Ex Student Archive




Home About Browse Advanced Search


Aggeryd, Bengt (2009) Optimalt råvarulager för biobränsleföretaget. Other thesis, SLU.

Full text available as:
[img]
Preview
PDF
aggeryd_b_090625.pdf

Download (540kB) | Preview

Abstract

This report was written as a final thesis within the Department of forest economics at the Swedish University of Agricultural Sciences as a completion of the M.SC Forestry programme. The aim of this study was to create a model that, with respect to given stock levels, temperature, fuel usage and fuel price, would identify the optimal levels for purchase of bio fuels, optimal levels of incineration of bio fuels and the sum of the present values of all costs. The model would be applicable for a larger district heating plant. Data regarding temperature were received from SMHI. Information considering fuel prices was taken from the Swedish Forest Agency and the facts about fuel usage were collected from a large district heating plant, located in the southwest of Sweden. For the temperature, a sinus curve was adapted to describe the temperature over a year. With the use of temperature and fuel use, a regression analysis resulted in an equation that described the fuel use as a function of temperature. The divergence from the average weekly temperatures were divided into classes so the variations in temperature, and thus variations in fuel usage, easily could be described in the model as discrete alternatives. Then a model was created in the program QBASIC. A model that, with respect to temperature, fuel usage and entering stock levels, would minimize the sum of the present values of all costs, determine the optimal purchase levels of bio fuels and optimal levels of bio fuel incineration. The model implemented these calculations for all weeks during one year. The model is based on the theory of stochastic dynamic programming. With the method you consider all options in all periods. Due to limitations in QBASIC the number of fuel assortments was restricted to two: waste products from the forest industry and forest fuels. The results from week 1 (January) and week 26 (July) were presented in this report. This was done to indicate the differences between a cold period and a warm period. The results gave at hand that the sum of the present value of all costs declined with increasing entering stock levels. Regarding the purchase levels, results showed that in cases of low entering stock levels the purchases of forest fuels should be kept high and then decrease with increasing stock levels. With low entering stock levels, it was also made clear from the results that purchase levels should be higher than the incineration. This indicates that it is optimal to create a buffer until the next period. The purchase levels for waste products was at a constant low level. A reason was that the purchase cost function for this fuel was higly sensitive to the purchase volume. Results regarding optimal incineration levels proved to be a bit hard to interpret, but they indicated that it in most cases were equally optimal to combust either fuel. One indentified source of error is that the sinus curve fitted to the temperature doesn't match the observed temperature perfectly. Deviations between calculated fuel use and real fuel use depend on the errors in temperature. Calculated data regarding fuel prices may differ from the real prices due to price increases when very large amounts are bought. Empirical data for estimation of such functions have however not been found. The relations between prices and volumes are now described via linear functions that could be transformed to nonlinear functions in future versions of the model in case empirical data supporting such functions will become available. Finally the model itself is a result of the entering parameters and should not be blamed for errors. During this process a number of factors were brought to the attention that may be of interest to investigate further. This applies to the inclusion of substance losses and handling costs in the model, and the competition effect on the price of fuel.

Item Type: Thesis (Other)
Keywords: Biobränsle, lagring,
Subject (faculty): Faculty of Veterinary Medicine and Animal Science > Dept. of Forest Economics
Divisions: SLU > Faculty of Forest Sciences
Depositing User: Bengt Aggeryd
Date Deposited: 26 May 2009
Last Modified: 18 Aug 2015 10:14
URI: http://ex-epsilon.slu.se/id/eprint/3262

Actions (login required)

View Item View Item

Downloads

Downloads per year since May 2015

View more statistics