Modeling Energy and Climate Policy in the Finnish Forest Sector Lauri Hetemäki, Hanna-Liisa Kangas, Jani Laturi, Jussi Lintunen and Jussi Uusivuori Working Papers of the Finnish Forest Research Institute 146 http://www.metla.fi/julkaisut/workingpapers/2010/mwp146.htm ISBN 978-951-40-2221-0 (PDF) ISSN 1795-150X www.metla.fi Working Papers of the Finnish Forest Research Institute 146 http://www.metla.fi/julkaisut/workingpapers/2010/mwp146.htm 2 Working Papers of the Finnish Forest Research Institute publishes preliminary research results and conference proceedings. The papers published in the series are not peer-reviewed. The papers are published in pdf format on the Internet. http://www.metla.fi/julkaisut/workingpapers/ ISSN 1795-150X Office Post Box 18 FI-01301 Vantaa, Finland tel. +358 10 2111 fax +358 10 211 2101 e-mail julkaisutoimitus@metla.fi Publisher Finnish Forest Research Institute Post Box 18 FI-01301 Vantaa, Finland tel. +358 10 2111 fax +358 10 211 2101 e-mail info@metla.fi http://www.metla.fi/ Working Papers of the Finnish Forest Research Institute 146 http://www.metla.fi/julkaisut/workingpapers/2010/mwp146.htm 3 Authors Hetemäki, Lauri, Kangas, Hanna-Liisa, Laturi, Jani, Lintunen, Jussi & Uusivuori, Jussi Title Modeling energy and climate policy in the Finnish forest sector Year 2010 Pages 78 ISBN 978-951-40-2221-0 (PDF) ISSN 1795-150X Unit / Research programme / Projects Vantaa Research Unit / Project 50168 Accepted by Leena Paavilainen, Director of Research, 8.2.2010 Abstract The development and structural changes that the forest sector in Finland is going through emphasizes the need to modify the forest policies in Finland. A shift from a focus on timber production and traditional forestry products toward new products as well as new services is called for. New services include the climate and bioenergy potential services, landscape, travelling and recreation, and ecological services. This report describes the structural context and principles of an ongoing modeling work in Metla. This work is part of the research project The Future Development of the Finnish Forest Sector, which is sponsored by Metla and Metsämiesten Säätiö Foundation. The purpose of the modeling work is to build a policy simulation model to be able to analyze policies relevant for the Finnish forest sector. Keywords forest sector, energy sector, climate change, partial equilibrium, policy simulation model Available at http://www.metla.fi/julkaisut/workingpapers/2010/mwp146.htm Replaces Is replaced by Contact information Jussi Uusivuori, Finnish Forest Research Institute, P.O. Box 18, FI-01301 Vantaa, Finland. E-mail jussi.uusivuori@metla.fi Other information Taitto: Maija Heino Working Papers of the Finnish Forest Research Institute 146 http://www.metla.fi/julkaisut/workingpapers/2010/mwp146.htm 4 Contents Foreword...................................................................................................................................... 5 1 Introduction................................................................................................................6 2 Wood demand and supply in Finland......................................................................8 2.1 Structure of the demand..................................................................................................... 8 2.1.1 Wood categories........................................................................................................ 8 2.1.2 Demand due to processing........................................................................................ 9 2.1.3 Demand due to energy generation.......................................................................... 12 2.1.4 Conclusions and future outlook.............................................................................. 17 2.2 Timber supply.................................................................................................................. 19 2.2.1 Forests in Finland................................................................................................... 19 2.2.2 The ownership of forests and timber supply in Finland......................................... 20 3 Models of the forest sector.....................................................................................27 3.1 Demand models............................................................................................................... 27 3.1.1 Objectives of the demand models........................................................................... 27 3.1.2 Dichotomy of the models....................................................................................... 27 3.1.3 Policy cost analysis................................................................................................. 29 3.1.4 Examples of energy sector models......................................................................... 30 3.1.5 Examples of forest sector models........................................................................... 32 3.2 Forest management and forestry models......................................................................... 33 3.2.1 The optimal management of forest......................................................................... 33 3.2.2 The forest simulation and forestry models............................................................. 33 4 Climate, energy and forest policies........................................................................36 4.1 Climate policies............................................................................................................... 36 4.1.1 Energy and emission taxes..................................................................................... 37 4.1.2 Emission trading..................................................................................................... 37 4.1.3 Carbon sequestration policies................................................................................. 39 4.2 Renewable electricity promoting policies....................................................................... 40 4.2.1 Investment and electricity subsidies....................................................................... 41 4.2.2 Feed-in laws............................................................................................................ 41 4.2.3 Tradable green certificates...................................................................................... 44 4.3 Finnish forestry subsidies................................................................................................ 46 4.4 Climate, energy and forest policy modeling ................................................................... 47 4.4.1 Climate policy economics...................................................................................... 48 4.4.2 The economics of RES-E policies ......................................................................... 52 4.4.3 The economics of forestry subsidies...................................................................... 54 4.5 Conclusions..................................................................................................................... 55 5 Framework of the model.........................................................................................57 5.1 Introduction ................................................................................................................... 57 5.2 Defining the structure of the model................................................................................. 58 5.2.1 Partial vs. general equilibrium model..................................................................... 58 5.2.2 The scope of the sectors in the model..................................................................... 59 5.2.3 Structure of the demand side.................................................................................. 62 5.2.4 Structure of the roundwood and energywood supply............................................. 63 5.2.5 Dynamics of the model........................................................................................... 64 5.2.6 Policy measures .................................................................................................... 65 References....................................................................................................................66 Appendix. Forestry center level data for section 2.2.................................................73 Working Papers of the Finnish Forest Research Institute 146 http://www.metla.fi/julkaisut/workingpapers/2010/mwp146.htm 5 Foreword This report is part of an ongoing project that examines the impacts of climate, energy and forest policies on the Finnish forest and energy sectors. One of the objectives of the project is to build a partial equilibrium simulation model, in order to run policy simulations and analyze the effectiveness and impacts of policy measures. The purpose of the present paper is to provide the background for the model building process. It describes the main features of the Finnish forest and energy sectors. In doing that, it discusses the demand and supply structure for the various forest products, energy products, and roundwood and energy wood. In addition, the relevant simulation modeling literature, and the modeling of different climate, energy and forest policies are discussed. The study concludes by outlining the main features that the simulation model being built in the project will have. The study is part of the research project “The Future Development of the Finnish Forest Sector” being carried out at the Finnish Forest Research Institute (Metla). This project was made possible by funding from Metsämiesten Säätiö Foundation. This work is an outcome of joint effort by the authors. Of the particular chapters or sections the main responsibility is the following: Hetemäki & Uusivuori chapters 1 and 5; Hetemäki section 2.1.4; Kangas chapter 4; Laturi chapter 2.2 and 3.2; Lintunen chapter 2.1 and 3.1. Vantaa, 25.9.2009 Jussi Uusivuori and Lauri Hetemäki Working Papers of the Finnish Forest Research Institute 146 http://www.metla.fi/julkaisut/workingpapers/2010/mwp146.htm 6 1 Introduction National forest sectors around the globe are in the process of being reshaped dramatically. Forests and forest sectors with their products and services are one of the focuses in the national and international climate policies world wide. This is true also in terms of energy policies: in many countries forests provide substantial potentials for bioenergy and in building up renewable energy sectors. Finland, as many other industrialized countries with sizable forest sectors, is undergoing a process where the traditional wood utilizing industries are cutting down their operations in the country and redirecting their new investments to mainly those markets where the product consumption is increasing the most and/or where the wood fiber supply is based on fast growing plantation forestry. Along with the forest industrial changes, there are also public attitude changes toward forests and their use in developed countries. These changes are reflected in the land ownership behavior of nonindustrial private forest owners. The private forest owners, who own most of the forests in Finland, are being urbanized, and are becoming to some extent less production minded. This will influence the forest sector in the upcoming decades. The development and structural changes that the forest sector in Finland is going through emphasizes the need to modify the forest policies in Finland. A shift from a focus on timber production and traditional forestry products toward new products as well as new services is called for. New services include the climate and bioenergy potential services, landscape, travelling and recreation, and ecological services. The Government of Finland launched its National Forest Programme 2015 in spring 2008. The Finnish Forest Research Institute (Metla) was involved in providing background work for this programme. In particular, Metla helped the Ministry of the Agriculture and Forestry in preparing scenarios for the development of the forest sector. One aspect of this work was using