Appendix E

ORIENTED STRANDBOARD - SOUTHEAST

July 17, 2002

Prepared by:

D. Earl Kline¹
Department of Wood Science and Forest Products
Virginia Tech University

¹ Kline is associate professor , Department of Wood Science and Forest Products, Virginia Technical and State University, Blacksburg, VA 24061-0503

EXECUTIVE SUMMARY

The objective of this study is to develop a life cycle inventory (LCI) for the production of Oriented Strand Board (OSB) as manufactured in the Southeastern U.S. (Alabama, Arkansas, Florida, Georgia, Kentucky, Louisiana, Mississippi, North Carolina, South Carolina, Tennessee, Texas, and Virginia). OSB is considered a structural panel product, and is used for roof, wall and floor sheathing and sub-flooring in residential and commercial construction. Roundwood used for flakes to make OSB in the Southeast include 73% softwoods (southern pine) and 27% hardwoods (yellow poplar and a mix of other hardwood species). OSB plants were surveyed in this region to record the quantity of all inputs and outputs associated with the production process. Input data collected included transportation distances and the use of wood, bark and other wood co-products, electricity, fuel and resin. Output include statistics regarding quantities of the OSB product, co-products of screening fines, bark, trimmings, and sawdust, as well as land, water and air emissions.

Four OSB manufacturing plants were surveyed in the Southeast. The Year 2000 production for these plants was approximately 18% of the total production for the survey region. The surveyed plants produced 1.41 billion square feet (MMMSF) 3/8-in basis.

Southeast OSB manufacturing had a wood recovery of 78.6 % as determined by the output of wood in the OSB product as a percentage of the wood input to the plant. It took 49.8 ft.³ (1,540 lb oven-dry wood excluding bark) of wood in the form of logs to produce one Msf 3/8-inch of OSB (1,210 lb oven-dry wood). The amount of bark generated during the debarking process was 10.8% based on the weight of bark to weight of wood.

A complete listing of emissions, as determined from the OSB survey are presented in this report. Emissions were allocated to products and co-products (i.e. burdens were assigned) based on a mass basis. Emissions reported for OSB had a burden 89.3% and the remaining 10.7% are assigned to the co-products of bark and sawdust. Bark and other wood waste used for hogged fuel within the plant were not allocated any emissions. CO₂, a greenhouse gas of international interest, is generated by combustion of fuels. A major portion of the heat generation for the production of OSB was based upon wood fuel; however, an analysis of CO₂ emissions have not yet been completed as of this writing.

The quality of the OSB LCI data is considered accurate and representative. A comparison of data for the four plants surveyed, which indicated a high level of consistency, provided one measure of validity. Additional data analysis (i.e., mass and energy balances) further supported the integrity of our findings.

Proposed research in the coming year includes expanding the OSB LCI model into a sub-unit process approach. This model will be useful in analyzing ways to improve production efficiency subject to environmental impacts. Also, the development of an LCI model for Parallel Strand Lumber will be completed. These models will be used in conjunction with other LCI models developed in the CORRIM project. A number of scenarios will be studied to provide insights as to how to optimize building materials design to minimize environmental impact. Output of the analyses will be compared to benchmark data to assess improvements.

TABLE OF CONTENTS

EXECUTIVE SUMMARY
1.0 INTRODUCTION
     1.1 OVERVIEW
     1.2 MILL SURVEYS
     1.3 DESCRIPTION OF OSB UNIT PROCESSES
     1.4 MATERIAL FLOWS
     1.5 TRANSPORTATION
     1.6 ASSUMPTIONS
2.0 RAW MATERIAL INPUTS AND PRODUCT YIELDS
3.0 MANUFACTURING ENERGY SUMMARY
     3.1 ELECTRICITY USE SUMMARY
     3.2 WOOD FUEL AND FOSSIL FUEL UTILIZATION FOR HEAT GENERATION
     3.3 ADHESIVE USE AND ENERGY/ELECTRICITY TO PRODUCE RESINS
4.0 POLLUTION CONTROL EMISSIONS TO AIR
5.0 UNIT FACTOR ESTIMATES FOR MAIN PRODUCTS
6.0 PROPOSED RESEARCH FOR COMING YEAR
7.0 REFERENCES

LIST OF FIGURES

Figure 1.1. Description of the OSB manufacturing processes.

