Reiss SP Incremental maintenance of software artifacts. Robinson WN Seeking quality through user-goal monitoring. IEEE Software 26 5 : 58— J Syst Software 81 5 : — Carnegie Mellon University.
J Soft Maint 80 1 : d—d Taylor RN, Hoek Avd Software design and architecture: the once and future focus of software engineering.
Thompson CW Smart devices and soft controllers. In: International conference on embedded software and systems, pp — In: International symposium on information engineering and electronic commerce, pp — Werbos PJ Putting more brain-like intelligence into the electric power grid: What we need and how to do it.
In: International joint conference on neural networks, pp — Yang L, Schopf JM, Foster I Conservative scheduling: using predicted variance to improve scheduling decisions in dynamic environments. In: IEEE transactions on services computing. Mehboob Z, Zowghi D, Lowe D An Approach for comparison of architecture level change impact analysis methods and their relevance in web systems evolution.
In: Australian software engineering conference, pp — J Soft Maint 14 5 : — Download references. You can also search for this author in PubMed Google Scholar. Correspondence to Norman Schneidewind. Reprints and Permissions. Schneidewind, N. Representativeness models of systems: smart grid example.
Innovations Syst Softw Eng 7, 23—41 Download citation. Therefore, the SDN-microSENSE frameworks mentioned earlier focus on the following cybersecurity-related research areas: a Risk assessment, b intrusion detection and correlation and d self healing and recovery.
Therefore, according to the above specifications, the con- ceptual frameworks are placed within the Data, Controller, Application, and Management Planes. The conceptual frameworks and their components are placed within the Application Plane. In this plane, the most important operational decisions take place, such as the detection of a cyberattack or the decision to isolate a malicious network flow. It is worth men- tioning that the Management Plane is placed vertically since it provides complementary services to all planes.
Figure 2. Thus, following this methodology, S-RAF receives the security events and alerts coming from XL-EPDS and incorporates into this information a cumulative risk value for each involved asset and the corresponding connections. In other words, honeypots are commonly used as an extra security layer in order to act as a decoy, which lures the cyberattackers and captures useful information about their identity and activities [37].
In more detail, the IEC honeypot emulates real intelligent electronic devices usually located in circuit breakers of the substations by parsing the Intelligent Capability Description ICD files. To this end, the framework integrates a SIEM system especially designed for the energy sector.
The security detec- tors are deployed throughout the EPES infrastructure and undertake to recognise various EPES cyberattacks and anomalies, generating the respective security logs. First, Suricata is used with the Quickdraw ICS signatures and specification rules devel- oped during the project. Moreover, there is a detector called Nightwatch, which is able to discriminate poten- tial anomalies related to the entire SDN network based on the statistics given by the SDN-C.
Finally, the Discovery tool constitutes a visual-based anomaly detector, which provides the appropriate visual interfaces through which the security administrator can distinguish the presence of an anomaly that possibly cannot be detected by the aforementioned detectors.
They are also accompanied by the risk information calculated by S-RAF. Therefore, ARIEC allows storing and sharing technical details of the cybersecurity incidents among different EPES organisations belonging to a trusted network without identifying the victim identity or other sensitive information that can affect the reputation of the EPES organisation.
In comparison to existing state of the art, refs. Supposing the communication quality is degraded in a manner that criteria of minimum latency cannot be satisfied. In that case, EDAE employs the PaDe [45] genetic algorithm in order to decompose the multi- objective problem of path reconstruction to multiple single-objective ones that are resolved using the asynchronous generalised island model to distribute the solution process [46].
The final solution i. In this case, a Mixed-Integer Linear Programming algorithm chooses and applies the best PMU reallocation scheme to minimise the overall network latency.
This problem is also studied by [40]; however, authors are limited to maximising observability, while EDAE also addresses QoS and security requirements. Activated as a response to specific security incidents received from S-RAF, IIM collects information regarding the triggering event, as well as the current status of the grid, and delivers appropriate islanding recommendations, which are evaluated and applied by the system operator.
More specifically, the islanding solutions aim to partition the grid into several segments, creating islands that isolate the affected assets and at the same time minimise the power imbalance while maintaining supply to the maximum number of consumers.
IIM employs two different methods for calculating the islanding schemes, namely: 1 a genetic algorithm, which provides the optimal solution at the cost of increased time-complexity and 2 a deep learning architecture [31] which addresses the islanding problem by utilising graph convolutional neural networks, able to provide the solution in real-time.
