 - M Waste to recycling kg 1.13E+00 - M * measured (M) / calculated (C) / estimated (E) / literature (L)

Data provided by the participating sites compares well with the CORRIM study (Bowers, et al., 2017) in terms of recovery rates and electricity use. However, natural gas usage is higher in the current study. Resin usage is approximately double that of the CORRIM study. However, many factors affect the resin loading of the glulam produced, such as size of members, type of resin used and structural property requirements of the glulam member.

Hardwood glulam

Inputs and outputs of the production process for hardwood glulam are given in Table 4-69. Polyurethane resin is used as the adhesive by three of the four manufacturers in this study, with phenol resorcinol formaldehyde (PRF) used by one manufacturer. As a result, industry-average glulam includes a mix of both polyurethane and PRF, even though only one resin type is used in any specific product.

Table 4-69 Hardwood glulam – inventory per m3 of glulam

Type Flow Unit Average CORRIM SE DQI* USA (softwood) Inputs Hardwood KD untreated m3 1.05E+00 1.14E+00 M Cyprus Pine KD untreated m3 8.10E-02 - M PRF amount kg 4.03E-01 - M PU amount kg 2.26E+00 - M MUF amount kg - 4.31E+00 M Electricity kWh 1.14E+02 8.45E+01 M Natural gas MJ 1.99E+01 1.60E+02 M Diesel transport kg 6.31E-01 - M LPG transport kg 1.35E-01 - M Water municipal kg 2.28E+00 - M Outputs Glulam m3 1.00E+00 1.00E+00 M Shavings and trimmings m3 1.32E-01 1.40E-01 M Water vapour kg 2.13E+00 - C Waste to landfill kg 1.44E+00 - M Waste to recycling kg 9.85E-02 - M * measured (M) / calculated (C) / estimated (E) / literature (L)

LCI and EPDs for Australian Wood Products 110 of 159

The hardwood glulam inventory shows a higher amount of electricity is used in glulam production than in the softwood glulam production LCI. Natural gas usage however, is lower for hardwood than for softwood. The site using PRF resin in their glulam production accounts for a smaller share of the total hardwood glulam output than for softwood glulam. Therefore, less PRF is used on average for hardwood glulam and more PU is used compared to softwood glulam production.

4.10.4. Packaging

Packaging materials used for glulam are presented in Table 4-70 for softwood glulam and in Table 4-71 for hardwood glulam. Packaging data were included within the scope of the CORRIM study; however, different packaging materials were used so it is difficult to make a meaningful comparison. Packaging materials are assumed to be used evenly on all products types produced per site per cubic metre of glulam output.

Table 4-70 Softwood glulam packaging inventory per m3 of glulam

Flow Unit Average CORRIM DQI* SE USA (softwood) Glulam packaged m3 1.00E+00 1.00E+00 M Softwood gluts m3 1.04E-02 - M Hardwood gluts m3 - - M LDPE wrap kg 1.42E-01 3.72E-03 M Plastic strapping kg 1.78E-01 - M Steel strapping kg - 6.90E-04 M Paper labels kg - - M Inks kg 5.45E-02 - E Cardboard kg - 4.04E-03 M * measured (M) / calculated (C) / estimated (E) / literature (L)

Table 4-71 Hardwood glulam packaging inventory per m3 of glulam

Flow Unit Average CORRIM DQI* SE USA (softwood) Glulam packaged m3 1.00E+00 1.00E+00 M Softwood gluts m3 5.96E-04 - M Hardwood gluts m3 8.14E-03 - M LDPE wrap kg 5.78E-02 3.72E-03 M Plastic strapping kg 1.83E-01 - M Steel strapping kg 7.25E-02 6.90E-04 M Paper labels kg 2.25E-03 - M Inks kg 5.45E-02 - E Cardboard kg - 4.04E-03 M * measured (M) / calculated (C) / estimated (E) / literature (L)

LCI and EPDs for Australian Wood Products 111 of 159

4.10.5. Allocation

Allocation to the co-products of glulam production is based on economic value as the difference in the value of co-products is large (>25%, in line with IEPDS (2017, section 7.7)). Most co-products have some economic value and therefore cannot be considered as wastes. Economic values have been provided by participants and are displayed as volume-weighted averages in Table 4-72.

Table 4-72 Glulam co-product allocation factors

Glulam Units Amount Average Economic Allocation (m3) price value ($) (%) ($/m3)

Glulam m3 1 $1,753 $1753 99.7% Production residues* m3 0.163 $39.9 $6.52 0.37%

Total $1760 100% *Given at density and MC of glulam

Sales prices for glulam have been provided by three of the four participating sites. A volume- weighted average price has been used for the forth site. Where no value was available for a site’s residue outputs, a mass-weighted average of the other participating site’s residues has been used due to differences in the density of the residues produced. The glulam and residue prices have been taken as a volume-weighted average of the produced glulam varieties (softwood, hardwood and cypress pine). This approach has been taken as insufficient price data were available to distinguish the price of softwood, hardwood and cypress pine residues. It is assumed that the residues of these products follow the same relative price as the corresponding glulam products.

4.10.6. Assumptions

Assumptions made during the construction of the glulam LCI are listed within the following section.

General assumptions

All data received from manufacturers have been accepted as accurate (after cross-checks for mass balance, plausibility, etc.) Energy used is reported by energy source (e.g. coal, diesel fuel and electricity), but the upstream profiles of energy sources are modelled based on Australian averages. Where a water input exceeded water discharged to a drain or treatment facility (water imbalance), remaining water is assumed to be evaporated, including water contained in wood and residue inputs

Computational assumptions

Glulam production data has been modelled using the production-weighted average of inputs from each manufacturer. Softwood and hardwood inputs have been treated equally with respect to all other inputs such as energy and water and outputs such as scrap rate for each manufacturer. For sites producing both hardwood and softwood glulam, the volumetric ratio of hardwood and softwood production has been used to allocate all other inputs for each site to hardwood or softwood glulam production. Upstream resin manufacture has been modelled based on available GaBi datasets. The composition of resins within the product has been calculated according to Annex F.

LCI and EPDs for Australian Wood Products 112 of 159

Packaging

Packaging materials such as wooden gluts are assumed to be carbon neutral with respect to sequestered carbon, as the installation phase (module A5), where the sequestered carbon would be released, is not declared within this study. When modelling the production of paper, any scrap paper used has been assumed to be burden free. Inherent properties: biogenic carbon and primary energy have been accounted for within the dataset. Similarly, no credit has been applied for any waste paper that is recycled in the product life cycle. This methodology of cutting-off impacts has been applied throughout the model as it was not possible to credit recycled paper as no consistent, high-quality data were available for 100% primary paper production and 100% secondary paper production. The carbon balance has been corrected via CO2 emissions (biotic), assuming decomposition or incineration in the time frame of 100 years.

Electricity grid mix

The electricity grid mix for glulam has been modelled based on actual electricity consumption by each site included in this study and the state in which they are located. Electricity for sawn softwood and sawn hardwood production has been modelled at the state-level using total cubic metres of production per state (see sections 4.5.9 and 4.6.9). Electricity at end-of-life (module C) has been modelled using an average Australian electricity mix as the location where the product reaches end- of-life is unknown. The electricity grid mixes used for each product are given in Annex E.

4.11. Wood preservative treatment

Data on the types of wood preservative treatments used by the sites were requested as part of the survey for all product types. Concentrations and volumes were also requested as part of the questionnaire; however, these questions were often only partially completed or not completed at all. As wood preservative must be approved for use by APVMA (Australian Pesticides and Veterinary Medicines Authority) and any product claiming to be preservative treated must have preservative applied according to relevant approval and Australian Standards (TPAA, 2017a), the authors have included the treatment types in-use by manufacturers within this project, but have modelled the material composition and concentration from public information available within the relevant Material Safety Datasheets (MSDSs), the APVMA’s Public Chemical Registration Information System Search PubCRIS (Australian Government, 2017b) and/or the relevant Australian Standards on timber preservative treatments (AS 1604.1-2010; AS/NZS 1604.2:2012). The only exception is the chromated copper arsenate (CCA) treatment, which was modelled using an existing GaBi process.

Table 4-73: Covered treatment types and sources

Treatment Source chemical composition Source for concentration Bifenthrin (water MSDS, Osmose Determite ULO Pubcris label, Determite ULO borne) Timber Framing Insecticide rtu Timber Framing Insecticide rtu Boron MSDS, Diffusol Wood preservative Pubcris label, Diffusol Wood concentrate preservative concentrate Copper + DDAX MSDS, Tasco Micropro Treated Pubcris label, Osmose Micropro Timber based timber preservative Copper + azole MSDS, Tanalith E Copper Based MSDS, Tanalith E Copper Based Wood Preservative Concentrate Wood Preservative Concentrate

LCI and EPDs for Australian Wood Products 113 of 159

Treatment Source chemical composition Source for concentration LOSP (Permethrin) Average between Vascol T Timber MSDS, Vascol T Timber Insecticide Insecticide and Country Permethrin 500 EC Insecticide LOSP (Azole + MSDS, Vascol Azure Clear MSDS, Vascol Azure Clear permethrin) Ammoniacal copper MSDS, Koppers ACQ Australian Standards on timber quaternary (ACQ) preservative treatments Copper chromium GaBi process: “US: Chromated Australian Standards on timber arsenic (CCA) copper arsenate ts” preservative treatments

Table 4-6 shows the GaBi processes used for the modelling of the treatments. Data gaps were filled using proxy datasets that are similar to the required chemicals with the exception of copper acetate, where the main reactant, acetic acid, was used as a proxy. The non-chemical inputs are listed in Table 4-74. The numbers are based on the usage from four different sites. For the modelling a timber volume-weighted average was used. The waste outputs are not considered for the modelling as the standard deviation is too high to make appropriate estimates and the sludge was not considered due to its low value.

Table 4-74: Energy inputs for treatment per m3 of timber

Type Flow Unit Average DQI* Inputs Electricity kWh/m3 10.70 M Diesel kg/m3 0.24 M * measured (M) / calculated (C) / estimated (E) / literature (L)

Table 4-75 shows the LCI for the preservative treatments. It is assumed that all water and volatile organic compounds evaporate during or after the treatment process and are released to air. This is a worst-case assumption as the exact procedures during and after the treatment are unclear.

Table 4-75: Bill of materials for the different treatment types

Material [kg/L] Treatment type

borne)

(water Boron Copper+ DDAX Copperazole LOSP (Permethrin) LOSP(Azole +permethrin) ACQ Bifenthrin Ammonia ------0.015 Didecyl dimethyl ammonium ------0.056 chloride Bifenthrin 0.006 ------Boric acid - 0.039 - 0.011 - - - Copper ammonium carbonate - - - - - 0.090 Copper acetate - - - 0.063 - - - Copper (II) carbonate hydroxide - - 0.200 0.245 - - - Dispersant - - 0.030 - - - -

LCI and EPDs for Australian Wood Products 114 of 159

Material [kg/L] Treatment type Dye 0.005 ------Aromatic hydrocarbon 0.030 - - - 0.339 0.771 - Imidacloprid ------Methanol - - 0.008 - - - - Monoethanolamine - - - 0.572 - - - N-Methyle-2-Pyrolidone ------Permethrin - - - - 0.500 0.003 - Propiconazole - - - - - 0.005 - Propylene glycol (propane-1,2- - - 0.015 - - - - diol) Sodium Nitrite - - 0.001 - - - - Sodium tetraborate pentahydrate - 0.040 - - - - - Tebuconazole - - - 0.006 - 0.005 - N,N-Didecyl-N,N- - - 0.080 - - - - Dimethylammonium carbonate Copper acetate - - - 0.063 - - -

Didecyl tertiary amine - - 0.080 - - - -

Water 0.959 1.000 1.000 - - - 1.000

Outputs Water vapour 0.959 1.000 1.000 - - - 1.000 VOC to air 0.030 - - 0.572 0.339 0.771 -

The different types of treatments can be used to treat wood for use in different hazard classes. Table 4-76 lists the concentrations for the different treatments according to the different hazard classes used by manufacturers according to the data collection. The penetration rate is based on industry practises and products and was only necessary when the recommended concentration was given as a retention rate. For the Hazard class H2F for framing products the standard framing timber size of 75mm x 35 mm was assumed to calculate the concentration.