available modeling tools to study possible scenarios and policy impacts within the forest sector. This work showed that there is a discrepancy between the demand side and supply side models describing forest sector in Finland. From the policy perspective, the description of the private forest owner behavior should be strengthened in the modeling work. Metla was also involved in the National Climate and Energy Strategy work led by the Ministry of Employment and Economics during 2007–2008. This program outlines the policy responses that Finland will follow to meet the EU, Kyoto and IPCC commitments in climate mitigation and adaptation and in terms of bioenergy targets One lesson from that work has been that for policy purposes the forest and energy sectors in Finland should be viewed within a more integrated framework than currently is typical. Therefore also, simultaneous analysis and modelling of the forest and energy sectors is becoming a necessity, due to the increasing role of forest biomass based energy production. The models should also be more conducive to integrating policy tools in order to study their impacts. These are some of the central issues that are addressed in this paper. In particular, the work analyses the links between the roundwood supply side and demand side by basing the supply side on the behavioral optimization of the nonindustrial private landowners. In addition, the energy sector and the role of energywood is discussed. Working Papers of the Finnish Forest Research Institute 146 http://www.metla.fi/julkaisut/workingpapers/2010/mwp146.htm 7 In summary, the modeling initiative described in this report has three main ‘drivers’ or objectives within which it tries to make a contribution to existing modeling work: – Linking forest and energy sectors – Linking forest industry and accurate nonindustrial private forest owners’ behavioural descriptions – Enable improved integration of policy tools in the model, in order to analyze policy effectiveness and impacts. The purpose of this report is to describe the background for building of a new partial equilibrium model for the Finnish forest and energy sectors, which incorporates the three features listed above. In doing this, we discuss e.g. the various approaches that have been adopted in the literature for building sectoral partial equilibrium models. The outline of the report is as follows. First, the structure of wood demand and supply in Finland is described. This lays the background for what one needs to model, in order to capture the salient features of the Finnish forest sector and wood using energy sector. Next, the forest, climate and energy policies are discussed. That is, the content of the commonly used polices and their implications are presented. Next, the different approaches and more technical issues related in modelling demand and supply, as well as different policy measures are discussed. This discussion leads us to consider the essential question of this report, namely: What type of structure and features should a partial equilibrium model have, which would allow to study the effectiveness and implications of various forest, climate and energy policies to Finnish forest and energy sector? Working Papers of the Finnish Forest Research Institute 146 http://www.metla.fi/julkaisut/workingpapers/2010/mwp146.htm 8 2 Wood demand and supply in Finland 2.1 Structure of the demand Wood is a versatile raw material which can be classified into several sub-categories. Here, we use a classification of wood based on tree species and suitable raw material uses. The demand for wood consists of the demands for these separate categories. There are three main users of the wood: forest industries, energy industries and households. The forest industries use wood as a raw material for various types of final goods. All the users, including forest industries, utilize wood in thermal energy and/or electricity production via different kinds of processes. In addition, there is a wood demand that does not involve harvesting, for example climate services. In this section, we present the wood categories and the level of their demand and the industries that use them. A special emphasis is given to the forest bioenergy generation, where the forest biomass demand is expected to grow significantly due to the climate and energy policies. 2.1.1 Wood categories The usage of the wood depends on its physical properties. These properties depend on species of the tree, types of forest site, silviculture etc. From the viewpoint of the consumer, the main characteristics of the wood are species of the tree and size and quality of the tree stem. We classify the timber by these attributes. Due to the climate, coniferous forests are widespread in Finland, as approximately 80 % of the growing stock volumes consist of softwood (Metla 2007a, 67). The main softwood species are Scots pine and Norway spruce. The prevalent hardwood species is birch. These three species dominate the Finnish wood markets although there are some other species that are used in smaller scale. As their names suggest, the softwood is less dense (on average about 400 kg/m3 basic) than hardwood (490 kg/m3 basic). However, the effective heating values of the wood species are quite similar (dry matter: 19.2 MJ/kg) (Alakangas 2000). The fibers in softwood are longer than in hardwood which is reflected in the characteristics of various pulp and paper categories. Trees of all species are divided into two timber grades. These grades are saw-timber trees and pulpwood. The grading is based on the size and quality of the yielded roundwood. The trees with large diameter, so that they yield at least one log, are considered to be saw-timber trees (abbreviated here as logs).1 Trees with smaller diameter, poor quality large trees and top stems of the larger trees are pulpwood. There are also more rarely used timber grades that are more strictly related to the suitable raw material use (e.g. post, small-diameter log, veneer log etc.). However, here we focus on the two main timber grades. The timber grades do not cover all the wood consumption. Firstly, a share of the trees cannot be categorized as logs nor pulpwood. These include mainly small-sized trees, whose diameter size is too small for pulpwood. Secondly, the logging residues and stumps and roots of all the harvested trees do not belong to any of the above-mentioned groups, yet they are utilized. These two groups can be burnt as such or after chipping into forest chips. Therefore, we call this category as forest chips. Thirdly, there is roundwood that is used in energy generation. This wood is mostly chopped fuelwood for the small-sized dwellings and some of it is utilized by heating and power plants. Since industry chips their otherwise unusable roundwood, we include its portion of energy roundwood 1Log is a straight stem with top end diameter at least 15 cm. ‘Small-diameter log’ is thinner. Working Papers of the Finnish Forest Research Institute 146 http://www.metla.fi/julkaisut/workingpapers/2010/mwp146.htm 9 into the forest chip category. We label the fuelwood of dwellings as fuelwood. The forest chips and fuelwood are jointly labeled as energy wood. The wood types above (logs, pulpwood, fuelwood and forest chips) are harvested from the forests. In addition there are the wooden residues of forest industry processes. These industry residues contain bark, dusts and chips and they are utilized as raw material but also in energy generation. 2.1.2 Demand due to processing Forest industries Forest industries use wood as their main raw material. The industry is divided into wood-product and pulp industries. The four main sub-industries of the wood-product industry are sawmilling, plywood and veneer, particle board and fiberboard industries. The other intermediate level industries are manufacturing of builder’s carpentry and joinery, wooden containers and other wooden products. The pulp industry is divided in to mechanical, semi-chemical and chemical pulp industries. Sawmilling industry has the largest roundwood input volumes of all the wood-product industries. It produces sawn and planed sawn timber and other machined wood. The second largest industry in volume is plywood and veneer industry. As the name suggests it produces veneer sheets and their upgrades. Cutting, barking, sawing and planeing of wood produces chips and dusts of various grain sizes in all the wood-product industries. These by-products are used by particle board and fiberboard industries which produce various wood based boards with or without adhesives. The pulp industries produce several pulp types for different paper categories. There are two processes for making mechanical pulp: grinding and refining. In the grinding process small logs (bolts) of pulpwood are pressed against rotating stone. In the refining process chipped pulpwood is ground up in refiner plates. Of the two, the refining process consumes more electricity, but it is gentler to fibers. Spruce is the dominant raw material for mechanical pulp. When wood is pretreated with chemicals prior to refiner plates, the pulp is called chemi-thermomechanical pulp (abbreviated as semi-chemical pulp). In the chemical pulp process chemicals and heat separates the fibers of the wood from lignin. The yield is low since the lignin is separated from fibers, but the fibers remain longer than in mechanical pulp. Chemical pulp processes need thermal energy which is received from burning the lignin of the black liquor in a recovery boiler. The pulping processes that use chipped wood can also use the by-product chips of the forest industries.2 A rough outline of forest industries’ inputs and outputs is presented in Table 2.1.1. (Finnish Forest Industries 2007). Production of sawmilling industry is an order of scale larger than in other wood product industries. In 2008 the wood-product industries produced 9.9 million m3 of sawn goods, 1.3 million m3 of plywood and veneer sheets and 0.3 million m3 of other wood-based boards in Finland (Metla 2009a). Production of sawn goods nearly doubled during the 1990’s but has leveled since that. Plywood industry is yet growing, even if its production levels are still small compared to the sawmilling. Production levels of chemical pulp are about one and a half times the levels of mechanical pulp, including semi-chemical pulp. In 2008 pulp industries produced total of 7.2 million tons of chemical pulp and 4.5 million tons mechanical and semi-chemical pulp. Production of chemical pulp increased quite steadily for thirty years and production doubled in twenty years from mid-eighties. Production of mechanical pulp grew at a similar rate. Since 2006 the production of pulp products has declined by 2.5 million tons (Finnish Forest Industries 2009). 2Chips are here defined as wooden residue that is or can be chipped to a size large enough to be used as a raw material for pulp industries. Dust contains finer residues. Working Papers of the Finnish Forest Research Institute 146 http://www.metla.fi/julkaisut/workingpapers/2010/mwp146.htm 10 A large share of Finnish made pulp is also used in Finland (67 % in 2008). The volumes of paper production are large (2008 figures): magazine paper 5.9, fine paper 2.9, other papers 1.4 and cardboard 2.9 million tons (Finnish Forest Industries 2009). Production over time is shown in Figure 2.1.1. Growth was steady until the year 2000, after which it has stopped. In fact, in recent years there have been significant capacity cut-downs, due to which the paper and paperboard capacity has declined from the 15.5 mill. t in 2005 to 12.7 mill. t in 2009, and the pulp capacity from 14.9 mill. t in 2006 to 12.2 mill. t in 2009. Since the forest industries in Finland are large compared to the size of the nation, a significant share of produced goods is exported. In 2008, the total value of forest industry product exports was 11500 million € (including wood products and pulp and paper). Wood products cover about a quarter of the export value while pulp and paper industry covers the rest of the value. The export volumes of forest industry products follow the same trends as production levels: exports of sawnwood increased rapidly in 1990s, but have levelled after that. However, exports of plywood have been increasing also in the last decade. One-fifth of wood pulp is exported from Finland, mainly to the companies’ own mills in Central and Western Europe. The exports consist mainly of sulphate pulp. In case of paper, about 90 % of Table 2.1.1. A rough outline of forest industries’ inputs and outputs. Industry Main input Main products By-products/residues Sawmilling Logs: soft wood Sawn and planed wood Chips, sawdust, bark Plywood & veneer Logs: spruce & hardwood Plywood and veneer Chips, dusts, bark Boards Chips, sawdust Particle & fiberboard Cutting residues, dust Other wood products Logs: pine Wooden products Cutting residues, dust Mechanical pulp Pulpwood: spruce, chips Mech. pulp Heat, bark Semi-chemical pulp Pulpwood: hardwood, chips Semi-chem. pulp Bark Chemical pulp Pulpwood: pine & hardwood, chips Chem. pulp Waste liquors, bark Figure 2.1.1. Pulp and Paper Production in Finland 1980−2008 (Finnish Forest Industries 2009). 