Long term assessment of critical factors affecting the reliability of transmission grids

The critical nature of electricity transmission grids is an established fact that supports requirements of permanent availability and very high reliability. To meet such requirements, investments in reserve capacity and grid redundancy are acknowledged by regulators on each market while maintaining pressure on Transmission System Operators [TSO] to demonstrate the need of such investments. Improving the availability and reliability of the grid involves investments of two different natures: investments on how to prevent failure (usually line tripping) and investments on how to cope with failure by mitigating its consequences. The paper focuses on a long-term effort to assess technical and economic benefits of the former, so that the latter becomes less important. An alternative reasoning on this approach is that identifying effective measurements to improve reliability will extend grid capacity in the future, thus postponing the need for investment in additional generation capacity, line redundancy and transmission reserve capacity. Generation from distributed, renewable resources presents a special case since regulations often require TSO to dispatch all available power and, as most of it comes from wind generation and is difficult to schedule, the reliability of the grid is tested on remote locations that are often at the end of transmission grids. Environmental factors are a major cause for reliability issues as transmission grids spread across the territory, often crossing remote areas to connect to hydro and wind generation, but the interpretation of phenomena that leads to reliability issues is seldom straightforward. The authors participate on a long-term project to optimize maintenance decisions and scheduling according to field findings on the Portuguese electricity transmission grid operated by REN [1]. The grid operated by REN is mature in the sense that most of its assets are beyond half lifetime and quite a significant number of them are near the end of their estimated lifetime. One of the main challenges found, is that the Portuguese transmission grid already has a high quality record for reliability; therefore it is imperative to learn the maximum from each issue. Also, it is difficult and uncertain to wait for years to test a hypothesis on the explanation of a given set of line failures. Finally, it is deemed unacceptable to degrade the quality of service and reliability to empirically find the optimum balance with decreasing total cost of ownership. The authors begin with a ten year list of recordings of line faults where each occurrence is tagged with an immediate cause determined by experience or some heuristics. Then they match these occurrences in time and space with many phenomena related to equipment, operation, natural hazards and third-party human interaction (the latter two are considered environmental in the broad sense that people living and working around power overhead lines are part of its environment), to find whether the field data corroborates the immediate explanation. While the project is still on-going, some interesting cases have been highlighted: how vegetation management affects mean time between failure in individual lines; how reliable is the binding between incidents associated with lightning and lightning referencing in space and time provided by weather services; how storks seem to be near towers with broken insulators; how human activities on the ground may help or hinder grid maintenance and exploitation. The authors use an information system platform to characterize the environment and land use [2] along transmission lines and massively cross different factors to find plausible explanations to line incidents and how they could relate with the cause mentioned in the incident recordings. The near goal is to evaluate the effectiveness of maintenance practices implemented to improve the reliability of the existing grid and where and how resources ought to be applied to reach optimum efficiency. The longer term goal is to supply such data to grid planners so that future grids could be more economical and more reliable as well.