%A"Malinen, H" %D1997 %I; Lappeenranta Univ. of Technology (Finland) %2 %J[] %K29 ENERGY PLANNING AND POLICY, 32 ENERGY CONSERVATION, CONSUMPTION, AND UTILIZATION, PAPER INDUSTRY, POWER DEMAND, ELECTRIC POWER, POWER GENERATION, CARBON DIOXIDE, EMISSION, ENERGY CONSUMPTION, FORECASTING %PMedium: ED; Size: 106 p. %TForecasting energy demand and CO{sub 2}-emissions from energy production in the forest industry %XThe purpose of this study was to develops new energy forecasting methods for the forest industry energy use. The scenarios have been the most commonly used forecasts, but they require a lot of work. The recent scenarios, which are made for the forest industry, give a wide range of results; e.g. from 27,8 TWh to 38 TWh for electricity use in 2010. There is a need for more simple and accurate methods for forecasting. The time scale for the study is from 1975 to 2010, i.e. 36 years. The basic data for the study is collected from time period 1975 - 1995. It includes the wood use, production of main product categories and energy use in the forest industry. The factors affecting energy use at both industry level and at mill level are presented. The most probable technology trends, which can have an effect on energy production and use and CO{sub 2}-emissions are studied. Recent forecasts for the forest industry energy use till the year 2010 are referred and analysed. Three alternative forecasting methods are studied more closely. These methods are (a) Regression analysis, (b) Growth curves and (c) Delphi-method. Total electricity demand, share of purchased electricity, total fuel demand and share of process-based biofuels are estimated for the time period 1996 - 2010. The results from the different methods are compared to each other and to the recent scenarios. The comparison is made for the results concerning the energy use and the usefulness of the methods in practical work. The average energy consumption given by the forecasts for electricity was 31,6 TWh and for fuel 6,2 Mtoe in 2010. The share of purchased electricity totalled 73 % and process based fuels 77 %. The figures from 1995 are 22,8 TWh, 5,5 Mtoe, 64 % and 68 % respectively. All three methods were suitable for forecasting. All the methods required less working hours and were easier to use than scenarios. The methods gave results with a smaller deviation than scenarios, e.g. with electricity use in 2010 from 29,5 TWh to 34,7 TWh. Regression analysis and Growth curve -method could be tested using a time period 1975 - 1986 for forecasting the energy consumption in 1993 - 1995. The error in most of the forecasted variables remained lower than 5 %. The share of purchased electricity was most difficult to forecast, and errors were cat 7 % with Regression analysis -method and cat 22 % with Growth curve -method. For the Delphi-method similar analysis could not be made. The methods developed for energy forecasting seemed not reliable for estimation of CO{sub 2}-emissions. (orig.) 55 refs. %0Technical Report %NLTKK-TJ-65;Other: ON: DE98764048; ISBN 951-764-182-6; TRN: FI9803411 %1 %CFinland %Rhttps://doi.org/ Other: ON: DE98764048; ISBN 951-764-182-6; TRN: FI9803411 FI %GEnglish