ATC assessment and enhancement of integrated transmission and distribution system considering the impact of active distribution network
2020; Institution of Engineering and Technology; Volume: 14; Issue: 9 Linguagem: Inglês
10.1049/iet-rpg.2019.1219
ISSN1752-1424
AutoresDevesh Shukla, S. P. Singh, Amit Kumar Thakur, Soumya R. Mohanty,
Tópico(s)Power Systems and Renewable Energy
ResumoIET Renewable Power GenerationVolume 14, Issue 9 p. 1571-1583 Research ArticleFree Access ATC assessment and enhancement of integrated transmission and distribution system considering the impact of active distribution network Devesh Shukla, Devesh Shukla Department of Electrical Engineering, IIT(BHU), Varanasi, IndiaSearch for more papers by this authorShiv P. Singh, Corresponding Author Shiv P. Singh spsingh.eee@itbhu.ac.in Department of Electrical Engineering, IIT(BHU), Varanasi, IndiaSearch for more papers by this authorAmit Kumar Thakur, Amit Kumar Thakur orcid.org/0000-0002-2871-5433 Department of Electrical Engineering, IIT(BHU), Varanasi, IndiaSearch for more papers by this authorSoumya R. Mohanty, Soumya R. Mohanty Department of Electrical Engineering, IIT(BHU), Varanasi, IndiaSearch for more papers by this author Devesh Shukla, Devesh Shukla Department of Electrical Engineering, IIT(BHU), Varanasi, IndiaSearch for more papers by this authorShiv P. Singh, Corresponding Author Shiv P. Singh spsingh.eee@itbhu.ac.in Department of Electrical Engineering, IIT(BHU), Varanasi, IndiaSearch for more papers by this authorAmit Kumar Thakur, Amit Kumar Thakur orcid.org/0000-0002-2871-5433 Department of Electrical Engineering, IIT(BHU), Varanasi, IndiaSearch for more papers by this authorSoumya R. Mohanty, Soumya R. Mohanty Department of Electrical Engineering, IIT(BHU), Varanasi, IndiaSearch for more papers by this author First published: 08 June 2020 https://doi.org/10.1049/iet-rpg.2019.1219Citations: 3AboutSectionsPDF ToolsRequest permissionExport citationAdd to favoritesTrack citation ShareShare Give accessShare full text accessShare full-text accessPlease review our Terms and Conditions of Use and check box below to share full-text version of article.I have read and accept the Wiley Online Library Terms and Conditions of UseShareable LinkUse the link below to share a full-text version of this article with your friends and colleagues. Learn more.Copy URL Share a linkShare onFacebookTwitterLinked InRedditWechat Abstract Assessment and enhancement of transmission network capability have been one of the prime interests of power system monitoring, operation and control. Conventionally, the issue of available transfer capability (ATC) assessment has been tackled considering the transmission and distribution networks as segregated entities because they had little or no influence on each other's performance owing to the unidirectional flow of power from the transmission to distribution networks. Now, with increasing interest in smart transmission and distribution in amalgamation with a paradigm shift in sources of generation from centralised to decentralised even at distribution levels mandating the analysis of power grid while inculcating both the hierarchies simultaneously. In this study, a platform for quasi-static operation of power transmission and distribution networks, employing a multi-agent based system has been developed. The developed multi-agent systembased system has been utilised for assessment of ATC considering active distribution network (ADN) whereby Modified IEEE 24 bus system at transmission and Modified IEEE 123 node system at distribution level has been considered as a test system. It has been observed that the presence of ADN considerably affects the ATC of the system. Nomenclature TSO, DSO transmission, distribution system operator TN, DN transmission, distribution network DER distributed energy resources VVC volt-VAR control ADN active distribution network TNO transmission network optimiser agent component containing the equivalent load of ADN power output of PV and wind at th bus at time t PV inverter power output and losses reactive Power, efficiency and maximum rating of inverter PV inverter reactive power, efficiency and maximum rating. velocity, cut in, cut out, nominal wind velocity susceptance of SVC, reactance of TCSC admittance angle of line connecting buses to . voltage angle of buses i and j active power demand and injections of transmission level at th bus reactive power demand and injection of transmission level at th bus Base Case Active power demand in transmission level of th bus at time t. load at th node in the ADN at time t c set of credible contingencies voltage, nominal voltage at DN and TN levels of th bus minimum voltage at DN and TN levels of th bus maximum voltage at DN and TN levels of th bus power flowing through line i, j and its minimum and maximum values at TN q,F feeder number and last feeder acceleration factor