Determining preferred substation configurations based on reliability and cost

• Main Author: Van der Merwe, Johanette
• Other Authors: Gaunt, C Trevor
• Format: Doctoral Thesis
• Language: English
• Published: University of Cape Town 2016
• Subjects: Electrical Engineering
• Online Access: http://hdl.handle.net/11427/20914

Many different substation configurations exist, each with its own space requirements, reliability and cost. Distribution planners need to choose between these numerous layouts and suggest the preferred substation configuration for future networks, but often don't understand the advantages and total life cycle cost associated with each. The aim of this research is to identify a simplified approach that can be used to determine a set of preferred substation configurations that minimises the total life cycle cost of the substation. An analytical approach is considered which uses deterministic reliability assessment to determine the load-based and economic performance indicators for different substation layouts. This analytical approach builds on the simplified reliability estimation approach developed by Van der Merwe (2014). This simplified approach is sufficiently complex to ensure that errors are not one-sided, while at the same time minimising the number of calculations of the reliability evaluation. A cost estimation tool was developed to determine the utility cost for each of the substations. The purpose of this tool is to prepare preliminary estimates of substation life cycle costs during the planning stage, based on conceptual design information. The value based reliability planning approach (VBRP) is used to determine the substation configuration with the lowest life cycle cost, given the specified design criteria. This substation is then the optimal configuration for the specific design criteria. The research considers 432 different substation configurations, and 720 different design criteria. This results in 311 040 different substation layouts. The approach was programmed into an MS Excel model and this model was used to compare the 432 different substation configurations for each of the 720 different design criteria and 18 different customer damage functions, i.e. repeating the comparison of the 432 different substations 12 960 times. The time required to perform one comparison of the 432 different substation configurations for one given design criteria and customer damage function is approximately 2 seconds. The process was automated and the time required to run through all the design criteria and customer damage functions was 8 hours 46 minutes on a standard laptop with an i5 processor. The optimal substation configuration obtained for each of the 12 960 scenarios were analysed and clustered in order to derive the preferred set of substation configurations. Through this process the 432 different substation configurations were reduced to only 17 different substation configurations.