1980 1985 1990 1995 2000 2005 6 8 10 12 14 Paper and paperboard Wood pulp million metric tons Working Papers of the Finnish Forest Research Institute 146 http://www.metla.fi/julkaisut/workingpapers/2010/mwp146.htm 11 the production is exported. The paper and paperboard exports grew for thirty years almost hand in hand with pulp and paper production. The main export products are nowadays magazine and fine paper totalling over 8 million tons. Paperboard is also important export product with quantity of 2.6 million tons. Newsprint has lost its significance and the exports of kraft and other paper have been quite constant. Exports of these paper grades totalled in about 0.9 million tons. Wood products industry is less export orientated industry than paper industry, since about 40 % of the production is consumed in domestic markets. Finnish forest sector products are exported to all around the world, as can be seen from Table 2.1.2. In value, most important import countries of Finnish wood products are EU-countries, Japan, Russia and Egypt (Metla 2009d). Roundwood demand The consumption of wood in forest industries is presented in Table 2.1.3. The division by timber grades follows the industry classes: logs are mostly utilized in wood-product industries and pulpwood in pulp industries. In the sawmilling industry softwood is the main raw material and in the plywood and veneer industry spruce and hardwood. In the pulp industries the main raw material for mechanical pulp is spruce, for semi-chemical pulp hardwood and for chemical pulp, pine and hardwood. The board industries and pulp industries use also imported chips as well as chip and dust by-products from all branches of the forest industries. Consumption of logs has been quite stable for a decade in Finland. Log consumption was almost at the same level in 2008 as it was in 1997. Pulpwood use, however, has been increasing, especially Table 2.1.2. Wood products exports from Finland in 2008 (million euros) (Metla 2009d). Euro Asia North America Africa South America Oseania Total Sawn goods 666 240 1 245 – 2 1154 Plywood 543 48 15 2 0 1 610 Veneer sheets 34 4 1 0 – – 39 Particle board 22 0 0 – 0 – 22 Fiberboard 16 2 0 – 0 0 17 Wood pulp 837 217 2 14 5 – 1 076 Paper 4 120 433 455 78 264 157 5 507 Paperboard 1 333 250 90 47 33 25 1 778 Other wood products 972 119 68 6 11 10 1 187 Total 8 641 1 323 633 413 314 194 11 518 Table 2.1.3. Consumption of wood in forest industries in Finland in 2008 (million m3) (Metla 2009a). Wood-product industries Pulp industries Total Saw- milling Plywood/ veneer Boards Other Mech. Semi- chem. Chemical Logs Pine 10.48 0.00 – 0.37 – – 0.27 11.13 Spruce 9.84 2.05 – 0.02 0.54 0.01 0.06 12.51 Hardwood 0.17 1.45 – – – 0.00 0.01 1.63 Pulpwood Pine 1.20 – – – 0.14 – 13.99 15.33 Spruce 0.38 – – – 7.20 0.08 2.66 10.31 Hardwood 0.00 0.00 – – 0.89 0.94 10.91 12.75 Imported chips – – 0.06 – 0.17 0.00 2.38 2.61 Sawmill chips & dust – – 0.62 – 1.95 0.64 6.40 9.61 Total 22.08 3.50 0.68 0.38 10.88 1.67 36.68 75.87 Working Papers of the Finnish Forest Research Institute 146 http://www.metla.fi/julkaisut/workingpapers/2010/mwp146.htm 12 in the case of hardwood. In the period 1996–2006, the total pulpwood use increased by 38 % and hardwood by 50 %. By 2008, the total pulpwood use has decreased 10 % and hardwood 15 % from peak rates of 2006. Some of the Finnish roundwood and wood residues are also exported. In 2008 exports were 1.1 and 0.37 million m3 of roundwood and wood residues respectively. The joint value of these exports was 130 million €. Roundwood is mainly sold to Sweden (72 % in 2008) and wood residues to Sweden, Denmark and The Great Britain (Metla 2009d). 2.1.3 Demand due to energy generation Wood-based energy generation Demand for wood by the energy sector is generated by the use of the wood fuels. Wood fuels are mainly utilized by forest industries, energy industry and households. Fuels are used in order to produce heat and electricity (power). Heat is needed for industrial processes and heating of buildings. Therefore, heat production is divided into industrial steam and district heating segments. Heat and power can be produced separately or by combining heat and power production (CHP). Wood based industrial heat is mainly produced by the forest industries whereas energy industries produce separate power and most of the district heating. Households’ wood fuel use is mostly heating of their dwellings. In Finland 21 % of the consumed energy in 2008 was produced by wood-based fuels (Statistics Finland 2009a). The distribution of energy generation by wood fuels is presented in Table 2.1.4.3 Most of the wood-based energy comes from industrial CHP, which practically means the forest industries. Also the heating of the small-sized dwellings has a large share. In fact, at the present moment energy industry generates quite a small share of wood-based energy. The climate change sets increasing pressures on energy industries. The EU policies targeted to mitigate the change, expand the use of renewable energy sources (RES). For Finland, the target share of renewable energy in final energy consumption is 38 % by the year 2020 (EC 2008b). In 2006 the share was 28.9 % (Statistics Finland 2009a). Renewable energy is based on biomass, wind, hydro, solar and geothermal sources. In Finland, most promising renewable energy sources are forest biomass and wind power. Whatever the Finnish solution for reaching the renewable energy targets will be, it is clear that wood-based energy generation will increase. Some remarks on wood-based energy generation technologies In this section we present a rough outline of the energy generation technologies when wood fuels are utilized. Although the technology is highly evolving the basic guidelines can be given. This section draws on VTT (2004), Helynen et al. (2002) and Jalovaara et al. (2003). Heat and power can be produced separately or they can be co-generated. Separate heat production of heat has high efficiency of 80-90 %, while power production has electricity efficiency of 30–55 3The figures of Table 2.1.4 are partly inaccurate since part of the small heating and industrial plants are not included. The total wood energy produced in Finland in 2007 was 295 PJ (Statistics Finland 2009c, 59). Table 2.1.4. Wood fuel consumption in energy production in Finland in 2007 (PJ). Separate power production includes CHP power production with auxiliary condensers and separately recovered condensing power. DH refers to district heating (Statistics Finland 2009c). Separate production CHP Total Power District heat Industrial steam Heating of dwellings In DH In industry 10.5 5.0 10.7 48.6 18.8 191.1 285 Working Papers of the Finnish Forest Research Institute 146 http://www.metla.fi/julkaisut/workingpapers/2010/mwp146.htm 13 % depending on technology used. While the total efficiency (thermal efficiency) of cogeneration (CHP) is typically high, nearly 90 % or more, the separate production is needed also, since the heat and power loads differ in size as well as in timing. In cogeneration the total efficiency is divided between electricity and heat production efficiencies determined by power-to-heat ratio of the plant. Typically, electricity efficiency cannot be made in cogeneration as high as in separate production. Wood fuels can be used in direct combustion, in combustion after gasification or after further processing. The case of processed wood fuels is discussed separately in next section, here we focus on combustion and gasification. There are four major types of combustion technologies used: pulverized fuel burners, grate firing, fluidized bed combustion and recovery boilers. Pulverized fuel burner is least suitable for wood fuels and it is in Finland mainly used with coal and peat. There are estimates that up to 5 % of input power could be wood in pulverized fuel burner, without any modifications in fuel injection processes (e.g. Helynen et al. 2002). This kind of cofiring has been experimented in Finland. Grate firing is typical in households and small-scale heat production. With traveling grate the technology is suitable for peat use and also for cofiring of peat and wood. More large scale combustion of wood fuels is done in fluidized bed and recovery boilers. There are different kinds of fluidized bed technologies and more is under development. Typical models utilized are atmospheric bubbling and circulating fluidized bed combustion, BFBC and CFBC respectively. FBC technologies have some favorable properties. Firstly, combustion is done in low temperature and the process mixes the fuel with air which reduces emissions. Secondly, the process is suitable for multi-fuel use. In Finland these fuels are typically peat and wood. The other large scale combustion technology is the recovery boiler. These boilers are used in chemical pulp processes for recovering the process chemicals. While the recovery takes a part of the heat produced, the rest is used in paper processes or as district heating. Gasification is another way to produce energy from wood fuels. These technologies are developing rapidly and only some main principles are presented here. In gasification fuel is heated in low oxygen environment, which doesn’t allow proper burning. While some of the energy in fuel is consumed in the heating process, most of it is separated in various gases which constitute syngas (synthesis gas, also product gas). After cleaning, the syngas can be used in similar processes as natural gas, even though syngas has lower heating value. Typical direct uses of syngas are combustion in gas engine or in gas turbine. Syngas is also easily incorporated in cofiring in existing plants. The gas engines are suitable for small scale power production and they have good total efficiencies. Gas turbines have been used as source of peak power, but this has changed due to the development of combined cycle technologies. In combined cycle exhaust gases of gas turbine are reheated and they power a set of steam turbines. This combination raises the electricity efficiency over 50 % and it can also be utilized in CHP. The technology for integrated gasification and gas turbines with combined cycle (IGCC) is under rapid development. It is good to notice that combined cycle technology for boilers (pulverized fuel and fluidized bed) needs pressurized combustion and it is one line of research also. Working Papers of the Finnish Forest Research Institute 146 http://www.metla.fi/julkaisut/workingpapers/2010/mwp146.htm 14 Processed wood fuels Wood fuels can be refined. Instead of burning the wood and syngas, both of them can be processed to have higher energy density and additional uses, e.g. liquid transport fuels. There are multitude of processing technologies and they are under intense research. Therefore the field is evolving rapidly. We illustrate here some of the prominent current and future technologies. Good introduction to the topic in Finnish can be found in (VTT 2004). The pellets and briquettes are compressed wood dust and chips. The raw materials are typically by-products of sawmills and other mechanical forest industries. This way the humidity of the dusts is low enough (10-15 %) for direct processing. In the case of moist raw materials they need to be dried first, which consumes energy and raises production costs. Since pellets have high energy density and uniform quality they are a usable fuel for many applications. In Finland the production of pellets has grown in recent years (see Figure 2.1.2). Most of the pellets are exported (60 % in 2008) but domestic use has been increasing. Also the other by-products can be processed. For example bark can be powdered to be used in cofiring with fossil fuels in a burner. The abundant by-product in chemical pulp industry is lignin of black liquor and it can be utilized in several ways. The usual burning in recovery boiler can be replaced by gasification and the product gas is usable for further processing. It is also possible to extract the lignin and utilize it in energy production or as a raw material. Raw soap and tall oil are also raw materials for fuel products. Wood fuel can be processed to liquid fuel. Ethanol is received via hydrolysis and following fermentation of sugars. The hydrolysis of wood is more difficult than that of starch containing biomass and lignin cannot be utilized in the process. Ethanol can be used as such or further upgraded. Another way to liquidify wood is via pyrolysis. The yield of the product called pyrolysis oil is high and it is suitable for replacing fuel oil. Syngas received from gasification of wood biomass opens new ways to upgrade wood fuels. These methods are numerous (e.g. Spath and Dayton 2003). Most promising technologies from pulp and Figure 2.1.2. Production and exports of pellets in Finland (Metla 2009b). 