Recent advances in the field of operations research and systems science have yielded rigorous multi-criteria optimisation techniques that can be successfully applied to processes having numerous conflicting criteria. One such method, referred to as the Rough Set Method, incorporates the knowledge of an expert within the optimisation algorithm. The result of this process is a set of all possible candidates for the optimal solution, called the Pareto domain, that has been ranked according to the expert's preferences. In addition to providing the optimal solution to the problem at hand, the ranked Pareto domain can yield useful information regarding the robustness of the optimal operating point. This paper presents a novel approach for evaluating the ranked domain, referred here as the drift group analysis, which is applied to the optimisation of a thermomechanical pulping process. This analysis may also prove useful in the design of control systems for such processes. (c) 2005 Elsevier Ltd. All rights reserved.