Towards this goal, EMO continuously observes the grid status, aiming to identify islanding cases and automatically commences the required restoration and management processes, ensuring the real-time operation through the optimal allocation of the network capacity. At its core, EMO consists of two modules, the first responsible for the economic management of the power flow between the DERs and the second undertaking to control the voltage-reactive and the frequency-active power, based on a hybrid multi-agent system that optimally allocates the requested energy between the units.
The e-auction module establishes secure and trustworthy networks among the parties involved in energy transactions, including consumers and prosumers and Energy Service Company Organisations that manage the financial transactions. The communication among the participants is performed through a fabric blockchain network based on the Hyperledger Fabric. Finally, the status of each participating device e. The loop-free topology relies on EDAE in order to apply optimisations and enable redundant paths.
The SDN-C undertakes to program the underlying intermediary network devices i. The SDN-microSENSE platform acts as a decision support system for the TSO in order to decide on intentionally islanding segments of the affected grid or to shed redundant load in order to balance energy demand and supply [49].
Unarguably, the SDN technology is one of the main enablers that pave the way to a holistic cybersecurity solution that addresses detection and mitigation of cyberthreats.
However, it should be noted that SDN introduces new organisational and technical chal- lenges for potential end-users. First of all, the required technologies e. On top of that, compatibility and vendor inte- gration issues may arise due to vendor-specific implementations that deviate from the standards. To sum up, despite its benefits on network management, SDN may introduce unforeseen technical and manage- rial complications, increase financial costs during adoption, and possibly be rejected by the management if the drawbacks outweigh the benefits [50].
However, this progression creates severe cybersecurity and privacy issues. S-RAF applies a collaborative and dynamic risk assessment, thus determining the risk related to each security event and alert. Author Contributions: Conceptualization, P. Achilleas Sesis , A.
Antonios Sarigiannidis , R. All authors have read and agreed to the published version of the manuscript. Institutional Review Board Statement: Not applicable. Informed Consent Statement: Not applicable. Data Availability Statement: Not applicable. Conflicts of Interest: The authors declare no conflict of interest.
The funders had no role in the design of the study; in the collection, analyses, or interpretation of data; in the writing of the manuscript; or in the decision to publish the results. Tan, S. Survey of security advances in smart grid: A data driven approach. IEEE Commun. Alshamrani, A. A survey on advanced persistent threats: Techniques, solutions, challenges, and research opportunities.
Stellios, I. Advanced persistent threats and zero-day exploits in industrial Internet of Things. Di Pinto, A. Radoglou-Grammatikis, P. Implementation and Detection of Modbus Cyberattacks.
Darwish, I. Cyber Secur. Securing the Internet of Things: Challenges, threats and solutions. Internet Things , 5, 41— Kumar, P. Smart grid metering networks: A survey on security, privacy and open research issues. A survey of iot-enabled cyberattacks: Assessing attack paths to critical infrastructures and services. Hassan, M. Differential privacy techniques for cyber physical systems: A survey.
Karimipour, H. A deep and scalable unsupervised machine learning system for cyber-attack detection in large-scale smart grids. IEEE Access , 7, — Nguyen, T. IEEE Access , 8, — Securing the smart grid: A comprehensive compilation of intrusion detection and prevention systems. Rehmani, M. Software defined networks-based smart grid communication: A comprehensive survey. Musleh, A. A survey on the detection algorithms for false data injection attacks in smart grids.
IEEE Trans. Smart Grid , 11, — It practically provides one-to-one mapping to make sure that the creation of two messages that produce the same digest is impossible. The approach of signing with the message digest is explained in Figure 4. The standard proposes different mechanisms to protect IEDs. The password should be a minimum of 8 characters with at least one upper and lower cases, one number and one alpha-numeric character. References [1] Tanenbaum, A. This overloading will worsen as large numbers of electric vehicles, heat pumps and other new loads use low-carbon energy from the electric power system.
Therefore, demand-side programmes have been introduced widely to make better use of the existing power supply infrastructure and to control the growth of demand. The dual aims of reducing CO2 emissions and improving energy security energy policy goals in many countries coincide in the increasing use of renewable energy for electricity generation.
