A Statistical Data Processing Scheme for Electrostatic Precipitator Experiments

When investigating the collection efficiency of electrostatic precipitators, the model of the process is only roughly known and the procedures for collecting data lead to large deterministic and random errors. The measurements are very expensive and the decisions arising from the results are economically very important so that even complex statistical procedures are worth to be applied for the best use of tjie collected data. In this paper a survey is given of the statistical procedures which can be applied to the results of experiments perfomed in order to assess the benefit of a specific treatment; as a case study these procedures are applied to the results obtained during an actual experimental campaign performed on an ESP installed at a thermal power unit. It is believed that the analysis can provide guidelines for the design of rational testing policies in future experiments