Feed Point in Feeder Power System

Load Feeder

The nonsensitive-load feeder is the feeder that may be shut down if there is a disturbance or there are power quality problems on the utility;

From: Microgrid , 2017

Handbook of Conveying and Handling of Particulate Solids

A.W. Roberts , in Handbook of Powder Technology, 2001

5 FEEDER LOADS DURING FILLING AND FLOW

The determination of feeder loads and drive powers requires a knowledge of the stress fields generated in the hopper. The relationship between the vertical pressure p_v generated in a mass-flow bin during both filling and flow and the feeder load Q is illustrated in Fig. 6. Under filling conditions, a peaked stress field is generated throughout the entire bin as shown. Once flow is initiated, an arched stress field is generated in the hopper and a much greater proportion of the bin surcharge load on the hopper is supported by the upper part of the hopper walls. Consequently, the load acting on the feeder substantially reduces as shown in Fig. 5(b).

Fig. 6. Vertical pressure and load variations on feeder.

It is quite common for the load Q_f acting on the feeder under flow conditions to be in the order of 20% of the initial load Q_i. The arched stress field is quite stable and is maintained even if the flow is stopped. This means that once flow is initiated and then the feeder is stopped while the bin is still full, the arched stress field is retained and the load on the feeder remains at the reduced value. The subject of feeder loads and performance is discussed in more detail in Refs.[3, 4].

Consider the mass-flow hopper and feeder of Fig. 7. The procedures used to determine the feeder loads are now summarised. The loads acting on the feeder and corresponding power requirements vary according to the stress condition in the stored bulk mass. The general expression for the load Q is

Fig. 7. Loads on feeder.

(13) $\text{Q}={\text{p}}_{\text{vo}}{\text{A}}_{\text{o}}$

where

p_vo = vertical pressure on feeder surface

A_o = area of hopper outlet

For convenience, following the procedure established by Arnold et al [5], the load may be expressed in terms of a non-dimensional surcharge factor as follows:

(14) $\text{Q}=\text{q}\text{γ}{\text{L}}^{\left(1-\text{m}\right)}{\text{B}}^{\left(2+\text{m}\right)}$

where

q = non-dimensional surcharge factor

ρ = bulk density

B = width of slot or diameter of circular opening

m = hopper symmetry factor

m = 0 for plane-flow hopper

γ = ρ g = bulk specific weight

L = length of slotted opening

m = 1 for conical hopper

It follows from (13) and (14) that the non-dimensional surcharge factor is given by

(15) $\text{q}={\left(\frac{\text{π}}{4}\right)}^{\text{m}}\frac{{\text{p}}_{\text{vo}}}{\text{γ}\text{B}}$

There are two non-dimensional surcharge factors q_i and q_f and corresponding two loads Q_i and Q_f which are determined for the initial filling and flow cases respectively. The methods for determining these loads are described in Refs.[3]. For the initial loads, the surcharge pressure p_s has an important influence and depends on the type of storage system, such as a mass-flow bin, expanded flow bin or gravity reclaim stockpile. In all cases, the geometry of the feed zone, the clearance between the hopper and feeder, the stiffness of the feeder and the compressibility of the bulk solid will have an influence.

Considering the flow case for a belt or apron feeder with skirtplates, the redistribution of the stress field in the clearance space between the hopper and the feeder is illustrated in Fig. 8. The pressure p_vof is determined using the procedures for wall load analysis [3] and the design pressure on the feeder p_vod is then estimated using

Fig. 8. Stress fields at hopper and feeder interface.

(16) ${\text{p}}_{\text{vod}}={\text{k}}_{\text{Fm}}{\text{p}}_{\text{vof}}$

where the recommended pressure multiplier k_Fm for a plane-flow or wedged hopper is given by

(17) ${\text{k}}_{\text{Fm}}=(1+\sin \delta )$

$\text{δ}=\text{effective}\text{angle}\text{of}\text{internal}\text{friction}$

δ = effective angle of internal friction

Read full chapter

URL:

https://www.sciencedirect.com/science/article/pii/S0167378501800233

Switchgear and Controlgear Assemblies

K.C. Agrawal , in Industrial Power Engineering Handbook, 2001

Example 13.1

Consider the power distribution system of Figure 13.14, having the following feeder details:

I/C feeder, 4000A

$\text{O/G}\text{feeders}\{\begin{array}{lll}1\text{No}.800\text{A}\hfill & =\hfill & 800\text{A}\hfill \\ 7\text{Nos}.630\text{A}\hfill & =\hfill & 4410\text{A}\text{and},\hfill \\ 4\text{Nos}.400\text{A}\hfill & =\hfill & 1600\text{A}\hfill \end{array}$

$\begin{array}{llll}\therefore \hfill & \text{Total connected feeder load}\hfill & =\hfill & 6810\text{A}\hfill \\ \hfill & \text{and diversity factor for}12\text{Nos}.\hfill & \hfill & \hfill \\ \hfill & \text{feeders as in Table}13.4\hfill & =\hfill & 0.5\hfill \end{array}$

$\begin{array}{llll}\therefore \hfill & \text{Maximum loading on the}\hfill & \hfill & \hfill \\ \hfill & \text{incoming feeder or the main}\hfill & \hfill & \hfill \\ \hfill & \text{busbars at any time}\hfill & =\hfill & 6810\times 0.5\hfill \\ \hfill & \hfill & =\hfill & 3405\text{A}\hfill \end{array}$

Accordingly we have selected the rating of the incoming feeder as 4000 A. 4000 A being the next standard rating after 3150A. The maximum loading on each vertical section is worked out in Figure 13.14. These ratings of vertical busbars are when the arrangement of busbars is to individually feed each vertical row. If one common set of busbars is feeding more than one vertical section, the rating of busbars can be further economized. But one must take cognisance that too many tapings from one section of the bus may weaken the bus system.