One solution to this increase in variability is to add large-scale energy storage devices to the power system. This is often not practical at present due to technical limitations and cost. Load control or load management has been widespread in power system operation for a long time with a variety of terminology used to describe it. To avoid the confusion caused by such overlapping concepts and terminologies, as recommended by CIGRE, Demand-Side Integration DSI is used in this chapter to refer to all aspects of the relationships between the electric power system, the energy supply and the end-user load.
However, the electro-mechanical meters that are presently installed in domestic premises have little or no Smart Grid: Technology and Applications, First Edition. Smart metering refers to systems that measure, collect, analyse, and manage energy use using advanced ICT.
The concept includes two-way communication networks between smart meters and various actors in the energy supply system. The most common type of meter is an accumulation meter, which records energy consumption over time. Accumulation meters in consumer premises are read manually to assess how much energy has been used within a billing period.
In recent years, industrial and commercial consumers with large loads have increasingly been using more advanced meters, for example, interval meters which record energy use over short intervals, typically every half hour. Smart meters are even more sophisticated as they have two-way communications and provide a real-time display of energy use and pricing information, dynamic tariffs and facilitate the automatic control of electrical appliances.
Figure 5. Since , there has been a dramatic increase in the performance of the metering infrastructure being installed. The Smart Grid vision represents a logical extension of these capabilities to encompass two-way broadband communications supporting a wide range of Smart Grid applications including distribution automation and control as well as power quality monitoring.
The differences between conventional metering and smart metering are shown schematically in Figure 5. The Gateway2 allows the transfer of smart meter data to energy suppliers, Distribution Network Operators DNOs and other emerging energy service companies.
They may receive meter data through a data management company or from smart meters directly. In order to integrate smart metering into the operation and management of the power system, interfaces to a number of existing systems are required, for example, the interface to the load forecasting system, the Outage Management System OMS , and a Customer Information System CIS see Chap- ter 7 for more details.
The disc is situated in between two coils, one fed with the voltage and the other fed with the current of the load. The torque is proportional to the product of instantaneous current and voltage, thus to the power. Electronic meters not only can measure instantaneous power and the amount of energy consumed over time but also other parameters such as power factor, reactive power, voltage and frequency, with high accuracy.
Moreover, electronic meters are not sensitive to external magnets or orientation of the meter itself, so they are more tamperproof and more reliable. Early electronic meters had a display to show energy consumption but were read manually for billing purposes. More recently electronic meters with two-way communications have been introduced.
In Fig- ure 5. The fundamental electrical parameters required are the magnitude and frequency of the voltage and the magnitude and phase dis- placement relative to the voltage of current. Current and voltage sensors measure the current into the premises load and the voltage at the point of supply. In low-cost meters the measuring circuits are connected directly to the power lines, typically using a current-sensing shunt resistor on the current input channel and a resistive voltage divider on the voltage input channel Figure 5.
The current rating of this shunt resistor is limited by its self-heating so it is usually used only in residential meters maximum current less than A. The voltage resistive divider gives the voltage between the phase conductor and neutral. The alloy Manganin is suitable for the resistive divider due to its near constant impedance over typical operating temperature ranges.
Example 5. When the load current is at the rated value of the meter, calculate: 1. Answer 1. A Current Transformer CT can also be used for sensing current and providing isolation from the primary circuit. A CT can handle higher currents than a shunt and also consumes less power. The disadvantages are that the nonlinear phase response of the CT can cause power or energy measurement errors at low currents and large power factors, and also the higher meter cost.
Some applications may require smart meters with high precision over a wide operating range. Detailed explanations of the Rogowski coil and optical methods are given in Chapter 6.
When it comes to physical implementation, the signal conditioning stages can be realised as discrete elements or combined with the ADC as part of an Integrated Circuit. To avoid inaccuracy due to aliasing, it is necessary to remove components of the input signal above the Nyquist frequency that is, half the sampling rate of the ADC. The sampling frequency is determined by the functions of the meter.
What should be the minimum sampling frequency used in the signal conditioning stage? Assume that the frequency of the supply is 50 Hz. This is shown in Figure 5. Since there are two signals current and voltage in a single phase meter, if a single ADC is used, a multiplexer is required to send the signals in turn to the ADC.
The ADC converts analogue signals coming from the sensors into a digital form. As the number of levels available for analogue to digital conversion is limited, the ADC conversion always appears in discrete form. The higher the number of bits used in the ADC, the lower the resolution. It uses a : 5 A CT for current measurements and 10 V potential divider for voltage measurements.