2002 2004 2006 2008 0 100 200 300 400 Production Exports 1000 tons Working Papers of the Finnish Forest Research Institute 146 http://www.metla.fi/julkaisut/workingpapers/2010/mwp146.htm 15 paper industries are Fischer-Tropsch (FT) fuels, dimethyl ether (DME) and alcohol fuels (Larson et al. 2006). FT synthesis of natural gas and gasified coal to liquid fuel is quite mature but growing industry and FT synthesis of biomass-based syngas has received wide interest. The product of FT synthesis are hydrocarbons that can be refined to diesel type fuels. DME is also usable in diesel- engines but it needs pressurized fuel and refueling systems in order to be in liquid state. Since its burning is more clean than normal diesel oil it has received interest as a fuel for centrally refueled urban fleet vehicles. Increasing use of ethanol as a transport fuel makes also the syngas- to-alcohol processes interesting. The methods studied are catalytic synthesis and fermentation. Through gasification also the lignin can be utilized in alcohol production (Larson et al. 2006). Wood fuel demand In Finland, a significant share of the consumed wood ends up in energy generation. The total consumption of wood is presented in Table 2.1.5. Fuelwood and forest chips account for some 13 % of the direct consumption of harvested trees. More than half of the solid by-products and residues are utilized in energy generation and roughly half of the barked raw material of chemical pulp is burnt as black liquor in recovery boilers. Wood spent for energy generation totals 21.6 million m3 of which forest chips have a share of 21 %. If the volume of black liquor is added, approximately 40 million m3 of wood was used in energy production in 2009, which is about one half of the total roundwood consumption. In order to analyze the energy use of wood in more detail, we study the segments of the wood fuels. In Table 2.1.6 these segments with their possible uses and tradability are shown. Although they are both forest chips, small-sized trees and logging residues are separated, since their collection costs are different and there are differences in their usability. Fuelwood for small-sized dwellings is typically traded in a small scale or not traded at all but used directly by the forest owner. Markets are, however, emerging. Industrial chips are suitable raw material for pulp and boards industries Table 2.1.5. Total consumption of different wood types in Finland in 2009 (million m3) classified by supply type and final user. Waste liquors are not included. Forest chips use by small-sized dwellings is estimated to be 0.6 million m3. (Metla 2009a). Harvesting and imports By-products and residues Forest industries Energy generation Forest industries Energy generationLogs Pulpwood1) Fuelwood Forest chips 25.27 40.99 5.40 4.63 9.61 11.60 1) Includes imported chips (2.61 million m3) Table 2.1.6. The main segments of wood fuels, their main use and tradability. Logging residues include stumps and roots. Source Segment Use Trading Harvesting & thinning Small-sized trees Energy Market (forest chips) Logging residues Energy Market (forest chips) Fuelwood energy Partly non-market By-products and industrial residues Industrial chips Multiple Market Sawdust multiple Market Bark Energy Mainly non-market Liquors Energy Mainly non-market Working Papers of the Finnish Forest Research Institute 146 http://www.metla.fi/julkaisut/workingpapers/2010/mwp146.htm 16 and sawdust is suitable for board industries only. Chips and dusts can also be used in pellet production. Bark and black liquor are typically burned at the production site. Other waste liquors can be processed and they are traded. Even though boards industry is quite small in comparison with the sawmilling and plywood industries, the fact that it uses dusts and chips as raw material, makes it important in wood fuel considerations. Most important wood fuel in Finland are waste liquors of pulp industry, especially black liquor. It covers half of the total wood fuel consumption. Other significant segments are fuelwood of small-sized dwellings and forest industries’ bark residue. They cover jointly 28 % of the wood fuel consumption. Forest chips cover 11 % and its share is bound to increase. Wooden by-products other than bark have small shares since they are mostly spent as raw material. The consumption of wood fuels is presented in Table 2.1.7. In recent years the amount of forests chips used has been rising rapidly. The utilization of forest chips in is presented in Figure 2.1.3. The forest chips consumption by the households is based on a study in 2000/2001 and 2007/2008 heating seasons but statistics on industrial use are compiled every year. The growth of use in CHP has risen strongly. The increase has been almost 25 % a year. Figure 2.1.3. The consumption of forest chips. Total use includes households and CHP and heating plants (Metla 2009b). Table 2.1.7. The consumption of wood fuels in Finland in 2008 measured in volumes and energy content. Waste liquors are mainly black liquor and other by-products include e.g. tall and birch oil and methanol. Other contains recovered wood, pellets and briquettes and other residues. Energy distribution between fuelwood and other wood fuels is partly estimated. Value of small-scale forest chips use is based on study on 2007/2008 heating season (Metla 2009a; Metla 2009b; Metla 2009c). Forest chips Wooden by-products Other by-products Fuelwood Other Total Industry Small- scale Chips Dusts Bark Waste liquors Other Million m3 4.03 0.60 0.76 1.61 7.09 – – 5.40 2.15 21.6 PJ 28.9 4.3 5.5 11.9 45.8 144.0 4.0 38.9 15.5 299 2000 2002 2004 2006 2008 0 1 2 3 4 5 Total use Use by CHP plants Use by heating plants million m3 Working Papers of the Finnish Forest Research Institute 146 http://www.metla.fi/julkaisut/workingpapers/2010/mwp146.htm 17 The use in heating plants has been also rapid (almost 19 % a year), but the use is in smaller scale. In year 2007 forest chip use in CHP plants decreased notably due to the reduction in emissions credit prices. Heating plants are not included in the emissions trade and there the growth has been more steady. The wood fuel prices are site dependent since the markets are local and evolving. There are, however, average estimates for the prices paid by heating and power plants. In 2006 the average price of forest chips was 11.95 euro/MWh. The industrial residues were cheaper, with price of 9.7 euro/MWh for industrial chips and dusts and 9.2 euro/MWh for bark (Metla 2007a). The prices have been increasing for several years and the rise is expected to continue as increased demand leads into the utilization of high cost wood sources. In 2009 forest chips price was 16.8 euro/MWh (Statistics Finland 2009b). 2.1.4 Conclusions and future outlook The discussion in chapter 2.1 has shown that woodfiber consumption and production structures in Finland are heterogeneous. There are many users of wood, and the supply of woodfiber originates from a number of different sources. However, the relative importance of different actors in the wood market varies considerably. The forest industry is the dominant user of wood. In 2007, out of the total wood consumption, 93% was due to forest industry demand, 6% due to small-sized dwellings, and 1 % by heating and power plants. Given that the share of exports in the Finland’s total production of paper is around 90 %, and for sawnwood about 60%, the Finnish forest products and roundwood markets are heavily dependent on EU and global forest products markets. The business cycles and structural changes in global forest products markets will directly influence the harvesting activity and roundwood production in Finland. The overall picture of wood consumption pattern obscures the fact that a significant part of the wood consumed by forest industry ends up in combustion. Figure 2.1.4 shows that almost half of the woodfiber consumed annually in Finland is used for combustion. A significant part of this combustion is related to forest industry processes. In chemical pulp production, the bark and the black liquor, or about 45% of the woodfiber coming to the pulp mill, is used for energy generation. In sawnwood industry, the bark and roughly a third of the sawdust ends into energy production. In mechanical pulp process, basically only the bark is used for energy generation. As a result of this close interconnection of forest industry output and woodfiber based energy generation, the level of forest products output is a significant determinant of woodfiber energy use. Therefore business cycles and the long term structural changes in the forest industry markets, have direct implications to the short-term and long-term consumption of woodfiber for energy purposes in Finland. This is one important feature indicating the need to analyze the forest and energy sectors simultaneously. The links between forest products outputs and woodfiber based energy production take place at many different levels, and at least from the policy analysis point of view, are also complex. First, in the demand side, there are large number of processes and users which consume woodfiber for energy generations. For example, industry process heat and power, municipal district heating plants, smaller scale heating and power industry, pellet industry, households, as well as exports of woodfiber for energy purposes. Similarly, the sources (supply) of woodfiber for energy purposes Working Papers of the Finnish Forest Research Institute 146 http://www.metla.fi/julkaisut/workingpapers/2010/mwp146.htm 18 are manifold. It may originate e.g. as forest chips and stumps from clear-cutting sites, bark, wood products industry logging chips or sawdust, black liquor, thinnings, or pulpwood harvests. What complicates the picture even more, is the fact that some of these woodfibers can be substitutes or complements to other raw materials in energy generation, such as for oil, coal, natural gas, or peat. Indeed, the co-firing of woodfiber and peat is of significant importance in Finland. If policy or markets affect any of the woodfiber demand or supply side factors, this is likely to have feed-back impacts on the other components. For example, a feed-in-tariff for wood energy is, ceteris paribus, likely to have impact on roundwood prices, as well as the price of and demand for peat. On the other hand, a demand shock to the paper or sawnwood markets would have direct implications to the level of forest products output, and therefore, also to the amount of woodfiber consumed in Finland. This would in turn affect the roundwood markets, as well as the possibilities in generating woodfiber based energy in Finland. Therefore, in