This essay has been submitted by a student. This is not an example of the work written by our professional essay writers.
This chapter will describe the whole process to choose a model to make distribution networks having two separate generators technologies such as SCIG and DFIG. These models will have same characteristics and conditions with the connection to the grid except their internal parameters etc. rotor, stator, and inertia. There will be some transformers involved as well as obvious in distribution network. The topology should also be specified either it's going to be radial or ring. If the ring topology would be suitable then power will find its way down to top to bottom via some transmission lines and loads which can be linear or non-linear. The whole grid components going to be used in the whole network should have technical information theoretical correct so that these values can be applied in to software environment so that it can be used it dynamic simulations.
3.2 Questions need to be addressed
There are some vital questions that need to be addressed before relevant model for study to analyse the objective which leads to goal of this thesis. Although this thesis will not concentrate of the system formation but these things are relevant to designing.
1. Transient stability in the power systems (distribution network).
2. Which type of model should be built to analyse the transient stability?
3. The technical information regarding parameters or grid components going to be used in model. (Topology identification)
4. Understanding the functionality of models of SCIG and DFIG w.r.t to grid.
5. The software selection where this dynamic simulation of respective model will be presented.
6. From how many methods or prospective we can observe the transient stability of the presented model?
7. Is it possible in software environment to design the system ideally and if there is some hurdles involve how to tackle those issues?
8. How many scenarios or ways of presenting the result will be adopted?
9. Which parameters will be crucial for the system stability overall and how we can give suggestions to provide evidence to minimize the power losses in the system?
3.3 Modelling of the distribution network regarding transient stability
For modelling power systems which are consist of many grid components as generators, motors, transmission lines, and loads. "For stability assessment, the power system is normally represented using a positive sequence model". Farmer (2001)The network is represented by a traditional positive sequence power flow model that defines the transmission topology, lines with their specification, connected loads and generation, and voltage profile before and after distribution. Generators can be represented with various levels of detail, selected based on such factors as length of simulation, severity of disturbance, and accuracy required.
Modeling has not been quite detailed in some recent years because of technology improvement in the available software and hardware available. But there is still need of modelling the grid components that are been using in the circuit by input the values require by the software. The analysing time for transient stability should be around 1 to 5 sec so one should keep in mind that won't be too long. Farmer (2001)
3.4 Methods of Analysis of Transient Stability
3.4.1 Simulation Studies
This section describes the process use for the modelling to analyse the transient stability of that particular system. These steps can differ for different software versions and basic thought is quite same. Usually in software environment user need in to input some data as the parameters of the components present in the network. Finally for getting results some parameters need to be observe for getting appropriate results .Farmer (2001)
3.4.2 Input data
1. Power flow: that tells the system's topology either it is radial or ring and initial operating data on the BB and other components. Farmer (2001)
2. Dynamic data: dynamic components in the system such as generators and motor, it is very important to mention their type and their input data specially regarding design. Farmer (2001)
3. Program control data: that defines the scenario under the system will work that mostly related to the time in which system will be observed for stability. Farmer (2001)
4. Switching data: Includes the details of the fault applied to the system. This consists on the time at which fault is initiated , location of fault is applied, the kind of fault and its fault impedance if required, the duration of the fault, the elements lost as a result of the fault, and the total length of the simulation scenario. Farmer (2001)
5. System monitoring data: as there are a lot of variables available related to the components present in the system and by mentioning and observing every variable makes the analysis so huge to cover up.So only limited number of variable always considered. Farmer (2001)
1. Simulation log: that describes the whole actions occurred during whole simulation process. It has details about the fault applied to the system with control strategy and it also mentions the any difficulty came across the whole simulation. Farmer (2001)
2. Results output: This is an ASCII or binary file that continuously runs for each fraction of time according to absolute time applied to duration of the fault and it contains the recording of each monitored variable over the duration of the simulation. These results are confirmed and checked, through graphical figures, to determine if the system remained stable and to assess the details of the dynamic behaviour of the system. Farmer (2001)
3.5 Sample case considered for design and modelling of distribution network
The model for impact of distribution generation on the stability of the distribution network explained in this case study by (Adel, 2007) is quite handy to use.it describes the sample distribution network with dispersed wind energy by using induction generator which can be modelled by choice as SCIG or DFIG. They have presented a single line diagram as shown in the figure3.1.
3.6 Experimental method
The system (sharaf, 2007) shown in Figure is an 11 kV distribution network with dispersed renewable wind energy, 4 linear loads at power factor 0.8 lagging, motorized load and converter type nonlinear load. Except the wind energy interfaced at bus 2 and the one at the main in-feed point representing an infinite bus as 138 kV, there is no other generation unit in the system. In this case, bus3, bus4, bus5 and bus 6 are meshed in radial structure. Two step-down transformers are used at the main in-feed point and at the bus 5 where a 4160V/600kVA motor is connected and a step-up transformer is employed for WECS grid integration. The detail parameters of the system under study are given in Table 1. As system simulation has done on the environment of MATLAB/Simulink (sharaf, 2007) by considering load switching and wind speed variations by creating and discussed different cases. The technical information of the components used in the single line diagram is also given in the case study (sharaf, 2007) which is the table can be seen in the appendices.