When the meter shows a current measurement of 50 A and a voltage measurement of V, what is the maximum possible error in the apparent power reading due to the quantisation of the voltage and current signals? There are many established methods for conversion of an analogue input signal to a digital output [3, 4, 5]. The majority of the methods involve an arrangement of comparators and registers with a synchronising clock impulse.
The most common ADCs for metering use the successive approximation and the sigma-delta method. The counter output is converted into an analogue signal using a Digital to Analogue Converter DAC and compared with the analogue input by a comparator.
If the analogue input signal is larger than the DAC output, then the up-down counter sets the MSB and the next bit to 1 and the comparison is repeated. This process is repeated until the analogue input signal is the same as the DAC output. At that point the DAC input will be same as the digitised value of the analogue signal.
Therefore, the integrator output reduces linearly rate determined by the difference of signals at A actual signal and E analogue value of the digitised signal. This negative feedback loop works such that the signal at D becomes the digitised signal of input signal at A.
In real implementations instead of the integrator in Figure 5. Further, the comparator is also in digital form, thus analogue errors are not accumulated [6]. The block diagram shown in Figure 5. Due to the relatively large number of arithmetic operations Table 5. In addition to routine arithmetic operations, a meter deals with a large number of other procedures that is, payment, tamper detection, system updates, user interactions as well as other routine tasks for example, the communication of billing information.
Table 5. Volatile memory is used for temporary storage of data to support the processor s as operations are undertaken. Additionally data related to energy consumption should be retained until successful communication to the billing company has been achieved. For this purpose a real-time clock is used. The accuracy of the real-time clock can vary with temperature.
In order to maintain this function during system power losses or maintenance, a dedicated clock battery is typically used. Both display types are available in seven-segment, alphanumeric and matrix format. Smart meters provide a small key pad or touch screen for human—machine interaction, for instance, to change the settings of a smart meter so as to select the smart appliance to be controlled or to select payment options. As smart meters require calibration due to variations in voltage references, sensor tolerances or other system gain errors, a calibration input is also provided.
Some meters also provide remote calibration and control capability through communication links. Energy consumption and tariffs may be displayed on a separate customer display unit located in an easily visible location within the residence for example, the kitchen. This is to encourage customers to reduce their energy use, either throughout the year or at times of peak demand when generation is short.
This is used with a Time of Use Tariff to control peak demand. It is hoped that customers will manage and reduce their energy consumption when they are provided with more accurate, up-to-date information, also that any reduction made soon after the display is installed will be maintained. These techniques are described in Chapter 2. A HAN uses wired or wireless communications and networking protocols to ensure the interoperability of networked appliances and the interface to a smart meter.
It also includes security mechanisms to protect consumer data and the metering system. A HAN enables centralized energy management and services as well as providing different facilities for the convenience and comfort of the household. Energy management functions provided by HAN include energy monitoring and display, controlling the HVAC system and controlling smart appliances and smart plugs.
Home-based multimedia applications such as media centres for listening to music, viewing television and movies require broadband Internet access across the HAN. A separate HAN used for energy services can coexist with the broadband Internet system but there is some expectation that the systems will be merged in the future. In order to provide demand management functions and demand response, two options are being actively considered in different countries Figure 5.
One option is to use the smart meter as the interface to the suppliers, network operators and other actors. The other option is to use a separate control box [7, 8] which is directly interfaced to the outside world through the NAN and WAN. The communication technology used for the NAN is based on the volume of data transfer. It manages the meters by automatically detecting them, creates and optimises repeating chains if required to establish reliable communication , coordinates the bi-directional delivery of data, and monitors the conditions of the meters.
In these European examples, PLC is used between the meter and data concentrator as the last mile technology and GPRS is used between the concentrator and gateway to the data management system. In Table 5. With local AMR, the meter readings are collected by staff using hand-held devices and with remote AMR the meter readings are collected from a distance through communication links. For most protocols listed in Table 5. The important factors for consideration when assessing communication protocols for smart metering are summarised in Table 5.
DSI can help defer investment in new infrastructure by reducing system peak demand. This encompasses the planning, implementation and monitoring of activities designed to encourage consumers to change their electricity usage patterns.
Load shifting is the movement of load between times of day from on-peak to off-peak or seasons. In Figure 5. Peak clipping Figure 5. As peak clipping reduces the energy consumed by certain loads in Fig- ure 5. However, for the most effective DSI, the utility needs to know not only which loads are installed in the premises but which are in use.
