Abstract: In a cascading power transmission outage, component outages propagate nonlocally; after one component outages, the next failure may be very distant, both topologically and geographically. As a result, simple models of topological contagion do not accurately represent the propagation of cascades in power systems. However, cascading power outages do follow patterns, some of which are useful in understanding and reducing blackout risk. This paper describes a method by which the data from many cascading failure simulations can be transformed into a graph-based model of influences that provides actionable information about the many ways that cascades propagate in a particular system. The resulting “influence graph” model is Markovian, in that component outage probabilities depend only on the outages that occurred in the prior generation. To validate the model, we compare the distribution of cascade sizes resulting from n-2 contingencies in a 2896 branch test case to cascade sizes in the influence graph. The two distributions are remarkably similar. In addition, we derive an equation with which one can quickly identify modifications to the proposed system that will substantially reduce cascade propagation. With this equation, one can quickly identify critical components that can be improved to substantially reduce the risk of large cascading blackouts.
Abstract: Large electric power systems are among the most complex systems created by humankind. One consequence of this complexity is that small unexpected disturbances trigger long chains of cascading component failures that can lead to massive power outages. Because of the many different mechanisms involved and the limited data available from historical cascades, modeling cascading failure is a challenging problem. Furthermore, among all the possible combinations of triggering events a relatively small number lead to large blackouts, which creates difficulties for conventional statistical approaches to risk analysis. In order to facilitate understanding about this complexity, this chapter describes several different approaches to modeling cascading blackouts and subsequently studying the risk of cascades, given a blackout model. Together these results suggest that when combined with good models, risk analysis can provide valuable and actionable insight into cascading failures in power systems.
Abstract: The potential for cascading failure in power systems adds substantially to overall reliability risk. Monte Carlo sampling can be used with a power system model to estimate this impact, but doing so is computationally expensive. This paper presents a new approach to estimating the risk of large cascading blackouts triggered by multiple contingencies. The method uses a search algorithm (Random Chemistry) to identify blackout-causing contingencies, and then combines the results with outage probabilities to estimate overall risk. Comparing this approach with Monte Carlo sampling for two test cases (the IEEE RTS-96 and a 2383-bus model of the Polish system) illustrates that the new approach is at least two orders of magnitude faster than Monte Carlo, without introducing measurable bias. Moreover, the approach enables one to compute the sensitivity of overall blackout risk to individual component-failure probabilities in the initiating contingency, allowing one to quickly identify low-cost strategies for reducing risk. By computing the sensitivity of risk to individual initial outage probabilities for the Polish system, we found that reducing three line-outage probabilities by 50% would reduce cascading failure risk by 33%. Finally, we used the method to estimate changes in risk as a function of load. Surprisingly, this calculation illustrates that risk can sometimes decrease as load increases.
Abstract: This paper describes a new approach, using "Random Chemistry" sampling, to estimate the risk of large cascading blackouts triggered by multiple contingencies. On a 2383 bus test case the new approach finds the expected value of large-blackout sizes (a measure of risk) two orders of magnitude faster than Monte Carlo sampling, without introducing measurable bias. We also derive a method to compute the sensitivity of blackout risk to individual component-failure probabilities, allowing one to quickly identify low-cost strategies for reducing risk. For example, we show how a 1.9% increase in operational costs reduced the overall risk of cascading failure in a 2383-bus test case by 61%. An examination of how risk changes with load yielded a surprising decrease in cascading failure risk at the highest loadings, due to increased locality in generation and less long-distance transmission. Finally, this paper proposes new visualizations of spatio-temporal patterns in cascading failure risk that could provide valuable guidance to system planners and operators.
Abstract: Plug-in electric vehicle (PEV) charging could cause significant strain on residential distribution systems, unless technologies and incentives are created to mitigate charging during times of peak residential consumption. This paper describes and evaluates a decentralized and ‘packetized’ approach to PEV charge management, in which PEV charging is requested and approved for time-limited periods. This method, which is adapted from approaches for bandwidth sharing in communication networks, simultaneously ensures that constraints in the distribution network are satisfied, that communication bandwidth requirements are relatively small, and that each vehicle has fair access to the available power capacity. This paper compares the performance of the packetized approach to an optimization method and a first-come, first- served (FCFS) charging scheme in a test case with a constrained 500 kVA distribution feeder and time-of-use residential electricity pricing. The results show substantial advantages for the packetized approach. The algorithm provides all vehicles with equal access to constrained resources and attains near optimal travel cost performance, with low complexity and communication requirements. The proposed method does not require that vehicles report or record driving patterns, and thus provides benefits over optimization approaches by preserving privacy and reducing computation and bandwidth requirements.
Abstract: This paper describes a method for estimating the impact of plug-in electric vehicle (PEV) charging on overhead distribution transformers, based on detailed travel demand data and under several different schemes for mitigating overloads by shifting PEV charging times (smart charging). The paper also presents a new smart charging algorithm that manages PEV charging based on estimated transformer temperatures. We simulated the varied behavior of drivers from the 2009 National Household Transportation Survey, and transformer temperatures based an IEEE standard dynamic thermal model. Results are shown for Monte Carlo simulation of a 25 kVA overhead distribution transformer, with ambient temperature data from hot and cold climate locations, for uncontrolled and several smart-charging scenarios. These results illustrate the substantial impact of ambient temperatures on distribution transformer aging, and indicate that temperature-based smart charging can dramatically reduce both the mean and variance in transformer aging without substantially reducing the frequency with which PEVs obtain a full charge. Finally, the results indicate that simple smart charging schemes, such as delaying charging until after midnight can actually increase, rather than decrease, transformer aging.