LIST OF TABLES

1.0 INTRODUCTION

1.1 OVERVIEW

Oriented Strand Board (OSB) evolved from waferboard in the late 1970s. OSB is manufactured by processing a log into strands of predetermined length and width. These strands are oriented, not randomly placed, to create a final panel product that can be used for structural applications. These applications include wall sheathing, roof sheathing, sub flooring, underlayment, structural insulated panels, I-joists, and rim boards. Today, all building codes in the U.S. and Canada recognize OSB panels as equivalent to for the same uses as plywood on a thickness-by-thickness basis. Although OSB comes in a variety of grades and thicknesses, its production is described in this report in terms of 1.0 thousand square feet (MSF) of 3/8-inch thickness. Panels are normally produced in 4- x 8-foot sheets.

OSB for the Southeast region is made from both softwood and hardwood species depending on the surrounding forest resources. Softwood species come from a group of wood species referred as southern pine, the dominant species in this group are slash and loblolly. Hardwood species are also used. Yellow poplar is by far the most commonly used hardwood, but the raw material mix can also include a range of other species such as oak and soft maple.

This report focuses on OSB production practices in the southeast region of the U.S. which includes Alabama, Arkansas, Florida, Georgia, Kentucky, Louisiana, Mississippi, North Carolina, South Carolina, Tennessee, Texas, and Virginia. The size of production facilities in the region range from about 340,000 to 380,000 Msf 3/8-inch basis annually.

1.2 MILL SURVEYS

In this study data was collected from representative manufacturing operations producing1.41 million Msf 3/8-inch basis. This represents about 18% of the production in the Southeast and 7.8% of total U.S. production.

To conduct the survey of OSB manufacturers, all OSB manufacturing plants in the Southeast Region (22 plants) were sent a LCI survey. Of these, four plants (18%) responded with complete data in terms of OSB and co-product production, raw materials, electricity and fuel use, and emissions.

This report documents the results of a life cycle inventory (LCI) of OSB manufactured in the Southeastern U.S.Reported herein are those impacts associated solely with the manufacture of OSB, including transportation of logs to the manufacturing site, production of phenol-formaldehyde resin, MDI, wax, electricity, natural gas, OSB and its co-products. Primary data was collected using a direct survey of OSB manufacturers as indicated previously. Supplemental secondary data was obtained for impacts associated with the manufacture and delivery of electricity and all fuels (Franklin and Associates, 1991; PRe´ Consultants, 2001; USDOE, 2000), phenol-formaldehyde resin (Athena, 1993), and MDI and slack wax.

1.3 DESCRIPTION OF OSB UNIT PROCESSES

The overall OSB manufacturing system consists of 10 primary sub-unit processes. The interrelationships between these processes are shown in Figure 1.1. The following describes these processes and discusses significant inputs and outputs.

Figure 1.1. Description of the OSB manufacturing processes.
System boundary for OSB processes is represented by dashed line.

1. Log Handling. The log handling process including sorting, storage, bucking, and debarking. The primary input to this process is roundwood from the forest. Other inputs include diesel fuel for transport vehicles and electricity for powering material handling systems and the debarker. The output from this process is debarked logs or log bolts that proceed to the next process. Bark and other wood residues are also an output of this process and are later used as fuel for drying and heating press oil. Sometimes this wood residue can also be sold as a co-product. A potential source of emissions to the environment is water runoff from the log yard.
   
   
2. Flaking. The flaking process strives to achieve a uniform thickness flakes up to 6-inches long and about 1-inch wide. The material inputs to the system are debarked logs. Energy is consumed in the form of electricity used to power the flakers. The material output is green flakes that are stored in wet bins until needed.
   
   
3. Drying. Green flakes with a moisture content (oven-dry basis) of about 100% are dried until the appropriate moisture content is reached, usually between 4 to 8 %. Removing this amount of moisture requires a large amount of heat. The primary heat source for this drying process is wood fuel, which consists of bark, screening fines from the screening process, OSB trimmings from the finishing process, and other wood residues. Heat from wood fuel is generated typically by direct-fired heating methods. Backup heating from fuels such as natural gas, propane, or fuel oil is often used when wood fuel is unavailable for drying. Since the drying process requires a large heating energy input, it contributes significantly to airborne emissions such as greenhouse gases and airborne particles.
   
