The occurrence of a Loss Of Offsite Power (LOOP) event can be a major threat to nuclear safety due to the dependence of auxiliary systems on electrical energy. Probabilistic safety assessments of nuclear power plants require, thus, estimates of the frequencies and durations of such LOOP events. These estimates are usually based on past statistical data, which is not always relevant. Model-based approaches are thus needed. This paper proposes an analytical method to estimate the frequency and duration of switchyard-centered LOOP events, which constitute one of the four main categories of LOOP events. The proposed method is mainly based on the identification of active minimal cut sets, considering the behavior of circuit breakers against faults according to their coordination and selectivity. Adapted versions of the Risk Reduction Worth and Fussel–Vesely importance factors are proposed to evaluate the impact of components on the switchyard-centered LOOP event frequency. Furthermore, uncertainty analysis is developed and performed. Various generic plant connection schemes are used for application. Results demonstrate the applicability of the methodology to estimate the frequency and duration of switchyard-centered LOOP events, and to identify optimal ways to reduce the risk by modifying the switchyard configuration.

Cascading outages in power systems can lead to major power disruptions and blackouts in power systems. By taking Manual Corrective Actions (MCAs), operators could be able to mitigate a cascading outage following an initiating event. However, due to stressful and time-constrained situations, they might not be able to take appropriate corrective actions in time and might even take counter-productive actions. Although numerous approaches have been developed to assess the risk of cascading outages in a probabilistic way, they generally do not consider in a realistic manner MCAs, including their imperfection. This paper aims to address that gap by proposing a Human Reliability Analysis-Optimal Power Flow (HRA-OPF) framework. The developed approach is applied to the New England Test System (NETS) and to the Reliability Test System (RTS). The risks of loss of supplied power are compared for different possibilities: no MCAs, perfect MCAs, and imperfect MCAs.

Monte Carlo sampling is frequently employed for uncertainty quantification in depletion calculations. Several assumptions are needed to perform this analysis. In this work, an assessment of these assumptions is proposed via sample convergence studies and perturbation of the sampling distribution. The Uncertainty Analysis in Best-Estimate Modeling (UAM) Pincell Hot Full Power and the Turkey Point reference cases were considered for this purpose. The U-235 thermal independent fission yield uncertainties evaluated in JEFF-3.3 and JEFF-4.0 were propagated to the nuclide vector and to the system multiplication factor. Using JEFF-4.0 data, a 75% reduction in the uncertainty of selected nuclide concentrations and an 80% reduction in the multiplication factor uncertainty were observed, showcasing the effect of full covariance evaluations. The presented results also prove that the uncertainty in the considered observables shows marginal dependence on the sampling distribution.

The most common reliability standard (RS) used to define the desired level of adequacy of power systems is the loss of load expectation (LOLE). Nevertheless, its relevance is questioned due to the massive integration of variable renewable energy sources (RES) and storage in current power systems. This paper studies the relevance of the loss of load expectation (LOLE) and of alternative adequacy indicators for such power systems. The different shortcomings of the LOLE are presented and other adequacy indicators being able to address these gaps are looked at. The main issue with LOLE is that it does not give any indication on the distribution and the magnitude of the different load sheddings. In order to fix these problems, the expected power unsupplied (EPU) and conditional value at risk (CVaR) are chosen to complement the LOLE. The former ensures that the shortfalls are not too consequent, while the latter checks if the average of the worst shortfalls do not reach vital structure. These indicators are based on national load shedding plans.

Cascading outages are the leading cause of major disturbances and blackouts. During the slow phase, timely manual corrective actions (MCAs) can halt the cascade. However, perfect execution of MCAs cannot be guaranteed, especially under stressful conditions. Furthermore, recent blackouts and disturbances have highlighted a range of consequences resulting from improper MCAs. In this study, we enhance our previously developed Human Reliability Analysis-Optimal Power Flow (HRA-OPF) framework, which integrates the estimated failure probability of MCAs in the risk analysis of outage scenarios. We expand the framework by introducing additional operator failures scenarios, distinguishing between failures in diagnosis and action phases, and accounting for a broader spectrum of consequences. We apply the HRA-OPF framework to the New England Test System grid. Finally, we evaluate the sensitivity of the results to various assumptions, such as the nominal probability of operator failure and islanding possibilities.

Various safety analysis methods have been used over the years to quantify safety margins: at the start by Deterministic Safety Analysis (DSA) with conservative models or best estimate models and “bounding” assumptions; then by the Best Estimate Plus Uncertainty (BEPU) approach which takes into account uncertainties related to physical and modelling parameters; and recently by the Extended BEPU (EBEPU) approach with the intention of incorporating information from Probabilistic Safety Assessment (PSA).By combining probabilistic and deterministic safety assessments while taking into consideration a wide variety of uncertainties, the current work aims at the development of an Integrated Safety Margin Quantification (ISMQ) technique to extend the scope of existing approaches. Multiple aspects of the “safety margin” can be handled by this ISMQ methodology, including pertinent Initiating Events (IE) and sequences, the distance between best-estimate load and safety limit, and the likelihood of exceeding the safety limit.Together with the suggested methodology, an industrial application on Design Extension Conditions (DEC) of a typical Generation II 1000 MWe Pressurized Water Reactor (PWR) is presented to illustrate the approach and the anticipated outcomes. This ISMQ methodology is also applicable for safety analyses of other reactor designs including Small Modular Reactors (SMRs).