conductance, susceptance and admittance of feeder qin p.u. conductance and susceptance matrix between nodes i and j of feeder q capacitor bank (CB) capacity, integer value for capacitor bank unit VAR value of each capacitor in CB off-nominal turn ratio of OLTC and voltage regulator voltage change for tap position. t,T time interval, last time interval 1 Introduction Assessment and enhancement of power transferring and handling capability of the power network, in the evolving scenario of power system restructuring would be playing a very vital role. ATC (available transfer capability) assessment and enhancement has been an important aspect of power system operation monitoring and control. Several methods have been reported in the literature for ATC assessment and enhancement [1-6]. Broadly speaking methods existing in literature can be classified as probabilistic, deterministic, stochastic, AI-based methods. The power system is being drastically restructured at transmission, distribution and generation levels. Conventionally, there used to be unidirectional power flow, directed from the source node (i.e. power system substations) to the feeder/consumer end in distribution networks. This facilitated separate analysis of transmission and distribution systems due to the negligible influence of the distribution network on the transmission system. With the increase in decentralised and distributed generation, the complexity and interdependency (in terms of influence) of the transmission and distribution system has increased. Occurrence of reverse power flows in distribution systems on account of increased distributed generation with the dominance of intermittent renewable sources and mobile virtual power plants capable of acting as sink or source depending on the market scenario (electric vehicles) further aggravated the issue by adding to the interdependency of (integrated transmission & distribution) systems. The existing literature on ATC estimation covers in great detail about the conventional systems, but little work is done while considering the presence of active distribution network (ADN). The problem of ATC assessment for integrated system lies with lack of proper method/technique to deal with the complicacy arising because of considering the ADN. Therefore, this paper first attempts to develop a MAS (multi-agent system) based system for integrated analysis of system and provides a pattern search optimisation-based solution for ATC assessment. Co-simulation platforms for integrated analysis of systems are being developed for assessing the impacts of probable phenomenons at DN on TN and vice-versa. In [7] authors presented an architecture for transmission and distribution integrated monitoring and analysis system and employs the developed method for analysing higher penetration of DER's (distributed energy resource) at DN levels. In [8] authors investigated the economic aspect of (integrated transmission & distribution (ITD) and presented a two stage optimisation problem for analysing economic models using transmission constrained residual supply curve along with aggregated residual demand curve. Impact of ADN on the risk analysis of transmission system has been assessed in [9] where the risk indices are derived using an iterative calculation between . Network equivalent models have also been utilised to integrate system into a common electrical model [10]. In [11] the authors presented master–slave splitting based global power flow method for integrated analysis of system. Authors have proposed heterogeneous decomposition for coordinated economic dispatch of coupled systems in [12]. A decoupled co-simulation approach [13] whereby existing transmission and distribution simulators are linked through FNCS (framework for network co-simulation), a middleware framework, to assess the dynamic behaviours of ITD networks. Co-simulation has been achieved through federation, in which each process runs on its own simulator with FNCS facilitating the data exchange and clock synchronisation between the simulators. Three sequence models for transmission system and three phase models for distribution networks are utilised and an integrated power flow and dynamic simulation have been proposed in [14]. The method was capable of handling both unbalanced, balanced and mixed scenarios of ITD. Attempts at developing OPF (optimal power flow) solutions for ITD operation have also been made in [15] whereby heterogeneous decomposition algorithm inspired by heterogeneous characteristics have been proposed to solve the OPF in a distributed manner. The methods for estimating the flexibility in active and reactive power exchange at the TSO-DSO interface have also been presented in [16] where authors have suggested optimisation-based methods for estimation of the flexibility. Furthering, the research in this direction, this paper proposes ITD framework which involves