Alocação de fomentos florestais sob condições de incerteza

Abstract

One of the modalities of forestry outgrower schemes is the partnership between rural producers and forestry companies of which objective is the production of planted forests for wood supply. The contract of outgrower schemes by purchasing wood at a contracted price has been a common form of contracting of large Brazilian forestry companies. Thus, the producer benefits from the guarantee of wood sale and the company guarantees the supply of wood without the need for land acquisition. The definition of optimal regions for hiring these outgrower schemes is part of the strategic planning of the companies' wood supply. The knowledge of the financial risks that can increase the cost of production in these selected regions increases the chances of success of the venture. In this context, the main objective of this work was to test the use of stochastic linear programming (LP) scenarios to direct the strategies of allocating forestry outgrower schemes to wood supply. The specific objectives of the study are (i) to verify whether it is possible to model financial risk by employing a probability distribution function and associate it with the challenges of operating on outgrower scheme properties; (ii) compare the recommendation of the candidate regions for the allocation of new outgrower schemes between a deterministic and a stochastic linear programming model; (iii) quantify the impact of uncertainty variables on the cost of allocating new outgrower schemes; (v) compare the results obtained between the deterministic and stochastic LP models and indicate which is best to assist the company's decisionmaking. Two LP mathematical, one deterministic and one stochastic, were tested regarding the allocation of forestry outgrower schemes to meet the transport restrictions and volumetric wood supply goals for six years of cellulose production. As a result, we obtained a model capable of indicating, to which regions forestry outgrower schemes should be directed without exceeding the established maximum costs. In conclusion, the use of the stochastic linear programming model was efficient in generating scenarios and additional strategic information when compared to the deterministic model. The stochastic model allowed more expanded decision-making since it presents probability distributions regarding the probable values of the objective function.