order, to be able to fully assess the policy or market impacts, one would need to model all the possible linkages between the different factors and sectors. Modeling the different feed-back linkages between forest and energy sector is a very demanding task, which however, appears to be coming ever more important. For example, the need to regulate CO2 emissions and the objective to increase renewable energy production will increase the importance of this in the coming decades. Also, the concerns related to oil prices and energy security, drive the development towards increasing the role of forest biomass energy. Evidently, also new polices which enhance the forest and energy sector linkages will be put in place in the coming years. In order to illustrate what type of implications could result to Finnish forest and energy sectors, we take up few possible examples of the foreseeable development. This will also help to point out some of the key features that a policy relevant simulation model of the Finnish forest and energy sector needs to incorporate. Combustion (= energy) 46% Other 7% Pulp exports 5% Ply- wood 4% Sawnwood 14% Paper & paperboard 24% Figure 2.1.4. The wood consumption in Fin- land in 2004 (Data Source: Statistics Finland). Working Papers of the Finnish Forest Research Institute 146 http://www.metla.fi/julkaisut/workingpapers/2010/mwp146.htm 19 Future outlook Finland’s forest sector is currently undergoing a structural change. The largest impacts of the structural change are probably still to be seen in the coming decade or so. Some of the essential features of this structural change are the following: – Forest industry is cutting its capacity in Finland, and redirecting its investments mainly to those markets where the product consumption is still increasing, or where the wood fiber supply is based on fast gro- wing plantation forestry. Basically, the industry is continuously looking at the most competitive invest- ment cites globally, and less so in Finland. As a result, the pulp and paper and sawnwood output level is most likely to decline in Finland in the coming decades. – Forest industry, energy industry and investment companies are developing new forest based energy and chemical products, some of which may be available at commercial scale in the coming 3-5 years. For example, pulp and paper industry and sawnwood industry are likely to increase significantly their energy related operations in the coming decade. However, the commercialization of some new products may take a considerably longer time. – Energy companies are increasing district heating and electricity production based on forest biomass. – The wood consumption allocation for different purposes, shown in Figure 2.1.4, is likely to change significantly due to the above developments. In particular, the importance of combustion is increasing, while that of pulp and paper and sawnwood is declining. – The attempts to cut back CO2 emissions will most probably lead to forest playing more important role in climate policy. For example, forest owners may earn monetary benefits for providing climate serv- ices through their forests. – Attitudes and values related to forests are evolving. For example, the private forest owners that own most of the forests in Finland, are being urbanized, and to some extent may become less production minded. – All of the above changes will most likely affect e.g. forest, energy, environmental and economic poli- cies. For those planning the new polices, there is a great need for information about the likely impacts of the different polices that could be put in place. It is the last point, which is one of the major motivations for the current work. 2.2 Timber supply 2.2.1 Forests in Finland Forestry land covers about 68% of the total land area, i.e. 26.3 million ha, including all land which is not classified as agricultural land or built-up areas. Finland’s forests are almost entirely in the boreal coniferous forest zone. The most common species in Finnish forests are Scots pine (Pinus sylvestris L.), Norway Spruce (Picea abies) and birch (Betula spp.)(Metla 2007a). Forestry land is divided into four categories upon the growth conditions of trees. The forest land is the most important category covering 20.1 million ha. The scrub lands cover 2.8 million ha. Bald or near bald areas are categorized as waste lands (3.2 million ha) and to the other forestry land (0.2 million ha) which consist of for example forest roads. 17,1 million ha of the total forestry area in the three first categories are on the mineral soil sites and 9 million on the mires. Forestry land also includes forestry lands which are out of wood production as nature conservation areas. Working Papers of the Finnish Forest Research Institute 146 http://www.metla.fi/julkaisut/workingpapers/2010/mwp146.htm 20 For the purposes of this report and the modeling work, we will focus mostly on the forestry land which is available for wood production (FAWP) and is classifield to the forest land (19.1 million ha) or to the scrub lands (2.0 million ha). The forestry area available for wood production in total is 21.0 million ha. The total size of the Finnish forestry land has been almost invariant since Word War II. However, between the 1950s and 1980s the forest land increased by about 15% and both scrub and waste land areas decreased by about 30 %. Since the 1980s the forestry land measured as forest land has been almost invariant and the changes in scrub and waste lands have been small. The FAWP area is mainly owned by nonindustrial private forest (NIPF) owners (59 %). The state owned area covers 26 % of FAWP, the remaining area belongs to companies and institutions such as local communities and parishes. The state owned forestry areas are typically in the northern Finland especially the nature conservation and wilderness areas. Therefore the state owned forestry areas have on the average smaller growing stock volume and the yearly increment of the growing stock is also below the forestry land owned by others. The growing stock on forest land and scrub lands is in total 2189 million m3 and the growing stock on FAWP lands is 2054 million m3. The average growing stock in Finland is nowadays 105 m3/ ha, and it varies between 160 m3/ha in the South-coast to 64 m3/ha in Lapland. The average stock typically decreases when going from South to North in Finland. The annual addition to the of growing stock has been increasing strongly since 1970s. In the latest National forest inventory the annual growth of the stock has been estimated as 99 million m3, of which the annual increment on FAWP lands part was 96 million m3. The average annual growth of trees in Finland on the forest land is 4.8 m3/ha and 4.3 m3/ha if the scrub lands are also included. On the forest land the difference on the annual increment of growing stock is minor between mineral soils and mires 4.9 m3/ha and 4.6 m3/ha respectively. Taking into account also scrub lands the average annual growth of trees in the mineral soils is 4.6 m3/ha /a and in the mires 3.5 m3/ha /a. The regional differences on the annual increment of growing stock in forest land differ between the Häme-Uusimaa region’s 7.4 m3/ha/a and Lapland’s 2.3 m3/ha/a. The total drain of growing stock in Finland have been 65–70 million m3/a in recent years which is about 70% of the annual increment of the growing stock. The total drain consists of roundwood harvests, logging residues and the naturally drained trees in the forests. In recent years, the growing stock in the Finnish forests has been incresing about 30 million m3/a (Korhonen et al., 2007). National forest inventories (NFI) have been carried out in Finland since 1920s to collect national and regional data of forest resources; for example volumes, growth, health, mean diameter of trees, and the of ownerships of forestry lands in Finland. The field measurements of the latest NFI10 started in 2004. Recent inventories are based on systematic cluster sampling and field measurements. Since the NFI9 (1996–2003) persistent sample plots have also been used and will be measured again in the coming inventory. About 80 000 sample plots were measured in the NFI9. One fourth of those plots is stated as permanent. The results of NFI give reliable estimates for areas over 200 000 hectares, as the size of Finnish Forestry Centres (Korhonen et al. 2007, Metla 2007b). Working Papers of the Finnish Forest Research Institute 146 http://www.metla.fi/julkaisut/workingpapers/2010/mwp146.htm 21 2.2.2 The ownership of forests and timber supply in Finland The forestry land area in Finland is maily owned by non industrial private owners and the state, as shown in the Figure 2.2.1. The role of NIPF owner’s is even higher in the timber supply while they owns 59 % of the forestry land area available for wood production. Also their round wood harvest covers over 80 % of the total roundwood removals in Finland in last 10 year, as presented in Figure 2.2.2. The significance of state forest is opposite in the timber supply because state forest’s are mainly in the Northen Finland and due the conservation areas. In the Appendix Table A.2.2.1 is presented ownership of forestry land at regionally at Forestry Center level. The Finnish Forestry Centres are regional administration organisations, which are controlled by the Finnish Ministry of Agriculture and Forestry. The function of the 13 Forestry Centres is to provide forest owners with information about forestry practices and their environmental impacts. The Forestry Centres also carry out administrative regulation based on the Forest Law. The statistics of Finnish forestry are usually collected and published at Forestry Centre level. The regional timber supply is presented in the Table A.2.2.5. Figure 2.2.2. Roundwood removals by forest owner in Finland 1998-2008 (Metsähallitus has been inclu- ded to forest industy since 2008) (Metla 2009a). Figure 2.2.1. The ownership of forestry land in Finland at in NFI9 (Metla 2005). State 35% Others 5% NIPF 52% Forest industry 8% 1998 2000 2002 2004 2006 2008 0 10 20 30 40 50 60 Total NIPF Metsähallitus+FI Forest Industry (FI) Metsähallitus (MH) million m3 Working Papers of the Finnish Forest Research Institute 146 http://www.metla.fi/julkaisut/workingpapers/2010/mwp146.htm 22 Non industrial private forest owners The non industrial private forest owners’ area is 13.8 million ha covering 52 % of the forestry land area and 59 % of the forestry land area available for wood production. The roundwood removals from those forest has been between 45 to 55 millions m3 during the last 10 years. As presented in Figure 2.2.3 total removals has been slightly declining due the decrease of log removals. Pulp wood removals have been quite stable. the fuel wood removals have been increasing and the share of that have increased from 9 % to 13 % of the total harvesting volume. Studying the NIPF owners’ objectives in forestry, Kuuluvainen et al. (1996) found four categories among the Finnish NIPF owners. According to that study, over half of the private forest land is owned by multiobjective owners and recreationists gaining non-monetary benefits and values from their forests. The other categories, the investors and self-employers were considered to receive mainly monetary values from their forests. Table 2.2.1 presents the share of land area and the shares of forest owners in each category. Favada et al. (2007) found five categories to classify NIPF owners’ objectives. Four of them correspond to Kuuluvainen et. al. study. The fifth group, called indifferent owners, was formed by forest owners who could not specify any specific objectives in forestry. Figure 2.2.3. Roundwood removals from non industrial private forests in Finland 1998–2008 (Metla 2009a). Table 2.2.1. The distribution of Finnish forest area and amount of owners by the objective of forestry (Kuuluvainen et al. 1996). Multiobjective owners (%) Recreationists (%) Self-employed owners (%) Investors (%) Share of forest land area 33 21 31 14 Share of owners 26 31 30 13 1998 2000 2002 2004 2006 2008 0 10 20 30 40 50 60 Total Logs Pulpwood Fuelwood million m3 Working Papers of the Finnish Forest Research Institute 146 