software in loop simulation whereby multi-agent based system (MAS) has been used for data exchange between the at the boundaries for integrated analysis of and pattern search-based method for assessment of ATC. Although in the existing literature methods for coordinated optimisation of systems, co-dynamic simulation and application of the developed framework to different applications can be found, the technique for efficient quasi-static analysis over a larger span along with ATC assessment in the integrated framework are still under development. The main contributions of the paper are as follows: • MAS driven approach has been proposed for ITD development and implementation using MATPOWER and OpenDSS co-simulation based models. • A coordinated scheme using MAS for ITD framework has been developed to assess the impact of ADN on ATC of the system at the transmission level. • Impact of deploying CVR in ADN at DN on ATC of the system at TN has been analysed. • Impact presence and absence of DERS (PV, WIND, PV & WIND ) in ADN at DN on ATC of the system at TN has been analysed. 2 ITD framework The developed framework for integrated analysis of transmission and distribution system is schematically represented in Fig. 1. The ITD platform relies on agent-based communication between the TSO and DSO. The method is implemented by dividing the system into two levels. The TSO performs the transmission system analysis, control and monitoring function while being in continuous coordination with the energy management system (EMS). Analogously the DSO performs the functions of distribution system and is in continuous coordination with advance distribution management system (ADMS). That regularly monitors measurements with the help supervisory control and data acquisition system and AMI ( advanced metering infrastructure). Fig. 1Open in figure viewerPowerPoint ITD framework The first step towards establishing the ITD interaction begins with the identification of nodes/buses/substations acting as PCC (Point of Common Coupling) for the interface. Defining the cardinality of such nodes would lead to identify the size of agents required for facilitating the ITD interaction. 2.1 PCC node The PCC node can be defined as a floating bus. It can be considered as a floating bus because there happens to be a range of operating points at which the states of PCC node may exist [16]. Once the states of the floating bus are obtained it can be set as PV or PQ bus depending upon the conditions prevailing in the network. 2.2 Multi-agent based system Multi-agent based system has been developed to materialise the interaction between the . The architecture of the MAS system has been shown in Fig. 1. The step-by-step implementation algorithm of MAS has been given in Table 1. The MAS consists of several agents which are employed for exchanging the information between the DSO and the TNO. The proposed MAS system can function in two different modes: • Top to Bottom mode: Here the TN is solved first and then after the states of the PCC are communicated to the ADN where DSO solves the ADN. • Bottom to top mode: In this mode of operation DSO solves the ADN depending upon the scenarios prevalent in the DN and states of the ADN are communicated via the agents to the PCC. Table 1. Implementation algorithm of MAS system Steps Description of the steps step 1 Specify the ADN nodes in transmission network. step 2 Surf the TN Topology and designate the ADN nodes as PCC. step 3 Allocate agents for designated PCC's in MAS (Multiagent System). step 4 Select the operating mode of MAS. step 5 Access the measured states (P, Q, V, , , ) from DSO using protocol into MAS agent vectors. step 6 Communicate the data from the MAS agent vectors to the PCC of TN through protocol. The number of agents in the MAS would be equal to the number of ADN's (i.e. number of PCC) considered. The agent represents the th state of th agent. Each agent has two major components TSO agent vector and DSO agent vector. Depending upon the architecture deployed for analysis these vectors would be determinantal for TSO or DSO side variables (1) (2) (3) Here, na is the number of agents is the number of component of th agent. In , agentware is a function containing two subscript having the protocols of information exchange for TN/DN to agent and vice-versa, the mode (1 or 2) indicates whether tr2agent (mode 1) or (mode 2) in case of top to bottom () and (mode 1) or (mode 2) in case of bottom to top () architecture would be invoked. The input parameters for agentware are transmission network information (TN), agent (initially a set of null matrix), mode. The agent contains the information pertinent to the TN node i, P & Q flexibility region of th interface, state variables at the th interface in the purview of ADN. 2.3 Interface architecture The ITD interface can be achieved by using any of the two different architectural topologies which are top to bottom and bottom to top topology. In bottom to top topology, the ADN demands and DERs drive the interface parameters through the DSO while in the top to bottom architecture the interface parameters are governed TNO through the TSO. The flexibility of controlling the ADN resources either by dispatching controls from TSO or DSO through a coordinated mechanism could be enabled for enhancing the functioning of the overall system. 