Use this URL to cite or link to this record in EThOS:
Title: Dynamic rating for improved operational performance
Author: Huang, Rui
ISNI:       0000 0004 5916 7948
Awarding Body: University of Southampton
Current Institution: University of Southampton
Date of Award: 2015
Availability of Full Text:
Access from EThOS:
Full text unavailable from EThOS. Please try the link below.
Access from Institution:
Many power transmission systems are under pressure from increasing load demand as well as changes in power flows due to the evolution of the power market and the integration of renewable energy generations. At the same time, limited finance for installing new cables and the difficulties in reinforcement of existing circuits in urban areas incentivize transmission operators around the world to find ways to maximize the flexibility and usage of their existing transmission network. As a result, it is crucial to adopt new current rating methods which are able to optimize asset utilization, minimize risk and reduce the constraint costs incurred by transmission system operators. Historically, most cable thermal ratings are continuous ratings, with fixed seasonal values for a certain cable circuit. They are based on worst-case assumptions and are not able to consider the real-time environmental conditions. The ignorance of the real-time change in environmental conditions, which control the rate of heat dissipation from the cable, makes continuous ratings generally conservative. However, the rating values can also be optimistic for some extreme situations such as thermal runaway in the soil around the cables, which might cause overheating. Several dynamic rating systems have been applied to the existing underground cable in practice by using online monitoring data. Some worst case assumptions used in conventional cable rating standards have been removed. Such systems have been reported to deliver increases of 5-20% in cable current capacity. However, most existing dynamic rating systems can only determine a short-term rating at the current time step. It would be valuable for transmission operators to know the short-term rating in advance to assist in day-ahead planning. To solve this problem, a predicted rating system, which is capable of providing network operators with accurate short term current ratings at the day ahead stage, has been developed in this work. This novel cable rating concept integrates a day-ahead load forecasting system into the dynamic rating system to provide the time-limited short-term rating calculated forward from any point within the next 24 hours. Some shortcomings of existing rating methods for different kind of insulated cable installations have been detected and overcome. More suitable models have been built, compromising between accuracy and solution speed to fit them into the predicted rating system. A day-ahead load forecasting system has been built by using the Support Vector Regression (SVR) method. Dynamic thermal models are used to translate the load prediction into thermal prediction 24 hours ahead. Thus, the time-limited short-term ratings can then be calculated 24hrs ahead, based on the predicted load data and cable temperature data. In addition, an error estimation system has been integrated to estimate the predicted conductor temperature error quickly, thus increases the reliability of the predicted rating system. Utilizing this predicted rating system has the double benefit of reducing variations in dynamic ratings (which makes them difficult to plan with), while reducing the risk of thermally overloading the cable, thus prematurely ageing the dielectric. For a large scale transmission network, the dynamic rating and predicted rating systems for all the cable circuits might require huge amounts of computation and very long solution times, which make their application impractical and infeasible. The idea of using a machine learning method, such as Support Vector Regression, has been shown to dramatically reduce the solution time for dynamic rating calculations.
Supervisor: Pilgrim, James Sponsor: Not available
Qualification Name: Thesis (Ph.D.) Qualification Level: Doctoral
EThOS ID:  DOI: Not available