3.7 Analysis of the sample case study:
This case study describes the basic requirements of the distribution system to analyse the stability under the transient stability. The parameters information is also quite complete and the simulation results that have been given in the case study are performed in the environment of MATLAB/Simulink. They have created a component name Modulated Power Filter Compensator (MPFC) and this is included in my model (Sharaf, 2007).
The results describes the whole sequence of faults created by tripping of the model components such as induction motor ,linear load and wind speed for different amount of time in the different scenarios. But according to this thesis objective the model will be based of dynamic simulation in the environment of DIgSILENT power factory.
3.8 DIGSILENT Power Factory:
DIgSILENT power factory presents many functions in same software environs such as load flow ,short circuit calculation, optimal power flow, modal analysis ,contingency analysis, stability analysis which includes RMS (electromechanical simulations) ,EMT transients (electromagnetic simulations) and harmonic load flow.(Hasen, 2007)
Digsilent provides a wind range of library which comprises of almost all the components of the power system. These components are transmission lines, many types of terminals, circuit breakers, fuse ,grounding switch, surge arrester, many types of bus bars, synchronous machines and asynchronous machines which can be converted in to generators as well as wind generators, static generator, wind generator separately, external grid, ac voltage source, ac current source, battery storage, fuel cell, photovoltaic, general load ,low voltage load, many types of shunt filters, many types of transformers, capacitors, converters, dc voltage source, dc current source and inductive coupling .For modelling wind turbine there are some functions available in the work space to input the data according to the requirement of the user. For example if we have to convert asynchronous generator to a wind generator then we have to model both of them separately with given options. (Hasen, 2007)
3.8.1 Electrical machinery available
DIgSILENT has built in models for both SCIG and DFIG with their predefined input and output data and user can use these models as it is. Because asynchronous machine can be converted in to generators and these generators can be converted in to wind farms. So these models can be used as existing project types no modification is needed to use them. (Hasen, 2007)
3.8.2 Squirrel cage induction generator model (ElmAsmo asynchronous machine built in block model) has some parameters constant fixed such as mechanical power, number of poles and kind of connection of generator with the network. There are some defaults values for each parameters such stator resistance and reactance and same for rotor but for getting accurate results user should input data according to need of the hour. On the output side generator speed and active power of the generator is crucial to observe. Generator active power and voltage of the generator bus bar should be specified specially during load flow calculation. (Hasen, 2007)
3.8.3 Doubly-fed induction generator model (ElmAsmsc slip controlled asynchronous machine block model) has the mechanical power, PWM converters inputs and rotor resistance as its inputs. Although speed and active power are major element of the output but for DFIG mechanical angle of the rotor and stator flux can be obtained. While performing the load flow active and reactive power of the stator and rotor slip should be considered. The whole functions of the grid side converter and rotor side converter solved internally by the software. (Hasen, 2007)
So the software has authority in many aspects such as stability analysis which consists of the load flow and calculation of transients, model analysis, contingency analysis, harmonics, reliability analysis, generation adequacy analysis, optimal capacitor presence, calculating short circuits and calculating optimal power flow as described by ( Hasen,2007).
3.9 Some problems and story of modelling the design
DIgSILENT power factory version 14.1 available in lab that I have used has limitations. These limitations are related to limited access to library and the built in models. These models are available in the sample cases present in the data manager and it was great help to see them and observe them in the sample cases. But the problem came in the new project where the every component I have used in the model from generators to transmission line is have to make from beginning as new project type. The whole modelling of the system in DIgSILENT power factory is very important to look at and how the components been modelled and reach to the point of destination to record the results. The process will tell the whole process the problem I faced during the whole journey.
According to the system I choose to make the models of both induction generators based wind turbines with the distribution network ,I have to model and design the components in the software. According to figure above it starts from infinite bus bars which means the external grid we are using in the system.
3.10 Model outlook in DIgSILENT power factory
This is the basic sample distribution network single line diagram that I have made through the case study (sharaf, 2007).This power system has included 6 bus bars with the two connecting bus bars to generator( which either can be SCIG or DFIG )and induction motor. This same model will be used for both technologies of the generators for the sake of comparison between two technologies on stability of the distribution network. System presented by (sharaf, 2007) is built on 60 Hz frequency but I designed it by 50 Hz.
3.11 The power system specifications:
This design can't be done without the authentic parameters values for all the components used in this network. As I have used basically (Sharaf, 2007) for all the components present in the distribution network model presented above in the figure excluding the SCIG model specification which have been taken from the (Energinet.dk,2007) and (EWIS,2008).
There is whole list of values and parameters which have been used in the whole network can be seen in appendices.
3.11.1 Grid modelling
For modelling that grid we can use external gird component available in the equipment [provided in the software. But the compulsory thing is that to attach with the bus bar which can be said in other words infinite bus bar as shown in the figures below. The others features don't do any kind of modifications as those sections have already default values provided by the software.as grid is been used in this model as grid so it should be regarded as slack.