3 Modelling of ITD & DER ATC assessment and enhancement would be affected by the uncertainties in the power produced by the DER'S. ATC has been conventionally evaluated while considering the predicted/forecasted value of the demand/generation. When the DER's are taken into consideration the degree of uncertainty increases substantially, this would adversely affect the ATC assessment. Thus appropriate models of DER's should be considered while evaluating the ATC. Further, the ATC enhancement could be achieved by utilising the FACTS devices, therefore this section presents the model of DER's and FACTS devices which have been used in this work. 3.1 Solar PV & smart inverter modelling The Solar PV has been modelled as a PQ load where P will be negative for generation (i.e. injection to grid) and Q would be positive or negative depending upon the operating conditions of the smart inverter through which the PV is connected. To consider the uncertainties in PV power output, the Beta distribution have been taken to realise the probability density function of solar irradiance for each hour [17]. The active and reactive power support available from the smart inverter [18] sourced by the PV module is determined by the following equations: (4) (5) (6) The maximum value of is a function of real power generation and is determined by (7) (7) 3.2 Wind modelling Synthetic wind power model developed from the data measured directly in the wind power domain is used for modelling the uncertainties of power generated from the wind farms [19]. The model uses Markov Chain Monte Carlo method for generating the time series samples for which the analysis is done. The static wind power characteristics of the wind turbines are typically non-linear and the relation between the wind speed v and power output is mathematically given by (8) The wind turbine characteristics within the range to has been modelled by linearising the non-linear curvature portion of the turbine characteristics, here and are nominal power output and function expressing output of wind power in terms of wind speed respectively. The linearisation has been done in the following manner: Identify the point at which there is an abrupt change in slope in terms of a, b, n ( and being starting and end points of the th slope ranging from 1 to (), n is the number of abrupt change). Divide the characteristic into different zones. Linearly model each zone in the form of ( constant; independent variable and y dependent variable) (9) The model fitted has an MSE (mean squared error) of 0.023. The MSE for the fitted model can be increased by increasing the number of regions in which the non-linear curve is broken for modelling the characteristic. 3.3 Modelling of FACTS Modelling of the devices namely SVC (static VAR compensator) and TCSC (thyristor controlled series compensator) as FACTS devices which are considered in this work are given hereunder. 3.3.1 TCSC modelling Diagrammatic representation of TCSC have been shown in Fig. 2. Equivalent reactance of TCSC can be represented as a function of its capacitive, inductive reactances along with the firing angle (10) where (11) (12) (13) Transfer admittance matrix of TCSC between nodes '' and '' is given by the following equation: (14) Power injections (active & reactive power) at bus 'k' are modified due to the presence of TCSC and are given as (15) (16) The power flow equations considering the TCSC model can be linearised as (17) In the above equation, is the active power mismatch. Fig. 2Open in figure viewerPowerPoint Diagrammatic illustration of TCSC modelling 3.3.2 SVC modelling Modelling of SVC has been materialised by representing the SVC characteristics with an adjustable reactance. An equivalent circuit representation of the model has been diagrammatically shown in Fig. 3. The current, drawn by the equivalent model of SVC is which is and is the reactive injected. Taking as a state variable, the linearised power balance equation as given in (18) would be obtained (18) Fig. 3Open in figure viewerPowerPoint Diagrammatic illustration of SVC modelling 3.4 ITD modelling The transmission and distribution system has been modelled by identifying the nodes at which the active distribution system is considered as a PCC (point of common coupling). The ADN has been represented by a combination of several DN feeders. (19) (20) (21) Here, is the number of feeders used to form ADN. To analyse the impact of the increase in the total