The neutronics design of new reactors has been historically relying on zero power research reactors for the simulation of the core irradiation conditions as well as for nuclear data and code validation. In this framework, a study on the possible MYRRHA representativity improvement coming from the loading of MOX fuel in VENUS-F zero power reactor is under investigation. This work analyses the effect of the nuclear data used in the sensitivity and representativity calculations, highlighting the need for a comprehensive set of nuclides and reactions to be further studied.

The variability of renewable energy sources (RES), threatening the adequacy of power systems, could be dampened using storage. Concretely, RES would be stored during periods of excess and used when needed. However, the impact of the operating strategy on the loss of load expectation (LOLE) remains an open question. Therefore, this paper studies the impact of different storage operating strategies on adequacy indicators. To do so, five strategies are defined: yearly costs minimization with single and depth-dependent value of lost load (VoLL), LOLE minimization, and weekly costs minimization, with and without water values. Using Monte Carlo methods, these are applied to three case studies. The various operating strategies give spread results in terms of LOLE, with up to a factor 10 between the lowest and highest value. Weekly minimization gives poorer adequacy indicators. This paper recommends weekly costs optimization, with a depth-dependent VoLL and with water values, in adequacy studies.

Grid stability affects the progression of loss of offsite power scenarios to the nuclear power plants, as the most dominant core-damage-inducing initiating event. The ever-increasing tendency toward integration of large-scale wind farms into the transmission systems drastically challenges their dynamic behavior. In this study, we proposed a dynamic-probabilistic framework to evaluate the effect of wind penetration scenarios on the frequency of occurrence of LOOP to the NPPs. We developed 30 initial grid states, including 3 wind penetration percentages, 5 operational loading levels, in both optimal and normal power flows. We estimated 150 GR-LOOP frequencies for 5 NPP locations upon 3Ph-Shc faults on the transmission lines. Extending from 1.3 x 10-7 to 9.6 x 10-5 (year)-1 for the safest to the riskiest situations, the mean GR-LOOP frequencies have been assessed as 1.74 x 10-5, 2.96 x 10-5, and 2.77 x 10-5 (year)-1, for grids with 0%, 10%, and, 20% penetration percentage of installed wind generation, respectively. Our results indicate the undeniable influence of variations in the operational loadings and the wind pene-tration percentage on the GR-LOOP frequency, in addition to the NPPs' locations. Such intensification effect was more pronounced in instantaneous frequencies of unique states than in the averaged ones. Furthermore, we observed more outstanding growth in the corresponding values of optimal cases. Our findings suggested that the reliability of the offsite grid to the NPPs be considered as a decision-making criterion, in addition to the existing security and economic issues in the grids with wind penetration.

With the transition towards a smart grid, Information and Communications Technology (ICT) infrastructures play a growing role in the operation of transmission systems. Cyber-physical systems are usually studied using co-simulation. The latter allows to leverage existing simulators and models for both power systems and communication networks. A major drawback of co-simulation is however the computation time. Indeed, simulators have to be frequently paused in order to stay synchronised and to exchange information. This is especially true for large systems for which lots of interactions between the two domains occur. We thus propose a self-consistent simulation approach as an alternative to co-simulation. We compare the two approaches on the IEEE 39-bus test system equipped with an all-PMU state estimator. We show that our approach can reach the same accuracy as co-simulation, while using drastically less computer resources.

Cascading outages in power systems are complex phenomena that are difficult to model and not yet fully understood. When attempting to simulate cascading outages, close attention should thus be paid to modelling uncertainties. This is especially true for fast cascading outages (i.e. cascading outages that are driven by electromechanical transients) during which many protection systems can operate in close succession. Indeed, small variations in the timing of a protection operation can vastly impact the cascade propagation and its consequences. A classical approach to account for modelling uncertainties is to perform Monte Carlo (MC) simulations which is computationally challenging. To alleviate this challenge, we propose an indicator based on sequences of tripping events to predict which contingencies are more sensitive to uncertainties and which are less. MC simulations can be avoided for the latter, saving computation time. The performance of our indicator is demonstrated using a modified version of the IEEE 39-bus test system. We show that analysing the sequences of tripping events leads to better performing indicators than looking only at the consequences (e.g. load shedding) of cascades. We also show how the proposed indicator can be used to speed up a simplified probabilistic security assessment.

In a power system featuring a large share of distributed generations (DGs), the variability of power supply results in various issues in the implementation of more DG units incorporation to the existing distribution networks, particularly, congestion risk.Active Network Management (ANM) could provide (almost) real-time control, by possibly curtailing their production in case of grid congestion so as to allow more DG units integration, while deferring costly and time-consuming network upgrades.This paper provides a methodology for the fast assessment of the connection capability of DG units to a grid in ANM scheme, based on efficient Monte Carlo sampling.Besides, resorting to correlated sampling, it is possible to simultaneously estimate the congestion risk with and without connecting a new DG unit of variable capacity.This significantly reduces the computation burden in assessing the connection capability of a grid.The effectiveness of the proposed method is demonstrated on a test power grid.