Table 4-76: Concentration of treatments according to the different hazard classes

Name Penetration Concentration Unit Hardwood H1 – Boron - 6.00 l/m3 Hardwood H3 – ACQ - 43.22 l/m3 Hardwood H3 – Copper azole - 16.10 l/m3 Hardwood H4 – CCA 10% 0.70 kg/m3 Hardwood H5/6 – CCA 20% 2.40 kg/m3 Softwood H2 – Bifenthrin (water borne) - 5.50 l/m3 Softwood H2F – Bifenthrin (water borne) - 3.12 l/m3 Softwood H2 – LOSP (Permethrin) - 0.22 l/m3 Softwood H2F – LOSP (Permethrin) - 0.13 l/m3 Softwood H3 – LOSP (Azole + Permethrin) - 40.00 l/m3

LCI and EPDs for Australian Wood Products 115 of 159

Name Penetration Concentration Unit Softwood H3 – Copper + DDAX 100% 10.73 l/m3 Softwood H3 – Copper azole - 7.03 l/m3 Softwood H4 – Copper azole - 12.79 l/m3 Softwood H3 – CCA 100% 3.80 kg/m3 Softwood H4 – CCA 40% 2.52 kg/m3

4.12. End-of-life

When a wood product reaches the end of its useful life, it may either be reused, recycled, landfilled or combusted to produce energy. Table 4-77 below provides a breakdown of which scenarios are available within each EPD. The final row of this table indicates which primary product is offset by the reuse, recycling or energy recovery of each wood product at end-of-life. The choice of offsets for recovery and recycling are consistent with those recommended in EN 16485 on EPDs for wood products as shown in Table 4-78. The offset for reuse is discussed in section 4.12.3.

Table 4-77: End-of-life scenarios per product group

EPD EoL scenario Reuse Recycling into Recycling into Energy Landfill particleboard landscape recovery mulch or animal bedding Softwood no yes yes yes yes Hardwood yes no yes yes yes Particleboard no yes yes yes yes MDF no no no yes yes Plywood - formwork yes no yes yes yes Plywood no no yes yes yes Glulam - softwood yes yes yes yes yes Glulam – hardwood yes no yes yes yes Treated products yes no no no yes

Offset against Production of Forest Forest Thermal None new timber management management energy from thinnings/ thinnings/ natural gas harvest harvest residues and residues and residues from residues from new timber new timber production production

LCI and EPDs for Australian Wood Products 116 of 159

Table 4-78: End-of-life scenarios from EN 16485:2014

This choice of scenarios is supported by a report by Forsythe Consultants (2007), which analysed handling of wood from demolition waste in Australia. Its most relevant findings were:

• Most timber demolition waste is sent to landfill. This is particularly true for low value waste, such as particleboard and MDF, and for small sections of higher-value solid woods that are costly to separate from the rest of the waste stream. Despite their large size, pine framing and pine roof trusses are typically classified as low value and mostly go to landfill or energy recovery as they are too difficult to pull apart for reuse of the individual members and too specific to reuse as an assembly in a new building. • In some locations, mixed waste sent to landfill may be separated so that biomass goes to energy recovery and only inert waste goes to landfill. Depending on the type of building, separation can also occur on-site, though this is not the most common approach. • There is currently only a small amount recycling of wood products into wood chips, though markets do exist in some locations and this option holds promise for the future. • There is a reasonable recovery and reuse rate of high-value timbers, particularly large section hardwoods and Douglas fir. Each scenario assumes that 100% of the wood is sent to that scenario. That is, they do not represent a single average value, but are instead the range of possible options in the Australian market today. As such, each scenario assumes that 100% of the waste is treated using the given pathway. This allows the end user to define their own waste treatment mix, by applying a percentage to each waste treatment option they consider to be reasonable for their own building (e.g. 75% landfill + 25% energy recovery). This approach was chosen as the split of waste to each

LCI and EPDs for Australian Wood Products 117 of 159

end-of-life pathway depends on many factors, such as the value of the wood in the building, the state where waste treatment occurs, the size of the building, the amount of wood used, etc. (Forsythe Consultants, 2007). When the user is in doubt, they should apply the “landfill (typical)” scenario, as this is the most common throughout Australia. End-of-life for treated timber products depend on the product type and may be reuse or landfill. No additional impacts are considered for the treatment of landfilled timber, as any preservative treatment will act to reduce the decomposition rate of wood in the landfill, and reduce any associated carbon emissions. The effects on other assessed impact categories are assumed to be insignificant due to the very small amount of preservative per cubic metre of timber, and the assumption of a sealed landfill.

4.12.1. Landfill

Wood consists of three types of organic (carbon-based) components: carbohydrates (cellulose and hemicelluloses), lignin, and extractives as well as some inorganic matter (Forest Products Laboratory, 2017). 50% ±2% of the dry weight of the wood of native species and non-native Pinus radiata is carbon (Gifford, 2000). The long-term sequestration and/or decomposition of carbon inherent within wood, as well as other organic materials, when deposited in landfill has been the subject of much research in the last decade. Much of this research is driven by governments increasingly having to account for carbon emissions from waste in their countries as well as some countries also accounting for storage of carbon in harvested wood products under agreements such as the Kyoto Protocol. Due to the ongoing nature of this research, two scenarios for landfill are included in the EPDs, each with a different rate for the degradable organic carbon fraction (DOCf) of wood. The two values are based on bioreactor laboratory research. This experimental work involves the testing of a range of waste types in reactors operated to obtain maximum methane yields. As the laboratory work optimises the conditions for anaerobic decay, the results can be considered as estimates of the

DOCf value that would apply over very long time horizons (Australian Government, 2014a, p. 17).

• Landfill (typical): DOCf as per Table 4-79. This is based on bioreactor laboratory research by Wang et al. (2011) and Ximenes et al. (2013). These values can be considered as an upper limit for degradation of carbon in solid wood products placed in a landfill.

• Landfill (NGA): DOCf = 10%. This is the value chosen for Australia’s National Greenhouse Accounts (NGA) (Australian Government, 2016a). This is a reduction from the previous value of 23% (Australian Government, 2014b) that was derived from early bioreactor laboratory research from the 1990s (Barlaz, 1998) that investigated the degradability of wood tree branches ground to a fine powder under anaerobic conditions (Australian

Government, 2014a, p. 17). This DOCf value can be considered extremely conservative when compared to values from later research (as used in the typical scenario above) and effectively assumes that at least part of the wood waste is ground into a powder to accelerate degradation. The impacts associated with the landfill are declared in module C4. All landfill gas that is combusted for energy recovery is assumed to be used to generate electricity via an alternator (module C4) and the resulting electricity receives a credit for offsetting average electricity from the Australian grid (module D).

LCI and EPDs for Australian Wood Products 118 of 159

Table 4-79: Degradable organic carbon fractions (DOCf) of timber and wood products from bioreactor research (Wang, et al., 2011; Ximenes, et al., 2013)

Timber Product DOCf typical (%) DOCf NGA (%) Australia grown softwood / Radiata 0.1 10 Australia grown hardwood / Blackbutt 0 10 Particleboard 1.6 10 MDF 0.7 10 Plywood 1.4 10 Softwood glulam (based on softwood) 0.1 10 Hardwood glulam (based on hardwood) 0 10

Why include two scenarios?

As part of its submission to the United Nations Framework Convention on Climate Change, the Australian Government (2014a, p. 17) stated that:

Estimates of DOCf that are applicable to very long term time horizons (3-5 half lives) can be estimated from investigations into the carbon storage under anaerobic conditions of a range of waste types under laboratory conditions (Barlaz, 1998; Barlaz, 2005; Barlaz, 2008). This experimental work involves the testing of a range of waste types in reactors operated to obtain maximum methane yields. As the laboratory work optimises the conditions for

anaerobic decay, the results can be considered as true estimates of the DOCf value that would apply over very long time horizons. These estimates could also be considered to represent an upper limit of the decay processes found in landfills under anaerobic conditions over more restricted time horizons. A review was undertaken by an external expert (Guendehou, 2010). Guendehou concluded that “Australia should take advantage of the availability of good waste composition data to apply waste type specific DOCf in order to improve the accuracy of the emissions estimate”.

In 2011 the Australian Government reviewed these figures and concluded that the DOCf value, for inventory purposes, for wood deposited in landfill should be 23%. The Australian Government further stated that additional research then currently underway was indicating that certain timber classes may be displaying much lower rates of degradation in ideal anaerobic conditions. The

Government committed to reviewing the DOCf values again once that research was published. This research was subsequently published by Wang et al. (2011) and Ximenes et al. (2013). The DOCf values listed in Table 4-79 are those values arrived at experimentally from such studies by Wang et al. (2011) and Ximenes et al. (2013) for a range of specific Australian-grown timber species and manufactured wood products. Since the time of the original study, the Australian Government has revised the DOCf for wood products down to 10% (Australian Government, 2016a, Table 43). Both the experimental and the governmental values are provided within each EPD, as the “typical” and “NGA” scenarios respectively.

Landfill modelling assumptions

Within the LCA model, the landfill is assumed to be a typical municipal waste landfill with surface and basic sealing meeting European limits for emissions. While the landfill is sited in Australia, it is assumed that there are only small differences between Australian and European landfills. The site includes leachate treatment together with sludge treatment and deposition. Landfill gas capture and combustion is also included based on the predicted fraction of Australian landfills fitted with this technology in 2020 (Hyder Consulting, 2007). Impacts associated with sealing materials (clay,

LCI and EPDs for Australian Wood Products 119 of 159

mineral coating and polyethylene film) and diesel for the compactor are included. The two landfill datasets for wood and wood waste available in the Australian LCI database appeared to be much less detailed than the European dataset. Key assumptions: 1. Landfill height: 30 m 2. Landfill area: 40,000 m2 3. 100 year time horizon 4. For degradable organic carbon (see Figure 4-11): o Decomposition rates of organic carbon to landfill gas as per Table 4-79. o Of the landfill gas produced from decomposition of wood, 50% forms methane and 50% forms carbon dioxide (Australian Government, 2016a, Table 43). o All carbon dioxide is released directly to the atmosphere. o 36% of the methane is captured, based on forecasted average methane capture in Australian landfills by 2020 (Hyder Consulting, 2007). Of this, one quarter (9% of the total) is flared and three quarters (27% of the total) is used for energy recovery (Carre, 2011). o Of the 64% of methane that is not captured, 10% (6.4% of the total) is oxidised (Australian Government, 2016a, Table 43) and 90% (57.6%) is released to the atmosphere. o In summary, for every kilogram of carbon converted to landfill gas, 71.2% is released as carbon dioxide and 28.8% is released as methane. 5. Landfill gas is combusted in an engine with an electrical efficiency of 36% to generate electricity to be supplied to the national grid mix (Australian Government, 2016b). 6. Electricity is fed into the electricity grid mix. A credit is given for replaced electricity in module D.

Key assumptions for leachate: 7. 60% transpiration/run-off rate 8. Exponential solubility of fluids in landfill leachate 9. Leachate and landfill body are homogeneous 10. Landfill body is saturated 11. No circulation of leachate 12. 70% basic sealing effectiveness for leachate 13. Leachate treatment includes active carbon and flocculation/precipitation processing 14. Sludge treatment and deposition is included

LCI and EPDs for Australian Wood Products 120 of 159

Figure 4-11 Key assumptions for biogenic carbon in landfill (reproduced from Carre (2011))

The background system is addressed as follows: the sealing contains gravel, sand, clay and polyethylene film as most relevant processes. Gravel and sand are used as filter layers, PE film as waterproofed sealing and clay as mineral coverage in the surface and basic sealing. Gravel, sand and clay are mined from dry quarry. The basis for the production of polyethylene film is crude oil. All manufacturing processes of the sealing materials are considered.

4.12.2. Energy recovery

This scenario includes the preparation of wood chips (chipping and drying to 12% MC, depending on initial moisture content). The wood chips are then used as a fuel to provide thermal energy to a boiler. Environmental impacts of the wood fuel preparation are reported in module C3 and are treated as a secondary material (EN 16485:2014). Carbon sequestered in wood is assumed to leave the system boundary at module C3 and is exported to future product systems in line with EN 16485. The combustion of wood chips is included in Module D along with an offset against thermal energy from natural gas (EN 16485:2014).

LCI and EPDs for Australian Wood Products 121 of 159

4.12.3. Reuse

According to a survey of Australian demolishers in 2008 (Kapambwe, et al., 2008), 40% of timber products from demolished residential structures in Australia are salvaged for reuse. In the EPDs, the product is assumed to be removed from a building manually and reused with no further processing (i.e. direct reuse). Transport and wastage are excluded and only one reuse cycle is considered. The second life is assumed to be the same (or very similar) to the first, meaning that a credit is given for production of one unit of the input product in module D (excluding biogenic carbon). Carbon sequestered in wood is assumed to leave the system boundary at module C3 and is exported to future product systems in line with EN 16485. Any further processing, waste or transport would need to be modelled and included separately, e.g. transporting old, large dimension hardwood beams offsite for sawing to make furniture. Reuse is not provided for in EN 16485. In this case, the chosen offset is virgin wood. As can be seen in Table 4-77, a reuse scenario is only provided for hardwood, plywood formwork and glulam. An offset of virgin wood is applied as it is assumed that wood is only reused once. This assumption has been made as (1) reuse is currently relatively limited; (2) reuse focuses on larger members that are higher value and more easily separated from the rest of the waste stream; and (3) the wood will often be sawn further before being reused in a new building, e.g. a large beam being sawn down into floorboards, making it less favourable to reuse in future (Forsythe Consultants, 2007).