load at the PCC has been decomposed into two parts as (i). Fixed PQ load and (ii) . The equivalent schematic representation of model has been shown in Fig. 4. ATC is usually computed for transaction from one area (source) to another (sink). In this work, the impact on ATC has been assessed therefore the presence of ADN in the sink area has been considered. The and its percentage in the sink area has been determined using the (19)–(21). The ADN comprises of several DERS such as PV DER, WIND DER and PV&WIND DER. The analysis presented in this work considers the deployment of CVR in the DN as a tool for optimal operation of ADN and subsequently assess its impact on ATC at TN level. Fig. 4Open in figure viewerPowerPoint Diagrammatic representation of aggregated Model of T&D System 4 ATC & CVR: a brief overview 4.1 ATC The amount of power which can be transmitted over and above the existing transmission commitments while providing for the capacity benefit margin (CBM) and transient reliability margin (TRM) is defined as ATC [17]. ATC can be mathematically given as: (22) In (22), TTC is the total transfer capability and ETC is the existing transmission commitments. Contingencies play a major role in determining the appropriate ATC value, the credible contingencies should be considered for assessing the ATC of the system. At transmission level, the transmission system operators (TSOs) strive to materialise a secure, economic and reliable operation of transmission network (TN) with a flat voltage profile and stringent frequency regulation. On the other hand, the objective of distributed system operator (DSO) is to operate the distribution network (DN) in such a way that the cost of operation and losses are minimum with redundant and reliable supply to the consumers. The DSO employs volt VAR optimisation, CVR, network reconfiguration, demand response management and other available techniques at their disposal to achieve their objectives. Some of the objectives may be contradictory at the transmission and distribution levels, for instance, CVR strives to operate the DN at the lower permissible limit of voltage deviation to facilitate demand reduction [20] on the contrary, the TSO strives for flat voltage profile through TNO (transmission network optimiser). To ensure proper functioning, the coordinated operation of the TSO and DSO should be performed so that the contradicting objectives could be dealt with appropriately. Also, the control techniques adopted in ADN for VVO (volt VAR optimisation) might as well affect the ATC of the transmission network, therefore the VVO control implemented through CVR has been considered while formulating the problem for assessment of ADN impact on ATC. 4.2 CVR The distribution feeders mostly have voltage-dependent type loads which are very sensitive to voltage variations. Many practical studies reveal that feeder load demand can be indirectly controlled by varying the voltage profile [21]. In this respect, the concept of (CVR) has been exploited to reduce the feeder peak load demand. Various studies on CVR have been conducted to analyse its effect [22]. However, most of the reported work deals with CVR effect on energy savings in the distribution system. It is worth mentioning that very few has demonstrated the impact of CVR in the transmission system. The authors of [23] have studied the CVR effect on voltage stability of the transmission system and the study reveals that voltage stability margin can be increased with the integrated operation of CVR. Motivated from the impact of CVR in voltage stability, in this study the effect of CVR on ATC has been addressed. Various schemes are already available in the literature to enable the CVR operation. In this study, a smart grid-enabled CVR methodology has been utilised. Basically, this scheme executes the CVR operation through optimal settings of Volt/Var regulation devices. However, the optimal set points have been determined using Volt/VAr optimisation engine operating with CVR as an objective function. In this work, the impact of enabling CVR is measured based on CVR savings that is the total reduction in power demand as defined under: (23) where is percentage savings achieved, is total power consumptions in the base case or without CVR and is the total power consumption after the CVR execution. The CVR savings achieved based on (23) in distribution feeders will be aggregated and its impact of ATC is further analysed using ITD platform. 4.3 Load model impact Savings achieved through CVR is greatly influenced by the type of load especially voltage-dependent loads. To model the loads, a method has been proposed in [24, 25] considering the practical aspects of user end behaviour. This study also utilises the similar load model to build a relationship among load power demand, voltage and CVR factor with exponential equations as delineated under (24) (25) where & are CVR factors in kW and kVAR and have been taken from [24]. 