In addition to its important metallurgical role, the coke accounts for more than half of the total production cost of the pig iron and around 20% of the total cost of steel production. For this reason, the integrated steelmaking mills seek minimum cost coal blends for producing coke within the quality requirements. To guarantee this quality the steelmaking mills impose to their coal vendors some contractual clauses regarding humidity, ashes, sulfur, phosphor, volatile matter, fluidity, reflectance and dilatation, and sometimes, still others. Since these requirements are not always met by the vendors, contractual penalties are specified with the purpose of avoiding that the steelmakers end up with the burden of the costs caused by such specs deviations. Nonetheless, such penalties do not always reflect the actual costs they cause, because the latter derive not only from the reduction of the usable material, but also from the blend adjustments necessary to keep the coke within the required quality. This thesis proposes a system for determining the burden effectively caused by specs deviations. A detailed presentation of the problem, its formulation as a linear programming model and a procedure for analysis are proposed. A functional prototype system was developed and numerical illustrations of the analyses are discussed.