08 - Stability Exercises v1.5
08 - Stability Exercises v1.5
PowerFactory
Training Exercises
INTRODUCTORY REMARKS
• The purpose of these exercises is to introduce the basic methods of the analysis of different stability phenomena.
• The instructions in the exercises are deliberately brief. However, more detailed instructions are provided in the Appendix.
• During the exercises there will be two supervisors, who will support and help you with the tasks. Additionally they can
provide answers to general questions regarding the topic of the training (or different problems from your own practice).
Please do not hesitate to address the supervisors at any time to any topic!
• Before starting to create the network in the single-line graphic, we have to build up the library
including the different types with the electrical data of the devices. This is necessary before
entering the elements into the network.
Model definition
Model definition
Input data for the types:
• In the following tables the electrical data for an overhead line type, a 2-winding transformer type
and a synchronous machine type are shown. These are inserted into the local library of the active
project, and used when inputting the elements of the system.
Model definition
Model definition
Input the elements in the single-line graphic:
1. Setup the case according to the single line diagram below. Insert the elements with the
data given in the figure.
2. The generator is set to a constant PV type. It should be dispatched at 800 MW, 1.00 p.u.
voltage.
1. Define an event to apply a short-circuit at time 0s, and to clear the short-circuit at time 0.1s.
2. Selecting the following variables to be recorded:
o Generator Active Power, s:P1
5. Determine the critical fault clearing time of the system. Change the event that clears the short-
circuit manually and repeat the simulation until you find the critical clearing time within 10ms.
Transient stability analysis
2. Simulate a short-circuit in the middle (50%) of the overhead line „CCT1‟ with a duration of
100ms. Clear the fault by opening the line.
4. Apply the script “Critical Fault Screening” to find the critical fault clearing time automatically.
5. You may also analyse the current through the lines CCT1 and CCT2 during the simulation and
the voltage at Bus Bar HV.
Transient stability analysis
Short-circuit simulation at the end of the line CCT1:
1. Simulate a short-circuit at the end (99%) of the overhead line „CCT1‟, i.e. near the HV
busbar, with a duration of 100ms. Clear the fault by tripping the line as mentioned above.
3. Compare the results and the critical fault clearing time of these cases with the case of a
shortcircuit directly on the bus „HV‟.
Transient stability analysis
Short-circuit simulation at the end of the line CCT1:
1. Simulate a short-circuit at the end (99%) of the overhead line „CCT1‟, i.e. near the HV
busbar, with a duration of 100ms. Clear the fault by tripping the line as mentioned above.
3. Compare the results and the critical fault clearing time of these cases with the case of a
shortcircuit directly on the bus „HV‟.
Transient stability analysis
Effect of generator operating point:
2. Modify the voltage of the external grid so that the generator operates at a power factor of
0.95 leading (under-excited).
3. Calculate the critical fault clearing time for a fault on CCT1 close to the HV busbar
(including the tripping of CCT1).
4. What effect does the operating point have on the critical fault clearing time? Why?
Transient stability analysis
5. Define a new scenario “Lagging” and activate it.
1. Insert events to apply a torque step, i.e. an additional torque of 0.02 p.u. for a period of 100 ms. Hint:
use “Event of a Synchronous Machine”: Run the simulation, taking care to use an appropriate time
step. Observe the speed of the generator in a subplot.
2. Determine, and make a note of, the frequency of oscillation, ratio of amplitude deviations and
damping of the generator speed signal.
3. Now switch off the line „CCT1‟ and investigate the differences in the generator speed oscillation.
Instead of investigating the system using time-domain simulations, we will now perform a modal analysis.
Modal Analysis
1. Run a load-flow calculation.
For this exercise we will add a voltage controller (AVR) and a power system stabilizer (PSS) to the
system. The control of the generator is no longer neglected, since it influences the stability of the
investigated system.
• Copy and paste the project “Rotor-angle stability_1” and rename the new project “Rotor-angle
stability_2”.
Automatic Voltage Regulator (AVR)
A voltage controller is created and the step response of the generator system is analysed (closed-
loop response).