If two sections are joined together to have a common vertical bus system, say, Sections 3 and 4, then the rating of the common bus will be:

$\begin{array}{lll}\text{Total connected load}\hfill & =\hfill & 2060+2230\hfill \\ \hfill & =\hfill & 4290\text{A}\hfill \end{array}$

Diversity factor for 8 numbers of feeders = 0.6

$\begin{array}{lll}\therefore \text{Maximum rating}\hfill & =\hfill & 4290\times 0.6\hfill \\ \hfill & =\hfill & 2574\text{A}\hfill \\ \text{or}\text{say}\hfill & =\hfill & 2500\text{A}\hfill \end{array}$

as against 1600 A + 1800, i.e. 3400 A, worked out in Figure 13.14, when both the sections were fed from individual busbars.

Read full chapter

URL:

https://www.sciencedirect.com/science/article/pii/B978075067351850091X

Microgrid Control Problems and Related Issues

M.S. Mahmoud , in Microgrid, 2017

2 Microgrid Review

Following the standards of the Consortium for Electric Reliability Technology Solutions (CERTS) architecture [19, 20], a basic microgrid architecture is shown in Fig. 1.1.

Fig. 1.1. Wide-sense microgrid architecture.

A microgrid is an interconnection of [3]:

•

Distributed energy sources, such as microturbines, wind turbines, fuel cells, and photovoltaics.

•

Storage devices for energy integration, such as batteries, flywheels, and power capacitors on low-voltage distribution systems.

•

A group of radial feeders, which could be part of a distribution system. There are three sensitive-load feeders (feeders A–C) and one nonsensitive-load feeder (feeder D):

1.

The sensitive-load feeders contain sensitive loads that must always be supplied; thus each feeder must have at least a microsource rated to satisfy the load at that feeder.

2.

The nonsensitive-load feeder is the feeder that may be shut down if there is a disturbance or there are power quality problems on the utility; the nonsensitive-load feeder will be left to ride through the disturbance or power quality problems.

3.

When there is a problem with the utility supply, feeders A–C can be islanded from the grid with use of the static switch that can separate them in less than a cycle to isolate the sensitive loads from the power grid to minimize disturbance to the sensitive loads.

•

The energy manager, which is responsible for managing system operation through power dispatching and voltage setting to each microsource controller. Some possible criteria for the microgrid to fulfill this responsibility are as follows [19]:

1.

ensure that the necessary electrical loads and heat are fulfilled by the microsources;

2.

ensure that the microgrid satisfies operational contracts with the utility;

3.

minimize emissions and/or system losses; and

4.

maximize the operational efficiency of the microsources.

Remark 1.1

In islanded operation, a microgrid will work autonomously; therefore, it must have enough local generation to meet the demands of the sensitive loads [5, 19]. Furthermore, a disturbance requiring a feeder to operate individually may also occur. Each sensitive-load feeder in the microgrid design must have enough local generation to supply its own loads, while the nonsensitive-load feeder will rely on the utility supply.

Remark 1.2

After a disturbance the microgrid will reconnect to the utility and work normally as a grid-connected system. In this grid-connected system, excess local power generation, if any, will supply the nonsensitive loads or charge the energy storage devices for later use. The excess power generated by the microgrid may also be sold to the utility; in this case, the microgrid will participate in the market operation or provide ancillary services.

Remark 1.3

The disconnection or reconnection processes must be specified by the point of common coupling, a single point of connection to the utility located on the primary side of the transformer. At this point the microgrid must meet the established interface requirements, such as defined in the IEEE Standard 1547 series. Furthermore, the successful disconnection or reconnection processes depend on microgrid controls. The controllers must ensure that the processes occur seamlessly and the operating points after the processes are satisfied.

Remark 1.4

In grid-connected mode, the microgrid is supposed to follow the rules of the distribution network without being involved in the operation of the main power system. The microgrid operation based on this approach is significant for the stable operation of the power system. In this mode the microgrid can draw power from the main grid or can supply its power to the main grid, thus functioning similarly to a controllable load or source. By supplying or drawing power, the microgrid should be able to control the active and reactive power flows and keep an eye on the energy storage [21, 22]. However, in this mode, owing to the small size of the distribution units, the system dynamics have to be fixed to a wide extent. Another issue is the slow response of the control signals whenever there is a change in output power. Furthermore, because of the lack of synchronous machines connected to the low-power grid, virtual inertia has to be incorporated in the control loops of the power electronic interfaces [23].

Remark 1.5

The islanded mode is an operating condition in which the microgrid isolates itself from the main grid in the case of a fault. However, the transition from the grid-connected mode to the islanded mode must be stable [24]. If the microgrid is consuming or supplying power to the main grid before disconnection, a power imbalance occurs. This is compensated by the energy storage units because the microsources have low inertia and slow dynamic response [25–27]. Further research can be found in [18, 28–38].

Read full chapter

URL:

https://www.sciencedirect.com/science/article/pii/B9780081017531000012

Coordinated Control System

Swapan Basu , Ajay Kumar Debnath , in Power Plant Instrumentation and Control Handbook (Second Edition), 2019

2.1.1 Generation of Unit Load Demand

In the coordinated control mode, the unit load demand is generated from the LDC as the particular unit would be connected to the electrical grid (Fig. 10.2). The LDC is a computer-controlled center that collects electrical data/parameters from the grid, such as voltage, frequency, individual feeder load, and generation of various generators connected to a particular grid. Depending on the health of each generating unit, demands of various feeders connected to it, and the software algorithm, LDC generates optimum demand pertinent to each unit connected to the system grid, which would then process the demand and pass on the relevant demand for proper execution by the boiler and turbine load control systems. Over a communication network, the LDC conveys the load demand to the particular unit in the form of electrical pulses, which are then integrated to generate demand.

Fig. 10.2. Unit load demand control.

Read full chapter

URL:

https://www.sciencedirect.com/science/article/pii/B978012819504800010X

Communications and internet of things for microgrids, smart buildings, and homes

Gianluca Fadda , ... Virginia Pilloni , in Distributed Energy Resources in Microgrids, 2019

9.2 The transition toward the smart grid

Several power grid evolution strategies are still under study due to many crucial challenging aspects that are evolving at the same time but in different ways. Fig. 9.1 shows the evolutionary character of smart grids (SGs) [1].

Figure 9.1. The evolutionary process of smart grids.

