Study on the performance indicators for smart grids: a comprehensive review

This paper presents a detailed review on performance indicators for smart grid (SG) such as voltage stability enhancement, reliability evaluation, vulnerability assessment, Supervisory Control and Data Acquisition (SCADA) and communication systems. Smart grids reliability assessment can be performed by analytically or by simulation. Analytical method utilizes the load point assessment techniques, whereas the simulation technique uses the Monte Carlo simulation (MCS) technique. The reliability index evaluations will consider the presence or absence of energy storage elements using the simulation technologies such as MCS, and the analytical methods such as systems average interruption frequency index (SAIFI), and other load point indices. This paper also presents the difference between SCADA and substation automation, and the fact that substation automation, though it uses the basic concepts of SCADA, is far more advanced in nature.

evaluations will consider the presence or absence of energy storage elements using the simulation technologies (i.e., MCS), the analytical methods such as systems average interruption frequency index (SAIFI), and other load point indices. And also, the load flow analysis and other parameter check are done by using the online available data and control using fuzzy controller. Various network configurations will be considered. A number of indices can be computed for the load point assessment [14]. Some of these indices are presented next:

System Average Interruption Duration Index (SAIDI)
SAIDI is calculated by dividing the sum of total customer interruption durations per year (∑ ) to total number of customers (∑ ). This index is expressed by using: where is outage duration, is total number of customers.

Customer Average Interruption Duration Index (CAIDI)
CAIDI is calculated by dividing the sum of total customer interruption durations per year (∑ ) to total number of customers affected (∑ λ ). This index is expressed by using: where λ is failure rate.

Expected Energy Not Served (EENS)
An optimization problem for planning and/or operational purposes using RERs with integrated energy storage technologies has to be considered for the development of SG. The objective of an optimization problem is to minimize EENS or Expected Unserved Energy (EUE), and cost minimization, subjected to various constraints such as the capacity of RERs, network capacity, voltage margin, storage, and reliability margin [15]. EENS is sum of each load ( ) times its outage duration ( ). Mathematically, it can be expressed as [16]: Typical distribution networks will be chosen for both the grid tied and stand-alone power system network. A special distribution load flow program will be utilized to analyze the network under different operating scenarios. The reliability of these systems will be computed based on the variability of RERs. The load flow study and reliability will be recomputed in the presence of energy storage systems (ESSs) and MCS techniques [17].

Vulnerability Assessment
Vulnerability assessment is used to determine, identify and rank the contingencies of the system. Power system is a very complex and vulnerable system. Vulnerability assessment in power system provides information on state of system which indicates the system's inability to be stable in any abnormal condition or an unforeseen catastrophic contingency. Vulnerability index is used to determine the strengths and weaknesses of the system against undesired events. There is a list of contingencies that may lead the power system to major blackouts and cascaded failures. Moreover, operating conditions are different according to the location and the system. Vulnerability assessment is a slow process, therefore the real time assessment is very difficult. In a new power system environment, there are lots of measurements are taken, and the verification of these measurements and the development of correlation with vulnerability assessment is quite complex. Evaluation of specific and accurate border line for vulnerability is also a difficult process [18].
There are different approaches used by various researchers for vulnerability assessment. In time domain approach, stability is determined through simulating the generator behavior. However, it is quite time-consuming process as it involves plenty of nonlinear and differential equations. We can also use a direct method approach using the energy functions. Stability of the system can be determined by comparing the energy value of the system with the critical energy value. This method is being widely used by utilities and researchers as it has a very high speed of convergence [19].
A new method of vulnerability border tracking is by using Partical Swarm Optimization (PSO) along with the Artificial Neural Networks (ANNs). ANN is used for the purpose of increasing the speed of convergence of PSO. PSO is used for better search technique, therefore, it has the benefits of both the systems through one single algorithm. This method is useful for the real time evaluation of vulnerability border. Similarly, many methods can be used together for different applications and combine them to get best optimization results. One such method is to use ANN for vulnerability assessment as it has very good speed of convergence and Fuzzy logic for vulnerability control applications. Fuzzy logic can also be used for locating short circuit faults, which is used for vulnerability contours. This method is useful both for online and offline applications [20]. Fault resistance is also taken into account. Fuzzy logic reasoning is applied to cope with the inherent uncertainty in the problem. The above methods can also be used as a combination of one or more methods and one can develop a hybrid method for vulnerability assessment. Moreover, adaptive dynamic programming (ADP) is also a new approach for vulnerability assessment.
Vulnerability assessment using phasor measurement unit (PMU) is also a new approach. PMUs are used to provide time synchronized data in the form of signals which can later be converted into data (voltage, angle) using various softwares, which contain dynamic information for voltages and angels, and even precursor signals for system collapse. A scheme can be developed to warn the system operator about severe conditions, vulnerabilities and to predict cascading failures using a pattern recognition and phase-space visualization using dynamic data received from PMU [21].
Nowadays, a situational awareness tool based on google maps is used for the advanced power system. It gives the latest system topology and helps the system operator to understand operational conditions not only his own region but also of the neighboring regions to avoid major blackouts [22]. This kind of visualization includes line descriptions, power flows and the status of outage lines, transportation and infrastructure impacts, geo-spatiotemporal information and impacts-population, weather impacts and analysis and predictions results. Moreover, the related data can be overloaded on the system topology. Hence, for various analyses, respective data is readily available. Two types of vulnerability indexes namely power system loss (PSL) and anticipated loss of load (ALL) are used for different contingences. Application of Geographical Information System (GIS) also helps in developing the vulnerability assessment [23].