1. Define and activate a new study case “Exercise 3.1”. Ensure that all variations and
scenarios of exercise 3 are inactive.
2. Create, within the “Grid” folder, a new “Composite Model” and a new “Common-Model”.
Therefore right-click the generator and select “Define -> Automatic Voltage Regulator
(avr)”
3. Select the Block-Diagram “avr_EXAC4” from the global library “Standard Models”.
4. The Frame “SYM Frame_no droop” from the global library is selected automatically.
5. Calculate a load-flow and the initial conditions to check your project for errors.
Automatic Voltage Regulator (AVR)
Check the model by a step-response (closed-loop) test on the voltage set-point:
1. Test the model with a step response to the input voltage “usetp” (closed-loop test) and change it from its
steady-state value of 1 p.u.
2. To check the setup of the controller, show the block diagram of the controller.
3. Define a variable set to monitor for the AVR element and visualize the results of the simulated step response.
o Short-circuit in the middle (50%) of the overhead line „CCT1‟. Clear the fault by tripping the line
as mentioned above.
o Determine the critical fault clearing time for both cases.
2. Compare the results and the critical fault clearing times of these cases with those from
Example 1 with no AVR.
Automatic Voltage Regulator (AVR)
3. Execute a „Modal Analysis‟ for the system with AVR.
4. Compare the frequency and damping of the local mode of oscillation to the results obtained
in Exercise 2:.
5. What happens to the damping of the local mode if the AVR gain is changed?
6. Change the AVR parameters back to their original values (as tabulated above).
Power System Stabiliser (PSS)
After we analysed the operation of the generator combined with an AVR, now the system
should be stabilized using also a power system stabilizer. The PSS is connected into the frame
2. Create a new controller model for the power system stabilizer in the existing composite model.
Therefore right-click the generator and select “Define -> Power System Stabilizer (pss)”. The
reference inside the existing composite model for the PSS slot is created automatically.Select the
Block-Diagram “pss_STAB2A” from the standard IEEE-Library in the global library of PowerFactory.
3. Calculate a load-flow and the initial conditions to check your project for errors.
Power System Stabiliser (PSS)
Verify the behaviour of the power plant model using a closed-loop test similar to the AVR before:
1. Test the model by applying a step change to the AVR voltage reference.
2. Define a variable set to monitor for the PSS element and visualize the results of the
3. To check the setup of the controller, show the block diagram of the controller.
4. Calculate a load-flow and the initial conditions to check your project for errors.
Power System Stabiliser (PSS)
Now the influence of the additional PSS is to be analysed on the small and large signal stability of the
single-machine problem. Therefore put the frame model, the AVR and PSS into service.
1. Simulate the disturbances as above:
3. Short-circuit in the middle (50%) of the overhead line „CCT1‟. Clear the fault by tripping the line as mentioned
above.
5. Compare the results and the critical fault clearing times of these cases and the cases without the PSS in service.
6. Execute a „Modal Analysis‟ for the system with AVR and PSS.
7. Output the eigenvalues, damping and eigenfrequencies and analyse the differences.Determine the damping
We want to investigate the effect of adding a new load at the HV busbar. We will modify our small
power system model and use QV and PV curves to investigate its voltage stability. PowerFactory
includes scripts that automate the construction of these curves.
1. Make a copy of the project “Rotor-angle stability_1” and change the name of the new
project to “Voltage stability”.
3. Connect a new load, with an active power of 800 MW and a power factor of 0.9 lagging to
the HV busbar.
1. Create a new study case “Exercise 4.1a” by copying and pasting an existing study case.
2. Does the load flow calculation converge with the new load? At what level of power (within 100 MW)
does the load flow not converge anymore?
3. Import the script “P-V-Curve” into the study case of the project.
4. Construct a PV curve for this load using the script “PV-Curve”: Mulit-select the HV busbar and the
load, right-click the selection, choose Execute DPL Script and select the script “PV-Curve”. The power
should be varied from about 10 MW.
First PV-curves
5. Does the voltage remain above 0.90 p.u. for all active powers up to 800 MW? At what load power does the
voltage drop below 0.95 p.u.?
