Search

Voltage sag mitigation utilizing dynamic voltage restorers (DVRs) can be classified as a common-pool resource (CPR) good. However, the ability of DVRs in improving voltage sag performance of only downstream customers provides the ability to exclude the non-contributors selectively. Therefore, unlike traditional CPR goods, the DVR allocation problem can give rise to partial excludability. Here, the non-collocated customers have been divided into feasible clusters through suitable positioning of DVRs using the proposed graph-partitioning principle. In the absence of trustworthiness, especially with the participation of electricity supply companies, the participants may ask an external agent and share their willingness-to-pay information to design the optimal set of clusters. Alternatively, the customers, including electricity supply companies, can also share internal information as an open system. Strategies for sharing internal information to avoid free-riding are also discussed. The utility distribution would ensure the viability of contribution group formation. Here, the results from three different utility-distributing solution concepts, such as an alternative definition of the core, the nucleous, and Shapley value, are compared. The simulation results are numerically verified for a small scale system and validated utilizing a large scale system. The proposed methodology can be used for any real-world system that follows similar properties. [ABSTRACT FROM AUTHOR]

With ever increasing penetration of inverter-interfaced generators and consequent reduction in available inertial reserve, traditional allowable delay in the provision of the primary frequency reserve will threaten the security of the future grids. However, droop control with fast-enough inertia-less generators can limit the post-contingency frequency excursion. To conform with the fast-enough reserve provision requirement in a reduced and varying-inertia system, an economical clearing of the reserve requires an allowable delay heuristic corresponding to the desired system frequency excursion characteristic; while alleviating the causality dilemma introduced by the unavailability of maximum delay heuristic itself before clearing. The proposed heuristic establishes a relationship among the system parameters, such as the aggregated system inertia, regulation factor, system frequency, desired frequency response, and allowable delay margin. The Padé approximation technique is utilized for deriving the transfer function of such a system, allowing calculation of the allowable delay with various damping ratios corresponding to the dominant poles. The impact of different Padé order on the determined heuristic is analyzed. The regulation factor of the generators is calculated based on the reserve bid volume to ensure proportional injection during contingencies. The impacts of regulation factor and associated delay on the frequency response are also studied. [ABSTRACT FROM AUTHOR]

In an open system, system resources are managed by the contributors themselves. Because the participation needs to be voluntary and contributors true utility will remain unmonitored, proper communication among the participants is essential. In the discussed common-pool resource (CPR) problem, all the members need not be contributors, but the non-excludable component of the resource is required to be multiplied with each of the member's rivalrous component, and all these products are needed to be summed up to calculate the overall resource requirement. This characteristic applies to a typical power system optimization problem, where, if a customer group installs a common dynamic voltage restorer, voltage improvement can be treated as non-excludable quantity, while the peak load of individual customers can be treated as rivalrous quantity. In this work, we consider, the participants sharing the CPR contribute to form an open system to capitalize on ‘economy of scale’, while discouraging the unilateral free-riding benefit. Considering the benefit and average production cost curve represented by piecewise linear functions we have shown that the utility function is convex. Furthermore, for the given problem, we have numerically calculated the utility distribution scheme by solving an optimization problem. [ABSTRACT FROM AUTHOR]

There is an increase in fault current levels seen by utilities due to grid expansion and the addition of new resources. Installation of superconducting fault current limiter (SCFCL) at transmission voltage levels is being explored to reduce the fault levels. The SCFCL limits the fault current by inserting an additional impedance in the network, which may impact the protection system. This paper discusses the impacts of SCFCL on the operation of line distance protection. It is observed that operation of SCFCL may not impact all the distance relays in the substation and near‐by substation in the same way, thus a boundary definition method is proposed in this study to identify the affected relays. The definitions of boundary condition based on steady‐state fault current values and critical current of the SCFCL are proposed in this study. The impact of SCFCL on the line distance relays which manifest as underreaching of the relays and may also lead to coordination issues and erroneous operation. An implementation method for mitigation which is compatible with existing substation relays is proposed and demonstrated to ensure proper operation of the relays. The proposed mitigation ensures that the system integrity is maintained during faults. [ABSTRACT FROM AUTHOR]

Renewable energy (RE) resources are non-dispatchable due to their intermittent nature, and battery storage devices (BSDs) play an important role to overcome their inherent variability. Therefore, for optimal operability, BSDs must be appropriately sized. Historical RE generation data can be used for sizing, with the objective to minimize the annualized planning cost. Application of high and low pass filters about a given cut-off frequency on the frequency spectrum of the historical generation data, calculated using discrete Fourier transform approach, segregate the fast and slowly varying components. The proposed methodology is based on ${3} {\sigma }$ principle and will ensure minimum injection of RE-generation variability into the grid for day-ahead scheduling with both fast and slowly varying components. The analysis shows that the batteries with the minimum unit capacity cost to throughput ratio provide minimum annualized planning cost for both slow and fast varying components. Determination of sizing of BSDs for a given cut-off frequency is numerically “costly” and to obtain the optimal cut-off frequency, a derivative-free mode-pursuing sampling method is applied. Exponential reduction in the daily injection of variability with increasing statistical significance in sizing is observed. Impact of unit-capacity cost to throughput-ratio on the sizing of BSDs is also studied. [ABSTRACT FROM AUTHOR]