   
4. Screening. This process screens out fine wood materials from dried flakes that are considered too small for OSB mat formation. This process uses electricity to power the process. The output of this process is flakes of appropriate size for subsequent manufacturing. The screening fines are used as fuel for drying and heating press oil or are sold as a co-product.
   
   
5. Blending. This process blends strands with resin binders and a small amount of wax, which improves the efficiency of the resin binder and enhances the panel's resistance to moisture and water absorption. Material inputs in this process include phenol-formaldehyde (PF) and/or methyldiisocyanate (MDI) resin, and wax. Electricity powers the process.
   
   
6. Mat Formation. Strands go through the forming line where cross-directional layers are formed in the OSB mat. Electricity is used to power the mat forming line.
   
   
7. Pressing. The OSB mat is pressed under intense heat and pressure to form a rigid, dense structural panel of oriented strand board. Presses are typically multiple opening allowing 8 to 16 master panels to be pressed in one operation. Continuous pressing technology is also available to form a continuous ribbon of OSB. A significant source of energy is required to supply heat for the press. As with the drying process, most of this heat is generated using wood fuel. Often, the same wood-fueled heating system provides heat for both the dryer and heated thermal oil for the presses. The emissions from the press include VOC and formaldehyde emissions from the heated OSB.
   
   
8. Finishing. In this process, panels are cooled, cut to size, grade stamped, stacked in bundles and packaged for shipping. Energy inputs include electricity for powering material handling and processing systems. Also, fuel for powering forklifts is consumed in this process. Waste material output is generated in this process as OSB trimmings, sawdust, and rejected boards; this material is used as a heating fuel or sold as a co-product.
   
   
9. Heat Generation. In this process, wood fuel or fossil fuel input is converted to heat to supply the drying and mat pressing heat demand. The output of this process includes air emissions from combustion and solid waste (ash). This process could also include co-generation of electricity, however, none of the plants responding to the survey reported any electricity co-generation capability.
   
   
10. Air Pollution Control. In OSB manufacturing operations, significant processing is devoted to air pollution control including Regenerative Thermal Oxidizer (RTO), Electrostatic precipitator (ESP), Regenerative Catalytic Oxidizer (RCO), and other such technologies. While these processes minimize levels of air emissions, they require significant inputs such as electricity and natural gas.

1.4 MATERIAL FLOWS

Those materials considered in the LCI analysis included those listed in Table 1.1. Input materials considered were logs (includes wood and bark), phenol-formaldehyde (PF) resin, polymeric methyldiisocyanate (MDI) resin, and wax. Outputs were OSB and the co-products bark and sawdust. All bark and wood fuel were considered green, at a moisture content of 50% wet-basis. To derive the wood and bark weights and to determine how much water was removed from the wood and bark, the following assumptions were made: bark was at 50% moisture content (MC) on a wet-basis, wood at an average of 100% MC oven-dry basis, and dry flakes and wood waste at 6% MC on an oven-dry basis.

Table 1.1 Listing of Input Materials and Co-products Associated with Production of OSB.

1.5 TRANSPORTATION

Delivery of the input materials was by truck, with an average one-way delivery distance of 89 miles. Delivery distances for PF and MDI resin and for wax were not determined in this study.

The survey data provided wood input data in either cords or green weight. Data given in cords was converted to cubic feet (ft³) of wood by multiplying by 75 ft³/cord (Toennisson and Hadden, 1993). A final conversion was then made from ft³ to mass (lb) by multiplying by the average weighted densities as determined by their percentage use as given by the survey, and the densities for hardwood or softwood species as provided by Toennisson and Hadden (1993). The average wood density used was 30.9 lb/ft³ oven-dry for the mix of softwood and hardwood logs (see Table 1.2). Conversion of green log weight to dry wood weight was carried out using the moisture content assumptions specified in the previous section.

Table 1.2 Average Density of Wood Species Used to Calculate Mass of Wood from Logs.

¹ Wood density values are from Toennisson and Hadden (1993).

1.6 ASSUMPTIONS

• All data from the survey was weight averaged for the four plants based on their production in comparison to the total production for the year.
  
