Transient-state real-time thermal rating forecasting for overhead lines by an enhanced analytical method

Fan, Fulin and Bell, Keith and Infield, David (2019) Transient-state real-time thermal rating forecasting for overhead lines by an enhanced analytical method. Electric Power Systems Research, 167. pp. 213-221. ISSN 0378-7796

[img] Text (Fan-etal-EPSR-2018-Transient-state-real-time-thermal-rating-forecasting-for-overhead-lines)
Fan_etal_EPSR_2018_Transient_state_real_time_thermal_rating_forecasting_for_overhead_lines.pdf
Accepted Author Manuscript
Restricted to Repository staff only until 16 November 2019.
License: Creative Commons Attribution-NonCommercial-NoDerivatives 4.0 logo

Download (2MB) | Request a copy from the Strathclyde author

    Abstract

    The majority of published approaches to real-time thermal rating (RTTR) deal with continuous or steady-state ratings for overhead lines. Less attention has been given to short-term or transient-state RTTRs, partly due to the increased computation time required. This paper describes a fast-computational approach to providing a transient-state RTTR in the form of percentiles based on the predictive distributions modelled for the measured weather variables that are combined with Monte Carlo simulation. An analytical method developed in IEEE Standard 738 calculates the transient-state conductor temperature after a step change in line current only and additionally requires the conductor to be in thermal equilibrium before the step occurs. The IEEE analytical method is enhanced here through inference of an equivalent steady-state initial line current from the initial conductor temperature and weather conditions over a specified time period. Numerous transient-state RTTR forecasts for a particular span are estimated via weather inputs randomly sampled from predictive distributions for a number of time steps ahead combined with the secant method to find the transient-state RTTR. Along with an enhanced analytical method, this yields a maximum allowable conductor temperature for a specified time period under each set of weather samples. The percentiles of transient-state RTTR forecasts are then determined from their sampled values using kernel density estimation. The approach developed here considers variations in weather forecasts at each 10-min time step.