Power systems are not designed to survive a large number of simultaneous contingencies and are thus vulnerable to extreme events that can cause considerable societal and economical losses. Power systems resilience against adverse weather events could be enhanced with the implementation of Intentional Controlled Islanding (ICI). This paper presents a strategy design of ICI to improve system survivability in case of severe windstorms. It includes the design of a predetermined network split applied with a preventive and an adaptive approach, and a decision-making process based on the forecast wind speed. The network response to storms is studied dynamically, in order to model cascading failures and protection scheme effects. Simulations are conducted on the IEEE 39-bus system, then the resilience enhancement is assessed with a stochastic approach. ICI is demonstrated to be effective, as the results show an significant decrease of blackout risk while limiting unnecessary load shedding.

An empirical potential study was performed for the americium (Am), neptunium (Np) containing uranium (U) and Plutonium (Pu) mixed oxides (MOX). The configurational space of a complex U1-y-y′-y″PuyAmy′Npy″O2 system was predicted by the rigid lattice Monte Carlo method. Based on the computing time and efficiency performance, the method was found to rapidly converge towards the optimal configuration. From that configuration, the relaxed lattice parameter of Am, Np bearing MOX fuel was investigated and compared with available literature data. As a result, a linear behaviour of the lattice parameter as a function of Am, Np content was observed.

In the field of power systems, a fast cascade is defined as the part of a cascading outage that is driven by electromechanical transients. During a fast cascade, the time interval between two protection system operations can be as small as a few milliseconds. The order of events and the global evolution of the cascade are thus very sensitive to measurement errors and epistemic uncertainties. The stochastic failures of protection devices are another source of uncertainty. Therefore, a dynamic probabilistic security assessment is necessary to fully understand and better mitigate fast cascading outages. Dynamic security assessment methodologies have already been applied to the study of the fast cascade, but they require the use of a custom-built simulator and their results are difficult to interpret and to exploit. In this work, we propose to apply the MCDET (Monte Carlo Dynamic Event Tree) methodology to the probabilistic dynamic security assessment of transmission systems. The methodology is implemented using the DIgSILENT PowerFactory simulator and tested on the IEEE 39-bus system. Importance measures are used to identify assets whose maintenance or replacement should be prioritised, and sensitivities are used to identify inadequate or sensitive protection settings.

Loss of Offsite Power (LOOP) is the most dominant initiating event in the Core Damage Frequency (CDF) of Nuclear Power Plants (NPPs). Since the frequency of occurrence of LOOP depends on the grid configuration, outage schedule, load flow, etc., even a relevant generic frequency may result in significant uncertainty. In this study, an analytical hybrid methodology, which combines the probabilistic and deterministic modules, is introduced to systematically estimate the frequency of Grid-related LOOP. The proposed probabilistic module develops the relevant event-trees and fault-trees in SAPHIRE7 to identify the LOOP sequences upon the response of the distance protection system to a fault on the transmission lines. Meanwhile, the identified scenarios are confirmed based on the simulation results from the deterministic module in DIgSILENT-PowerFactory15.1. 15-out-of-48 scenarios (100-out-of-300 sequences) are identified as grid-related LOOP upon faults on the four transmission lines of the test grid, and result in 1.76E-06/yr for the grid-related LOOP frequency. The developed methodology enables us to trace a LOOP chain, from a primitive presumed fault through the progress of the cascade. The grid-related LOOP scenarios in a certain network can be ranked probabilistically, with the desired level of details in both deterministic and probabilistic modules.

Generic frequencies are used for Grid-related Loss of Offsite Power (GR-LOOP) in the Probabilistic Safety Analysis (PSA) of Nuclear Power Plants (NPPs). When historical databases are used, the influence of variation in the grid conditions over time and the impact of the NPP location are not considered in the estimation of GR-LOOP frequency. Therefore, relying purely on historical data induces uncertainty in total Core Damage Frequency (CDF). This paper aims at assessing the influence of the dynamic load behavior on the GR-LOOP frequency, for different locations. For that purpose, sensitivity cases are defined. The occurrence frequency of GR-LOOP following the occurrence of a three-phase-short-circuit fault on a transmission line is evaluated via a probabilistic-deterministic methodology for all cases. The minimum and maximum estimated values of the GR-LOOP frequency, 2.98E-04/year and 9.95E-03/year respectively, indicate the dependence on the variation in either the dynamic load behaviors or the NPP locations. The results reveal that GR-LOOP frequency varies with load model in the initial grid. In the presented probabilistic-deterministic methodology it is beneficial to consider the influence of timely changes of the grid conditions on the GR-LOOP occurrence frequency for different NPPs’ locations, especially where relevant or sufficient generic databases are not available.