http://www.metla.fi/julkaisut/workingpapers/2010/mwp146.htm 23 The Forest and Park Service, Metsähallitus Metsähallitus is a state-owned enterprise that controls about 9 million ha of land area and 3.5 million ha of water area in Finland. For the wood production Metsähallitus has about 3.5 million ha of forest land and 1.5 million ha of waste- and scrub lands. The rest, about 4 million ha land area consists of e.g. conservation areas and wilderness areas. In the statistics of Finnish forestry Metsähallitus have been included to group industry since 2008 (Metla 2009a). Regionally forests owned by Metsähallitus are emphased strongly to Northen Finland. Regional roundwood removals from Metsähallitus forests in 2007 are shown in the Table A.2.2.3. The Northen-Finland Forest Centers Kainuu, Pohjois-Pohjanmaa and Lappi have the main role in the Metsähallitus timber removals covering about 70 % of the total Metsähallitus harvests. The harvests from Metsähallitus forests consist mostly of pulpwood which cover over 60 % of total their harvests, and in Lappi Forest Centre the share is even 76 %. Forest industries The forest industries’ forests account for about 8 % of the forestry land area in Finland. The largest forestry areas owned by forest industries are in the Keski-Suomi, Pohjois-Savo, Pohjois-Karjala and Kainuu regions. Table A.2.2.4 presents commercial roundwood removals from forest industries’ forests by Forestry Centres in 2007. The share of log removals was about 43 % and pulpwood removals about 57 % of the total forest industries removals in 2007. In the Rannikko, Pohjois- Pohjanmaa and Lappi Forestry Centres the share of log removals are lowest, covering about 29 %, 28 % , 25 % of the total removals in those areas respectively, but the total harvests in those areas covers only 4 % of the removals from forest industry forests. The most important regions for the forest industries harvests are Keski-Suomi, Pohjois-Savo and Pohjois-Karjala, whose total share is about 50 % of the total harvests in the industry owned forests in 2007. Other owners The other forest owners consists on e.g. municipalities, parishes, Finnish Forest Research Institute and jointly owned forests, various communities and other companies than forest industry. Municipalities, parishes and common forests are the most important owners in this owner class, owning 425 000, 168 600, 520 000 ha forestry area respectively (Kirkkohallitus, 2007; Metla 2007a; Metsätalouden kehittämiskeskus Tapio, 2006). Altought the recreation and outdoor activities are an important use of the municipality forests, more than half of the municipality forests were classified as timber producing forests in a study, which covered 390 of the 432 Finnish municipalities (Monimuotoisuuden turvaaminen..., 2006). In the parish-owned forests, the total income from forestry was 16,6 million € in 2006. In the statics of timber supply in Finland are municipality, parishes, other societies owned forest and those state owned forest which are not owned by Metsähallitus included in the non industrial private owned forest statistics. Wood imports Finland used about 20 million m3 of imported roundwood in 2008 (see Table 2.2.2). In 2005 Finland was world’s third biggest roundwood importer after China and Japan. Roundwood is imported mainly for Eastern Finland’s pulp industry. The amounts of imported softwood and hardwood are in the same magnitude. Birch, spruce and pine are the major imported wood species. Russia is by far the biggest roundwood exporter in the world. Between 2005 and 2008 Over 70 % of the Finnish roundwood imports come from Russia. The imported amounts of roundwood from Russia declined in 2008 and they are expected to decline more, because of the uncertainties Working Papers of the Finnish Forest Research Institute 146 http://www.metla.fi/julkaisut/workingpapers/2010/mwp146.htm 24 concerning the export duty increases by the Russian authorities. As a WTO member Russia could perhaps become a less risky source of wood imports into Finland. Besides Russia, the other imported wood into Finland (2005-2008) originated from Latvia, Sweden and Estonia, with shares of 8 %, 6 % and 6 % respectively (Metla 2009a). Energywood supply An overall objective exists to increase the use of bioenergy in Finland, to mitigate climate change and to increase the self-sufficiency of energy production. The main raw material of used wood based bioenergy in Finland is the proseccing residuals from forest industry. The roundwood supply to energywood use was about 10,9 million m3in 2008 of which about 60 % was firewood and the share of forest chips was about 40 % (Metinfo 2009, Mäkelä 2009). Table 2.2.3 presents the shares of sources of energywood in Finland at 2006. In the tending of seedling stands and the first thinning the share of energywood potential of the total wood supply of forest is notably higher than in the later thinnings and final fellings. Kiema et al. (2005) estimated energywood potential in some central Finland communities. According to them, the energywood potential from the final fellings was about 60 m3/ha to 80 m3/ha and from the logging residues and stumps about 50 m3/ha. Table 2.2.2. Imports of roundwood and wood residues in 2008 in million m3 (Metla 2009a). Roundwood imports Softwood 6.6 Pine 3.4 Spruce 3.2 Hardwood 8.7 Birch 8.2 Other 0.6 Fuelwood 0.3 Chips 4.1 Wood residues 0.5 Total 20.2 Table 2.2.3. The energywood (roundwood) consumption shares in Finland (%). Firewood (2000–2001) Forest chips (2006) Total Total Heat and power plants Small properties Total Final fellings Logging residues** 0–50 64 25 57 25–55 Stump 0 14 0 12 5 Thinnings Logs size trees 0–50 – – – 0–30 Small size trees 50 20 75 26 40 Import 0 3 0 3 * Small properties uses also about 1 million m3 of waste wood as firewood ** Including harvested roundwood which are not used as market roundwood (Sevola et. al. 2003, Ylitalo 2006, Ylitalo 2007, Laitila et al.2008 ) Working Papers of the Finnish Forest Research Institute 146 http://www.metla.fi/julkaisut/workingpapers/2010/mwp146.htm 25 Hakkila (2004) presents the energywood potential of pine and spruce logging residues in the first and second thinnings and final fellings. The energy wood potential is typically higher (or the share of energywood potential to industrial roundwood is higher) in the spruce forests than in the pine forests. In the table 2.2.4 is an example of the energywood supply from thinnings and final felling in Southern Finland. The share of residual roundwood in pine and spuce forests is about 25 % of the roundwood removals in the first thinnings. In the second thinning the share is about 15 % and in the final felling only about 5 % of the total removals are residual roundwood. The dry mass of crown residuals consists mainly of living branches. The total volume of crown residuals is about 20 % of roundwood removals in final fellings and second thinning in pine forests. In the first thinning of pine forest the crown residuals make about 35 % of the total volume of round wood. In the spruce forest this share is about 50 %. The relation of dry weight and energy content of pine and spruce stumps increases with the dimension of stump. For example when the dimension of spruce stump increases from 20 cm to 40 cm, the energy content increases about 300 % ( Hakkila 2004). Several studies have estimated the future use and and supply potentials of energywood at the local or at the state level in Finland. Pöyry (2007) evaluates that the theorethical potential of the energywood supply in Finland in 2020 is 26.3 million m3. The shares of stumps and small size trees are notable higher in the theoretic potential than in the energywood consumption shares in Finland at 2006 (see Table 2.2.3 and 2.2.5). Hakkila (2004) estimated that the annual supply potential in the Finnish forest is 15 million m3 energywood. The main potential of energywood orginates from the final fellings, which consist of logging residues and stumps 8 mill. m3 and 2 mill. m3 respectively. The potential of energywood from young forest felling was 5 mill. m3. The potential of energywood from the later thinning is observed to the zero, because it is assumed that energywood inflict the growth of forests so that it is not economical to collect residues. The total energywood harvesting potential is supposed to be 33 % from the total energywood potential of forests fellings. Table 2.2.5 presents the theoretical potential of energywood supply in Finland in 2020. The techno- economic potential is lower than the theoretical potential, due to the transportation costs and higher costs to collect energy wood from smaller diameter forests . Pöyry (2007) calculated techno- economic potential of energy wood supply in Finland, which is presented in Table 2.2.6. The techno-economic potential is totally only 38 % of the theoretical potential. Table 2.2.4. An example potential of forests energywood (roundwood) supply in Southern Finland based on Hakkila (2004) study. Fellings Age of forest Industrial wood (m3/ha) Energywood (m3/ha) Share of energywood (%) Min, Min-Max, Max Totally Maintenance of seedlings 10–20 – 15–50 100 100 1. thinning 25–40 30–80 30–50 50–38 28–63 2. thinning 40–60 50–90 20–40 29–31 18–44 3. thinning 50–70 60–100 20–40 25–29 17–40 Final felling 70–100 220–330 70–130 24–28 18–37 Total 360–600 155–310 28–34 19–46 Working Papers of the Finnish Forest Research Institute 146 http://www.metla.fi/julkaisut/workingpapers/2010/mwp146.htm 26 Total wood supply The total supply of wood was 83.3 million m3 in 2008 of which domestic supply covered 63.2 million m3 and imported wood 20.2 million m3 as shown in Table 2.2.7. In Appendix Table A.2.2.5 is presented total wood supply by forestry centres. Multiple-use forestry The most significant economic values of Finnish forest orginates from wood production. Forests provide also possibility to carry out other commercial or recreational activities such as hiking, camping, hunting, or picking of berries and lichen. The valuation of those activities are estimated at a coarse level and the values are not always commeasurable (Metla 2007a). Energywood harvesting and collecting logging residues decreases the amount of decayed wood in the forests. The increased drive on foresry machines due the energy wood harvesting and collecting logging residues affects negatively to habitat of saproxylic species (Siitonen 2008). Table 2.2.5. The theoretical potential of energywood supply by categories in Finland in 2020 (Pöyry 2007).* Felling residuals Stumps Small size tree Total Volume (million m3) 8.4 9.3 8.6 26.3 Share (%) 32 35 33 100 * Assumed that energy content of energywood is 2 MWh/ m3 Table 2.2.6. The techno-economic potential of energywood supply in forestry centers in 2020 (Pöyry 2007).* Felling residuals Stumps Small size tree Total Volume (million m3) 3 905 2 670 3 515 10 090 Share (%) 39 26 35 100 Of theoretical potential (%) 47 29 40 38 * Assumed that energy content of energywood is 2 MWh/ m3 Table 2.2.7. Total wood supply in 2008 in Finland (million m3) (Metla 2009a, Mäkelä 2009). Domestic Import Total Logs 22.3 1.8 24.2 Pine 10.2 0.6 10.8 Spruce 11.0 0.7 11.7 Hardwood 1.2 0.6 1.8 Pulpwood 30.2 13.5 43.7 Pine 14.6 2.9 17.5 Spruce 8.2 2.5 10.7 Hardwood 7.3 8.0 15.4 Fuelwood 6.6 6.6 Pine 1.6 1.6 Spruce 1.4 1.4 Hardwood 3.6 3.6 Other 4.6 4.6 Energywood 4.0 2.8* 4.3 Total 63.2 20.2 83.3 * Imported fuelwood Working Papers of the Finnish Forest Research Institute 146 http://www.metla.fi/julkaisut/workingpapers/2010/mwp146.htm 27 3 Models of the forest sector 3.1 Demand models The demand of wood consists of two primary uses, namely raw material use in forest industries, and fuel use in various energy producing industries. Therefore, the modelling of wood demand needs representations of wood product, pulp and paper and energy industries and their markets. This has been done in various ways, and various approaches have been used in previous models. In this section, we present some basic features in sectoral modelling, and how the issues analyzed are addressed in sectoral models. 