In this case two-way communication between the smart meter and network operators is necessary. However, the immediate consequence is the increase in electricity load causing potentially serious operational problems to both distribution and generation systems. Appropriate control and management are required if plant is not to be overloaded. Demand-Side Integration describes a set of strategies which can be used in competitive electricity markets to increase the participation of customers in their energy supply.
Customers are able to sell energy services either in the form of reductions in energy consumption or through local generation. Traditionally electric power systems were designed assuming that all loads would be met whenever the energy is requested. The coincidence of domestic demand follows the shape of Figure 5.
Consider a peak clipping control that sends a signal to switch off one hundred 3 kW water heaters that operate under thermostatic control.
Although water heaters have been installed, only, say, 20 will be drawing power at any one time. Thus the peak will be reduced by 60 kW. When, after, say, two hours, the water heaters are reconnected, all the water tanks will have cooled and a load of kW will be reconnected. Thus DSI measures must consider both the disconnection of loads but also their reconnection and the payback of the energy that has not been supplied.
It is much easier to manage both the disconnection of loads and their reconnection with bi-directional communications whereby the state of the loads can be seen by the control system. Calculate the percentage reduction in energy loss in the 33 kV line if load shifting shown in Figure 5. Price-based DSI encourages customer load changes in response to changes in the electric- ity price. Various DSI programs are deployed and integrated within the power system core activities at different time scales of power system planning and operation, as shown in Figure 5.
The basic rate structure is ToU. The various pricing schemes are illustrated in Figure 5. Two-part RTP tariff designs include a historical baseline of customer use, added to hourly prices only for marginal use above or below the baseline. Customers thus see market prices only at the margin.
CPP uses real-time price at major system peaks. The CPP prices are restricted to a small number of hours per year, where electrical prices are much higher than normal peak prices, and their timing is unknown ahead of being called. In price-based systems, the response of demand to price signals determines the DSI perfor- mance. Price elasticity is a measure used in economics to show the responsiveness, or elasticity, of the quantity demanded of a good or service to a change in its price.
Elasticity of substitution: is a measure of the percentage change in the ratio of the peak to off-peak demand as a result of a percentage change in the ratio of the peak to off-peak price. Long-term price elasticity: is the annual energy consumption response to an average change in energy price. In addition, load control switches, controllable thermostats, lighting controls and adjustable speed drives are required.
Such equipment receives signals such as alarms or price signals and controls loads accordingly. It is wired into the control circuit of an air conditioning system, a water heater or a piece of thermal comfort equipment. The time that the appliance will remain disconnected is generally pre-programmed through an inbuilt clock. The DSI program operator or a HAN can increase or decrease the temperature set point through the communication module, changing the functioning of the equipment and hence the electricity load.
Lighting control strategies for energy consumption reduction are listed in Table 5. Estimated energy savings are presented for each case. These savings are based upon estimated average consumption, the time of use and user behaviour.
The loads of the majority of motorised appliances change over time and equipment is often operated at less than full load. ASDs allow the motors to satisfy the required functioning conditions and to economise power and energy use when the system is not functioning at its maximum load. Replacement of this system by an ASD can yield considerable saving of energy.
Such customers not only consume electricity, but also are able to manage their capacity to supply power to the grid. Hence they are also called prosumers. The controllability of active power is fundamental for the commercial integration of prosumers. DER connected to low and medium voltage levels will have more opportunities to provide local network services than generation connected at higher voltage distribution networks. It is triggered by under-frequency relays when the frequency drops under a certain threshold, for example, Load shedding is planned by the TSO but is implemented by the DNOs who arrange the tripping of distribution feeders and choose which feeders are tripped.
In order to maintain frequency, NGET buys frequency response services. When the frequency goes up, high frequency response is used to reduce the power output of the large generators and hence the frequency. A sudden drop in frequency is con- tained using primary response Figure 5. This should be delivered within 10 seconds and maintained for another 20 seconds [15]. The system frequency is brought back to normal using secondary response which lasts from 30 seconds to 30 minutes.
If the frequency continues to drop below Primary and secondary response is usually provided by partially loaded generators increas- ing their output. Primary response requires energy to be used that is stored as high pressure steam in the boiler drum and so only a fraction of the de-loaded power can be used. Large loads that are contracted to provide frequency response are typically steelworks or aluminium smelters though hospitals and banks that have their own generators can also take part in this market.