4.12.4. Recycling

Wood may be recycled in many different ways. The scenario included in the EPDs is effectively downcycling into wood chips in line with EN 16485 (see Table 4-78). Wood waste is chipped (module C3) and assigned credits relative to the avoided production of woodchips from virgin softwood (for all products except hardwood) or hardwood (for hardwood products only) (module D). In line with the reuse scenario, carbon sequestered in wood leaves the system boundary at module C3 (EN 15804:2012+A1:2013). Whether the wood chips replace landscape mulch, animal bedding or are used to produce particle board does not matter as in all scenarios recycled woodchips replace virgin wood chips and no further processing is considered.

LCI and EPDs for Australian Wood Products 122 of 159

5. LCIA Results

This chapter contains the results for the impact categories and additional metrics defined in section 3.7. It shall be reiterated at this point that the reported impact categories represent impact potentials, i.e., they are approximations of environmental impacts that could occur if the emissions would (a) follow the underlying impact pathway and (b) meet certain conditions in the receiving environment while doing so. In addition, the inventory only captures that fraction of the total environmental load that corresponds to the chosen functional unit (relative approach). LCIA results are therefore relative expressions only and do not predict actual impacts, the exceeding of thresholds, safety margins, or risks. Within sections 5.1 to 5.6, all graphs presenting a comparison to other studies display net sequestration as a negative for clarity, with a more negative comparison indicating a higher net sequestration. GWPBC and GWPEB have also been used, instead of GWPB and GWPF respectively for comparison purposes, as these were used in the baseline study. GWPBC is the GWP impacts resulting from the uptake or release of carbon-dioxide originating from biogenic sources. This differs to GWPB used elsewhere in this study which also includes any GWP impacts from methane originating from biogenic sources. GWPEB is the total GWP (GWPT) - GWPB. Similar to GWPF used elsewhere in this study, which is calculated as GWPT - GWPB. ODP has also not been displayed in graphs presenting a comparison to other studies due to large differences compared to the current study. The scale of these differences can be seen in the environmental impact tables within sections 5.1 to 5.6. There are no major releases of ozone depleting substances in the primary data. The large decrease in ODP relative to the baseline study seen for most products occurs because of updates in background data within the GaBi database since the baseline study. This difference does not reflect a major change within the foreground system since the baseline study.

5.1. Sawn softwood

This section displays the LCIA results for 1 m3 kiln-dried, dressed softwood. For detailed assessment results of all products and life cycle stages included within the study please refer to Annex B

5.1.1. EN 15804 results (A1-A3)

The LCIA results for environmental impacts of kiln-dried, dressed softwood are displayed in Table 5-1. These results only show the production phase impacts (A1-A3) as no comparison was available for other life cycle stages from other studies. Note that GWPF and GWPB were not assessed within any of the comparison studies. For completeness, the EN 15804 LCI indicators for resource use and waste are shown in Table 5-2 and Table 5-3 respectively for the production stage (A1-A3). Baseline (2006) refers to data collected by the Commonwealth Scientific and Industrial Research Organisation (CSIRO) that inventoried data from the 2005/06 Australian financial year (1st July 2005

LCI and EPDs for Australian Wood Products 123 of 159

to 30th June 2006) (CSIRO, 2009a), as updated by EPD Registration No. S-P-00560, Version 1.1, Revised 13 July 2015, Valid until 22 June 2020.

Table 5-1 Environmental impacts for production stage (A1-A3) for 1 m3 kiln-dried, dressed softwood

Environmental Unit Average Baseline CORRIM - Wood for impact (2006) Softwood Good - lumber Softwood GWPT kg CO2-eq. -6.99E+02 -6.31E+02 -4.26E+02 -7.12E+02 GWPF kg CO2-eq. 1.83E+02 - - - GWPB kg CO2-eq. -8.82E+02 - - - ODP kg CFC11-eq. 4.72E-11 4.23E-09 3.09E-09 2.52E-09 AP kg SO2-eq. 1.10E+00 1.33E+00 7.73E-01 6.44E-01 EP kg PO43-- eq. 2.75E-01 3.65E-01 7.43E-02 1.26E-01 POCP kg C2H4-eq. 6.80E-01 6.72E-01 1.48E-01 4.53E-02 ADPE kg Sb-eq. 7.86E-05 4.31E-05 4.36E-07 5.13E-06 ADPF MJ 2.25E+03 3.11E+03 1.07E+03 1.42E+03

Table 5-2 Resource use for production stage (A1-A3) for 1 m3 kiln-dried, dressed softwood

Resource use Unit Average Baseline CORRIM - Wood for (2006) Softwood Good - lumber Softwood PERE MJ 3.05E+03 3.04E+03 4.95E+03 2.27E+03 PERM MJ 9.29E+03 9.29E+03 - 8.44E+03 PERT MJ 1.23E+04 1.23E+04 4.95E+03 1.07E+04 PENRE MJ 2.26E+03 3.13E+03 6.24E+03 1.57E+03 PENRM MJ - - - - PENRT MJ 2.26E+03 3.13E+03 6.24E+03 1.57E+03 SM kg - - - - RSF MJ - - - - NRSF MJ - - - - FW m3 1.36E+00 1.21E+00 9.50E+01 2.04E-01

Table 5-3 Waste categories and output flows for production stage (A1-A3) for 1 m3 kiln-dried, dressed softwood

Waste Unit Average Baseline CORRIM - Wood for categories and (2006) Softwood Good - output flows lumber Softwood HWD kg 1.33E-06 5.17E-04 - 7.54E-06 NHWD kg 2.31E+01 6.32E+01 - 1.37E+00 RWD kg 2.53E-03 6.56E-03 - 6.15E-02 CRU kg - - - - MFR kg - - - - MER kg - - - - EEE MJ - - - -

LCI and EPDs for Australian Wood Products 124 of 159

Waste Unit Average Baseline CORRIM - Wood for categories and (2006) Softwood Good - output flows lumber Softwood EET MJ - - - -

5.1.2. Detailed results

The results from the LCIA show that most of the environmental impacts arising from the production of kiln-dried, dressed, untreated softwood occur during either the kiln-drying stage or from burning of residues during the forestry stage, depending on the impact category. Figure 5-1 displays the percentage contribution of each production area to the total production impact, for each environmental impact indicator.

01 Seeds 90.0%

11 Packaging 80.0% 02 Stand establishment 70.0% 60.0% 50.0% 40.0% 10 Planing 03 Forest residue burning 30.0% 20.0% 10.0% 0.0% -10.0%

09 Kiln drying 04 Forest management

08 Storing, debarking and sawm 05 Thinning

07 Log haulage 06 Clear felling

Global warming potential (fossil) Depletion potential of the stratospheric ozone layer

Acidification potential of land and water Eutrophication potential

Photochemical ozone creation potential Abiotic depletion potential – elements

Abiotic depletion potential – fossil fuels

Figure 5-1 Softwood, KD, dressed A1-A3 impacts by production stage

Kiln-drying contributes over 40% of the module A1-A3 impacts for four of the nine assessed environmental impact categories (GWPF, AP, POCP, ADPF) and over 35% for five of the nine assessed impact categories (additionally EP). The main cause of these impacts from kiln-drying originate from the combustion of natural gas, combustion of wood residues or use of electricity. Due to its high energy demand, kiln-drying is also the primary contributor towards many of the assessed energy-related inventory categories, namely, PERE, PERT, PENRE and PENRT. This stage is also the largest contributor towards the net use of fresh water.

LCI and EPDs for Australian Wood Products 125 of 159

Chemical application within the forest, such as herbicides and fungicides, as part of forest management also has significant impacts for the impact categories ODP and EP. The additional types of chemicals modelled is also the primary cause of the large increase in ADPE. Seed production has little impact in any of the assessed impact categories and inventory categories except for water, for which it contributes approximately 30% of the net use of fresh water. Stand establishment, thinning, clearfelling, and log haulage have minor significance and together contribute less than 15% of the total to each of our assessed impact and inventory categories. An increase in POCP impacts relative to the baseline study is seen in Figure 5-2, which displays the comparison studies relative to the average impacts from the current study normalised to 100%. This originates from the burning of residues in the forest despite an overall decrease in the total mass of residues burnt per cubic metre of dressed timber. This is due to a revision of the emission factors from burning forest residues and the introduction of NMVOC emissions as described in section 4.3.3 “Forest residue burning”.

200.00%

150.00%

100.00%

50.00%

0.00%

-50.00%

-100.00%

-150.00% GWPT GWPEB GWPBC AP EP POCP ADPE ADPF

Average Baseline (2006) CORRIM - Softwood lumber Wood for Good - Softwood

Figure 5-2 Softwood, KD, dressed A1-A3 comparison to other softwood studies

5.2. Sawn hardwood

This section displays the LCIA results for 1 m3 kiln-dried, dressed hardwood. For detailed assessment results of all products included within the study please refer to Annex B

LCI and EPDs for Australian Wood Products 126 of 159

5.2.1. EN 15804 results (A1-A3)

The LCIA results for environmental impacts of kiln-dried, dressed hardwood are displayed in Table 5-4. For completeness, the EN 15804 LCI indicators for resource use and waste are shown in Table 5-5 and Table 5-6 respectively for the production stage (A1-A3). These results only show the production phase impacts (A1-A3) as no comparison was available for other life cycle stages from other studies. Note that GWPF and GWPB were not assessed within any of the comparison studies. For completeness, the EN 15804 LCI indicators for resource use and waste are shown in Table 5-4 and Table 5-5 respectively for the production stage (A1-A3). Baseline (2006) refers to data collected by the Commonwealth Scientific and Industrial Research Organisation (CSIRO) that inventoried data from the 2005/06 Australian financial year (1st July 2005 to 30th June 2006) (CSIRO, 2009a), as updated by EPD Registration No. S-P-00561, Version 1, Issued 13 August 2015, Valid until 13 August 2020.

Table 5-4 Environmental impacts for production stage (A1-A3) for 1 m3 kiln-dried, dressed hardwood

Environmental Unit Average Baseline (2006) CORRIM - impact Redwood Decking (California) GWPT kg CO2-eq. -7.31E+02 -6.73E+02 -6.79E+02 GWPF kg CO2-eq. 3.27E+02 - - GWPB kg CO2-eq. -1.06E+03 - - ODP kg CFC11-eq. 8.92E-11 1.21E-09 3.57E-06 AP kg SO2-eq. 2.54E+00 3.58E+00 8.69E-01 EP kg PO43-- eq. 5.65E-01 7.78E-01 1.03E-01 POCP kg C2H4-eq. 3.88E+00 3.03E+00 1.18E-01 ADPE kg Sb-eq. 1.14E-05 5.14E-05 4.52E-07 ADPF MJ 3.83E+03 5.24E+03 1.32E+03

Table 5-5 Resource use for production stage (A1-A3) for 1 m3 kiln-dried, dressed hardwood

Resource use Unit Average Baseline (2006) CORRIM - Redwood Decking (California) PERE MJ 1.19E+03 3.10E+03 1.34E+02 PERM MJ 1.26E+04 1.32E+04 - PERT MJ 1.38E+04 1.63E+04 1.34E+02 PENRE MJ 3.84E+03 5.26E+03 2.30E+03 PENRM MJ - - - PENRT MJ 3.84E+03 5.26E+03 2.30E+03 SM kg - - - RSF MJ - - - NRSF MJ - - - FW m3 1.92E+00 2.10E+00 8.50E+05

LCI and EPDs for Australian Wood Products 127 of 159

Table 5-6 Waste categories and output flows for production stage (A1-A3) for 1 m3 kiln-dried, dressed hardwood

Waste Unit Average Baseline (2006) CORRIM - categories and Redwood output flows Decking (California) HWD kg 1.29E-06 9.90E-04 - NHWD kg 4.64E+01 4.35E+01 - RWD kg 3.20E-03 8.00E-03 - CRU kg - - - MFR kg - - - MER kg - - - EEE MJ - - - EET MJ - - -

5.2.2. Detailed results

The results from the LCIA show that most of the environmental impact indicators which are linked to the burning of wood (AP, EP and POCP) originate from the burning of forest residues during the forestry stage for the production of kiln-dried, dressed, untreated hardwood. The impacts arising from each of the three sawmill unit processes (storage debarking and milling, kiln-drying, and planing) are relatively evenly distributed, with the exception of ODP, which occurs mainly during the sawing, de-barking and milling or packaging stages. Figure 5-3 shows the percentage contribution of each production stage to the total production impact, for each environmental impact indicator.