5 Problem formulation: ADN impact on ATC The overall problem of assessing the impact of deploying CVR in ADN on ATC involves the coordinated approach whereby the CVR at the distribution level and ATC at transmission level are simultaneously evaluated. Therefore the problem involves a platform which could facilitate the integrated operation and control of transmission and distribution systems. The ITD platform proposed in Section 2 has been used for this purpose. For simplicity of understanding the problem is segregated into three sections as:- 5.1 ATC: transmission level 5.1.1 Stage 1: Real Time Contingency Analysis Real time contingency analysis (RTCA) is a vital function of the modern energy management system. The RTCA is performed to identify the critical contingencies that would adversely affect the performance and reliability of the power system [26]. As optimisation is not required for performing RTCA, the computing process does not enforces any constraints. In this work an algorithm for RTCA is developed using MATPOWER and (26) is applied for contingency ranking as used in [27] (26) (27) In the above equations is the contingency index, are the normalised upper and lower limit violations for the alarm and security limits, and are the normalised power flow limits, n is the normalisation factor. The n is exponent used in hyper ellipse equation [27] and is taken as 2. The composite security index (26) is used for obtaining the contingency ranking. The contingency with the highest index is most severe and with least value of the index is least severe. Hence a set of credible contingencies (C) is formed using the function which forms a set of 'k' credible contingencies that would be used for ATC assessment. 5.1.2 Stage.2: ATC assessment and enhancement problem formulation The ATC as given in (22) can be formulated as an optimisation problem [17] and is given as . The maximisation is achieved while maintaining the power system operational constraints within limits. The ATC in the system is affected by the presence of FACTS devices in the system. The ATC assessment formulation considering the impact of FACTS devices and considering the set of credible contingencies is given in (28). (28) (29) (30) for to (number of credible contingencies considered from RTCA output) The optimisation is achieved while maintaining the equality and inequality constraints and inequality constraints given in (31) to (33). The equality and inequality constraints inadvertently represent the impediments of the power flow (i.e. power balance at each node along with the maximum and minimum limits on Voltages, Voltage angles, line flows and generators) along the upper and lower limits of compensation that could be provided by the FACTS devices (31) (32) (33) The Min (Max) optimisation given in (28) strives to maximise the ATC corresponding to each credible contingency available in the set of credible contingencies, provided by the RTCA and at the same time yields the minimum value of ATC. This is done because the ATC corresponding to the contingency which would yield the minimum value of ATC would be considered as the final value. If the system is scheduled considering minimum ATC, and any other credible contingency happens then the system would be able to operate reliably and securely. Contrarily, if the higher value of ATC is scheduled and the contingency corresponding to the minimum value of ATC happens than the system will not be able to satisfy its operational commitments. 5.2 CVR: distribution level The CVR is achieved in the distribution level through the ADMS by the DSO. The objective of the ADMS server is to minimise the load demand while enabling the CVR operation as given in (34) followed by system constraints as given here under. 5.2.1 Objective function (34) here (35) (36) here . Further, in the following equation . (37) 5.2.2 Constraints The objective mentioned in (34) have been met while adhering to the impediments delineated in the following equations (38) (39) (40) (41) (42) (43) (44) In addition to the above impediments, the constraints discussed in ADS Modelling have also been met. 5.3 Combined transmission & distribution: ITD objective The objective function for this stage is an amalgam of the two parts discussed in the previous sections and can be defined in (45) (45) (46) The objective of overall ITD operation (45) has been met by adhering to the impediments of both ATC evaluation and CVR deployment problems (31) to (33) and (38) to (44). 6 Optimisation technique The objective function of ITD (i.e. ATC assessment in the presence of ADN) has been solved by employing pattern search optimisation [28-30]. Being a direct search optimisation technique the pattern search optimisation does not require information pertinent to the gradient of the objective function. The pattern search
Referência(s)