Using load in this way reduces the system operating cost and, depending on the alternative generation being used, CO2 emissions. DLC directly switches off loads to balance supply and demand during emergency conditions such as sudden loss of generation. References [1] Kreith, F. International Journal of Distributed Energy Resources, 4 4 , — In larger power systems, regional control centres serve an area, with communication links to adjacent area control centres.
In addition to this central control, all the generators use automatic local governor and excitation control. Traditionally, the distribution network has been passive with limited communication be- tween elements. Some local automation functions are used such as on-load tap changers and shunt capacitors for voltage control and circuit breakers or auto-reclosers for fault manage- ment.
These controllers operate with only local measurements and wide-area coordinated control is not used. Over the past decade, automation of the distribution system has increased in order to improve the quality of supply and allow the connection of more distributed generation. The connection and management of distributed generation are accelerating the shift from passive to active management of the distribution network.
Network voltage changes and fault levels are increasing due to the connection of distributed generation [1]. Without active management of the network, the costs of connection of distributed generation will rise and the connection of additional distributed generation may be limited [2].
Traditionally, the secondary circuits of the circuit breakers, isolators, current and voltage transformers and power transformers were hard-wired to relays.
Relays were connected with multi-drop serial links to the station computer for monitoring and to allow remote interrogation. Figure 6. Two possible connections marked by boxes of the substation equipment are shown in Figure 6. Although it may vary from design to design, generally it comprises three levels: r The station level includes the substation computer, the substation human machine interface which displays the station layout and the status of station equipment and the gateway to the control centre.
In connection 2, analogue and digital signals1 received from CTs and VTs are digitised by the interfacing unit. The process bus and station bus take these digital signals to multiple receiving units, such as IEDs, displays, and the station computer that are connected to the Ethernet network.
To increase reliability, normally two parallel process buses are used only one process bus is shown in Figure 6. The station bus operates in a peer-to-peer mode. The hard-wiring of traditional substations required several kilometres of secondary wiring in ducts and on cable trays.
In modern substations as inter-device communications are through Ethernet and use the same communication protocol, IEC , both the cost and physical footprint of the substation have been reduced. When a short circuit fault occurs, the current may increase to more than 20 times the normal load current. Current transformers CTs are used to transform the primary current to a lower value typically 1 or 5 A maximum suitable for use by the IEDs or interfacing units.
The majority of CTs, which are now in service, are iron cored with a secondary winding on the core. The primary is often the main circuit conductor forming a single turn. The operating principle of these transformers can be found in [3, 6, 7].
Measurement CTs are used to drive ammeters, power and energy meters. They provide accurate measurements up to per cent of their rated current. In contrast, protection CTs provide measurement of the much greater fault current and their accuracy for load current is generally less important.
The accuracy limit can be 5, 10, 20 or This number indicates the secondary terminal voltage that the transformer can deliver to a standard burden at 20 times the rated current without exceeding an accuracy of 10 per cent. More details about these current transformers can be found in the respective standards [8, 9].
Example 6. Using the equivalent circuit of the CT [3, 10], obtain the percentage current magnitude error and phase displacement error for the rated current and for the current at the accuracy limit when the rated burden is connected to the secondary. Multiple use of the same digitised measurement requires high accuracy CTs that measure both load and fault currents. While iron cored CTs and hybrid CTs an iron cored CT with an optical transmit- ter remain the most widely used CTs in the power system, high accuracy designs such as the Rogowski coil formed on a printed circuit board and optical CTs [3, 11, 12] are becoming available.
The secondary winding of the coil is a multi-layer printed circuit board as shown in Figure 6. The voltage induced on the secondary windings due to primary current see Box 6. Box 6. Optical CTs use the Faraday effect Box 6. Some other designs are being developed based on the Faraday effect that use a disc of an optically active material around the conductor [3, 7, 11]. The light enters the disc from one side and travels around it thus around the conductor and is collected at the other end.
The secondary voltage used is usually V. At primary voltages up to 66 kV, electromagnetic voltage transformers similar to a power transformer with much lower output rating are used but at kV and above, it is common to use a capacitor voltage transformers CVT.
As the accuracy of voltage measurements may be important during a fault, protection and measuring equipment are often fed from the same voltage transformer VT. Accuracy classes such as 0. For example, Class 0. The basic arrangement of a high voltage CVT is a capacitor divider, a series reactor to compensate for the phase shift introduced by the capacitor divider and a step-down transformer for reducing the voltage to V.
