Prediction of Availability for new power plant in the absence of data

This paper is geared in presenting a model to predict the Availability of the Generating Units (GU´s) of New large Coal Fired Stations based on an expert opinion elicitation process, when the data for the Availability is not directly obtainable. The Availability used here is the Energy Availability Factor (EAF, as a percentage) as defined internationally by UNIPEDE (International Union of Producers and Distributors of Electrical Energy) [1]. The EAF is an inferred indicator, i.e. it is derived from the PCLF which is the Planned Capability Loss Factor (maintenance in %), the UCLF (Unplanned Capability Loss Factor) and the OCLF (Other Capability Loss Factor) which are the random percentage losses, the latter being out of management control. After a preliminary statistical analysis, the indications are that the UCLF is at least 300% larger than the OCLF. Therefore, it stands to reason to utilise the UCLF as a determinant for the EAF, since management can adjust the PCLF at will. The results provided are fictitious and for illustrative purpose only.