Source: Adapted from N. Andreadou, M.O. Guardiola, G. Fulli, Telecommunication technologies for smart grid projects with focus on smart metering applications, Energies (2016) [1].

Indeed, the continuous growth of renewable energy sources (RESs) of different types must fit within the requirements imposed in terms of emission reduction and efficient energy usage [1]. The SG of the future will exploit modern and reliable telecommunication technologies to implement energy management systems able to cope with all the new scenarios.

Monitoring and control of the energy generation, as well as transmission and distribution, are the main issues when dealing with the "smartization" of the power grid [2], requiring the interaction between different disciplines with a coordinated design of all the layers in the SG architecture [3], with interactive information and technology (ICT) devices and smart meters (SM) for empower an advanced metering infrastructure.

From a general point of view, three different kind of networks should be considered, depending on the specific function domain of the power grid: the transmission of the electrical energy refers to the high voltage (HV) network, while its distribution is enabled through the medium voltage (MV) network, and the low voltage (LV) network provides such energy to end users [1].

This chapter focuses on some key aspects regarding some LV and MV network arrangements and management system architectures, like the ones related to MGs, smart homes, and buildings.

The US Department of Energy (DOE) defines an MG as follows: "A MG is a group of interconnected loads and DERs within clearly defined electrical boundaries that acts as a single controllable entity with respect to the grid. A MG can connect and disconnect from the grid to enable it to operate in both grid-connected or island mode." As such, the possibility of having more load/customers within the MG boundary is intrinsically intended. Then, among the various definitions and arrangements, an MG can be used, for instance, like an integrated energy system located downstream of a main distribution substation through a point of common coupling (PCC) containing production and consumption users [4] but the concept of an MG can be applied also to the small power system destined to feed smart homes and buildings.

A possible system architecture referring to the MG concept, for MV networks, is shown in Fig. 9.2. An MG can support two different operating modes: when is running as "grid-connected," the MG is linked to the main grid through the distribution substation transformer, while when running as "islanded" (or autonomous), it results in being isolated from the main grid, typically during a blackout or brownout [5,6].

Figure 9.2. Microgrid system architecture.

Source: Adapted from W. Su, J. Wang, Energy management systems in microgrid operations, Electr. J. (2012) [7].

All the most common MG arrangements consider five different typical components [4,8]:

1.

PCC is the connection point for the power production, distribution network, and customer interface.

2.

Distributed generation (DG) usually denotes a small-scale electric power supply directly connected to the distribution system at or near the load feeder, which supplies power in an intermittent way, through one or more RES, based on the network demand.

3.

Energy storage systems (ESS), which may allow the implementation of energy buffering strategies when the energy price from the main grid is cheaper or an over-generation from the local DGs occurs. ESS can also be employed as an additional power generator during peak demand periods.

4.

Controllable loads, such as plug-in electric vehicles (EVs) or thermostatically controlled loads, which are smart things able to modify their own electric energy usage based on real-time set points.

5.

Critical loads, such as schools or hospitals, to which MG have to guarantee the highest levels of power supply availability and reliability.

The SG transition for MGs requires the distribution of the communication, computation, and storage resources and services to be on or close to devices and systems in the control of end-users, as well as a distribution management system (DMS) as fundamental part of the advanced grid modernization, for controlling all the distributed devices and components, including protection, intelligent electronic devices (IED) and microprocessor-based controllers of power system equipment. For instance, by exploiting the concept of "fog computing," introduced by Cisco in 2014 and claimed as "a standard that defines how edge computing should work, and which facilitates the operation of compute, storage, and networking services between end devices and cloud computing data centers" [9]. The edge computing, often just referred to as edge, "brings processing close to the data source, and it does not need to be sent to a remote cloud or other centralized systems for processing" [10].

In July 2018, the IEEE Standards Association approved the OpenFog Reference Architecture, developed by the OpenFog Consortium, an association of high-tech industry companies and academic institutions, including Cisco Systems, Intel, Microsoft, Princeton University, Dell, and ARM Holdings, as an official standard, under the name of IEEE 1934 [11]. The IEEE 1934 is a technical reference framework, "designed to enable processing to be distributed across things-to-cloud continuum" [12].

From a general point of view, all smart-grid scenarios considered in this chapter exhibit the same three-layer-architecture composed by the smart things tier (i.e., the lower layer), the fog tier (i.e., the middle layer), and finally the macrostation tier (i.e., the upper layer) as shown in Fig. 9.3.

Figure 9.3. Fog computing architecture in a SG.

The middle layer, namely the "Fog tier," could include all the network edge devices, such as smart meters, MGs, or any other smart appliance which could implement an energy load balancing application, by switching in a suitable way among alternative distributed energy resources (DERs) at the lowest layer, namely the "smart thing tier," based on data generated by their grid sensors and devices. In terms of communication signals, based on the received information, the so-called "fog collectors" at the fog tier, make actuations at the SG tier, process data so as to drop those references that have to be consumed locally, and send the filtered data to the higher layer, specifically the "macrostation tier," for management purposes, that is, visualization and reporting for real-time or transaction analytics [13].

The fog computing paradigm enables the implementation of applications of big data (BD) analytics in real-time, since by this way a dense distribution of data-collection points can be managed. The smart thing tier refers to machine-to-machine (M2M) interactions, which allow local devices, that is, protections or controllers, to communicate with a remote system and for making a suitable response to a particular event or situation within a processing time from milliseconds to seconds, that is in real-time [14]. The fog tier and the macrostation tier refer to human-to-machine-interactions, that is, visualization and reporting, but also supervision and control of systems and processes. Such communication exchange over the Fog platform can span from seconds to minutes for real-time analytics, and up to days for transactional analytics. Due to this fact, the Fog must support different kinds of storage, from transient at the lowest tier to semi-permanent at the highest tier. Then, when dealing with the internet of everything for streaming and real applications implemented by a fog computing approach, the following classification of interactions can be considered:

•

M2M: Any machine can send/receive data to/from any other one, by exploiting the networking capabilities of IoT and sensors.

•

People-to-Machine: Any person can send/receive data to/from any other person or machine, and the connection relies on capabilities in data and analytics.