6. Construct additional curves to study the effect of different power factors. In the end, reset the load to 800
MW, 0.9 pf lagging.
7. Create a new study case “Exercise 4.1b” by copying and pasting an existing study case.
8. Create scenarios “CCT1 switched out” and “CCT2 switched out” for use in subsequent parts of the exercise.
9. Investigate, by means of additional PV curves, the effect of one of the two lines being switched out.
QV-curves
• In this exercise we will determine the reactive power requirement, considering different operating scenarios.
1. Create a new study case “Exercise 4.2a” by copying and pasting an existing study case.
4. Connect a static generator to the HV busbar. Ensure that the static generator‟s active power dispatch
is 0 MW and that it is set in voltage control mode. Place the static generator out of service, since we
will only use it for the construction of QV curves.
5. Define a new scenario “QV Base Case”, activate it and place the generator into service.
8. Create additional scenarios for the different combinations of loading conditions and circuits in service.
9. Determine, by means of QV curves, the reactive power requirement for 100% voltage and 95% voltage of all
scenarios.
11. Make some notes about the reactive power requirements considering the minimum load power of 0MW,
the maximum load power of 800 MW and the contingencies. What range of reactive power is required,
considering the different operating scenarios?
Additional PV-curves
• In this exercise we will add switched shunts in order to keep the voltage within the range 0.95
p.u. and 1.05 p.u., considering variations in the load and considering the contingency “CCT2
switched out”.
1. Create a new study case “Exercise 4.3a” by copying and pasting an existing study case.
2. Deactivate any active scenario. The temporary generator should be out of service.
3. Add a capacitive shunt to the HV busbar. The shunt should be set to switch automatically to
try to keep the voltage in the range 0.98 p.u. and 1.02 p.u.
4. Edit the load flow options and enable “Automatic shunt adjustment”. Also set the tap
adjustment method to “stepped” and the maximum tap changes per iteration to 1.
Additional PV-curves
5. Determine (by means of PV curves) a suitable rating and number of steps of the shunt, so that a
load of 800 MW, 0.9 pf lagging can be supplied without causing the voltage to fall below 0.95
p.u.
6. Design the shunt to limit the step change in voltage during switching below 3%.
7. Create a new study case “Exercise 4.3b” by copying and pasting an existing study case.
8. Create an additional scenario for the case where CCT2 is switched out. Check the performance
of the shunt for this condition.
Exercise 5
Case studies of a multi-machine power
system
Case studies of a multi-machine power system
• The studies will be carried out using a system having the structure shown in Figure 3. It consists of
four areas that are interconnected by a number of tie lines, plus a 2 pole HVDC link to an external
area, represented by static generator elements. Areas South-West, North-West and North-East
are closely interconnected by shorter tie lines, whereas Area South-East is loosely connected to
the rest of the system. Most generators are equipped with typical voltage controllers and power
system stabilizers as well as governor models.
Case studies of a multi-machine power system
• Two scenarios and two study cases are given. These define an importing and exporting power
flow conditions of the South-East system, as shown in Figure 3 and Figure 4. The area interchange
is controlled by means of a Power Controller “Power Exchange Controller South-East”. Please
familiarize yourself with the model, identifying the plant- and system controllers, power
interchanges etc.
Case studies of a multi-machine power system
Case study 1: Frequency Stability
Case study 1: Frequency Stability
• The network consists of four areas. This exercise will deal with different modelling issues
regarding frequency stability: The effect of inertia, primary reserve, governors, loads, under-
frequency load shedding etc. Various critical contingencies are analysed: The tripping of a single
generator, the tripping of one or both interconnector lines between the areas North-East and
South-East, and the tripping of an entire power plant in the South-West area (generators 1-4).
Case study 1: Frequency Stability
2. Simulate the tripping of this generator. Plot the speed of at least one generator in each area and scale
the plots to represent the frequency in Hz.
3. Does the location of the disconnected generator or the transfer condition affect the resulting
frequency deviation?
4. Will the frequency remain above 49.7 Hz, following a single generator trip, without requiring any load
shedding.