  
• Density for the wood species used to make the OSB were obtained from Toennisson and Hadden (1993), and based on their weighted percentage of use as reported by manufacturers; the weighted average density was calculated to be 30.9 lb/ft³ oven-dry (see Table 1.2).
  
  
• Log inputs that were provided in cords were converted to ft3 by multiplying the factor 75 ft³/cord solid wood (excluding bark).
  
  
• All conversion units for forestry and forest products type conversions were taken from the Wood Products Engineers Handbook by Toennisson and Hadden (1993).
  
  
• An unaccounted wood mass of 2.1% was established by the difference between reported input and output wood material flows (see Table 1.4 for material balance analysis).
  
  
• SimaPro5, a software package designed for analyzing the environmental impact of products during their whole life cycle, was used to input the OSB life cycle inventory (LCI). Developed in The Netherlands by PRé Consultants B.V., SimaPro5 contains a U.S. database for a number of materials, including paper products, fuels, and chemicals. The U.S. database is provided by Franklin Associates (FAL).

2.0 RAW MATERIAL INPUTS AND PRODUCT YIELDS

The input to produce 1.0 thousand square feet (Msf) of OSB 3/8-inch basis includes 49.8 cubic feet (ft³) or 1,540 lb of logs (based on volume and wood densities given in Table 1.2), 42.4 lbs of PF resin, 8.16 lbs MDI resin, and 19.3 lbs of wax . Table 2.1 summarizes all inputs required to produce 1.0 Msf of OSB including electricity and fuel inputs.

Table 2.1 Inputs to Produce 1.0 Msf 3/8-inch Basis of OSB in the Southeast.

¹ All materials unless noted, are given as weight oven-dry or as solids content.

² Includes volume of only wood (bark excluded)

³ This material includes bark, screening fines, and OSB trimmings

A complete wood mass balance for woody raw materials used in OSB production is given in Table 2.2. The difference between the total wood input and output was 43 lb., and it is this quantity that is referred to as"unaccounted for wood". The unaccounted for wood is 2.4% of the total wood input. Only one plant was able to report the exact amount of bark generated separate from other wood residues. Based on this report, the percentage of bark generated on an oven dry weight basis was 10.8 % of the roundwood input at the plant. This percentage was used to estimate the bark input of in the mass balance. Most likely the unaccounted for wood of 43 lb. is due to inaccurate/incomplete reporting of bark from the debarking process or from the wood per cord conversion factor used.

Table 2.2 Wood Mass Balance for OSB Production from the Southeast Region per 1.0 Msf 3/8-inch Basis

¹ Includes weight of both bark and solid wood from Table 2.1.

² OSB (wood only) based on estimated weight of OSB, 1260 lb, minus weight of resin and wax (50 pounds or 4% by weight basis).

³2.4% unaccounted for wood

3.0 MANUFACTURING ENERGY SUMMARY

Energy for the production of OSB comes from electricity, diesel, liquid propane gas (LPG), natural gas, and wood fuel from bark, fines and other wood residue. The electricity is used to operate all the systems described in Section 1.3. Diesel fuel use is assumed to be by log loaders in the "log handling" process. Forklift trucks used small amounts of LPG primarily in the "finishing" process.

3.1 ELECTRICITY USE SUMMARY

The source of fuel used to generate the electricity used in the manufacturing process is very important in determining the type and amount of impact in the LCI analysis. The breakdown of electricity use in the Southeast by fuel source is given in Table 3.1. The source of this data is the U.S. Department of Energy (DOE). In 1998, the dominant form of fuel in the region was coal, representing 49.2% of the total, followed by nuclear at 25.6%, natural gas at 9.6%, petroleum at 3.7% and hydro at 3.4%. In the SimaPro 5 analysis using the FAL database, combusting of coal contributes significant impact values, as does nuclear and petroleum, whereas natural gas contributes relatively less.

Table 3.1. `Electric Power Industry Generation of Electricity by Primary Energy Sources and State for the Southeast Region as Defined by the U.S. Department of Energy (2000).