LCI and EPDs for Australian Wood Products 128 of 159

01 Seed collection 80.0%

11 Packaging 70.0% 02 Stand establishment 60.0%

50.0%

40.0%

10 Planing 30.0% 03 Forest residue burning 20.0%

10.0%

0.0%

-10.0%

09 Kiln drying 04 Forest management

08 Storing, debarking and sawm 05 Fire prevention

07 Log haulage 06 Harvest

Global warming potential (fossil) Depletion potential of the stratospheric ozone layer

Acidification potential of land and water Eutrophication potential

Photochemical ozone creation potential Abiotic depletion potential – elements

Abiotic depletion potential – fossil fuels

Figure 5-3 Hardwood, KD, dressed A1-A3 impacts by production stage

For the production stages at the sawmill, electricity is a key contributor to the impacts for almost all of the assessed environmental indicators. The electricity consumption for storage, milling and debarking and planing is similar per cubic metre of dressed timber as shown in Table 4-35 and Table 4-37, therefore resulting in similar environmental impacts for these two unit processes. Kiln- drying uses less electricity (see Table 4-36) than for storage, milling and debarking and planing, however the combustion of wood residues, natural gas and LPG significantly contribute to a range if indicators for kiln-drying. A large increase in POCP impacts relative to the baseline study is seen in Figure 5-4, which displays the comparison studies relative to the average impacts from the current study normalised to 100%. This originates from the burning of residues in the forest despite an overall decrease in the total mass of residues burnt per cubic metre of dressed timber. This is due to a revision of the emission factors from burning forest residues and the introduction of NMVOC emissions as described in section 4.3.3 “Forest residue burning”. Aside from POCP, improvements are seen across all impact categories and are primarily caused by an overall increased recovery rate at the sawmill and a decrease in residues burnt during the forestry stage. This improvement in the recovery rate results in a decrease in log inputs per cubic metre of dressed timber, and hence a decrease in impacts coming from the forestry stage. The large decrease in ADPE is due to a change in the modelling of municipal water supply from “de-ionised water” to “tap water” as this was seen to be more representative of the Australian municipal water supply.

LCI and EPDs for Australian Wood Products 129 of 159

500.00%

400.00%

300.00%

200.00%

100.00%

0.00%

-100.00%

-200.00% GWPT GWPEB GWPBC AP EP POCP ADPE ADPF

Average Baseline (2006) CORRIM - Redwood Decking (California)

Figure 5-4 Hardwood, KD, dressed A1-A3 comparison to other hardwood studies

5.3. Particleboard

This section displays the LCIA results for 1 m2 MR, E1, Melamine 18mm particleboard. For detailed assessment results of all products included within the study please refer to Annex B

5.3.1. EN 15804 results (A1-A3)

The LCIA results for environmental impacts of MR, E1, Melamine 18mm particleboard are displayed in Table 5-7. These results only show the production phase impacts (A1-A3) as no comparison was available for other life cycle stages from other studies. Note that GWPF and GWPB were not assessed within any of the comparison studies. For completeness, the EN 15804 LCI indicators for resource use and waste are shown in Table 5-8 and Table 5-9 respectively for the production stage (A1-A3). Baseline (2006) refers to data collected by the Commonwealth Scientific and Industrial Research Organisation (CSIRO) that inventoried data from the 2005/06 Australian financial year (1st July 2005 to 30th June 2006) (CSIRO, 2009a), as updated by EPD Registration No. S-P-00562, Version 1, Issued 21 October 2015, Valid until 21 October 2020.

LCI and EPDs for Australian Wood Products 130 of 159

Table 5-7 Environmental impacts for production stage (A1-A3) for 1 m2 MR, E1, Melamine 18mm particleboard

Environmental Unit Average Baseline CORRIM - Wood for impact (2006) Particle Good - board Particleboard (untreated) GWPT kg CO2-eq. -8.76E+00 -1.63E+00 -8.54E+00 -1.11E+01 GWPF kg CO2-eq. 9.40E+00 - - - GWPB kg CO2-eq. -1.82E+01 - - - ODP kg CFC11-eq. 4.61E-11 1.44E-10 3.91E-08 3.55E-10 AP kg SO2-eq. 3.02E-02 3.84E-02 6.86E-02 1.90E-02 EP kg PO43-- eq. 7.76E-03 9.86E-03 4.39E-03 2.19E-03 POCP kg C2H4-eq. 9.76E-03 8.34E-03 8.69E-03 2.69E-03 ADPE kg Sb-eq. 3.55E-06 4.64E-06 5.53E-08 1.48E-06 ADPF MJ 1.37E+02 2.09E+02 1.03E+02 9.95E+01

Table 5-8 Resource use for production stage (A1-A3) for 1 m2 MR, E1, Melamine 18mm particleboard

Resource use Unit Average Baseline CORRIM - Wood for (2006) Particle Good - board Particleboard (untreated) PERE MJ 5.23E+01 4.41E+01 3.17E+02 1.33E+02 PERM MJ 1.91E+02 1.38E+02 - 1.85E+02 PERT MJ 2.44E+02 1.82E+02 3.17E+02 3.18E+02 PENRE MJ 1.39E+02 2.13E+02 8.41E+01 1.05E+02 PENRM MJ 2.14E+01 3.59E+01 - - PENRT MJ 1.61E+02 2.49E+02 8.41E+01 1.05E+02 SM kg 1.31E-01 - - - RSF MJ - - - - NRSF MJ - - - - FW m3 5.51E-02 7.55E-02 4.04E+01 3.18E-02

Table 5-9 Waste categories and output flows for production stage (A1-A3) for 1 m2 MR, E1, Melamine 18mm particleboard

Waste Unit Average Baseline CORRIM - Wood for categories and (2006) Particle Good - output flows board Particleboard (untreated) HWD kg 9.41E-08 8.52E-05 - 2.21E-03 NHWD kg 2.86E-01 2.13E-01 - 3.37E-02 RWD kg 9.62E-04 1.66E-03 - 2.12E-03 CRU kg - - - - MFR kg - - - - MER kg - - - - EEE MJ - - - -

LCI and EPDs for Australian Wood Products 131 of 159

Waste Unit Average Baseline CORRIM - Wood for categories and (2006) Particle Good - output flows board Particleboard (untreated) EET MJ - - - -

5.3.2. Detailed results

The LCIA shows the environmental impacts of moisture resistant E1, melamine-coated 18mm particleboard mainly occur from the production of wood fibres, resins and the use of energy and water during the production process. Figure 5-5 displays the percentage contribution of each production area to the total production impact, for each environmental impact indicator.

01 Wood particles and fibres 100.0%

90.0%

80.0%

70.0% 07 Packaging 02 Energy and water 60.0%

50.0%

40.0%

30.0%

20.0%

10.0%

0.0%

06 Additives 03 Resin

05 Lamination 04 Wax

Global warming potential (fossil) Depletion potential of the stratospheric ozone layer

Acidification potential of land and water Eutrophication potential

Photochemical ozone creation potential Abiotic depletion potential – elements

Abiotic depletion potential – fossil fuels

Figure 5-5 PB, MR, E1, Melamine 18mm A1-A3 impacts by production stage

Wood particles and fibres account for 20% of the POCP emissions for particleboard, 10% of EP, and 10% for ADPE. These impacts originate mainly from burning of forest residues in softwood plantations. Resins account for more than 45% of the total A1-A3 impacts for six of the nine assessed impact categories (GWPT, GWPF, ODP, EP, ADPE & ADPF) and accounts for 55% of net use of fresh

LCI and EPDs for Australian Wood Products 132 of 159

water. Resin production has especially high impacts for ODP and ADPE with 90% and 80% of total A1-A3 phase impacts respectively. The energy and water category accounts for more than 45% of the total A1-A3 impacts for three of the nine assessed impact categories (GWPT, AP & POCP) and more than 30% for six of the nine assessed impact categories (additionally GWPF, EP & ADPF). This category includes emissions from electricity production, thermal energy use and production emissions and other overheads such as waste disposal and manufacturing emissions. Lamination, including MF resin and paper production, has minor significance, contributing approximately 10% and 5% for ODP and ADPF respectively and less than 5% towards the A1-A3 total for the remaining impact categories. Wax, packaging and additives are insignificant and collectively account for less than 5% of the A1- A3 total for each of the assessed impact categories. Key differences to the baseline study include a reduction in resin and wax loadings. This contributes to decreases in ADPE, EP, AP and ODP as seen in Figure 5-6, which displays the comparison studies relative to the average impacts from the current study normalised to 100%. The decrease in ADPF is mainly because of the larger share of biomass used for thermal energy relative to natural gas used in the baseline study. The increase in POCP impacts seen in Figure 5-6 originates from the burning of residues in the forest despite an overall decrease in the total mass of residues burnt per cubic metre of log. This is due to a revision of the emission factors from burning forest residues and the introduction of NMVOC emissions as described in section 4.3.3 “Forest residue burning”.

250.00%

200.00%

150.00%

100.00%

50.00%

0.00%

-50.00%

-100.00%

-150.00% GWPT GWPEB GWPBC AP EP POCP ADPE ADPF

Average Baseline (2006) CORRIM - Particle board Wood for Good - Particleboard (untreated)

Figure 5-6 PB, MR, E1, Melamine 18mm A1-A3 comparison to other particleboard studies

LCI and EPDs for Australian Wood Products 133 of 159

5.4. MDF

This section displays the LCIA results for 1 m2 MR, E1, Melamine, 18mm MDF. For detailed assessment results of all products included within the study please refer to Annex B

5.4.1. EN 15804 results (A1-A3)

The LCIA results for environmental impacts of MR, E1, Melamine 18mm MDF are displayed in Table 5-10. These results only show the production phase impacts (A1-A3) as no comparison was available for other life cycle stages from other studies. Note that GWPF and GWPB were not assessed within any of the comparison studies. For completeness, the EN 15804 LCI indicators for resource use and waste are shown in Table 5-11 and Table 5-12 respectively for the production stage (A1-A3). Baseline (2006) refers to data collected by the Commonwealth Scientific and Industrial Research Organisation (CSIRO) that inventoried data from the 2005/06 Australian financial year (1st July 2005 to 30th June 2006) (CSIRO, 2009a), as updated by EPD Registration No. S-P-00563, Version 1, Issued 21 October 2015, Valid until 21 October 2020.

Table 5-10 Environmental impacts for production stage (A1-A3) for 1 m2 MR, E1, Melamine 18mm MDF

Environmental Unit Average Baseline CORRIM - Wood for impact (2006) Medium Good - MDF density fiberboard (MDF) GWPT kg CO2-eq. -3.55E+00 3.17E+00 -4.00E+00 -1.05E+01 GWPF kg CO2-eq. 1.57E+01 - - - GWPB kg CO2-eq. -1.92E+01 - - - ODP kg CFC11-eq. 6.16E-11 2.01E-10 4.54E-08 5.24E-10 AP kg SO2-eq. 5.68E-02 6.91E-02 1.14E-01 2.99E-02 EP kg PO43-- eq. 1.17E-02 1.52E-02 7.21E-03 6.15E-03 POCP kg C2H4-eq. 2.21E-02 1.35E-02 1.52E-02 5.76E-03 ADPE kg Sb-eq. 4.61E-06 6.74E-06 8.97E-08 1.01E-06 ADPF MJ 2.11E+02 3.09E+02 1.88E+02 8.91E+01

Table 5-11 Resource use for production stage (A1-A3) for 1 m2 MR, E1, Melamine 18mm MDF

Resource use Unit Average Baseline CORRIM - Wood for (2006) Medium Good - MDF density fiberboard (MDF) PERE MJ 5.69E+01 6.83E+01 3.53E+02 9.28E+01 PERM MJ 2.03E+02 1.60E+02 - 2.07E+02 PERT MJ 2.60E+02 2.28E+02 3.53E+02 3.00E+02 PENRE MJ 2.14E+02 3.14E+02 6.35E+01 9.90E+01 PENRM MJ 3.01E+01 5.03E+01 - -

LCI and EPDs for Australian Wood Products 134 of 159

Resource use Unit Average Baseline CORRIM - Wood for (2006) Medium Good - MDF density fiberboard (MDF) PENRT MJ 2.44E+02 3.64E+02 6.35E+01 9.90E+01 SM kg - - - - RSF MJ - - - - NRSF MJ - - - - FW m3 7.59E-02 1.13E-01 5.64E+01 3.79E-02

Table 5-12 Waste categories and output flows for production stage (A1-A3) for 1 m2 MR, E1, Melamine 18mm MDF

Waste Unit Average Baseline CORRIM - Wood for categories and (2006) Medium Good - MDF output flows density fiberboard (MDF) HWD kg 7.44E-08 1.37E-04 - 2.99E-02 NHWD kg 4.78E-01 4.57E-01 - 5.76E-03 RWD kg 1.26E-03 2.24E-03 - 1.01E-06 CRU kg - - - - MFR kg - - - - MER kg - - - - EEE MJ - - - - EET MJ - - - -

5.4.2. Detailed results

The LCIA shows the environmental impacts of moisture resistant, E1, melamine-coated, 18mm MDF mainly occur from the production of wood fibres, resins and the use of energy, during the production process. Figure 5-7 displays the percentage contribution of each production area to the total production impact, for each environmental impact indicator.