For applications up to 11 kV, optical CVTs are now available. Due to the lower voltage involved the inductor and transformer in Figure 6. In this arrangement there is no L-C circuit to resonate, and hence no oscillations, over-voltages or any possibility of ferro-resonance.
For example, the relay IED shown in Figure 6. A user may be able to read these digitised measurements through a small LED display as shown in Figure 6. Furthermore, a keypad is available to input settings or override commands. Various algorithms for different protection functions are stored in a ROM. For example, the algorithm corresponding to Type 50 continuously checks the local current measurements against a set value which can be set by the user or can be set remotely to determine whether there is an over-current on the feeder to which the circuit breaker is connected.
IEDs have a relay contact that is hard-wired in series with the CB tripping coil and the tripping command completes the circuit, thus opening the CB. A typical meter IED measures voltage, current, power, power factor, energy over a period, maximum demand, maximum and minimum values, total harmonic distortion and harmonic components.
Continuous event recording up to a resolution of 1 ms is available in some IEDs. These records are sometimes interrogated by an expert to analyse a past event. This fault recorder records the pre-fault and fault values for currents and voltages. The disturbance records are used to understand the system behaviour and performance of related primary and secondary equipment during and after a disturbance.
The bay controller facilitates the remote control actions from the control centre or from an on-site substation control point and local control actions at a point closer to the plant. The functionalities available in a bay controller can vary, but typically include: r CB control r switchgear interlock check r transformer tap change control r programmable automatic sequence control.
Modern RTUs, which are microprocessor-based, are capable of performing control functions in addition to data processing and communication. Some modern RTUs have the capability to time-stamp events down to a millisecond resolution.
Transmission systems use fast-acting protection and circuit breakers to clear faults within around ms. In contrast, the time-graded over-current protection of distribution circuits and their slower CBs only clear faults more slowly, typically taking up to ms.
Fast clearance of faults is important for industrial, commercial and increasingly for domestic premises. Many industrial processes rely on motor drives and other power electronic equipment which is controlled by microprocessors. This equipment is becoming increasingly sensitive to voltage dips [13, 14]. During a fault on the AC network, depending on the location of the fault, the voltage will drop. The subsequent operation of the ITE depends on the fault clearance time and the voltage dip.
For example, for a fault that creates a 40 per cent voltage dip 60 per cent retained voltage for ms, there is no damage to the ITE but its normal operation is not expected. However, for a fault that creates a 20 per cent voltage dip 80 per cent retained voltage for ms, the ITE should work normally.
Short-circuit faults are inevitable in any distribution system and so interruption in function of sensitive load equipment can only be avoided by doing the following: r ensuring the load equipment is robust against these transient voltage changes; r using very high speed protection and circuit breakers; r adding equipment to mitigate the voltage depressions for example, a Dynamic Voltage Restorer DVR or STATCOM [14, 16].
A sustained electricity outage may lead to severe disruption and economic loss, especially for industrial processes. Hence many Regulators impose penalties for the loss of electricity to customers in the UK for interruptions over 3 minutes. In these circumstances, distribution network operators are concerned to increase the speed of isolating the fault and restoring supply. Increasingly they are applying automatic supply restoration techniques which use automatic reclosers, remotely controlled switches, remote measurements and sometimes local Agents a piece of software running in a local computer.
This is achieved through a range of equipment generally known as switchgear. The term switchgear includes: a circuit breaker which is capable of making and breaking fault currents, a recloser which is essentially a CB with a limited fault-breaking capacity and variable pattern of automatic tripping and closing, a switch disconnector which has a limited fault-making capability and which is capable of making and breaking normal load current , and a sectionaliser which is capable of making and breaking normal load current but not the fault current.
The protection and metering equipment of substation switchgear is housed inside the substation control room. A typical RMU consists of two switch-disconnectors and a switch-fuse2 or circuit breaker Figure 6. For clarity, only switch-disconnectors are shown in Figure 6. The fuse switch either acts directly on the switch-disconnect trip bar or is automated by connecting a shunt trip coil.
In modern switchgear, instead of a motor wound spring, compressed gas or magnetic actuators are increasingly used [17].
Therefore a programmable logic controller which was hard-wired to the RTU was used to control the motorised switchgear. Hence either auto-reclose of the substations CBs or a self-controlled recloser, which can per- form a variable pattern of tripping and reclosing, is used on many overhead distribution circuits.