Though extensive work is reported on various catastrophic indicators for the assessment of power system stability, the comparative assessment has not received much attention by the researchers. With the advent of WAMS technology and near real-time tools available, these indicators can play an important role in operation and control of the system. The choice of the indicator(s) has to be based on the performance measures (accuracy, reliability, security, mis-detection, false-alarm, and correctness), which use a common platform for evaluation. This companion paper presents a critical review of the catastrophic indicators and computes the performance measures to assess the indicators, which is illustrated by an implementation architecture. The approach starts by generating a contingency dataset of a network, followed by obtaining the network measurements to the contingency dataset. These measurements are used as an input to the indicators for detecting stability/instability cases. The performance measures are computed to each indicator for their assessment. In addition to the above performance measures, robustness and sensitivity are obtained to each indicator. The IEEE 39 and IEEE 118 test networks are considered for comparing the performance measures. Additionally, 50-machine 145-bus system is used to demonstrate the applicability of indicators on practical systems. This work provides systematic taxonomy and insight into various indicators bringing out salient features on a comparative note to help decision makers to make suitable choice of indicator(s) for their system. [ABSTRACT FROM PUBLISHER]

The conventional power system domain includes various utility business processes that are carried out by multiple market participants. In the conventional setup, these processes are typically captured through various local representations understandable by the concerned actors in their respective subdomains. The heterogeneity in process representations can result in problems of traceability, transparency, time-trackability, and interoperability across actors and software systems. To orchestrate the utility business processes, the IEC has recently started an activity entitled “IEC 62325-451-2:2014” for representing the “energy scheduling business process” of the European market in a standardized way. Along similar lines, this paper presents a process ontology along with a systematic methodology for building process models of end-to-end process operations of power utility. The process ontology consists of power system process modeling artifacts such as process, participant, activity, gateway, event, and event-based subprocess. For representing events and chronology of task execution, the common information model (CIM) can be adopted, which is expected to act complementary to process ontology semantics. An energy scheduling application is considered as a case study, wherein the proposed process ontology artifacts and the CIM ontology are used for building process models for an “energy scheduling business process” in the Indian power grid context. [ABSTRACT FROM PUBLISHER]

Common Information Model (CIM) is widely adopted by many utilities since it offers interoperability through standard information models. Storing, processing, retrieving, and providing concurrent access of the large power network models to the various power system applications in CIM framework are the current challenges faced by utility operators. As the power network models resemble largely connected-data sets, the design of CIM oriented database has to support high-speed data retrieval of the connected-data and efficient storage for processing. The graph database is gaining wide acceptance for storing and processing of largely connected-data for various applications. This paper presents a design of CIM oriented graph database (CIMGDB) for storing and processing the largely connected-data of power system applications. Three significant advantages of the CIMGDB are efficient data retrieval and storage, agility to adapt dynamic changes in CIM profile, and greater flexibility of modeling CIM unified modeling language (UML) in GDB. The CIMGDB does not need a predefined database schema. Therefore, the CIM semantics needs to be added to the artifacts of GDB for every instance of CIM objects storage. A CIM based object-graph mapping methodology is proposed to automate the process. An integration of CIMGDB and power system applications is discussed by an implementation architecture. The data-intensive network topology processing (NTP) is implemented, and demonstrated for six IEEE test networks and one practical 400 kV Maharashtra network. Results such as computation time of executing network topology processing evaluate the performance of the CIMGDB. [ABSTRACT FROM AUTHOR]

In this study, a 'two‐stage' deterministic algorithm an for efficient transmission expansion planning with renewable energy (RE) resources under the assumption that existing conventional generators provide the reserve to mitigate RE generation forecast error has been proposed. Zero‐RE penetration has been considered as a 'reference scenario', as well as cost‐minimisation objective has been considered as a planning criterion in Stage 1. In the proposed algorithm, Stage 2 is required to be solved only if the network performance degrades in relation to the reference scenario. In Stage 2, congestion cost is also incorporated as a sub‐objective. Here, the locational marginal prices of the unit scheduling problem obtained from the solution of Stage 1 have been used to calculate the congestion cost. In addition to being computationally tractable, the proposed algorithm guarantees performance improvement in the network. The planning horizon is divided into smaller blocks to facilitate delayed investment. When implemented on an IEEE 24‐bus reliability test system, the proposed algorithm generated minimum cost plan with better performance. [ABSTRACT FROM AUTHOR]

An effective demand response (DR) model framework for DR management systems in a residential premise is proposed. The proposed model consists of dynamic price signals from load serving entity (LSE), benefit function, and loss of consumer comfort. The conventional approaches generally consider these factors as independent in their formulation. Since these factors have functional relationship, the proposed model points to a holistic approach to capture the relationship. The scheduling model can be integrated with the home energy scheduler coupled to the smart meter, which facilitates the flow of dynamic price and energy consumption information to and from the household premise. The objective is to maximize net consumer surplus, which is formulated as a function of all aforementioned factors. A method for developing a practical and dynamic consumer benefit model for matured DR programs is proposed in this paper. For the purpose of showing the effectiveness of dynamic benefit function, a set of analytically formulated benefit functions is also considered. The suitability of the dynamic benefit function, as compared with other alternatives, is illustrated with an example. This paper also reports the dynamic price design from the view point of LSE, in order to cover the risks of wholesale price volatility. [ABSTRACT FROM PUBLISHER]