3.1.1 Objectives of the demand models There exists a large number of models that describe forest and energy industries. These models vary widely with their modelling philosophy, but the objectives are typically the same. The main objective of these models is to give an estimate of how industries will adjust to policy and other economic shocks. In the short run, the adjustments are changes in input uses as the composition of inputs is changed and activity levels are adjusted for the new situation. In the long run the adjustments are done by altering the production technology by adaptation of new technologies. All these changes are costly and the models are used to assess the levels of these costs. During the recent years the climate change mitigation has been the driving force of energy policies. These policies target to the reduction of greenhouse gases (especially CO2), by decreasing the use of fossil fuels and encouraging the use of renewable energy and energy saving. The rising interest in renewable energy has increased its importance in energy sector models. Although forest industry has always been important factor in energy markets in Finland, the connection of forest and energy sectors is expected to tighten even more. This raises the need for joint modelling of energy and forest sectors with significant weight on wood-based energy issues. The objective of such a model is to present the effects of climate policies on forest and energy sectors: how the demand for wood changes, which industries benefit and which suffer and what are the costs of these adjustments. 3.1.2 Dichotomy of the models The objective of the modelling is to analyze adjustments originating from policy and other shocks. Therefore, it is of utmost importance to model the adjustment mechanisms of the studied industries as correctly as possible. In the short-run, the physical capital of the industry is invariant and the adjustments occur through other, variable, inputs. In the long-run, it is possible to invest in new technology (and disinvest old). This technological change represents an alteration of physical capital and allows for better adjustment to policies as the production possibilities change. There are two prominent methods for modelling the long-run technological change. The first is a traditional economic approach which uses abstract production functions that define the limits of production, given a set of input variables such as labour and capital. Technological change is modelled via substitution between inputs and a change of productivity parameters. The second method is an engineering approach of activity analysis. The production functions are a set of Leontief production technologies with constant coefficients that relate inputs to output. Technological change is modelled as mobilization of new production technologies through investments (Löschel 2002). These two modelling methods are called top-down (TD) and bottom-up (BU), respectively (e.g. Wene 1996, Hourcade et al. 2006). Working Papers of the Finnish Forest Research Institute 146 http://www.metla.fi/julkaisut/workingpapers/2010/mwp146.htm 28 The two approaches are very different and it is no wonder that the results of the models vary significantly (Grubb et al. 1993). Bataille (2005) and Hourcade et al. (2006) analyze the problem by defining three properties that a good policy model should have. The properties are technological explicitness, macroeconomic realism and microeconomic realism. Technological explicitness is self-explanatory and means proper description of production technologies. Macroeconomic realism consists of feedbacks between sectors and economy as a whole including growth and trade aspects. Microeconomic realism means realistic assumptions of production and investment behaviour. Conventional TD models have been strong in macro and micro realism while BU models have focused on technologies. Both methods have their pros and cons, but they are both currently in use. However, there has been a clear convergence of two modeling traditions and attempts have been made to combine the good properties of the two approaches. To differentiate these new models from conventional TD and BU models they are called hybrid models. In the rest of the section we illustrate pros and cons of the two modelling traditions and represent some of the hybrid models. Our presentation draws much on the useful discussions of these modelling traditions written by Bataille (2005) and Hourcade et al. (2006). Top-Down models Conventional TD models are economy-oriented and they are typically computational general equilibrium (CGE) models, yet some non-equilibrium models exist (Löschel 2002, Hourcade et al. 2006). In general, the focus is on economic relations instead of technological accuracy. Production technologies are aggregated into production functions, which are often assumed to be in constant elasticity of substitution (CES) form. As the name suggests, in CES production function the substitution of inputs is modelled via constant and exogenous elasticity of substitution parameters (ESUB). These parameters control the possibilities of input substitution as price-ratios change. Technical change is mostly assumed to be exogenous. The change is modelled through changes of an exogenously changing productivity parameter. For example in energy models this parameter is autonomous energy efficiency index (AEEI). ESUB parameters and the time trends of AEEI are estimated econometrically from past data or they are ‘guesstimated’ based on expert opinion. Therefore, their use in future projections can be unsound. This is partly because the estimation of AEEI is difficult as it includes only autonomous technological change (it is separated from changes due to the price-ratios variation) and because this kind of analysis does not take into account the induced technological change issues (e.g. Goulder and Schneider 1999, Goulder 2004). Policies affect price-ratios of the inputs. Adjustments are done by substitution between inputs, including capital. This is adjustment under given technological constraints, i.e. existing production technologies. Since in the CGE models the economy is efficient, there is no possibilities for Pareto improvements. Therefore, adjustments cause improvements in one sector but impairment in another. This causes costs and there are no “no-regrets” efficiencies available (Battaille 2005). Technological change of increasing efficiency parameters alleviates these costs. The results depend heavily on assumptions about technical change. Most of the TD models have exogenous technical change but there are numerous possibilities how to model it. For example, increases in efficiency parameters can be augmented by capital vintaging where different vintages of capital have different substitutability properties (e.g. Jacoby and Sue Wing 1999, Babiker et al. 2001). There are also TD models with endogenous technical change (see Working Papers of the Finnish Forest Research Institute 146 http://www.metla.fi/julkaisut/workingpapers/2010/mwp146.htm 29 Gillingham et al. 2008). Models with endogenous technological change are more complicated than the models with exogenous technological change they replace (Edmonds et al. 2000). Bottom-up models BU models are activity analysis models typically based on linear programming (e.g. Nyboer 1997). The models have a set of industry or sector level Leontief production technologies which are used in demand satisfying production. Demand can be inelastic or price dependent. In the former case the technologies are chosen by cost minimization criteria and in the latter by surplus maximizing criteria. The economic use of these methods is based on the papers by Hotelling (1938) and Samuelson (1952). Since the technologies are specified separately they can be presented rather realistically. Production possibilities are defined via technology snapshots that evolve through time. Technological change is based on these changes of snapshots when a new technology replaces an old one (Löschel 2002). Main criteria for the investing decision are the life-cycle costs and the performance characteristics of the technology. Technological change is quite rapid in BU models compared to the TD models. Grubb et al. (1995) see that technological change is fast in BU models because the engineering data includes future technologies, which causes optimistic implementation of new technologies. Note also that the guesstimation of parameters of future technologies is prone to an error so the long-run predictions may be unsound in BU modelling. While the description of the current technologies is accurate, there lies problems in BU models also. The models are technologically focused and the markets under study lack proper connection and feedback to the rest of the economy (e.g. Battaille 2005). Jacoby (1998) states three pitfalls of BU modelling: Confusion of market failures with other market barriers; Lack of attention to market structure; Failure to account for inter-market adjustments. These pitfalls, if ignored, may lead to a too fast technology adaptation and the lack of inter-market adjustments may cause faulty predictions if the policy changes are significant. The first two issues are related to the problems in microeconomic and the third in macroeconomic realism. Hybrid models Researchers have developed a new line of so called hybrid models that try to combine the good properties of the two modelling traditions. Hourcade et al. (2006) define hybrids as models that have at least one clear modification that removes them from traditional classifications. So the hybrids are TD-based models with more accurate technology description or BU-based models with increases in micro and macro realism. In Yatchew (2006) there is a good presentation of some of the recent hybrid models with various approaches. Böhringer and Rutherford (2008) see that there are three alternative routes to construct a hybrid model: (a) Combining existing top-down and bottom-up models with a recent example by Schäfer and Jacoby (2006); (b) Modeling thoroughly the other aspect and using a reduced form of the other (e.g. Manne et al. 1995); (c) Use of an integrated model and using mixed complementarity formulation for the CGE model (Böhringer 1998). The details of these approaches vary and technological explicitness, macroeconomic and microeconomic realism have been combined in various ways. In the examples subsection we present some of the hybrid models in more detail. Working Papers of the Finnish Forest Research Institute 146 http://www.metla.fi/julkaisut/workingpapers/2010/mwp146.htm 30 3.1.3 Policy cost analysis In the short-run, new physical capital cannot be built. Instead, adjustments are done through substitution of variable inputs. Utilization rates are also altered between sectors and technologies as some plants are shut down and others are put into operation. In the medium-term perspective the capital can also adjust to the changing price-ratios. New capital investments allow for increasing the substitutability of the inputs as well as increase efficiency. In medium-term perspective it is possible to name the technologies that can come to commercial use in that period. This allows the modeller to assume a set of BU technologies that are relevant for the time-period or to guesstimate the effects of these incoming technologies on elasticities of substitution as well as on efficiency factors. Also the use of history based econometric estimates can be justified. In the long-run, however, the technological change cannot be predicted very accurately. In fact many of technologies used in the distant future cannot be predicted at all. The costs and other properties of the future technologies are only very inaccurately, if at all predictable. There may also be some backstop technologies whose properties are developed beyond our imagination. Therefore, explicit technologies of BU models are not accurate in long-term calculations. Similarly, econometric methods cannot foretell the future advances of production technologies. Some kind of abstraction and projections are needed and in this respect the models of endogenous technical change are valued (e.g. Romer 1990, Löschel 2002). The assessment of the costs associated to above-mentioned adjustments is a major question of sectoral economic modeling. Adjustment costs are calculated comparing policy scenarios with a baseline scenario. For BU models these costs are mainly financial costs of production or total surplus changes in every market. TD models capture better the intangible costs such as option value and utility changes (see e.g. Jaccard et al. 2003). General equilibrium models also present all the costs in the economy while in partial equilibrium the costs are related on markets under study. Based on these observations Hourcade and Robinson (1996) have grouped the costs in four broad categories: engineering and financial costs, sectoral costs and macroeconomic costs (see also Edmods et al. 2000). The first two costs are observed in assessing technologies whereas the sectoral costs aggregate these costs taking into account the costs occurring within the whole sector. When inter-sectoral adjustments are taken into account one ends up with macroeconomic costs. 