•

People-to-People: People can exchange data through a cooperation mechanism.

Smart local grid (SLG) is another emerging concept, referred to a network and communication infrastructure consisting of multiple MGs, aiming to improve their effectiveness and reliability at a local level, by allowing communication among all devices [15]. By applying fog computing to SLG, the bandwidth and latency issues can be highly reduced by keeping costs to the minimum. Indeed, by allowing M2M direct connections, SLG enables real-time decisions without the need for a data exchange with the Cloud [16]. SLG is a smart network where all the devices are connected to the Cloud through open communication standards, but they are also autonomous in making decisions when some changes force them to reply in real time. The opportunity for devices to communicate with the Fog allow them to solve problems with a higher complexity. A classification of the application technologies is summarized in Table 9.1.

Table 9.1. Classification of the application technologies.

Technology	Applications	Fog computing applications	Smart grid features	Big data
Energy management	• MG management • Dynamic demand response operated within the MG • Real-time monitoring on applications for SG	• Data metric communication implementing private fog for small-size networks • Dynamic bandwidth increasing for fog applications to avoid congestions • Fog MG to MG interaction • Demand response model definition	• MG management • Dynamic pricing	• High QoS for real-time BD applications
Information management	• Smart meter data streams in Cloud • Dynamic data center operation	• Guaranteed work-flow latency and processing rates with the help of Fog data optimization • Dynamic pricing model in SG architectures according to load on Fog Data Services • Adequate data transfer framework from users to Fog and vice versa	• Cost optimization • Data Storage and processing	• For BD processing with Platform-as-a-Service and Infrastructure-as-a-Service
Security	• Security and protection system for electric power information • Privacy preserving over encrypted metering data for SG	• Fog as software as a service for data privacy issues in large scale deployment of SG • Security mechanisms definition while using fog computing applications • Effective and efficient security and privacy policies to support increasing data from smart meters	• Data security and privacy • Threat detection • Cyber security	• Important to ensure that all technology and application components include and maintain acceptable levels of security and privacy mechanisms

BD, Big data; QoS, quality-of-service.

The fog computing paradigm is in charge of supplying new SG models to manage response demand. The main advantage is the communication overhead reduction between devices compared to fully distributed SG models, where there is no possibility of improving power consumption due to the lack of sharing computing and communication information between devices and users.

A Cloud Computing approach in which all users (i.e., supplier and customer) are continuously connected to the Cloud can improve the development of centralized demand response management algorithms [17]. Considering macrogrids and MGs as Fog devices may reduce excessive and unnecessary communications between end users, typically due to distance in a fully distributed model. The idea is to allow customers to communicate frequently with close Fog devices, allowing them to interact periodically with the Cloud.

Another proposed solution in SG applications is demand response management [18], where a Fog device is in charge of coordinating a mutual power exchange between MGs as well as MGs and the main grid. A new power management algorithm has been proposed in Ref. [18] to minimize loss power, create exchange pairs among MGs and prioritize communications. In the demand response management solution, two main layers can be identified. The first one is related to the consumers connected to the same Fog device and the way to obtain local information from it. Likewise, in the second layer, several Fog devices are attached to the same Cloud server. Fog devices can also be interconnected and, thanks to the collected information, they are able to group themselves in order to minimize power losses and consequently limit communication cost.

Another interesting solution has been proposed in Ref. [19], based on two main steps: the first one is a classification step in which heterogeneous EVs are dynamically grouped; then each sub-group is scheduled in terms of charging demand considering a sliding-window iterative approach.

Read full chapter

URL:

https://www.sciencedirect.com/science/article/pii/B9780128177747000090

Electrical distribution and installation

I G Crow BEng, PhD, CEng, FIMechE, FIMarE, MemASME , R Robinson BSc, CEng, FIEE , in Plant Engineer's Reference Book (Second Edition), 2002

23.3 Distribution systems

The object of any electrical power distribution scheme is to provide a power supply system which will convey power economically and reliably from the supply point to the many loads throughout the installation. The standard method of supplying reliable electrical supplies to a load centre is to provide duplicated 100% rated supplies. However, there are a number of ways in which these supplies can be provided.

The standard approach for a secure supply system is to provide duplicated transformer supplies to a switchboard with each transformer rated to carry the total switchboard load. Both transformers are operated in parallel, and the loss of single incoming supply will not therefore affect the supply to the load feeders. With this configuration the supply switchboard must be able to accept the fault current produced when both transformers are in parallel. The system must, however, be designed for the voltage regulation to remain within acceptable limits when a single transformer supplies the load.

If the two incoming supplies are interlocked to prevent parallel operation, the power flow, fault current flow and voltage regulation are governed by a single transformer supply infeed. Although the configuration has the disadvantage that the loss of a single infeed will cause a temporary loss of supply to one set of load until re-switching occurs, the advantage is that each incoming transformer can be rated higher than the first, and thus a higher concentration of loads can be supplied.

A third alternative is to rate the transformer infeeds for single supplies but to arrange for automatic switching of the bus section and two incomers in the event of loss of supply to one section. A rapid transfer switching to the remaining supply in the event of loss of single supply can prevent the total loss of motor loads. In such a situation the effect of the current taken by motors to which the supply has been restored must be taken into account.

Since a group of motors re-accelerating together will draw an increased current from the supply, this current will affect voltage regulation and must be recognized when selecting protection relay settings. Figure 23.1 illustrates these arrangements of transformer operation. The effect of these combinations upon the loads which can be supplied is summarized in Table 23.1 for two typical supply voltages and fault levels, circuit breaker capabilities and transformer reactances.

Figure 23.1. Transformer arrangement. (a) Parallel operation; (b) 2 out of 3 interlock

Table 23.1. Load supply capabilities and single and parallel transformer combinations

Source Voltage kV	Load		Load circuit breaker				Max. supply transformer rating (MVA)
	Fault level MVA	Voltage kV	Full load, rating A			Fault rating kA (MVA)	Reactance (%)
			630	2000	3000		6	10	6	10
			MVA				Parallel		Single
33	1428	11	12	38.1	−	25(476)	21	35.7	38.1	38.1
11	476	3.3	3.6	11.4	–	25(143)	6.1	10.2	11.4	11.4
11	476	0.415	–	1.4	2.2	50(35.9)	12	1.9	2.2	2.2

In the case of parallel operation the maximum transformer rating is limited by the fault rating of the switchgear, while for a single transformer infeed the limitation is by the full-load current rating of the switchgear.