LCI and EPDs for Australian Wood Products 135 of 159

01 Wood particles and fibres 100.0%

90.0%

80.0%

70.0%

60.0%

07 Packaging 50.0% 02 Energy and water

40.0%

30.0%

20.0%

10.0%

0.0%

05 Lamination 03 Resin

04 Wax

Global warming potential (fossil) Depletion potential of the stratospheric ozone layer

Acidification potential of land and water Eutrophication potential

Photochemical ozone creation potential Abiotic depletion potential – elements

Abiotic depletion potential – fossil fuels

Figure 5-7 MDF, MR, E1, Melamine, 18mm A1-A3 impacts by production stage

Wood particles and fibres account for approximately 10% of the POCP emissions for MDF, 5% of EP, and 5% for ADPE. These impacts originate mainly from burning of forest residues in softwood plantations. Resins account for more than 35% of the total A1-A3 impacts for six of our nine assessed impact categories (GWPT, GWPF, ODP, EP, ADPE & ADPF) and accounts for 55% of net use of fresh water. Resins production has especially high impacts for ODP and ADPE with 95% and 85% of total A1-A3 phase impacts respectively. The energy and water category accounts for more than 40% of the total A1-A3 impacts for six of the nine assessed impact categories (GWPT, GWPF, AP, EP, POCP & ADPF). This category includes emissions from electricity production, thermal energy use and production emissions and other overheads such as waste disposal and manufacturing emissions. Lamination, including MF resin and paper production, has minor significance, contributing less than 5% towards the A1-A3 total for most of the assessed impact categories. Wax and packaging are insignificant and collectively account for less than 5% of the A1-A3 total for each of the assessed impact categories. Key differences to the baseline study include a reduction in resin and wax loadings. This contributes to decreases in ADPE, EP, AP and ODP as seen in Figure 5-8, which displays the comparison studies relative to the average impacts from the current study normalised to 100%. The decrease in

LCI and EPDs for Australian Wood Products 136 of 159

ADPF is mainly because of the larger share of biomass used for thermal energy relative to natural gas used in the baseline study. The increase in POCP impacts seen in Figure 5-6 originates from the burning of residues in the forest despite an overall decrease in the total mass of residues burnt per cubic metre of log. This is due to a revision of the emission factors from burning forest residues and the introduction of NMVOC emissions as described in section 4.3.3 “Forest residue burning”.

300.00%

200.00%

100.00%

0.00%

-100.00%

-200.00%

-300.00%

-400.00% GWPT GWPEB GWPBC AP EP POCP ADPE ADPF

Average Baseline (2006) CORRIM - Medium density fiberboard (MDF) Wood for Good - MDF

Figure 5-8 MDF, MR, E1, Melamine, 18mm A1-A3 comparison to other MDF studies

5.5. Plywood

This section displays the LCIA results for 1 m2 17mm, formply, A-Bond. For detailed assessment results of all products included within the study please refer to Annex B

5.5.1. EN 15804 results (A1-A3)

The LCIA results for environmental impacts of 17mm, formply, A-Bond are displayed in Table 5-13. These results only show the production phase impacts (A1-A3) as no comparison was available for other life cycle stages from other studies. Note that GWPF and GWPB were not assessed within any of the comparison studies. For completeness, the EN 15804 LCI indicators for resource use and waste are shown in Table 5-14 and Table 5-15 respectively for the production stage (A1-A3).

LCI and EPDs for Australian Wood Products 137 of 159

Baseline (2006) refers to data collected by the Commonwealth Scientific and Industrial Research Organisation (CSIRO) that inventoried data from the 2005/06 Australian financial year (1st July 2005 to 30th June 2006) (CSIRO, 2009a), as updated by EPD Registration No. S-P-00564, Version 1, Issued 14 October 2015, Valid until 14 October 2020.

Table 5-13 Environmental impacts for production stage (A1-A3) for 1 m2 17mm, formply, A-Bond

Environmental Unit Average Baseline CORRIM - Wood for impact (2006) Softwood Good - plywood Plywood GWPT kg CO2-eq. -4.22E+00 -5.50E+00 -1.20E+01 -1.16E+01 GWPF kg CO2-eq. 8.95E+00 - - - GWPB kg CO2-eq. -1.32E+01 - - - ODP kg CFC11-eq. 3.83E-11 1.38E-10 1.29E-09 1.43E-10 AP kg SO2-eq. 4.14E-02 5.86E-02 2.42E-02 2.59E-02 EP kg PO43-- eq. 1.07E-02 1.19E-02 2.13E-03 3.57E-03 POCP kg C2H4-eq. 2.58E-02 1.15E-02 3.50E-03 4.12E-03 ADPE kg Sb-eq. 4.81E-06 5.23E-06 1.66E-07 8.68E-07 ADPF MJ 1.60E+02 1.75E+02 3.49E+01 3.85E+01

Table 5-14 Resource use for production stage (A1-A3) for 1 m2 17mm, formply, A-Bond

Resource use Unit Average Baseline CORRIM - Wood for (2006) Softwood Good - plywood Plywood PERE MJ 1.17E+02 1.34E+02 1.12E+02 5.37E+01 PERM MJ 1.46E+02 1.38E+02 - 1.46E+02 PERT MJ 2.64E+02 2.72E+02 1.12E+02 2.00E+02 PENRE MJ 1.62E+02 1.78E+02 2.00E+01 4.12E+01 PENRM MJ 1.44E+01 1.93E+01 - - PENRT MJ 1.76E+02 1.97E+02 2.00E+01 4.12E+01 SM kg - - - - RSF MJ - - - - NRSF MJ - - - - FW m3 9.04E-02 8.00E-02 8.21E+00 2.13E-01

Table 5-15 Waste categories and output flows for production stage (A1-A3) for 1 m2 17mm, formply, A- Bond

Waste Unit Average Baseline CORRIM - Wood for categories and (2006) Softwood Good - output flows plywood Plywood HWD kg 1.96E-07 2.70E-04 - 6.79E-04 NHWD kg 8.04E-01 9.59E-01 - 1.06E-01 RWD kg 6.19E-04 8.91E-04 - 1.11E-03 CRU kg - - - - MFR kg - - - -

LCI and EPDs for Australian Wood Products 138 of 159

Waste Unit Average Baseline CORRIM - Wood for categories and (2006) Softwood Good - output flows plywood Plywood MER kg - - - - EEE MJ - - - - EET MJ - - - -

5.5.2. Detailed results

The LCIA shows the environmental impacts of 17mm, formply, A-Bond plywood mainly occur from the production of logs, the use of energy during the production of veneer or from the use of resins. Figure 5-9 displays the percentage contribution of each production area to the total production impact, for each environmental impact indicator.

01 Hardwood veneer 100.0%

90.0%

80.0%

70.0%

60.0%

06 Packaging 50.0% 02 Softwood veneer

40.0%

30.0%

20.0%

10.0%

0.0%

05 Glue, resin other 03 Energy and water (veneer)

04 Energy and water (plywood)

Global warming potential (fossil) Depletion potential of the stratospheric ozone layer

Acidification potential of land and water Eutrophication potential

Photochemical ozone creation potential Abiotic depletion potential – elements

Abiotic depletion potential – fossil fuels

Figure 5-9 17mm, formply, A-Bond A1-A3 impacts by production stage

The use of energy and water during the veneer production process was the main contributor to GWPF and AP. The reason for the high impacts arising from this production stage are the burning of wood fuel for thermal energy and the use of electricity. The key contributions to EP, POCP and ADPE were from the production of logs used to make the veneer. In the case of 17mm, formply, A-Bond, these impacts came mainly from the production of softwood logs as 90% of this product type was found to be made from softwood veneer. The main

LCI and EPDs for Australian Wood Products 139 of 159

contributor of EP and ADPE from these stages was chemical application, which does not occur in native hardwood forestry. POCP impacts arising from the forestry stage are due to emissions occurring during the burning of forest residues. The production of PF resin accounts for the majority of the contributions towards the ODP and ADPF indicators and is also significant for the ADPE indicator. When compared to the impact assessment results from the baseline study (see Figure 5-10), a small increase in GWPEB is seen. A small decrease in the amount of biogenic carbon sequestered is also shown by the GWPBC indicator due to the lower average density of the plywood in the current study. The decrease in the AP indicator is due to a decrease in emissions coming from the combustion of wood fuels in the forestry stage and the burning of wood fuels for thermal energy used on-site. Similarly, the decrease seen for the EP indicator is due to a decrease in the amount of biomass burnt during the veneer production process. The reduction for ADPE relative to the baseline scenario is due to preservative treatments not being included within the current study. This decrease is partially offset by the increase in chemicals used in the softwood forestry stage when compared with the baseline study. ADPF has decreased relative to the baseline study. This is due to a decrease in impacts from PF resin. Although the change in PF amounts remains similar, there has been a significant reduction of impacts within the upstream data for PF production. The increase in POCP impacts seen in Figure 5-10, which displays the comparison studies relative to the average impacts from the current study normalised to 100%, originates from the burning of residues in the forest, and an increase in the log input. The increase in residue burning POCP emissions is due to a revision of the emission factors from burning forest residues and the introduction of NMVOC emissions as described in section 4.3.3 “Forest residue burning”.

LCI and EPDs for Australian Wood Products 140 of 159

200.00%

150.00%

100.00%

50.00%

0.00%

-50.00%

-100.00%

-150.00%

-200.00%

-250.00% GWPT GWPEB GWPBC AP EP POCP ADPE ADPF

Average Baseline (2006) CORRIM - Softwood plywood Wood for Good - Plywood

Figure 5-10 17mm, formply, A-Bond A1-A3 comparison to other plywood studies

5.6. Glulam

This section displays the LCIA results for 1 m3 softwood glulam and hardwood glulam. For detailed assessment results of all products included within the study please refer to Annex B

5.6.1. EN 15804 results (A1-A3)

The LCIA results for environmental impacts of softwood glulam and hardwood glulam are displayed in Table 5-16. These results only show the production phase impacts (A1-A3) as no comparison was available for other life cycle stages from other studies. Note that GWPF and GWPB were not assessed within any of the comparison studies. For completeness, the EN 15804 LCI indicators for resource use and waste are shown in Table 5-17 and Table 5-18 respectively for the production stage (A1-A3).

LCI and EPDs for Australian Wood Products 141 of 159

Table 5-16 Environmental impacts for production stage (A1-A3) for 1 m3 softwood glulam and hardwood glulam

Environmental Unit Average Average CORRIM - Glued impact (softwood glulam) (hardwood laminated timber glulam) (softwood) GWPT kg CO2-eq. -6.12E+02 -4.08E+02 -7.09E+02

GWPF kg CO2-eq. 3.80E+02 5.27E+02 - GWPB kg CO2-eq. -9.92E+02 -9.35E+02 - ODP kg CFC11-eq. 4.02E-10 1.57E-10 3.29E-05 AP kg SO2-eq. 1.80E+00 3.41E+00 2.01E+00 EP kg PO43-- eq. 3.78E-01 6.87E-01 1.60E-01 POCP kg C2H4-eq. 8.12E-01 4.37E+00 3.91E-01 ADPE kg Sb-eq. 1.51E-04 5.00E-05 1.35E-03 ADPF MJ 4.96E+03 6.14E+03 3.06E+03

Table 5-17 Resource use for production stage (A1-A3) for 1 m3 softwood glulam and hardwood glulam

Resource use Unit Average Average CORRIM - Glued (softwood glulam) (hardwood laminated timber glulam) (softwood) PERE MJ 3.60E+03 1.43E+03 6.34E+03 PERM MJ 1.02E+04 1.12E+04 - PERT MJ 1.38E+04 1.27E+04 6.34E+03 PENRE MJ 4.98E+03 6.16E+03 3.42E+04 PENRM MJ - - - PENRT MJ 4.98E+03 6.16E+03 3.42E+04 SM kg - - - RSF MJ - - - NRSF MJ - - - FW m3 2.43E+00 3.00E+00 2.56E+04

Table 5-18 Waste categories and output flows for production stage (A1-A3) for 1 m3 softwood glulam and hardwood glulam

Waste Unit Average Average CORRIM - Glued categories and (softwood glulam) (hardwood laminated timber output flows glulam) (softwood) HWD kg 2.11E-06 1.75E-06 - NHWD kg 3.49E+01 5.35E+01 - RWD kg 1.10E-02 8.01E-03 - CRU kg - - - MFR kg 1.10E+00 9.70E-02 - MER kg - - - EEE MJ - - - EET MJ - - -

LCI and EPDs for Australian Wood Products 142 of 159

5.6.2. Detailed results

Softwood glulam

The LCIA shows the environmental impacts of softwood glulam mainly occur from the production of wood fibres, for the majority of the assessed indicators. Resins are significant for two of the indicators. The use of energy and water during the glulam production process is also significant for most indicators, however it is not the key contributor for any of the assessed indicators. Packaging has little significance for softwood glulam. Figure 5-11 displays the percentage contribution of each production area to the total production impact, for each environmental impact indicator. Softwood glulam was not included within the baseline study, therefore no comparison of LCIA results to the baseline study is possible.