5. Would the tripping of a major generator lead to the overloading (in steady-state) of the
interconnecting lines between the areas South-East and North-East?
Case study 1: Frequency Stability
• Load modelling:
The active power of loads within a typical power system varies with frequency. We will now study the importance of the
frequency dependence of loads in the network.
1. Study the load model(s) implemented in the network. Do the load models include frequency dependence?
2. Change the frequency-dependency of active power of all general loads to 0.1, and repeat the investigation of the
single generator trip above.
3. How does the modified load characteristic affect the transient and steady-state frequency?
4. Modify the frequency-dependence to 1.0. How does this affect the frequency in response to the generator trip?
5. Please leave the frequency-dependence of loads factor at 1.0 (for all loads) for all further investigations.
Case study 1: Frequency Stability
The inertia is important for the initial slope of the frequency drop/rise. The effect of the inertia
What is the maximum and minimum frequency in both remaining asynchronous parts of the network?
2. Include another generator in the dispatch and study the tripping of a single generator again. How is the
slope of the frequency and the overall minimum influenced? Please use a separate scenario for the new
dispatch.
4. Deactivate the new scenario, i.e. continue working with the original dispatch.
Case study 1: Frequency Stability
1. Make some notes about the gas turbine governor droop settings of the generators in the South-East
area.
2. Change the droop of these gas turbine governors in the range 2%-10%.
3. How are the frequency response and steady-state frequency deviation influenced?
of one pole tripping, how does the frequency respond? Does the system survive a surplus of 1000 MW? Consider
Area South-East is only connected by one double circuit interconnector to area North-East. Now we will look
asynchronous parts of the network? Consider both exporting and importing conditions.
3. Now a solid 3-phase fault is simulated at one end of the line. As malfunction of the protection both lines are disconnected
after 100ms. Is the transient frequency deviation affected by the short circuit?
5. Analyze the frequency behaviour for the case of 500 MW export when the interconnector is tripped with and without fault.
6. What differences in the response of frequency of both separated system islands can you observe for the Import and Export
condition?
Case study 1: Frequency Stability
• Influence of load-shedding:
To reduce the minimum frequency, a frequency-based load shedding scheme is used in the network:
1. Make two new study cases (by copying and pasting the original study cases). Activate them, and enable the protection relays for
all calculations (load flow options -> Advanced Simulation Options->Consider Protection Devices).
2. Study the response of frequency to the tripping of the NE-SE interconnector (both lines), and compare the results with the case
3. Will both parts of the system survive the contingency under all conditions? What is the minimum frequency and the steady-state
4. Check the messages in the output window. At what frequency is load-shedding started? Check the settings of the protection
5. Will all areas survive the contingencies? What are the minimum frequencies and the steadystate frequencies?
6. Study also the contingency of the simultaneous multiple unit trip in the South-West are (generators 1-4).
Case Study 2: Transient Stability
Case Study 2: Transient Stability
• To analyse the transient stability of a network, two indicators can be used:
o Critical fault clearing time – maximum duration of a short circuit at a given power flow, when no pole
slipping of generators occurs.
o Maximum power transfer – maximum active power transfer across a line at a fixed fault clearing time,
when no pole slipping of generators occurs.
• Both values cannot be calculated by one single simulation but have to be determined by an
iterative process.
• For the rest of the exercises below, please consider both exporting and importing conditions.
Case Study 2: Transient Stability
• Short-circuit simulation on line between area South-West and North-West:
1. Simulate solid 3-phase short-circuits at both ends of one of the interconnecting lines between these two areas.
2. Determine the critical fault clearing time (CCT) for both cases.
5. Run a time –domain simulation of the two faults with a clearing time very close to critical, and then with a clearing time
of half (or less) of the clearing time found. Compare the results.
6. Make the lines longer (l=300km). How does this affect the results from above?
7. Switch off one of the two lines, and repeat the time-domain simulations for the faults. How is the transient stability
affected?
8. Now set in service the out-of service line between area South-West and area North-East, and run again the fault
simulations. How is now the system response?