Table 23.1 takes into account only the fault contribution from the supply system. The contribution from rotating plant within an installation must also be considered when specifying switchgear and transformer ratings.

If a fully duplicated supply system is thought to be necessary, the transformer reactances can be increased in order to limit the fault level when operating in parallel mode. However, this will increase the initial capital cost, and voltage regulation with a single transformer in circuit will still need to be maintained within acceptable limits.

Read full chapter

URL:

https://www.sciencedirect.com/science/article/pii/B978075064452550078X

Stationary hybrid systems: Motivation policies and technical challenges

Sandip Deshmukh , Khalid Anwar , in Hybrid Technologies for Power Generation, 2022

1 Introduction to stationary hybrid systems

The high interest in power generation using Renewable Energy Sources (RES) demands a new power network which has the capability to integrate hybrid-RES with other distributed generators like diesel generator. Decentralized generation of power has the potential to integrate clean power resources within microgrid with high penetration of RES, which can be operated stand-alone or grid-connected mode with the help of Energy Management System (Fig. 1). Microgrid as a sustainable energy system, not only reduces greenhouse emissions but also has several other advantages. Advantages of microgrid are discussed by Lasseter and Chris [1, 2]. They have highlighted that introducing microgrids can improve local reliability, minimize feeder load, electrify remote locations, reduce transmission loss and provide cost-effective power to end-user. Microgrid can be operated either on stand-alone mode, which is also known as islanding mode, or grid-connected mode. When it operates as stand-alone, the storage system is a must component to normalize the variability in RES. Microgrid consists of five components—Distributed Generators (DGs), loads, storage devices, controls, and point of common coupling. Microgrid is expected to generate sufficient power to meet the local demand. It is, therefore, necessary to efficiently estimate DGs including RES.

Fig. 1. Typical schematic diagram of a microgrid.

Furthermore, efficient utilization of RES like wind and solar at the local level depends upon three factors; accurate perdition of RES, estimation of demand, and integration of sources into EMS. Available measured data can be used as a reference for the prediction and evaluation of wind and solar potential at the local level using efficient mathematical or prediction techniques. In order to predict the solar and wind potential accurately at a local level, different meteorological parameters need to be considered and analyzed. A numerical tool has to be used for handling large data and developing a model to predict the solar and wind potential. However, RES is intermittent in nature and site-specific, which makes the prediction problem complex. Hence, an efficient model like Artificial Neural Network (ANN) needs to be developed to predict solar and wind potential. Measured data of meteorological parameters obtained from Government agencies can be used to train and develop a model. Predicted solar and wind potential need to be mapped using a geographical mapping technique, such as Geographic Information System. Based on the predicted potential of RES and maps, regions with higher potential can be identified. Ideal types and areas of land within the regions with higher RES potential have to be investigated for the installation of the power generation system, such as wind farms and solar fields using land use and land cover analysis. Power generation capacity from available sources of energy has to be estimated and optimized, which can meet the demand of the targeted community. Therefore, actual estimation of energy demand for small population is also required, which has to be generated by microgrid. In addition, a storage system, such as battery and fuel cell, is also required for microgrid to improve its reliability. Along with a storage system, diesel generator can also be used for backup.

All the RES and backup DGs along with storage devices integrated into microgrid are required to be utilized efficiently to meet the power demand. This led to the introduction of Energy Management Systems (EMS) to microgrid. The EMS works on the primary function of monitoring the different energy resources and controlling energy consumption at a particular location. This way, EMS coordinates the DGs effectively, which are integrated with the microgrid, to ensure the power supply to loads with least possible operational cost. It helps the decision-maker to understand the limitations and advantages of a location and thereby control the usage accordingly [3]. All possible DGs need to be integrated into microgrid and optimized for high productivity. Researchers worldwide have developed several methods for energy management systems, yet there are some challenges to be addressed. Microgrid should deliver a quality of power with high reliability. Hence, managing a microgrid is challenging due to high geographical dispersion, limited location of distributed resources, and seasonal as well as intra-day variability in renewable resources.

Read full chapter

URL:

https://www.sciencedirect.com/science/article/pii/B9780128237939000085

Intelligent fault diagnosis technique for distribution grid

Md Shafiullah , ... A.H. Al-Mohammed , in Power System Fault Diagnosis, 2022

10.2 Four-node test distribution feeder modeling

This section presents the step-by-step modeling of the four-node test distribution feeder in the MATLAB/SIMULINK platform. It also discusses the load demand and fault information (inception angle and resistance) uncertainty modeling, data generation, and data recording techniques. Finally, it presents the feature extraction processes from the recorded signals employing DWT and ST.

10.2.1 Test feeder modeling in MATLAB/SIMULINK environment

The selected test distribution feeder consists of four nodes, two distribution transformers, one lumped load, one feeder, and one distribution line, as shown in Fig. 10.1. Detailed information regarding the test distribution feeder is tabulated in Table 10.1 and can be found in Refs. [33–35]. MATLAB/SIMULINK (R2020b) is used to develop the mentioned test feeder. In addition, the Simscape Electrical Toolbox that provides the component libraries (programmable power and load sources, motors, transformers, renewable energy resources, overhead transmission lines, underground cables, etc.) is used for modeling and simulating the electric power system networks [36,37]. It also supports C-code generation to deploy the models on other simulation environments, including hardware-in-the-loop systems. One can master himself/herself on this modeling platform by following different tutorials available online.

Fig. 10.1. Four-node test distribution feeder.

Table 10.1. Details of the four-node test distribution network.

