Search

In the past, cable system management has proceeded on the basis of cable system age, assuming that the oldest has the lowest reliability. It is now recognized that this constitutes a very coarse assessment and that a more targeted approach would bring a more efficient use of resources. The more targeted approach requires an assessment of the health of cable systems. It is increasingly common for the assessment of aged cable systems to be made through the application of diagnostic measurements. A recent study has shown that Very Low Frequency (VLF) Tan δ is the most commonly deployed cable system diagnostic. The practical use of this technique has been supported by the international standards IEEE Std. 400-2001 and IEEE Std. 400.2-2004. A key part of these standards is the guidance provided to a user that is detailed in the "Figures of Merit". These enable users to make practical improvements to the cable system as they help to identify cable systems that are more likely to fail in service in the near future. To aid these decisions a series of criteria have been developed. The benefit of the criteria described here is that the process for their determination is rational, reproducible, and transparent. The outcomes are supported by a probabilistic assessment of service performance.

North America has a significant underground electric distribution system that is nearing the end of its design and service life. The cost of global replacement of aging underground facilities is prohibitive. Consequently, utilities require an enterprise approach to managing underground power cable systems (cables, joints, and terminations) in order to develop prudent investment and maintenance practices.

The main contribution of this paper is to investigate the correlation between tan delta diagnostic measurements at 0.1 Hz (Very Low Frequency-VLF) and VLF breakdown performance for medium voltage (MV) cable samples through a laboratory test program. The cable samples used are a set of 15 kV, Cross-linked Polyethylene (XLPE), unjacketed cables removed from the field, the same service area, and having experienced similar operating conditions for almost four decades. The test program includes tan delta measurements at different voltage levels and a subsequent VLF extended step-withstand test. The VLF step test allows the evaluation of risk of failure during VLF tan delta testing and assessment of the ultimate performance of the cables. The tan delta diagnostic measurements are represented by the tan delta value and tip-up, which are considered the classical metrics. However, the paper also suggests the use of a new additional diagnostic feature that takes into account the scatter in the tan delta measurements for a particular test voltage level.

This paper discusses the experience of the National Electric Energy Testing Research and Applications Center (NEETRAC) with condition assessment of medium voltage power distribution cable systems using Partial Discharge (PD) measurements. Specifically, the main goal is to address the issue of selection of the appropriate PD diagnostic features. A large number of features for which a physical basis can be described are selected for analysis. The features are investigated using data obtained from field and laboratory tests. The investigation shows that two traditional diagnostic features are not enough to classify samples that are prone to fail in a four-year time window. Subsequently, cluster variable analysis is performed for a considerable number of PD diagnostic features. The analysis enables significant feature elimination and reveals that at least seven different types of PD diagnostic features are needed to achieve an acceptable level of dissimilarity between features. Future work, conclusions and remarks are also discussed.

When only partial information is available about equipment failures (installation date and amount, as well as failure and replacement rates), data on sufficiently large number of yearly populations of the components can be combined, and estimation of model parameters may be possible. The parametric models may then be used for forecasting of the system`s short term future failure and for formulation of replacement strategies. We employ the Weibull distribution and show how we estimate its parameters from past failure data. Using Monte Carlo simulations, it is possible to assess confidence ranges of the forecasted component performance data.

Hipot or voltage proof tests have long been used to assure the health of cable systems in the factory and when commissioning. One of the concerns with this approach is that there is no way to judge the quality of the Pass, ie did the system barely survive or was there a respectable margin. This paper shows how this problem has been practically addressed in US utilities with an approach which is termed "Monitored Withstand". The protocols and the test philosophies are discussed. A number of case studies are discussed together which some reflections on future activities in this arena.

Power system networks are complex systems that are highly nonlinear and non-stationary, and therefore, their performance is difficult to optimize using traditional optimization techniques. This paper presents an enhanced particle swarm optimizer for solving constrained optimization problems for power system applications, in particular, the optimal allocation of multiple STATCOM units. The study focuses on the capability of the algorithm to find feasible solutions in a highly restricted hyperspace. The performance of the enhanced particle swarm optimizer is compared with the classical particle swarm optimization (PSO) algorithm, genetic algorithm (GA) and bacterial foraging algorithm (BFA). Results show that the enhanced PSO is able to find feasible solutions faster and converge to feasible regions more often as compared with other algorithms. Additionally, the enhanced PSO is capable of finding the global optimum without getting trapped in local minima.

The ageing of the electrical infrastructure is a growing concern for utilities, regulators and customers; and there is no doubt that addressing this problem will become an ever more important priority. Any solution will have to address three basic issues: firstly where is the optimum place and time to start, secondly what is the most appropriate suite of actions that can be taken and finally is the solution going to deliver the expected life. Diagnostic programs play an important part in the first and the third issues. In the first they may be able to guide the identification and prioritization of assets to be addressed; here they operate on the ageing population. The third issue benefits from diagnostics by using them as part of the assurance process that determines that the replacements / repairs have been effective. Addressing the ageing infrastructure is a large, complex and interacting challenge; this paper focuses on the first issue, namely how to select the appropriate locations, timing and technique.

The components of electricity networks are ageing. It is expected that within a horizon of 15 years, the performance will deteriorate significantly, while the costs for operating the networks will increase enormously. The main problem is that a significant part of the population of the assets is installed in the same period, resulting in a highly concentrated number of failures in a short time. The currently applied replacement strategy has to be revisited, in order to accommodate the effects of ageing assets: higher maintenance costs, high failure rates, and a steep increase of capital expenditure (CAPEX).

The paper describes a simple replacement strategy developed on the basis of statistical failure parameter identification and a predictive replacement algorithm. The proposed approach is illustrated with examples showing both the effectiveness and limitations of the method

When large amount of statistical information about power system component failure rate is available, statistical parametric models can be developed for predictive maintenance. Often times, only partial information is available: installation date and amount, as well as failure and replacement rates. By combining sufficiently large number of yearly populations of the components, estimation of model parameters may be possible. The parametric models may then be used for forecasting of the system’s short term future failure and for formulation of replacement strategies. We employ the Weibull distribution and show how we estimate its parameters from past failure data. Using Monte Carlo simulations, it is possible to assess confidence ranges of the forecasted component performance data.