Systematic procedures for reducing wastes in complex chemical plants are needed to allow efficient optimisation. Here, a hierarchical procedure was applied to the optimisation of a real industrial plant to reduce wastes as well as energy and raw material consumption. In the case studied, the continuous production of methyl-butynol (MBI), acetylene reacts with acetone. The solvent ammonia and acetylene are recycled to the reactor. Unreacted substrates and by-products are separated from the product stream by distillation. Part of the unreacted acetone can be reused for other purposes after distillation. A substantial part of the unreacted substrates and by-products is delivered to a wastewater treatment plant. These waste streams constitute a substantial problem for the operation of this plant. First, waste streams were characterised and tracked back to their origin. Following the hierarchical design procedure, the overall input-output structure was fixed. The entrainer in the present process was critically examined and options were suggested. Then various recycle schemes were considered for later detailed study. The existing plant was simulated using ASPENPLUS. After adjusting the model to all important aspects of the real process scheme, excellent agreement between actual process performance data and simulation was obtained. The various process schemes were simulated and assessed for their economic and ecological performance. The objective functions used included utility, substrate and catalyst costs, as well as costs for wastewater treatment. Additionally, the environmental burden related to energy supply was accounted for by a carbon dioxide tax as suggested by the Nordic countries. The process changes included separation of unreacted acetone from the product stream and recycling to the reactor. By-products were converted back to substrates in an additional reactor separation system and recycled. In various simulated process configurations and operational schemes substantial economic and ecologic improvements were achieved. This study demonstrates the usefulness of hierarchical approaches combined with process simulation for plant optimisation.