¹ Source: Energy Information Administration/State Electric Profiles 2000, Department of Energy.

http://www.eia.doe.gov/cneaf/electricity/st_profiles/toc.html

² Abbreviations of Southeastern States

3.2 WOOD FUEL AND FOSSIL FUEL UTILIZATION FOR HEAT GENERATION

Most of the bark generated during debarking, OSB fines from the screening process, and other waste sources in the plants were combined to use as wood fuel in a direct-fired fuel cell. Wood fuel weight, following industry practice, was given as green weight and assumed to be 50% moisture content on a wet-weight basis. As such, the total wood fuel burned (774 lb.) at 50% moisture content on a wet basis, is equivalent to 387 lb. of oven-dry weight wood fuel. A very small amount of OSB wood waste (trimmings and sawdust) was burned in the boiler. Wood fuel was by far the dominant fuel source at 79.0% of the total energy for heat. In addition to wood fuel for heat generation, natural gas and some fuel oil were also used, representing 18.4% and 2.6%, respectively, of the total heat generation. Table 3.2 provides a breakdown of heat energy use for the boilers by fuel source.

Table 3.2. Southeast Weighted Data Conversion of Boiler Inputs into Heat Energy for 1.0 Msf 3/8-inch Basis of OSB

^1/ Weight dry wood fuel oven dry basis (includes bark, screening fines, and wood waste)

^2/ Weight of dry wood multiplied by 8500 BTU/lb

^3/ Volume of natural gas multiplied by 1100 Btu/ft<sup>3

4/</sup> Volume of fuel oil multiplied by 140,000 BTU/gal

3.3 ADHESIVE USE AND ENERGY/ELECTRICITY TO PRODUCE RESINS

Phenol-formaldehyde (phenolic) resin and/or MDI are the adhesives used in OSB production. The manufacture of these resins is particularly energy intensive. The total energy requirement for the production 42.4 lb of phenolic needed for Msf 3/8-inch basis OSB from the Southeast is 6.99E+05 BTUs. Electricity requirements for phenol-formaldehyde production per MSF 3/8-inch basis are 27.4 kWh in addition to that required in the OSB manufacturing process. The phenol-formaldehyde resin used is comprised of 65% formaldehyde and 35% phenol by weight. All the material, fuel, and electricity used to produce phenol-formaldehyde resin are listed in Table 3.3.

The total energy requirement for the production 8.16 lb of MDI needed for Msf 3/8-inchbasis OSB from the Southeast is 3.18E+05 BTU's. Electricity requirements for MDI production per Msf 3/8-inch basis are 6.16 kWh in addition to that required in the OSB manufacturing process. MDI resin used is comprised of many different materials. Table 3.4 lists some of the more significant materials along with fuel use, and electricity used to produce the MDI resin. A complete listing can be found in Boustead (1999).

Table 3.3: Production Requirementsfor the 42.4 lb of Phenol-formaldehyde Resin Needed to Manufacture 1.0 Msf 3/8-inch Basis OSB in the Southeast Region. ^1/

^1/Data obtained from Material, Energy & Environmental Unit Factor Emissions: Structural Wood Production, Athena, 1993.

Table 3.4. Production Requirements for the 8.16 lb of MDI Resin Needed to Manufacture 1.0 Msf 3/8-inch Basis OSB in the Southeast Region. ^1/

^1/Data obtained from I Boustead. Ecoprofiles of chemicals and polymers 1999

^2/Partial listing of Materials

4.0 POLLUTION CONTROL EMISSIONS TO AIR

Most emissions are generated from the heat generation process used to supply the dryers and the presses. Since all of the plants surveyed had direct-fired heating systems for dryers and presses, the emissions have components of CO, CO₂ (fossil), NO_x, SO₂, and others. Dryers are used to take the moisture content of green OSB flakes from about 100% down to about 5% (oven-dry basis) and have an inlet temperature ranging from 1,100 to 1,300 ^oF. Hot pressing is done in the OSB manufacturing process to provide intimate contact between oriented flakes while the phenol-formaldehyde and/or MDI adhesives cure as a result of temperature in the 360-400^oF range. Emissions are generated from the wood as a result of the high temperatures in the dryers and presses, and the adhesives also generate emissions during cure.

In OSB manufacturing, emissions from dryers and hot pressing are collected and processed through pollution control systems such as described in Section 1.3. Emissions are monitored through these pollution control systems. Table 4.1 lists emissions as determined by surveys of OSB manufacturers.

