Using normalised cross correlation and variance to determine the source of voltage unbalance exceedances in Eskom networks with wind farms

Abstract

During an exceedance of the voltage unbalance limit at a busbar there is a need to determine which entity is causing the problem between Eskom, wind farms and other entities that can influence the voltage unbalance at the busbar. There were voltage unbalance limit exceedances at Eskom-K, Eskom-C and Eskom-Z Eskom substations. There was a need to determine which entity was causing the voltage unbalance exceedances at these substations between Eskom, Transnet and wind farms. The normalised cross correlation was used to determine the source of voltage unbalance exceedances at Eskom-K and Eskom-C substation. The normal- ised cross correlation together with the variance was used to determine the source of voltage unbalance exceedances at Eskom-Z substation. The correlation value of Eskom-K voltage unbalance when correlated with the wind farm’s total active power was close to one. The correlation value of Eskom-C voltage unbalance when correlated with the Eskom loads was also close to one. There was a high variance of the voltage un- balance and corresponded to the high variance of the Transnet traction station loads. Based on the correlation and variance results, it was concluded that voltage unbalance at Eskom-K substation was caused by the wind farms. The voltage unbalance at Eskom-C substation was caused by the Eskom loads. The voltage unbalance at Eskom-Z was caused by the traction loads because the Eskom-Z voltage unbalance variance corresponded with the traction load variance.