	Item	Details
	Transformer I	12 MVA, 120/25 kV, Yg-Δ
	Transformer II	12 MVA, 25 kV/575 V, Yg-Δ
	Load	5 MW, 2.5 MVAR (inductive)
	Distribution line	30 km
	System frequency	60 Hz

This chapter picked different power system components to model the test feeder, as depicted in Fig. 10.2. Eventually, the four-node test distribution feeder model was developed, as shown in Fig. 10.3. Among different options of the "PSB option menu" block, the "discrete" option was selected where the sample time was 0.00001 seconds (equivalent to 100 kHz sampling frequency and 1667 samples/cycle). It is worth noting that this chapter modeled the distribution line with two blocks where the distances of the first block and the second block were "a" and "b" kilometers (km), respectively. These two parameters facilitated the faults in different locations on the 30-km long distribution line.

Fig. 10.2. Major components of the four-node test feeder in MATLAB/SIMULINK environment.

Fig. 10.3. Four-node test feeder in MATLAB/SIMULINK environment.

Likewise, the parameters of the load block were also set to variable numbers to facilitate dynamic loading conditions of the selected test feeder (load demand uncertainty). Other parameters were chosen in the three-phase fault block to incorporate different fault types and fault resistance uncertainty. All mentioned parameters and the fault inception time were generated from the specified ranges in a MATLAB script file. Moreover, the developed SIMULINK file was opened and simulated from the same script file. Finally, the recorded three-phase current signals from the SIMULINK file were exported to MATLAB Workspace using the 'To Workspace' block for feature extraction purposes.

10.2.2 Fault modeling and data generation

This study selected a total of 59 locations on the distribution line of the test feeder starting from 0.50 km to 29.50 km with a step size of 0.50 km. Different types of faults for four cycles were then applied to the selected locations by varying the prefault loading conditions, fault information (resistance and inception angle), and fault types. The IFD scheme picked fault resistance from "0 Ω" to "15 Ω" randomly and varied prefault loading conditions in a range of ±10% of the rated loading conditions. Besides, it applied different types of faults, including single-line-to-ground (SLG), line-to-line-to-ground (LLG), and three-phase-to-ground (LLLG) faults. The SLG faults included phase-A-to-ground (AG), phase-B-to-ground (BG), and phase-C-to-ground (CG) faults. In contrast, the LLG faults included phase-A-to-phase-B-to-ground (ABG), phase-B-to-phase-C-to-ground (BCG), and phase-C-to-phase-A-to-ground (CAG) faults, and LLLG faults included only phase-A-to-phase-B-to-phase-C-to-ground (ABCG) faults. Finally, the IFD scheme recorded three-phase faulty current signals (two-cycle: one cycle before and the other cycle after the fault occurrence) for feature extraction employing the DWT and the ST.

10.2.3 Feature extraction

From the recorded three-phase current signals, a total of 144 statistical measures, also known as features, were extracted through a seven-level DWT decomposition. Detailed illustrations of the DWT-based feature extraction process can be found in Chapter 4 and Refs. [33–35]. Likewise, the ST approach collected 36 features from the same signals. Detailed illustrations of the ST-based feature extraction process can be found in Chapter 4 and Refs. [38–42].

Read full chapter

URL:

https://www.sciencedirect.com/science/article/pii/B9780323884297000059

Maintenance, Troubleshooting, and Repair

George Patrick Shultz , in Transformers and Motors, 1989

What to Do

Be alert! Make use of all of your senses to determine what has happened. To often we look, but do not see; listen, and do not hear; touch, but do not feel; smell and taste, but do not comprehend. Someone once said, "You can see a lot by looking."

When this information gathered by your sensory perceptions is integrated with your previous experiences and knowledge about the system, the problem will usually be simple to determine. The one thing about power systems is that physical evidence of the malfunction is normally present. Be prepared to understand what the evidence is telling you. The time spent analyzing and diagnosing the fault will in itself save time in getting the transformer back on-line. Haphazard methods of troubleshooting are time consuming and should be avoided. Use a logical sequence.

If the transformer is still on-line when you arrive, it should be disconnected immediately if any smoke or liquid is coming from the tank. The same procedure should be followed if there is any unusual noise coming from the transformer. Do not re-energize until the cause of the malfunction has been established. Activate any ventilation systems to clear dangerous gases from the area and to lower the ambient temperature.

Take a temperature reading of the coolant, if liquid cooled, as soon as possible. This reading will be a good indicator if damage has occurred to the insulating quality of the transformer. If feasible, allow any cooling system to continue to operate until the temperature of the transformer is reduced to an acceptable level.

In the event the transformer is not in immediate danger of being destroyed or exploding, and can continue to be operated, make ampere readings. If the system is not metered, a clamp-on ammeter can be used for this purpose.

Feeder loads can be disconnected one at a time in order to determine which feeder has the problem. Individual loads can then be disconnected in a like manner. If the problem is not isolated in this way, then one of the conductors is at fault. Normal current readings on the feeders indicate that the problem is in the transformer itself.

With the transformer off-line and with the loads disconnected, check for any external damage to the transformer. This may be either electrical or mechanical. Examine the potheads, conductors, insulators, bushings, and surge arresters. Look for evidence that the coolant is leaking from the tank.

If some of the fuses on the transformer have blown, they can also provide valuable information. If the fuses are in the secondary circuit, the problem is more than likely being caused by the load. If the primary fuses are out, the transformer itself is most likely at fault.

How the fuse blows can also tell you something about the fault. Figure 4-5 shows a good single-element fuse linkage. Figure 4-6 depicts a fuse that has opened due to an overload—notice that only one segment of the element has opened. Figure 4-7 illustrates a fuse that has opened under short-circuit conditions. In this case all three segments of the fuse have melted.

FIGURE 4-5. Typical single-element fuse.

(Courtesy Bussman Division, McGraw-Edison Co.)

FIGURE 4-6. Single-element fuse opened under overload conditions.

(Courtesy Bussman Division, McGraw-Edison Co.)

FIGURE 4-7. Single-element fuse opened under short-circuit conditions.

(Courtesy Bussman Division, McGraw-Edison Co.)

Figures 4-8 through 4-10 show similar conditions when a dual-element fuse is used. The spring-loaded element is shown open in Figure 4-9. This indicates an overload on the circuit In Figure 4-10, the short-circuit element is shown opened. Note that the spring-loaded contact did not have time to open under this condition.