SCADA Systems
Supervisory Control and Data Acquisition (SCADA) pertains to automation and the concepts of automation borrow from SCADA. Before it gets into the SCADA part, it presents the important terminologies and components of SCADA which are used in the substation automation [24]. The functions of each of the components are presented next: Remote Terminal Unit (RTU): It processes the data input (both analog and digital), and converts it into digital output which can either be seen on a single screen Human Machine Interface (HMI) in the control room itself or it can be transmitted over the Ethernet to other places for remote control. The function of Front-End Processors (FEP) is to act as an interface between the computer system, and the RTUs located locally and at remote substations. There are two FEPs at each site, both functional simultaneously, and also any one FEP capable of fully taking over the functions of the other FEP. Each FEP has a dual LAN interface and houses multiple Remote Channel Controller (RCC) modules, according to the number of RTUs connected to the control center. These RCC modules provide RS-232 interface ports for connecting to the RTUs [25]. The RCC modules are of microprocessor-based design and are able to: -operate independently and support a different RTU protocol on individual channels -utilize drivers to establish the RTU communications -conduct simultaneous RTU communications on each channel, acquire data, perform message security checks, and decode the data -process the data. -buffer all data for transfer to the controlling server.
The FEP transfers data to the controlling server in a timely fashion. The FEP also responds to the controlling server's demands for performing the required functions at the RTUs. The RCC channel capacity covers the entire complement of RTUs which consists of the new RTUs and the existing RTUs. The channels are expandable in the future to ultimate quantity by acquiring and inserting RCC modules. All critical RTUs are provided with 2 communication channels right from the control center up to the RTU, and these channels are connected one on each FEP. The Remote Communication Controller (RCC) interfaces with the FEP through the VME bus. Redundant FEPs and RCCs are provided with automatic changeover from one to the other when any FEP or RCC fails. When a modem fails or a communication link to a critical RTU goes down, there is automatic changeover from the defective link/modem to an alternate link/modem. All critical RTUs are provided with redundant communication channels [26].
Critical RTUs have the capability to switch between redundant communication channels when the system detects a communication channel failure. To satisfy the redundancy requirements, each communication channel is switched between a primary and backup port under failure conditions. In the GE Harris Energy Control Systems implementation, redundancy is provided all the way up to the remote RTU communication interface, using redundant FEPs, RCCs and separate channels to connect to the RTU. Any single failure is protected against by this method [27]. a. Human Machine Interface (HMI): The HMI/SCADA industry was essentially born out of a requirement for a user friendly front-end to control system containing programmable logic controllers (PLC). b. Central Control Room Computer: Usually, the HMI/SCADA presents the information in the form of a mimic, which means that the operator can see a representation of the plant being controlled. c. Transducers: They are the energy converters which convert one from of energy to another. It is used in substations to convert the measured AC voltages and currents (measured from the CT and CVT inputs) and convert them into a common DC current. They can be self-powered or separately powered. In self powered transducers, the components of the transducer are powered by the source itself, where as in separately powered transducers, there is an external source which is used to run the circuit of the transducer. d. Multiplexer: It is a device that can interleave two or more activities. For example, a 16:1 multiplexer can take 16 different inputs and can create a time sharing mechanism that will allow it to give the required output of the 16 inputs based on a predefined logic. The simplest logic generally used is a clock pulse. Though automation basically involves all the concepts of SCADA and though they are terms that are used interchangeably, there exist certain basic differences between them [28], and they are listed below: -Conventional SCADA deals with data acquisition and control of most of the equipment in the substation, but it cannot be used for relays for the basic reason that they have to be manually operated. Whereas, the substation automation involves automation of relays and including them in the data acquisition and control process. The numerical relays used in substation automation systems act as a virtual RTU. In other words, the SCADA system applied to protection can be referred to as substation automation systems. -SCADA uses the concept of transducers, which are electronic devices that invariably cause a lag or delay in the transmission of data, whereas substation automation system aims at minimizing the data transmission time by removing the use of transducers. -Substation automation system needs to have higher accuracy than the conventional SCADA system as it involves the protection of switchgear of a substation. Thus, it removes the transducers, which approximate the values, and the obtained values directly from the devices. -Conventional SCADA has the concept of a single RTU, where the data is acquired and stored. The required control is performed form the centralized RTU itself. In such a case, a malfunction even in a part of the RTU affects the whole system. Whereas, in a substation automation system, the information and control are decentralized, i.e., each bay acts as a virtual RTU itself and sends the information into the control room.