01 Wood production 100.0%

90.0%

80.0%

70.0%

60.0%

50.0%

40.0%

30.0%

20.0%

10.0%

04 Packaging 0.0% 02 Energy and water

03 Resin

Global warming potential (fossil) Depletion potential of the stratospheric ozone layer

Acidification potential of land and water Eutrophication potential

Photochemical ozone creation potential Abiotic depletion potential – elements

Abiotic depletion potential – fossil fuels

Figure 5-11 Softwood glulam A1-A3 impacts by production stage

Figure 5-11 shows that the production of kiln-dried, dressed timber as an input to glulam production was the main contributor to all the assessed impact categories except for ODP. This is due to the glulam process being a relatively simple, low energy process with few material inputs. Processes related to energy and water contributed a significant proportion (14-37%) to the GWPT, GWPF, AP, EP and ADPF impact categories. The primary source of emissions within this category was electricity use. Resins contributed the largest share to the ODP impact category and were significant for ADPE and ADPF. The resin contributing the highest share towards the ODP impact category was PRF. For

LCI and EPDs for Australian Wood Products 143 of 159

ADPE, contributions from PU and PRF were similar per cubic metre of glulam, although 3.5 times more PRF was used for softwood glulam than PU. The main contributing resin towards ADP fossil was PRF.

Hardwood glulam

Figure 5-12 displays the percentage contribution of each production area of hardwood glulam to the total production impacts for each indicator. This shows that the production stage impacts follow a similar trend for both softwood and hardwood glulam. Some notable differences are that the kiln-dried, dressed hardwood has higher impacts for indicators such as POCP especially. This is due to more wood residues being burnt in native hardwood forestry than in softwood plantation forestry. The wood production stage therefore, accounts for over 60% of the total production impacts for all the assessed impact indicators except for ADPE, which the wood production accounts for 25% of the total production impacts. The use of energy and water in hardwood glulam production is significant for some of the assessed indicators. Energy and water contribute over 15% of the total production impacts for GWPF, AP and ADPF. The resin type mainly used in the production of hardwood glulam was found to be PU. The resin production accounted for over 65% of the production impacts for the ADPE indicator. It also accounted for over 20% of the production impacts for the ODP indicator. Packaging was found to not be very significant for all indicators except ODP. This was higher than the softwood impacts due to the use of steel strapping on some hardwood glulam products.

LCI and EPDs for Australian Wood Products 144 of 159

01 Wood production 100.0%

90.0%

80.0%

70.0%

60.0%

50.0%

40.0%

30.0%

20.0%

10.0%

04 Packaging 0.0% 02 Energy and water

03 Resin

Global warming potential (fossil) Depletion potential of the stratospheric ozone layer

Acidification potential of land and water Eutrophication potential

Photochemical ozone creation potential Abiotic depletion potential – elements

Abiotic depletion potential – fossil fuels

Figure 5-12 Hardwood glulam A1-A3 impacts by production stage

Glulam was not within the scope of the baseline study; therefore Figure 5-13 displays a comparison for hardwood glulam from the current study and softwood glulam from the CORRIM study relative to softwood glulam from the current study, normalised to 100%. Figure 5-13 shows that softwood glulam from the current study has lower impacts than hardwood glulam for all the assessed indicators except for ADPE, which was lower for hardwood glulam.

LCI and EPDs for Australian Wood Products 145 of 159

1000%

800%

600%

400%

200%

0%

-200% GWPT GWPF GWPB AP EP POCP ADPE ADPF

Softwood (average) CORRIM (softwood) Hardwood (average)

Figure 5-13 Softwood glulam and hardwood glulam comparison to CORRIM

5.7. Wood preservative treatments

For detailed assessment results of all wood preservative treatments included within the study please refer to Annex C.

5.8. Release of hazardous substances during the use stage

Particleboard, MDF, plywood and glulam are all made from formaldehyde-containing resins that emit small quantities of formaldehyde during use. Formaldehyde is a colourless strong smelling gas which occurs naturally in the environment. It is present in the air that we breathe at natural background levels of about 0.03 parts per million (ppm) and up to 0.08 ppm in urban air (EWPAA, 2012). Formaldehyde is used as an ingredient in synthetic resins, industrial chemicals, preservatives, and in the production of paper, textiles, cosmetics, disinfectants, medicines, paints, varnishes and lubricants. MDF and particleboard manufactured in Australia use one of two low formaldehyde emitting plastics: urea formaldehyde (UF) or melamine urea formaldehyde (MUF) (EWPAA, 2012). Australian plywood manufacturers use phenol formaldehyde or melamine formaldehyde (MF). Additionally, glulam manufactured in Australia typically uses either phenol resorcinol formaldehyde (PRF) or polyurethane (PU) which contains no formaldehyde.

LCI and EPDs for Australian Wood Products 146 of 159

The Engineered Wood Products Association of Australasia runs an industry-wide formaldehyde testing and labelling programme. All participating mills are required to forward samples to EWPAA's National Laboratory on a regular basis for formaldehyde emission testing. On the basis of laboratory tests, all Australian mills are permitted to brand a formaldehyde emission class on their products as detailed in Table 5-19 below. Table 5-19: Formaldehyde emission classes for Australian engineered wood products

Emission class Emission limit (mg/litre) Emission limit (ppm)* Super E0 Less than or equal to 0.3 Less than or equal to 0.03 E0 Less than or equal to 0.5 Less than or equal to 0.04 E1 Less than or equal to 1.0 Less than or equal to 0.08 * Based on a test chamber volume of 10litre, zero airflow during the 24hr test cycle, molecular weight of formaldehyde 30.03 and the number of microlitres of formaldehyde gas in 1 micromole at 101KPa and 298K. Participating Australian manufacturers listed in each EPD can supply test certificates to support their emission class. Engineered wood products with formaldehyde emissions of less than or equal to E1 are compliant with the basic Green Star Formaldehyde credit while E0 and Super E0 emission class products are compliant with the most stringent Green Star Formaldehyde credit. To achieve credit points all engineered wood products used in the project must be in accordance with these requirements.

LCI and EPDs for Australian Wood Products 147 of 159

6. Interpretation

6.1. General

The results in the EPDs were produced from a sample of data from the wood products industry in Australia. As such, the results do not necessarily reflect the whole industry, but are representative of the sample at the time data were collected. Furthermore, the reader cannot purchase the “average” product on the market, but can only purchase the product of a particular producer. The coverage of Australian production is shown in Table 6-1. The target set in section 4.1.1 of >50% coverage for all product groups except hardwood (>25%) was achieved for all products except sawn softwood,

Table 6-1: Coverage of Australian production Category Products Coverage of Australian production (approx. %) Logs – Softwood Large/Medium/Small saw logs, 61 Veneer logs, Pulplogs, Other Logs – Hardwood Veneer log, Saw log, Pulplogs, 88 Other Sawn timber – Softwood Rough sawn kiln dried timber, 33-35 Planed kiln dried timber Sawn timber – Hardwood Rough sawn green timber, 27-29 Rough sawn kiln dried timber, Planed kiln dried timber Plywood Exterior Plywood, Formply, 100 Tongue and Groove Flooring (2 thicknesses of each) Particleboard Standard melamine coated, 54 Moisture resistant melamine coated (2 thicknesses of each), Flooring (3 thicknesses). Medium density fibreboard Standard melamine coated, 100 (MDF) Moisture resistant melamine coated. (3 thicknesses of each) Glued-laminated timber Softwood glulam 64 (Glulam) Hardwood glulam

An indication of the variability in the data across most of EN 15804’s environmental impact assessment indicators is presented for each product in Annex B. It should be noted that the LCIA results and any interpretation of results are reflections of the underlying LCI data and assumptions described within this report. An indication of the variability of LCI data for each assessed product is given in Annex A. While interpreting the impact assessment results, it has to be borne in mind that all statements are relative expressions only and give no information about the endpoint of each impact category, or the exceeding of threshold values, safety margins or risk.

LCI and EPDs for Australian Wood Products 148 of 159

6.2. Identification of relevant findings

Sawn softwood: The carbon footprint has improved slightly relative to the previous study, with small to moderate improvements to some other environmental indicators (eutrophication and acidification) occurring due to a reduction in residues burnt during forestry, an improvement in sawmill recovery rates, and, to a lesser extent, a reduction in energy used in kiln-drying. Sawn hardwood: The carbon footprint has improved slightly relative to the previous study. This is largely due to a higher recovery rates achieved by the participating facilities, which has a positive effect across all indicators. Improvements were also found for acidification and eutrophication due to reduced energy use during kiln-drying. A revision to the emission factors from burning forest residues made in this study led to an increase for the summer smog indicator. The burning of residues during the forestry stage remains significant for a range of environmental indicators. Particleboard: Carbon footprint and most other environmental indicators have improved relative to the previous study due to lower resin loadings in the finished products and the substitution of natural gas for biomass energy. Revised emission factors resulting from forest residue burning cause the main increase in summer smog formation, with the substitution of natural gas for biomass also resulting in a small increase for this indicator. MDF: Environmental impacts have reduced across most environmental indicators due to lower resin loadings in the finished products and a reduction in natural gas use. All products included within the study now have a negative carbon footprint from cradle-to-gate (i.e., more carbon uptake during tree growth than carbon released during production). Summer smog formation has increased due to the revised emission factor from forest residue burning upstream. Plywood: A reduction can be observed across all assessed indicators except summer smog. This is due to a reduction in energy use within the veneer production process, including a reduction in electricity use, natural gas and biomass. The increase in smog formation is attributed to revised emissions from residue burning in the forestry stage. Glulam: The upstream process of producing sawn timber is the most significant across all assessed indicators except for ozone depletion for softwood glulam and depletion of elements for hardwood glulam. Higher forest burning and lower recovery rates for hardwood production result in higher environmental impacts across all assessed indicators compared to softwood glulam, except for abiotic depletion, which is higher for softwood glulam due to the use of chemicals in softwood plantations and the use of different resin types. Preservative treatment: These were found to be relatively low impact for the indicators included, despite their benefits for longevity. Copper-based treatments are found to be generally higher- impact for the assessed indicators.

6.3. Assumptions and limitations

6.3.1. Key assumptions

The literature source for emissions associated with the burning of forest residues has been updated in this study to a more recent source (Commonwealth of Australia, 2016). Emissions are taken for the burning of coarse residues in temperate forests for both softwood and hardwood forestry. See section 4.3.3 “Forest residue burning”.

LCI and EPDs for Australian Wood Products 149 of 159

Due to the rarity of wildfire events for softwood forestry and the lack of historical data to create an accurate estimate of occurrence rates, emissions from wildfire have not been included. See section 4.3.7. Carbon emissions from wildfires in native hardwood forests have been excluded from this study as there is no accepted method for the treatment of wildfire in LCA of temperate forests within Australia, due to the high variability (and hence high uncertainty) of emissions from wildfires, and difficulties in attributing the relative contribution to emissions from different forest management actions (S. H. Roxburgh, personal communication, 9 October 2017). See section 4.4.7. The electricity grid mix has been modelled to reflect the state level, production-weighted average for each product category. The electricity grid mixes for sawn softwood and sawn hardwood have been modelled as a production-weighted average, based on production volume by state, taken from 2012/2013 ABARES reported values (ABARES, 2017). MDF, particleboard and plywood coverage was high within this study and have therefore been modelled based on the actual electricity consumption from surveyed companies. Published production volumes were used for the remaining particleboard sites which did not take part in the study to estimate the electricity consumption by those sites. A mix of both approaches was used for glulam: electricity used to produce sawn inputs was based on the grid mixes used for those products. Electricity used at the glulam facilities was based on production-weighted values for the participating sites and the state where they were located. The electricity grid mix used for all end-of-life scenarios was the Australian average grid mix, as the location of disposal is unknown. Actual electricity inputs by state and individual product type were not known. The modelling of preservative treatments in this study covers the production and application of treatment chemicals. Landfill of treated timber has been excluded as any preservative treatment will act to reduce the decomposition rate of wood in the landfill, and reduce any associated carbon emissions. The effects on other assessed impact categories are assumed to be insignificant due to the very small amount of preservative per cubic metre of timber, and the assumption of a sealed landfill. Potential impacts from the use of treated timber products such as leaching of treatment chemicals over the service life of the product, was outside the scope of the study (module B1).