1. Simulate a single-phase fast reclosing scheme using the following event sequence:
o Single-phase short-circuit on the end of a line.
o Disconnection of the faulted phase at both ends of the line after 0.15s.
o Disconnection of the faulted phase at both ends of the line after 0.15s.
2. Set the fault clearing time for the 3-phase fault to 150ms.
3. Run the simulation for the predefined power transfer on this line (500MW). Is the network stable?
5. What is the maximum power you can safely import to area South-East without losing transient stability?
Case Study 2: Transient Stability
• Influence of generator controllers:
1. Neglect the power system stabilizer (pss) models for the simulation. Use the option the “Advanced
Simulation Options‟ page of the load-flow command.
What is the behaviour of the network now? How are the CCTs/transfer limits affected?
2. Neglect also the voltage regulators (vco) models for the simulation. Use again the option the
“Advanced Simulation Options‟ page of the load-flow command.
What is the behaviour of the network now? How are the CCTs/transfer limits affected?
Case Study 2: Transient Stability
• Load modelling:
Note that the load has been modelled as voltage-dependent.
1. Change the voltage dependency on P and Q of all general loads to zero (const. power).
4. Compare the critical clearing times? Which load type is giving better results?
5. Change the voltage dependency factor on P and Q of all general loads to 2 (const. impedance). Can
you reproduce the results above?
Case Study 3: Oscillatory Stability
Case Study 3: Oscillatory Stability
We now want to analyse the oscillatory behaviour of the network. Please follow the instructions below for both
scenarios.
1. Run a load-flow calculation, with the relays disabled.
3. Create an eigenvalue plot with the following scale, select the option “Damped Frequency” forthe y-axis:
- xmin = -5/-10; xmax = 0
- ymin = 0; ymax = 2/5
6. Identify local- and inter-area modes of oscillation using mode phasor diagrams.
7. Find the inter-area mode between area South-East and the rest of the system. Which frequency and damping does this mode have?
Create an event to excite the inter-area mode (e.g. add a temporary torque step to the generator which has the highest participation).
Check the line flow between the areas SouthEast and North-East. Measure the frequency and damping of the oscillation.
Case Study 3: Oscillatory Stability
• Increasing the impedance between areas:
1. Set the length of the lines between areas South-East and North-East from 200km to 300km
and then to 400km.
2. How are the damping factors and the frequencies of the modes affected? Check especially
the inter-area mode and note the damping for each line length.
1. Set the length of the lines between these areas back to 200 km and set one of the lines out
of service.
2. How are the mode frequencies and their damping affected? Check especially the inter-area
mode and note the change in damping and frequency of the mode.
2. How are the damping factors and the eigenfrequencies of the modes affected? Check
especially the inter-area and note the damping for each power transfer.
3. How does the line length influence the inter-are mode frequency and damping?
• Load modelling:
1. Change the type of the loads to constant power loads.
2. Are there any changes between the results with and without voltage dependency?
Case Study 4: Voltage Stability
Case Study 4: Voltage Stability
• We now want to analyse the system‟s voltage stability using active power-voltage (P-V) curves. Scripts are available for the construction of
these curves.
1. Set the „Power Exchange Controller South-East‟ out of service and reassign the generators from the South-East area to the
frequency controller. Why is this required?
2. Make sure that the option “consider reactive power limits” is enabled in the load flow command.
3. Mark all the loads and busbars from the South-East area (excluding the generator terminals) and execute the DPL script „P-V-Curve‟.
4. How large is the power transfer between the areas at the end of the voltage limits?
5. Which busbar/ terminal in area South-East seems to be affected most by increasing amount of load? Why this terminal?
6. Check the reactive power support of the generators in area South-East. Are they near/in their limits?
7. Change the load power factor for area South-East to 0.9. How are the results affected? Return the loads power factor to its original
value.
Case Study 4: Voltage Stability
• Considering reactive power limits:
1. Run the PV curve script again without considering the reactive power limits (in the load-flow command).
2. What is the new voltage stability limit? How is the maximum power transfer influenced?
o TSC = 3x50Mvar