FIGURE 4-8. Typical dual-element fuse.

(Courtesy Bussman Division, McGraw-Edison Co.)

FIGURE 4-9. Dual-element fuse opened under overload conditions.

(Courtesy Bussman Division, McGraw-Edison Co.)

FIGURE 4-10. Dual-element fuse opened under short-circuit conditions.

(Courtesy Bussman Division, McGraw-Edison Co.)

Figure 4-11 illustrates what you might face if the overcurrent devices on the transformer are overrated. Figure 4-12 depicts the explosive nature of short circuit, high-amperage current when the interrupting capacity of the fuse has not been properly selected. Figure 4-13 shows the resulting damage to a panelboard due to the explosion.

FIGURE 4-11. Effects of overload due to the use of too large ampere rating of the protective device.

(Courtesy Bussman Division, McGraw-Edison Co.)

FIGURE 4-12. Explosive nature of short-circuit, high-amperage current.

(Courtesy Bussman Division, McGraw-Edison Co.)

FIGURE 4-13. Results of a short-circuit condition on a panelboard.

(Courtesy Bussman Division, McGraw-Edison Co.)

These disasters can occur due to inadequate maintenance and carelessness on the part of the electrician who services this equipment Fuses may have deteriorated over the years and have never been checked as part of the preventative maintenance program. Fuses may have been blown and been replaced without regard to their ampere and voltage ratings or their interrupting capacity. In each case, the causes may be laziness, or a complete lack of knowledge on the part of the mechanic about the possible results of the negligence. Keeping the system in operation at all costs is not applicable in these circumstances.

Loss of life is a possible outcome of these actions. Loss of time, money, and electrical service can also be a major problem. No one except you and your family is going to worry about you losing your job and the possibility of a lawsuit. An exception may be your employer if he or she is involved in a lawsuit for hiring you.

Read full chapter

URL:

https://www.sciencedirect.com/science/article/pii/B978008051958650007X

Multi-time-scale energy management of distributed energy resources in active distribution grids

Bishnu P. Bhattarai , ... Birgitte Bak-Jensen , in Smart Power Distribution Systems, 2019

21.5 Results and discussion

This section presents results with multi-time-scale control algorithms considering EVs as flexible resources. In addition, different cases are presented for adaptive control considering EV only, EWH only, PV only, and EWH-PV.

21.5.1 Multi-time-scale energy management

21.5.1.1 Day-ahead predictive scheduling

The key idea of day-ahead predictive scheduling is to make optimum operational schedules for the flexible resources. It is worth mentioning that the operational schedules vary significantly depending upon consumer, DSO, and aggregator perspectives of scheduling. For instance, EV owners normally charge their EV during low price periods and discharge during high price periods irrespective of the network conditions. Therefore, day-ahead operational schedule from consumer perspective does not respect network limits. On the other hand, DSOs would have liked if the EVs were charged during low loading and discharged during high loading periods. Their key driver would be to avoid network congestion and improve its performance.

The day-ahead predictive scheduling of EVs looking from consumer perspective is illustrated in Fig. 21.15A . It can be observed that EVs are scheduled for charging at low price periods (2   am to 6   am) and for discharging during high price periods (8   pm to 10   pm). This leads to clear violations of feeder the capacity limit of 250   kVA. Moreover, the voltage at the farthest end node (as illustrated in Fig. 21.15C) is lower than the acceptable value (0.95   pu). More interestingly, the voltage and network limit violations are worse than the case with uncontrolled charging. Therefore, despite the consumer can significantly get economic benefit, such control schemes may not be technically feasible. Fig. 21.15C illustrates the EV charging profile while looking from DSO perspective. It can be observed that EVs are scheduled for charging during the lower demand period, which helps to improve the network utilization. DSO perspective respects both the voltage constraints as well as thermal constraints. Despite DSO perspective of scheduling improves the feeder utilization and respect network constraints, it allows EVs to be charged during higher price periods also. This leads to a higher EV charging cost compared with the EV charging case seen from consumer perspective. Such scenario demotivates the consumers to participate for network support only. Unlike consumer and DSO perspective, the aggregator perspective considers both charging costs and network limits. Fig. 21.16A illustrates the EV charging profiles for grid-to-vehicle (G2V) and vehicle-to-grid (V2G) cases. It is observed that the total feeder load is within the network limit of 250   kVA for both the cases. Further, it is observed from Fig. 21.16B that the voltage is also within the limit throughout the charging horizon. This is done by rescheduling some of the EVs during the time periods when network limits are violated.

Fig. 21.15. EV charging/discharging from different actor perspective. (A) V2G/G2V from consumer perspective. (B) EVs charging from DSO perspective. (C) Voltage at farthest end node.

Fig. 21.16. EV charging/discharging from aggregator perspective. (A) EV charging/discharging profiles. (B) Voltage at farthest end node.

Even though technical and economical matrices can be improved while solving the problem from aggregator perspective, the monitory benefits resulting from the EV participation to only day-ahead market is often insufficient to motivate EV owners to provide flexibility for grid services. Therefore, an intraday compensating approach is presented, which allows the resources to provide the total cost by providing a framework to participate in multiple markets with simultaneous assurance of consumer comfort and network constraints. The HCA amplifies the economic advantage by allowing the EVs to participate in BRM which offers significantly higher prices for the demand flexibilities.

21.5.1.2 Intraday/hour optimal adjustment of predictive schedules

Unlike a day-ahead predictive scheduling perspective where EVs are scheduled based on their individual objectives, intraday/hour adjustment allows the flexible resources to participate in BRM. At this stage, HCA minimizes the total energy cost of the flexible resources by considering day-ahead, balancing, and regulating prices; network limits; and EV owner's requirements. Therefore, one of the notable attributes of the hierarchical multi-time-scale energy management is that it provides a framework for EVs to participate in BRM with the available up/down-regulation capability of the flexible resources. This not only maximizes utilization of availability flexibility, but also maximizes the economic benefits to the consumer. Fig. 21.17 depicts the regulating capacity of EVs during G2V and V2G scenarios. It is observed that up-regulation capacity is significantly smaller compared with the down-regulation capacity in the G2V case. This is due to the fact that the EV can provide upregulation during charging periods only. As the EVs stay idle for most of the time, it provides higher down-regulation capacity. However, during the V2G case, both up/downregulations are significantly higher. This is due to increased regulation capacity resulting from V2G capability. One interesting observation from Fig. 21.17 is that the regulation capacity of EVs is greater than the feeder capacity limit especially during V2G case. Therefore, in reality the usable up/down-regulation powers are limited by the feeder capacity.