Non-local measurements from wide area measurement systems can be used to synthesize feedback signals for FACTS or HVDC based power oscillation damping controllers. Although non-local feedback signals may have better damping influence than local feedback signals, they involve greater complexity and reduced robustness due to the possibility of partial or complete loss of communication. A composite control strategy is proposed in this paper, which uses both local and non-local feedback channels to achieve enhanced damping and improved robustness. Special local signals and their non-local equivalents are identified for series reactive and shunt real power injection devices, which facilitate the design of controllers with a simple structure and fewer parameters. Unlike modal speed signals, synthesis of these signals does not require an online eigen-solution. Case studies are presented which demonstrate the efficacy and robustness of this strategy. [ABSTRACT FROM AUTHOR]

In this study, optimal scheduling of multiple non-colocated, price taker, independent wind power producers (WPPs) participating in forward day-ahead (DA) distribution electricity market is described; where, a WPP is comprised of multiple wind turbine generator (WTG) and battery storage device (BSD). Cost equivalent of reduction in network losses and improvement in voltage profile for non-colocated placement of WTG and BSD in Distribution Network (termed as ancillary benefit) is included in the objective function resulting in a scheduling strategy dependent upon location of WPP in the network. Objective function comprises of following sub-objectives: (i) maximize return from energy market, (ii) maximize benefit obtained from providing ancillary services, and (iii) minimize uncertainties in schedule by providing reserve from BSDs. Non-linear programming (NLP) technique is used for scheduling. Location of a WPP is varied to obtain a 'profit map'; which can be used as an 'offline-tool' to find out relative location of WTG and BSD for profit maximization. Proposed formulation is extended to participation of multiple WPP, where ancillary benefit is proportionally shared. Wind power forecast uncertainty leads to risk of not meeting the schedule and is probabilistically modeled in this work. Impact of reserve on DA energy schedule of is also studied. [ABSTRACT FROM AUTHOR]

Common information model (CIM) in vogue represents transmission and distribution components, and market operations in the perspective of electrical system, as unified modeling language (UML) objects. UML is object centric. CIM can offer interoperability and facilitate model exchange for large systems and is being adopted by many utilities. However, dealing effectively with multiple business processes is the backbone of the restructured power systems. Since CIM lacks chronological process sense, this has to be superimposed on CIM models, which is not easy due to inherent limitations of UML. To overcome this, this paper proposes to utilize business process modeling notation (BPMN) standards as a kind of pre-processor. BPMN is used to represent process centric interoperability requirements as business objects. UML object representation in CIM model is derived subsequently from it. The proposal is illustrated with an application use case—deriving CIM model of day-ahead market (DAM) transaction of distribution company, from its business process representation. Semantic model of DAM transaction using CIM IEC 62325 (Markets) and extensions required are presented with a sample Indian system. DAM models in BPMN and CIM can be utilized to build an efficient software application. [ABSTRACT FROM PUBLISHER]

Exchange of event information among multiple interconnected power system operators has become imperative for integrated and secure operation of the system. Considerable literature is reported on the standardization of power system data, while limited work has been done on the standardization of events. In this paper, we define a high level event ontology for power systems comprising seven concepts, namely, event, event extractor, event consumer, time, measurement, location, and level. These concepts are then developed with details extracted from the operating procedures followed by national and regional load dispatch centers in the Indian national grid. The methodology adopted for designing concrete sub-ontologies from the high level event ontology is also reported. The proposed event ontology has wide applications in the areas of i) event driven architecture (EDA) which facilitates the integration of event driven applications within and across the utilities, and ii) complex event processing (CEP) which facilitates development of sense-and-respond software capable of processing events extracted from large volumes of real-time data streams. [ABSTRACT FROM PUBLISHER]

In terms of scope and granularity of data, two kinds of connectivity models are broadly in vogue for representing power network data, namely, node-breaker model and bus-branch model. This paper examines both the connectivity models from the perspective of common information model (CIM). Although CIM is primarily designed for most detailed node-breaker representation, with certain extensions CIM can be leveraged to achieve a standard, interoperable bus-branch model representation and exchange format. To facilitate this, it is proposed to introduce new class called TopologyBranch. By decoupling the connectivity information and the equipment parameter information, a mechanism for efficient exchange of time varying system models with minimum information transfer between the power control centers is described in this paper. The proposed CIM bus-branch connectivity model has wide applications in an interoperable, decentralized, hierarchical structure of control centers. Three illustrative examples including one on a practical 20-substation model of the 400 kV transmission network of Maharasthra state, India, are then presented for demonstrating the benefits. [ABSTRACT FROM PUBLISHER]