6.3.2. Limitations of the study

All products declared within this study aim to represents the Australian industry average scenario. An Australian average product cannot be physically purchased, as each individual product will have a unique environmental profile based on the specific supply chain and region of production. This may result in specific products with an environmental profile higher or lower than the Australian industry average. Plywood production assumes the Australian average veneer manufacture process is applicable for all plywood products. However, certain facilities were found to specialise in the production of specific product types. Aggregation of veneer production data for all products was done to protect site confidentiality due to the small sample size. Considering individual veneer production for each product type and facility may result in alternative outcomes depending on the product. Plywood products which can be made from both softwood and hardwood veneer have been treated as a single product type. Therefore, inputs and outputs of the modelled process represent an average of the two. The distinction between hardwood and softwood products was not made to protect site confidentiality due to the small sample size, but such blended softwood and hardwood products are not typically available for sale and may have a different environmental profile. Glulam production assumes average kiln-dried dressed wood as an input, rather than the specific supply chain of each manufacturer. Glulam can use lower-quality, shorter-dimension wood as in

LCI and EPDs for Australian Wood Products 150 of 159

input; however, this was not considered in this study. Including this may result in a more favourable environmental profile for glulam. The assessment of toxicity indicators was outside the scope of this study. While many modern treatments are considered to be of low health risk when following recommended safety precautions (TPAA, 2017b), (CSIRO, 2011), preservatives may have toxicity impacts for certain treatment types (e.g. CCA). Water consumption does not account for relative water stress with respect to geographical location, meaning that it provides no information about the potential impacts of any water consumption that does occur. Further study is required to quantify the relative significance of water consumption for supply chains located within different geographical regions of Australia, which vary greatly in climate and annual precipitation (ABS, 2007).

6.4. Data quality assessment

Inventory data quality is judged by its precision (measured, calculated, literature or estimated), completeness (e.g., unreported emissions), consistency (degree of uniformity of the methodology applied) and representativeness (geographical, temporal, and technological). To cover these requirements and to ensure reliable results, first-hand industry data in combination with consistent background LCA information from the GaBi 2017 database were used. The LCI datasets from the GaBi 2017 database are widely distributed and used with the GaBi 8 Software. The datasets have been used in LCA models worldwide in industrial and scientific applications in internal as well as in many critically reviewed and published studies. In the process of providing these datasets they are cross-checked with other databases and values from industry and science.

6.4.1. Precision and completeness

Precision: As the majority of the relevant foreground data are measured data or calculated based on primary information sources of the owner of the technology, precision is considered to be high. Seasonal variations and variations across different manufacturers were balanced out by using yearly data and weighted averages. All background data are sourced from GaBi databases with the documented precision. Completeness: Each foreground process was checked for mass balance and completeness of the emission inventory. No data were knowingly omitted. Completeness of foreground unit process data is considered to be high. All background data are sourced from GaBi databases with the documented completeness.

6.4.2. Representativeness

Technological: All primary and secondary data are modelled to be specific to the technologies or technology mixes under study. Technological representativeness is considered to be good. Geographical: All EPDs represent an Australian average. The foreground data were collected from operations within Australia. For some materials, background data represent European conditions as no matching Australian LCI dataset was available within the GaBi databases. The specific reference location for all important background datasets used in this study can be found in section 4.1.2. Geographical representativeness is considered to be good. Temporal: Primary data were predominantly collected for the Australian financial year 2015-16. All secondary data come from the GaBi 2017 Database and are representative of the years 2013-2016.

LCI and EPDs for Australian Wood Products 151 of 159

Temporal representativeness is considered to be good for the current EPDs whose validity ranges from 2015 to 2020. The specific reference year for all important background datasets used in this study can be found in section 4.1.2.

6.4.3. Completeness

All relevant process steps are considered and modelled to represent each specific situation. The process chain is considered sufficiently complete with regard to the goal and scope of this study.

6.4.4. Consistency

To ensure consistency, all primary data are collected with the same level of detail, while all background data are sourced from the GaBi databases. System boundaries, allocation rules, and impact assessment methods have been applied consistently throughout the study.

6.4.5. Reliability

Primary data were collected using spreadsheets adapted for each product. Cross-checks concerning the plausibility of mass and energy flows were carried out by thinkstep and the external review team. Overall, the data quality can be described as good.

6.5. Conclusions and recommendations

Further improve recovery rates: This study showed that recovery rates improved between 2005/06 and 2015/16, which has helped to reduce environment impacts across the board. However, there remains considerable variation in recovery rates between mills: 49-59% for softwood green mills, 74-82% for softwood dry mills, 29-54% for hardwood green mills, and 66-90% for hardwood dry mills. While recovery rates are affected by factors such as log size and quality, there may be a potential for sawmills with lower recovery rates to improve. Alternative energy sources: Solar kilns were found to be an effective way of reducing the environmental impacts from kiln-drying. However, when considering the entire production phase of dressed hardwood timber, the improvements gained by the use of solar kilns were outweighed by other factors. The geographical location of a site was found to have more influence over the environmental impact of the product due to different electricity grid mixes. Substituting natural gas for biomass in particleboard production has reduced its carbon footprint. The move towards more renewable energy results in reduced dependency on increasingly expensive natural gas (Oakley Greenwood, 2016), but has some consequences for summer smog and, depending on the resin content, also acidification and eutrophication. This substitution may be favourable for Australia’s situation as GWP is an indicator which currently holds high global importance, whereas POCP is highly regional and dependent on the receiving environment. In most Australian towns, cities and rural areas, where most wood processing facilities are located, the amount of ozone in the air does not exceed the national standards. Only larger cities have occasions when there is enough ozone in the air for it to be a risk to human health (Department of the Environment and Heritage, 2005). Further work for hardwood: Given that the spread of results for hardwood products is larger than for softwood due to higher variability in species and forestry practises, it may be beneficial to analyse the results by state or region. While this was not done in this study, this is something that should be considered in future.

LCI and EPDs for Australian Wood Products 152 of 159

Improvements for forestry: For softwood, slash burning should be replaced with mechanical crushing where feasible, leading to reduced air emissions and retaining more sequestered carbon in the soil. For native hardwood forestry, the amount of residue burned through regeneration burns should be limited, however, it is recognised that burns are required as part of the natural regeneration process for many native hardwood species. Continue to improve resin loadings: Resin loadings and the type of resin used in engineered wood products was found to be significant for all the assessed indicators. It is recommended that a reduction in resin use and selection of resins based on their environmental credentials is undertaken where possible. The environmental profile for glulam, due to low resin loadings (<1% by mass), is similar to dressed kiln-dried wood. However, the use of resin allows for products with greater dimensions and more regular structural properties to be created from lower grade timber. Cover uncovered dams: Nine facilities included within the study used uncovered dams on site. Evaporation due to the use of dams, relative to evaporation which would otherwise occur from soil or pasture, was calculated to be approximately 5% of the total water use for each of the product categories with dam users. Assuming the sample is representative of the industry, water loss due to the use of uncovered dams for the product categories included in this study is estimated to be over 120 ML/a, enough to fill 50 Olympic-size swimming pools per year. It is therefore recommended that measures are taken such as dam covers to reduce water lost due to evaporation, especially in water stressed regions of Australia.

LCI and EPDs for Australian Wood Products 153 of 159

References

ABARES. (2013). Australia's State of the Forests Report 2013. Canberra: Australian Bureau of Agricultural and Resource Economics and Sciences. Retrieved from http://www.agriculture.gov.au/abares/forestsaustralia/Pages/SOFR/sofr-2013.aspx ABARES. (2017). Australian forest and wood products statistics: March and June quarters 2017. Australian Bureau of Agricultural and Resource Economics and Sciences. Retrieved November 2017, from http://www.agriculture.gov.au/abares/publications/display?url=http://143.188.17.20/anrdl/DA FFService/display.php?fid=pb_afwpsd9abfe20171107.xml ABS. (2007). FEATURE ARTICLE: WATER IN AUSTRALIA. Retrieved from Australian Bureau of Statistics (ABS): http://www.abs.gov.au/AUSSTATS/[email protected]/7d12b0f6763c78caca257061001cc588/330bc 8fdfd50bee4ca2573c6001049f9%21OpenDocument ABS. (2017). Average monthly rainfall abd temperatures, Capital cities and Alice Springs. Retrieved August 21, 2017, from Australian Bureau of Statistics: http://www.abs.gov.au/ausstats/[email protected]/Lookup/by%20Subject/1301.0~2012~Main%20F eatures~Australia%27s%20climate~143 AS 1604.1-2010. (2010). Specification for preservative treatment - Part 1: Sawn and round timber. Sydney, Australia: Australian Standard. AS 2082:2007. (2007). Timber - Hardwood - Visually stress-graded for structural purposes. Australian Standard. AS 2796.1:1999. (2016). Timber - Hardwood - Sawn and milled products. Part 1: Product specifications. Australian Standard. AS 2796.2:2006. (2006). Timber - Hardwood - Sawn and milled products. Part 2: Grade descriptioins. Australian Standard. AS 2858-2008. (2008). Timber - Softwood - Visually stress-graded for structural purposes. Australian Standard. AS 4785.1-2002. (2002). Timber-Softwood-Sawn and milled products. Part 1: Product specification. Australian Standard. AS 4785.2-2002. (2002). Timber-Softwood-Sawn and milled products. Part 2: Grade description. Australian Standard. AS/NZS 1604.2:2012. (2012). Specification for preservative treatment - Part 2: Reconstituted wood- based products. Australian/New Zealand Standard. AS/NZS 1748.1:2011. (2011). Timber – Solid – Stress-graded for structural purposes. Part: General requirements. Australian/New Zealand Standard. AS/NZS 1748.2:2011. (2011). Timber - Solid - Stress-graded for structural purposes. Part 2: Qualification of grading methods. Australian/New Zealand Standard. AS/NZS 1859.1:2004. (2004). Reconstituted wood-based panels - Specifications - Particleboard. Australian/New Zealand Standard.

LCI and EPDs for Australian Wood Products 154 of 159

AS/NZS 1859.2:2004. (2004). Reconstituted wood-based panels - Specifications. Part 2: Dry- processed fibreboard . Ausralian/New Zealand Standard. AS/NZS 1859.3:2005. (2005). Reconstituted wood-based panels - Specifications - Decorative and overlaid wood panels. Australian/New Zealand Standard. AS/NZS 1859.3:2005. (2005). Reconstituted wood-based panels - Specifications. Part 3: Decorative overlaid wood panels. Australian/New Zealand Standard. AS/NZS 1860.1:2002. (2002). Particleboard flooring - Specifications. Australian/New Zealand Standard. AS/NZS 2269:2004. (2006). Plywood - Structural (with 2006 Amendment). Australian/New Zealand Standard. AS/NZS 2270:2006. (2006). Plywood and Blockboard for Interior Use. Australian/New Zealand Standard. AS/NZS 2271:2004. (2006). Plywood and Blockboard for Exterior Use (with 2006 Amendment). Australian/New Zealand Standard. Australian Government. (2014a). National Inventory Report 2012, Volume 3: Australia data for waste, recalculations, improvements and Kyoto Protocol accounting requirements. Australian National Greenhouse Accounts. Retrieved August 07, 2017, from http://www.environment.gov.au/climate-change/greenhouse-gas- measurement/publications/national-inventory-report-2012 Australian Government. (2014b). National Greenhouse Accounts Factors. Department of Environment. Retrieved August 07, 2017, from http://www.environment.gov.au/system/files/resources/b24f8db4-e55a-4deb-a0b3- 32cf763a5dab/files/national-greenhouse-accounts-factors-dec-2014.pdf Australian Government. (2016a). National Greenhouse Accounts Factors. Department of Environment. Retrieved August 21, 2017, from http://www.environment.gov.au/system/files/resources/e30b1895-4870-4a1f-9b32- 3a590de3dddf/files/national-greenhouse-accounts-factors-august-2016.pdf Australian Government. (2016b). Technical Guidelines for the Estimation of Greenhouse Gas Emissions by Facilities in Australia. Department of Environment and Energy. Retrieved August 07, 2017, from http://www.environment.gov.au/system/files/resources/95cf8d59- dcf8-40c7-8bfc-5ceacc221700/files/nger-technical-guidelines-2016-17.pdf Australian Government. (2017b). Public Chemical Registration Information System Search. Retrieved July 2017, from https://portal.apvma.gov.au/pubcris Barlaz, M. A. (1998). Carbon storage during biodegradation of municipal solid waste components in laboratory‐scale landfills. Global Biogeochemical Cycles, 12 (2), 373–380.