Fig. 21.17. Regulation capability from EV during (A) grid-to-vehicle (G2V) and (B) vehicle-to-grid (V2G) cases.

21.5.2 Adaptive control

In addition to the day-ahead predictive scheduling and intra-hour adjusting, there is always chance of discrepancies during actual operating conditions. For this reason, an adaptive real-time control is desired.

21.5.2.1 Adaptive control of EV

An adaptive control is implemented to adjust the charging/discharging power of the EVs when network limits are violated during actual operating conditions. Fig. 21.18 illustrates the voltage and power profile at the farthest end node in the network for a portion of a time period where voltage got violated beyond the predefined limits. It can be observed that total EV charging power profiles in the network after realization of the adaptive control is significantly better compared with the one before implementing the control. Particularly, the P–V droop controls the actual charging/discharging power of the EVs whose POC voltage is violated to bring the voltage back to the acceptable limits. As long as the voltage is between V _th and V _min, the P–V droop decreases the EV power according to the observed POC voltage. Therefore, the droop locally improves voltage in real time which otherwise cannot be done by the SL and CL.

Fig. 21.18. Charging power and voltage from aggregator perspectives. (A) Flexible load at farthest node. (B) Voltage at farthest node.

21.5.2.2 Adaptive control of EWH

In order to demonstrate the performance of the adaptive control with thermostatic loads (i.e., EWH), 24   h time-sweep simulation is performed with and without the adaptive control. Fig. 21.19 demonstrates adaptive control of EWH connected to the farthest end node (node 13) in the network. It can be seen that the adpative control adjusts the active and reactive power consumption of the EWH whenever the monitored voltage goes below V _th (Fig. 21.19C). First, reactive power is injected as shown in Fig. 21.19B for providing voltage support. Since the reactive power is not sufficient to alleviate the UV problem, the active power consumption is deceased according to the predefined P–V droop. Following the voltge violation below V _min, the active power consumption of the EWH is decreased to zero. Followed by the V _min violation, SOE limits are decreased by 0.1   pu (Fig. 21.19A) so as to force the EWH to turnoff before reaching SOE _max. Active and reactive power adjustment significantly improves the voltage compared with the case without control.

Fig. 21.19. Simulation results of EWH of a consumer connected at farthest end node (node 13). (A) SOE with and without control. (B) Power consumption. (C) Voltage at farthest node. (D) Thermal consumption.

The aggregated power consumption of EWHs with and without adaptive control is illustrated in Fig. 21.20. When the adaptive control is implemented, the power consumed by EWHs is decreased significantly during the peak period to support the voltage. Both reactive power injection and active power reduction are realized to improve the voltage which was violated without control. This fact is reflected by decrement in total active power consumption and increment in reactive power injection as shown in Fig. 21.20B.

Fig. 21.20. Simulation results of EWH adaptive control for aggregated load. (A) Feeder apparent power. (B) Active and reactive power.

21.5.2.3 Adaptive control of PV

Adaptive control of PV is designed especially to address OV limit violations. This scenario is performed with minimum feeder loading so as to investigate the effectiveness of the proposed method during worst-case conditions. Fig. 21.21 depicts operations of the P–V and Q–V droops of the PV system connected to the farthest end node. From Fig. 21.21A and B, it can be observed that, whenever the POC voltage exceeds V _TH, Q–V adaptive control starts consuming reactive power. As long as the POC voltage is lower than V _m (i.e., midpoint between V _max and V _TH), the adaptive controller increase reactive power consumption. In this case, no active power curtailment occurs. However, when the voltage crosses V _m , the controller starts curtailing active power as well. As shown in Fig. 21.21, both active power curtailment as well as reactive power consumption are realized to keep the voltage within limit (i.e., below V _max). It can be seen that the proposed adaptive control significantly improves voltage compared with the case without the control.

Fig. 21.21. Simulation results of PV for a single unit (A and B) and aggregated (C and D). (A) Adaptive control of PV at farthest node. (B) Voltage at farthest node. (C) Feeder loading. (D) PV active/reactive power.

Fig. 21.21C and D illustrates the aggregated adjustment of active and reactive power to support the network congestion. First, reactive power consumption is increased to provide voltage support. The active power is curtailed only when reactive power control is insufficient to alleviate the OV problem. It can be seen from Fig. 21.21 that total active and reactive power are significanlty regulated. To reduce the active power curtailment, an integrated control of solar PV and EWH is performed.

21.5.2.4 Adaptive control of EWH and PV

Since future distribution networks will have flexible loads as well as distributed generation, they are susceptible to both UV and OV violations. Fig. 21.22 illustrates a simulation with both EWH and PV providing real-time adaptive support. It can be seen that the active power consumption of the EWH is adjusted first to support OV violations. Then, EWH and PV reactive power consumptions are increased to support the voltage. However, as shown in Fig. 21.22C, there is no active power curtailment of the PV because the adjustment of active power consumption of EWH and reactive power consumptions of EWH and PV are sufficient to alleviate the OV problem. As such, the OV problem is effectively solved by the integrated control without incurring any revenue loss to the PV owner that could have resulted from active power curtailment of PVs.

Fig. 21.22. Simulation results for EWH and PV with and without control. (A) Feeder power. (B) EWH power consumptions. (C) PV power consumption. (D) Voltage at farthest node.

Read full chapter

URL:

https://www.sciencedirect.com/science/article/pii/B9780128121542000213

howellwarst1944.blogspot.com

Source: https://www.sciencedirect.com/topics/engineering/load-feeder

Share this post