This paper presents an algorithm for the estimation of the time-dependent temperature evolution of power cables, when real-time temperature measurements of the cable surface or a point within its vicinity are available. The thermal resistivity and specific heat of the cable surroundings are varied as functions of the moisture content which is known to vary with time. Furthermore, issues related to the cooling effect during rainy weather are considered. The algorithm is based on the lumped parameters model and takes as input distributed temperature sensing measurements as well as the current and ambient temperature. The concept is verified by studying a laboratory setup of a 245 kV cable system.
I E E E Transactions on Power Delivery, 2013, Vol 28, Issue 3, p. 1909-1917