Table 4.1 Emissions from Pollution Control Systems of Southeast OSB Manufacturers as Reported in Surveys
(These are total emissions, no burden or allocation has been made to co-products).

^1/ Air emission data as reported from surveys
^2/ Not reported in survey

5.0 UNIT FACTOR ESTIMATES FOR MAIN PRODUCTS

Life-cycle inventory results to produce 1.0 Msf 3/8-inch of OSB in the Southeast is given in Table 5.1 (inputs) and 5.2 (outputs). Results include all processes within the system boundary defined in Figure 1.1. These inputs and outputs summarize all those that were obtained from the surveyed plants. These inputs and outputs exclude the emissions contributed by the transportation of logs to the mill or emissions contributed by the production, transportation and or utilization of resin, wax, fuel, and electricity.

Table 5.1 Life-cycle Input Inventory Results for Production of 1.0 Msf 3/8-inch Basis OSB in the Southeast Region. (Results are for OSB production only, no emissions included for production and delivery of fuel, electricity, resins or wax).

Table 5.2 Life-cycle Output Inventory Results for 1.0 Msf 3/8-inch Basis OSB Production from the Southeast Region.
(Results include OSB production only, no emissions were included for the production and delivery of electricity, fuels, resins, and wax).

^1/ Not reported in survey

6.0 PROPOSED RESEARCH FOR COMING YEAR

Proposed research in the coming year includes expanding the OSB LCI model into a sub-unit process approach. This model will be useful in analyzing was to improve production efficiency subject to environmental impacts. Also the development of an LCI model for Parallel Strand Lumber will be completed. These models will be used in conjunction with other LCI models developed in the CORRIM project. A number of scenarios will be studied to optimize building materials design to minimize environmental impact. Output of the analyses will be compared to benchmark data to assess improvements.

7.0 REFERENCES

APA The Engineered Wood Association (APA). 2001. North America Structural Panel
    Production by Geography 2000. March, 1 p.

Boustead, I. 1999. Ecoprofiles of chemicals and polymers.

Franklin Associates. 1998. Combustion of Wood in Industrial Boilers. SimaPro5 Life-Cycle
    Assessment Software Package, version 36, 2001.

Franklin Associates. 2000. Wood Precombustion. SimaPro5 Life-Cycle Assessment Software
    Package, version 36, 2001.

Franklin Associates. 1998. Natural Gas Combustion in Industrial Boilers. SimaPro5 Life-Cycle
    Assessment Software Package, version 36, 2001.

Franklin Associates. 1998. LPG Precombustion. SimaPro5 Life-Cycle Assessment Software
    Package, version 36, 2001.

Franklin Associates. 1998. Diesel Powered Industrial Equipment. SimaPro5 Life-Cycle
    Assessment Software Package, version 36, 2001.

Franklin Associates. 1998. Electricity from coal. SimaPro5 Life-Cycle Assessment Software
    Package, version 36, 2001.

Franklin Associates. 1998. Electricity from DFO. SimaPro5 Life-Cycle Assessment Software
    Package, version 36, 2001.

Franklin Associates. 1998. Electricity from natural gas. SimaPro5 Life-Cycle Assessment
    Software Package, version 36, 2001.

Franklin Associates. 1998. Electricity from uranium. SimaPro5 Life-Cycle Assessment
    Software Package, version 36, 2001.

Franklin Associates. 1998. Electricity from hydropower. SimaPro5 Life-Cycle Assessment
    Software Package, version 36, 2001.

PRé Consultants B.V. 2001. SimaPro5 Life-Cycle Assessment Software Package, version 36.
    Plotter 12, 3821 BB Amersfoort, The Netherlands. http://www.pre.nl/.

The Athena Sustainable Materials Institute (Athena). 1993. Raw Material Balances, Energy
    Profiles and Environmental Unit Factor Estimates: Structural Wood Products. Forintek
    Canada Corp, Ottawa, Canada. March 1993.

Toennisson, R.L. and S.W. Hadden. 1993. Wood Products Engineer's Handbook. Technical
    Note B66, Forest Resources, Tennesse Valley Authority, Norris, Tennessee. 63 p.

United States Department of Energy (USDOE). 2000. State Electricity Profiles 2000.
    http://www.eia.doe.gov/cneaf/electricity/st_profiles/.