3.1.4 Examples of energy sector models Many of the recent energy sector models are coupled with environmental issues. Therefore, some of the models presented here might be called as energy-environment models. In this section we present examples of models belonging to the different modelling categories. MIT-EPPA (MIT – Emissions Prediction and Policy Analysis) is a top-down energy-economy model that is a component of Integrated Global Simulation Model (IGSM).4 It is based on earlier OECD GREEN model (Burniaux et al. 1992). MIT-EPPA presents a good example of computational general equilibrium model that utilizes concepts presented in previous sections. The firms have mixed nested CES and Leontief production technologies that exhibit constant returns to scale. Technology mix differs between production sectors. The consumer utility is also represented by nested CES function. Therefore adjustments to shocks depend heavily on elasticity of substitution parameters. Inter-temporal decisions are made recursively period by period based on current values 4IGSM is a model used for climate change assessment, see Prinn et al. (1999). Discussion here covers version 3.0 of EPPA. Working Papers of the Finnish Forest Research Institute 146 http://www.metla.fi/julkaisut/workingpapers/2010/mwp146.htm 31 of variables (recursive dynamics). Therefore MIT-EPPA differs from truly dynamic forward-looking inter-temporal optimization models. Growth is based on accumulation of capital and technological change of which energy efficiency improvements are modelled via AEEI index. Accumulation of capital is divided into malleable and non-malleable parts meaning that new investments are malleable in a sense that they can be adjusted to current price ratios using CES structure. Older capital is non-malleable and has to be used in Leontief production with input shares determined by the period it became frozen. Therefore, there are several vintages of capital whose properties depend on previous periods’ economic setting (Babiker et al. 2002, Jacoby and Sue Wing 1999, Jacoby et al. 2004). A particularly good example of a bottom-up model is ETSAP-TIAM (ETSAP – Times Integrated Assesment Model). It is a global incarnation of TIMES model, which is a descendant of MARKAL and EFOM models. TIMES model is a partial equilibrium model with an extensive description of energy forms and process technologies covering both current and future technologies. The model solves equilibria of different markets by matching supply and demand. Endogenous supply is modelled via cost minimization while demand for a good is either endogenous or exogenous with constant elasticity structure. Under perfect competition and sufficient market independence this corresponds to the maximization of the joint consumer and producer surplus. The model contains multiple time periods ranging over 100 years. In basic setup inter-temporal decisions are made with perfect foresight as the agents observe future values perfectly at the decision time-period. However, it is possible to relax this assumption by modelling the decision making under uncertainities about the future by stochastic programming. Since the structure of the model is large there is a need for linearization. This allows the use of powerful linear programming tools. However, for example the production functions are typically non-linear but they are modelled in piecewise linear manner. While TIMES is a BU model, some of the macroeconomic effects are captured with demand relations that depend on macroeconomic parameters such as the GDP and the size of population. Technological change is represented by introduction of novel technologies and cost decreases due to the learning-by-doing (Loulou and Labriet 2008, Loulou 2008, Loulou and Lehtilä, 2007). As said, there are several ways to hybridize a model. Here we introduce briefly couple of models based on different approaches. CIMS (originally Canadian integrated modelling system) presents an example of a bottom-up model that has been augmented to better comply with the micro and macroeconomic realism. Especially the microeconomic realism of investment decisions is improved. This is done by simulating technological competition via logistic function. Also the life-cycle-costs of investments have been formulated with addition of non-financial costs and non- financial discounting. Both capital costs and intangible costs decline endogenously as the market share of given technology rises. Macroeconomic realism is added by adding demand modules that have feedbacks including foreign trade and disposable income effects of changing value-added of manufacturing. Iteration is used to even out supply and demand. The model is based on earlier ISTUM model which is technologically accurate bottom-up model (Jaccard et al. 2003, Bataille 2005 and Bataille et al. 2006) In the Second Generation Model (SGM) technological detail is inserted into a large top-down model by nesting the Leontief production functions via logit-function (e.g. Brenkert et al. 2004). Schumacher and Sands (2007) demonstrate the effect of nested Leontief production function in the case of steel industry compared to the usual CES-specification. It is, however, debatable whether SGM is a TD model or a proper hybrid model. The other approach is to include bottom-up technological detail for sectors of interest in a CGE model without any nesting. The technique is based on mixed complementarity problem (MCP) formulation of the general equilibrium problem. Working Papers of the Finnish Forest Research Institute 146 http://www.metla.fi/julkaisut/workingpapers/2010/mwp146.htm 32 This line of research has been promoted by Böhringer (1998) and Böhringer and Rutherford (2008). The method is demonstrated in an empirical application by Böhringer and Löschel (2006). 3.1.5 Examples of forest sector models Buongiorno (1996) classifies the forest sector models according to their general principles. The oldest class of models is based on econometric estimation of relevant market and production relations. This method faces many difficulties, including estimation and data problems as well as issues with technical change. Another method is linear programming approach where the optimal set of activities is calculated. The method is the same as in many of the bottom-up energy sector models. The difficulties of this approach are problems with getting good enough technological data and dynamic behavior that is unrealistically fast to mobilize new technologies. The third principle is to use system dynamics approach, where dynamics are presented with differential equations. The difficulties here are that models are loosely related to the data and differential equations may be invalid. The analysis is done without equilibrium which is problematic in the perspective of economic science. The more novel models try to capture the best parts of these principles. In general, forest sector models are bottom-up models that concentrate on accurate presentation of forest industry processes and giving less emphasis on wood supply considerations. A prominent line of modelling tradition is based on Global Trade Model (GTM) which was developed at IIASA (Kallio et al. 1987). It utilizes all the three principles described earlier. The model has been successful and there are a lot of descendants of GTM globally of which two more modern are EFI-GTM and NTM II (Kallio et al. 2004, Bolkesjø et al. 2006).5 Kallio et al. (2004) gives a good description of the GTM modelling structure and our summary is based on that paper. In the model the joint surplus of consumers and producers is maximized in every market in every region with inter-regional trade. Markets for goods are separate and consumer behavior in every market is based on constant elasticities. Production side, however, is presented with constant input-output coefficient Leontief technologies. Wood supply is in constant elasticity form with data-based growth functions. Optimization in wood supply and investments is myopic as the solution method is recursive programming. Bolkesjø et al. (2006) have included forest based bioenergy products into NTM II model. There are several models that represent Swedish forest sector which is relatively similar to the case of Finland. Nyström (e.g. 2000) has created a model of energy and material flows in the forest industry. The model is based on MARKAL structure with linear programming methods. The model consists of eight paper groups, six pulp mill types and four wood working product groups. Processes have been described in engineering accuracy. Lönnstedt’s (1986) model has a very different approach. The model is a set of equations that govern the relations between economic factors in a way that does not fully have microfoundations. The relations are plausible but functional forms used are partly unmotivated. The structure of the model resembles of system dynamic approach (e.g. Buongiorno 1996). However, it gives a good review of issues a forest sector model should represent. Summary There are several models that describe energy and forest industries with considerable precision. Yet models do not exist that include economically and biologically precise description of forest owner behavior. Hence the study of wood markets is flawed in a sense that supply side of markets 5Kallio et al. (2004) gives a more thorough list of the descendants as well as other forest sector models. Working Papers of the Finnish Forest Research Institute 146 http://www.metla.fi/julkaisut/workingpapers/2010/mwp146.htm 33 is not modelled properly. A model with modeling of forest owner behavior would allow study of policies that affect wood supply decisions and wood markets in general. The properties of existing models give insights on the properties needed for an adequate sector model. It seems that BU modeling of technologies is needed for accurate description of industries. Model with projections up to 30 years into future can be seen in forest and energy sector as a medium-term model. The expected lifetime of technologies is several decades and technologies that are commercially used in 30 years are already under development. Therefore, BU technology sets needed for modeling of technological change are already known and their properties can be predicted with reasonable accuracy. In a model based on inter-temporal decision making the use of BU technologies introduces a risk of unrealistically rapid technological change. This problem needs to be solved in a good model. 3.2 Forest management and forestry models The optimal management of forest, which can be seen as decision making over the alternative management activities, maximizes the benefits of a decision-maker. For comparing forest managements one needs to know the state of the forest and the estimates for the effects of different management activities (Pukkala, 1994). 3.2.1 The optimal management of forest The Faustmann model was the first forest model where the optimal rotation age of an even-aged forest was determined by using the assumption that a forest owner maximizes the net present values of forest. Hartmann (1976) presents a model based on the Faustmann model, where the forest owner benefits consist of both monetary and amenity values from forestry. The optimal management of forest is more complex if thinning