For the transformer modelling 2 winding transformer has been used on the three phase technology. It's quite easy to modelling just by mentioning the HV and LV sides respectively and mentioning the connection type if it's Y or Delta. That can be done as shown in the figures below described the procedure to design the transformer. As three transformers with the different ratings have been used in the system so all the settings can be regard as same among all the transformers.
3.11.3 Connecting lines
By connecting different bus bars to each other connecting lines have been used and in power factory lines are available for modelling the lines length of line should be specified. There are four different lines has been used and parameters are the same for all of them. For design it need to input date regarding rated current and voltage, nominal frequency, type of cable, resistance and inductance of the line.
3.11.4 Modelling and Designing Generator
As the whole thesis work depends on this part because modelling generator in the software with the correct values will leads to the appropriate results. The modelling of generator in Digsilent is so hard but its time taking as we need to model both generator and wind turbine together. In the software equipment asynchronous machines are available and we can attach it to bus bar. While designing it's built in option to weather to make the machine either motor or generator. There is also another option to make simple asynchronous generator in to win generator and also there is an option to turn the simple a synchronous machine in to doubly fed induction machine. There are lists of figures which describes the whole process in detail. But for modelling SCIG there is option in DIGSILENT to use either single cage rotor or double cage rotor and I choose the double cage rotor and use the different parameters for SCIG as compared to DFIG but the active and reactive power are same to both of generators. (Energinet.dk, 2007) , (EWIS, 2008)
3.11.5 Induction motor:
It's quite convenient to model induction motor in the system as by putting asynchronous machine from equipment given in to work area. They two things are important in to designing motor, first is to mentioning bus type such as PQ and second thing to mentioning the active and reactive power values related to it. These are the list of figures that describes the whole process.
We have used both non-linear and linear load in to this model. The only thing is to mention the while designing the loads are to mention the active and reactive power of that particular load. The figure describes the how to input load's data in to software.
3.12 System design:
The basic design of the sample distribution network as shown in the figure above is quite detail for DFIG and but for SCIG it's not difficult to design. The parameters values I used from the case study of (sharaf, 2007) are not that good for system because according that system was not behaving stable.so for remedy for this I had to change the ratings of the transformers and induction motor rated power to make the system more applicable.
3.12.1 System outlook with the original parameters values:
As it is clear from the figure3.5 in appendices by defaults values are not suitable for both transformers at the second bus bar and the fifth bus bar. Because of their more than hundred % present loading which is very harmful for system's life and it is been observed that if try to run this circuit with the same values the transient simulation doesn't occur fully because of the system breakdown.
So I made the changes in the both of the transformers rating from 3.6 MVA to 5.5 MVA and 0.6MVA to 0.85 MVA for both overloaded transformers respectively. The rating of the first transformer was suitable for system.so there was no need to change its value. This case is for the DFIG.
3.12.2The outlook of the system with the SCIG with original values
It is quite obvious from the system all of three transformers are overloaded according to their original values of rating. So by removing this overloading, the rating of the first transformer, second and third transformer is been increased. First transformer's rating is increased up to 9MVA and second transformer up to 5.5 MVA and third one to 0.85 MVA. Unlike the DFIG model in which overloading problem was with second transformer and third transformer .the SCIG model has over loading on the first ,second and third transformers. So after making the modification system looks is acceptable shape.
3.13 Low voltage level on the bus bars:
There is another issue regarding designing the model is the by using the data from the case study (sharaf, 2007) ,the voltage levels are at the second, third, fourth ,fifth, generator and motor bus bars are among lower voltage range <0.95 p.u except for the infinite and first bus bars. These lower voltage levels are not the sign of good system to discuss. The only reason I find out the resistance of the lines given in the case study (sharaf, 2007) seems so high that makes the major loss in first to second bus bar that makes the bus bars voltage levels low. But the analysis can be still carry on because the system still can be run through the transient simulation. (EMT transients)
3. 14 Stability Analysis in the power factory
In the DIGSILENT power factory before running the stability analysis to execute to run the transient simulation, it is compulsory to run the load flow. As load flow calculation will help to analyse the network voltage profile, losses on the branches, steady state conditions and balance between voltage and load.it can also be used for see active and reactive power flows, voltages at the nodes.so the power factory provides a comprehensive method for executing load flow. (Eurostag, 201)
The figures describe the results of the load flow for both models can be seen in appendices.
3.14.1 What is transient analysis?
"Transient analysis computes the circuit solution as a function of time over a time range. Since transient analysis is dependent on time, it uses different analysis algorithms, control options with different convergence-related issues and different initialization parameters than DC analysis. However, since a transient analysis first performs a DC operating point analysis. Some circuits, such as oscillators or circuits with feedback do not have stable operating point solutions. For these circuits, either the feedback loop must be broken so that a DC operating point can be calculated so the initial conditions must be provided in the simulation input." (Utddallas, 2012