Barlaz, M. A. (2005). Note to ICF consulting, dated June 29, 2005. Barlaz, M. A. (2008). Corrections to Previously Published Carbon Storage Factors, Note to Parties Interested in Carbon Sequestration from Municipal Solid Waste, dated February 28, 2008. Benyon, R., Carlyle, J., & Doody, T. (2006). Impacts of tree plantations on groundwater in south- eastern Australia. Australian Journal of Botany, 54: 181–192. Bergman, R. D., & Bowe, S. A. (2007). Life-cycle inventory of hardwood lumber manufacturing in the Northeastern and North Central United States. CORRIM Phase II Final Report. Seattle, WA: University of Washington. Retrieved https://www.fs.usda.gov/treesearch/pubs/52610

LCI and EPDs for Australian Wood Products 155 of 159

BoM. (2017). Climate Data Online. Retrieved from Australian Government Bureau of Meteorology: http://www.bom.gov.au/climate/data/index.shtml Bowers, T., Puettmann, M., Ganguly, I., & Eastin, I. (2017). CORRIM: Report – Module G2 Cradle- to-Gate Life-Cycle assessment of Glued Laminated Timbers US Southeast (Review Draft). Corrim. Bradstock, R. A., Boer, M. M., Cary, G. J., Price, O. F., Williams, R. J., Barrett, D., . . . Russell- Smith, J. (2012). Modelling the potential for prescribed burning to mitigate carbon emissions from wildfires in fire-prone forests of Australia. International Journal of Wildland Fire, 21, 629–639. Briggs, D. (1994). Forest products measurements and conversion factors: with special emphasis on the U.S. Pacific Northwest. Seattle, WA, USA: College of Forest Resources, University of Washington. Brown, A., Zhang, L., McMahon, T., & Western, A. (2005). A review of paired catchment studies for determining changes in water yield resulting from alterations in vegetation. Journal of Hydrology(310), 28–61. Bubb, K., & Croton, J. (2002). Effects on catchment water balance from the management of Pinus plantations on the coastal lowlands of south-east Queensland, Australia. Hydrological processes, 16(1), 105-117. Carre, A. (2011). A Comparative Life Cycle Assessment of Alternative Constructions of a Typical Australian House Design. Melbourne: Forest & Wood Products Australia. Retrieved July 2017, 13, from http://www.fwpa.com.au/images/marketaccess/PNA147- 0809_Research_Report_Compar Commonwealth of Australia. (2016). National Inventory Report 2014 (revised) Volume 2. Cornish, P., & Vertessy, R. (2001). Forest age induced changes in evapotranspiration and water yield in a eucalypt forest. Journal of Hydrology, 242, 43-63. CSIRO. (2009a). Life Cycle Inventory of Australian Forestry and Wood Products. Final report. (Tucker et al., Ed.) Commonwealth Scientific and Industrial Research Organisation CSIRO. CSIRO. (2009b). Life Cycle Inventory of Australian Forestry and Wood Products. Additional Modules B-E. Commonwealth Scientific and Industrial Research Organisation CSIRO. CSIRO. (2011). The facts about CCA-treated timber. CSIRO. Department of the Environment and Heritage. (2005). Air quality fact sheet. Retrieved from Department of the Environment and Energy: http://www.environment.gov.au/protection/publications/factsheet-ground-level-ozone-o3 ECN. (2017). Phyllis2, database for biomass and waste. Energy research Centre of the Netherlands (ECN). Retrieved July 2017, from https://www.ecn.nl/phyllis2 EN 15804:2012+A1:2013. (2013). Sustainability of construction works — Environmental product declarations — Core rules for the product category of construction products. Brussels: European Committee for Standardization. EN 16449:2014. (2014). Wood and wood-based products – Calculation of the biogenic carbon content of wood and conversion to carbon dioxide. Brussels: European Committee for Standardization. EN 16485:2014. (2014). Round and sawn timber. Environmental Product Declarations. Product category rules for wood and wood-based products for use in construction. Brussels: European Committee for Standardization.

LCI and EPDs for Australian Wood Products 156 of 159

Energy Information Administration. (2002). Emissions of Greenhouse Gases in the United States 2001. EWPAA. (2012). Facts About Formaldehyde Emissions from EWPAA Certified Products. Engineered Wood Product Association of Australia. Retrieved August 07, 2017, from http://www.ewp.asn.au/library/downloads/ewpaa_formaldehyde_facts.pdf Filho, W. L., Mannke, F., Mohee, R., Schulte, V., & Surroop, D. (2013). Climate-Smart Technologies: Integrating Renewable Energy and Energy Efficiency in Mitigation and Adaptation Responses. Springer Science & Business Media. Forest Products Laboratory. (2017). Wood Handbook: Wood as an Engineering Material. General Technical Report FPL–GTR–190. United States Department of Agriculture Forest Service. Retrieved June 14, 2017, from http://www.fpl.fs.fed.us/documnts/fplgtr/fpl_gtr190.pdf Forsythe Consultants. (2007). Mapping the Timber Waste Stream from Building Demolition. Retrieved August 07, 2017, from http://www.timberstewardship.org.au/images/pdf_documents/demolition_waste_barriers_fin al.pdf Gavran, M., Burns, K., Hug, B., Frakes, I., & Gupta, M. (2014). ABARES National Wood Processing Survey 2012–13. Research by the Australian Bureau of Agricultural and Resource Economics and Sciences. Research report 14.6. Gifford, R. (2000). Carbon Contents of Above-Ground Tissues of Forest and Woodland Trees. Technical report No. 22. National Carbon Accounting Scheme. Australian Greenhouse Office. Retrieved August 07, 2017, from http://pandora.nla.gov.au/pan/23322/20020220- 0000/www.greenhouse.gov.au/ncas/files/pdfs/tr22final.pdf GLTAA. (2017). Historical Review. Retrieved from Glued Laminated Timber Association of Australia: http://www.gltaa.com.au/historical-review/ Groover, M. P. (2010). Fundamentals of modern manufacturing: materials, processes and systems. Guendehou, S. (2010). Australia’s National Greenhouse Gas Inventory 2008: Solid Waste Review – Fraction of DOC dissimilated DOCf. report to the Department of Climate Change. Guinée, J. B., Gorrée, M., Heijungs, R., Huppes, G., Kleijn, R., de Koning, A., . . . Huijbregts, M. (2002). Handbook on life cycle assessment. Operational guide to the ISO standards. Dordrecht: Kluwer. Houghton, J. (2017). FWPA ABARES Softwood Market Coverage Discussion Paper. Hyder Consulting. (2007). Review of Methane Recovery and Flaring from Landfills. Australian Greenhouse Office, Department of Environment and Water Resources. IEPDS. (2017). PCR 2012:01 Construction products and Construction services, Version 2.2. The International EPD® System. IPCC. (2006). IPCC Guidelines for National Greenhouse Gas Inventories - Volume 4 - Agriculture, Forestry and Other Land Use. Geneva, Switzerland. IPCC. (2007). IPCC Fourth Assessment Report: Climate Change 2007. Retrieved from Intergovernmental Panel on Climate Change: https://www.ipcc.ch/report/ar4/ IPCC. (2016). IPCC Guidelines for National Greenhouse Gas Inventories - Volume 2 - Energy. Geneva, Switzerland.

LCI and EPDs for Australian Wood Products 157 of 159

ISO 14025:2006. (2006). Environmental labels and declarations — Type III environmental declarations — Principles and procedures. Geneva: International Organization for Standardization. ISO 14040:2006. (2006). Environmental management – Life cycle assessment – Principles and framework. Geneva: International Organization for Standardization. ISO/TS 14067:2013. (2013). Greenhouse gases – Carbon footprint of products – Requirements and guidelines for quantification and communication. Geneva: International Organization for Standardization. JRC. (2010). ILCD Handbook: General guide for Life Cycle Assessment – Detailed guidance. EUR 24708 EN (1st ed.). Luxembourg: Joint Research Centre. Kapambwe, M., Ximenes, F., Vinden, P., & Keenan, R. (2008). Dynamics of Carbon Stocks in Timber in Australian Residential Housing. Forest and Wood Products Australia. KPMG. (2016). Australian Pine Log Price Index (Stumpage), Updated to December 2016. Australia: KPMG. Lee-Archer, P. (2007). Water Monitoring in Tasmania. Retrieved June 14, 2017, from http://www.apvma.gov.au/community/downloads/WaterMonitoringTas.pdf Manzardo, A., Mazzi, A., Loss, A., Butler, M., Williamson, A., & Scipioni, A. (2016). Lessons learned from the application of different water footprint approaches to compare different food packaging alternatives. Journal of Cleaner Production, 112, pp. 4657-4666. Moran, R., & O'Shaughnessy, P. (1984). Determination of the evapotranspiration of E. regnans forested catchments using hydrological measurements. Agricultural Water Management, 8, 57-76. NCEA. (2017). Monthly evaporation calculator. Retrieved from Readyreckoner: http://readyreckoner.nceaprd.usq.edu.au/evaporationcalc.aspx Oakley Greenwood. (2016). Gas Price Trends Review. Canberra ACT 2601 Australia: Department of Industry, Innovation and Science. Puettmann, M., & Oneil, E. (2012). Cradle to Gate Life Cycle Assessment of Softwood Lumber Production from the Inland Northwest. Seattle, WA: CORRIM. Puettmann, M., Oneil, E., & Bergman, R. (2012). Cradle to Gate Life Cycle Assessment of Softwood Lumber Production from the Northeast-North Central. Seattle, WA: CORRIM. Quinteiro, P., Dias, A. C., Silva, M., Ridoutt, B. G., & Arroja, L. (2015). A contribution to the environmental impact assessment of green water flows. Journal of Cleaner Production, 93, 318-329. Rösemann et al. (2011). Berechnung von gas- und partikelförmigen Emissionen aus der deutschen Landwirtschaft 1990-2009. VTI-Sonderheft 342. thinkstep. (2017). GaBi Life Cycle Inventory Database 2017 Documentation. Leinfelden- Echterdingen, Germany. Retrieved from thinkstep AG: http://www.gabi- software.com/support/gabi/gabi-database-2017-lci-documentation/ Timber Queensland. (2014). Moisture in Timber. Technical Data Sheet 28. Brisbane: Timber Queensland. TPAA. (2017a). Timber treatment. Retrieved August 07, 2017, from Timber Preservers Association Australia: http://www.tpaa.com.au/timber-treatment/

LCI and EPDs for Australian Wood Products 158 of 159

TPAA. (2017b). FAQ. Retrieved from Timber Preservers Association of Australia: http://www.tpaa.com.au/faq/ Turner, P., Ximenes, F., Penman, T., Law, B., Waters, C., Mo, M., & Brock, P. (2014). Accounting for biodiversity in Life Cycle Impact Assessments of forestry and agricultural systems – the BioImpact metric. Research report for Forest & Wood Products Australia Limited. van Loo, S., & Koppejan, J. (2002). Handbook of Biomass Combustion and Co-Firing. Twente University Press. van Oers, L., de Koning, A., Guinée, J. B., & Huppes, G. (2002). Abiotic resource depletion in LCA. The Hague: Ministry of Transport, Public Works and Water Management. Vertessy, R., Watson, F., & O'Sullivan, S. (2001). Factors determining relations between stand age and catchment water balance in mountain ash forests. Forest Ecology and Management, 143, 13-26. Walters, R. (n.d.). Heats of combustion of high temperature polymers. Federal Aviation Administration. Wang, W., Padgett, J., De La Cruz, F., & Barlaz, M. (2011). Wood biodegradation in laboratory- scale landfills. Environmental Science & Technology, 45(16), pp. 6864-6871. Ximenes, F., Brooks, P., Wilson, C., & Giles, D. (2013). Carbon Storage in Engineered Wood Products in Landfills. Forest and Wood Products Australia. Retrieved August 07, 2017, from http://www.fwpa.com.au/rd-and-e/processing/181-carbon-storage-in-engineered-wood- products-in-landfills.html Zhang, L., Dawes, W. R., & Walker, G. R. (2001). Response of mean annual evapotranspiration to vegetation changes at catchment scale. Water Resources Research 37, 701–708.

LCI and EPDs for Australian Wood Products 159 of 159