These will prevent unnecessary sustained outages for temporary faults. Most reclosers have instantaneous and time-delayed tripping characteristics. Standard practice is to use a number of instantaneous trips followed by time-delayed trips. The upper sequence of Figure 6. The purpose of the delayed tripping see the lower sequence of Figure 6. In Figure 6. Note: Only one phase is shown [7, 18] some instances the delay time allows the fault to burn itself out so that a subsequent reclose restores the supply.
Pole-mounted reclosers are widely used in distribution circuits. They have different voltage ratings for example, 11, 15, 33 kV and interrupting currents of 8 to 16 kA.
The energy required to operate the reclosing arrangement is provided by a solenoid as shown in Figure 6. In this arrangement, whenever the line current is high due to a fault , the series trip coil opens the vacuum CB. As the CB is opened, the auxiliary contacts are closed automatically, thus providing energy for the reclosing operation. A distribution feeder which employs a pole-mounted recloser is shown in Figure 6.
The recloser characteristic is selected so as to make sure that its fast operating time is much faster than the operating time of the downstream fuses and its slow operating time is slower than the operating time of the fuses see Figure 6. An RTU can be incorporated with a pole-mounted recloser for remote switching and chang- ing its settings remotely see Figure 6.
A sectionaliser is an automatic isolator which can only be used to isolate a section of a distribution circuit once a fault is cleared by an upstream recloser. When the recloser opens, a self-powered control circuit in the sectionaliser increments its counter. The recloser is reclosed after a short delay to see whether the fault is temporary. If the fault is temporary and cleared by the recloser, the sectionaliser resets its counter and comes back to its normal state.
If the fault is downstream of the sectionaliser, the recloser then restores supply to the upstream section. This minimises the number of consumers affected by a permanent fault, and a more precise indication of the fault location is provided. Network A has fuses to protect each line section whereas network B has sectionalisers. The characteristic of fuse F4 solid curve and that of the pole-mounted recloser dotted curve is shown in Figure 6. The sectionaliser has a pick-up current of A.
Discuss the operation of the protection arrangements in the two circuits for a temporary fault as shown that produces a fault current of A and A. Then after a delay, the recloser is reclosed. If the fault is temporary, the supply will be restored to all the loads. As the fault is temporary, the recloser will close the circuit at its subsequent recloses, thus providing power to all the consumers.
When there is a fault on the network at the location shown, the over-current protection element in IED1 detects the fault and opens CB1. This will result in an outage at loads L1 to L5. Since there are no automated components in the network, supply restoration for a part of the network requires the intervention of a restoration crew and in some areas may take up to 80 minutes [20]. Supply restoration is normally initiated by phone calls from one or more customers in the area where outage occurred reporting a loss of supply to the electricity supplier.
Upon receiving these calls a restoration crew is dispatched to the area. It will take some time for the team to locate the fault and manually isolate it by opening SD3 and SD4. The normally open point NOP is closed to restore the supply to L5. Load L4 will be without supply until the fault is repaired. A simple method to reduce the restoration time of loads L1, L2, L3 and L4 is using a pole-mounted recloser and sectionaliser as shown in Figure 6.
When a fault occurs, the recloser trips. Upon detecting the interruption, the sectionaliser, S, increments its counter by 1. The counter of S increments again and it is then opened. The recloser then closes successfully. The operation of the sectionaliser facilitates restoration of supply to L1, L2, L3 and L4 within a couple of minutes. However, the restoration of supply to L5 requires the intervention of the crew. As this method does not need any communication infrastructure, it is reliable and relatively inexpensive.
A greater degree of automation may be introduced by using reclosers with RTUs, with communication infrastructure between them see Figure 6. In this scheme, an Agent is employed that gathers data from all the intelligent devices in the system. The Agent sends commands to RTU1 to RTU4 remote terminal units up to the normally open point to open them and requests current and voltage data from them in real time.
A possible automatic restoration method is: 1. Send a command to RTU1 to reclose R1. If the fault current prevails, initiate a trip but as there is no fault current, R1 remains closed. Then send a command to RTU9 to close the normally open point. Finally, send a command to RTU4 to close R4. Discuss the consequence of loss of Supply A when there is no automation. Discuss a possible automatic restoration scheme which employs an Agent and re- closers with remote terminal units that provide minimum interruption to all the loads.
0コメント