1 Selection of Dynamic performance Control Parameters for Classic HVDC in PSS/E Optimization of CCA and VDCOL parameters Master of Science Thesis Arunkumar Muthusamy Department of Electric Power Engineering Division of Energy and Environment Chalmers University of Technology 412 96-Göteborg, Sweden-2010. 2 Selection of Dynamic Performance Control Parameters for Classic HVDC in PSSE Optimization of CCA and VDCOL parameters Master of Science Thesis Conducted at: ABB Power Systems, Ludvika Examiner: Massimo Bongiorno Chalmers University of Technology Department of Energy and Environment Division of Electric Power Engineering 412 96 Gothenburg, Sweden. Supervisor: Per-Erik Björklund PSDC/DCTSM Department, Power Systems, ABB Ludvika, Sweden. Wenkan Huang PSDC/DCTSM Department, Power Systems, ABB Ludvika, Sweden. 3 4 Abstract The objective of this thesis work is to define the strategy of tuning current controller amplifier and voltage dependent current order limiter parameters to achieve good dynamic performance with a wide range of a.c. systems in PSS/E. The defined control parameter could be adapting independent of different power levels and configurations. Basically, three different HVDC transmission types are used such as overhead lines, cables and back to back. The dynamic control performances are varied with respect to transmission types. Hence, the control parameters could be different. The control parameters could be intended for stability improvement of HVDC system connected with wide range of AC system parameters. As a part of a HVDC project design, large efforts are made in tuning control system parameters for the specific project conditions. However, for customer support, in planning studies it is necessary to provide a set of robust control parameters giving a representative performance for “any” configuration. This thesis work mainly deals with dynamic response of HVDC system. The dynamic instabilities occur mostly at d.c. system connected to weak receiving a.c. network. It causes high ac voltage fluctuations as a consequence very difficult to recover the dc system after the fault clearing instant due to repetitive commutation failures. Therefore, good dynamic dc control strategy has to define by tuning the current control amplifier and voltage dependent current order limiter. This will be done by analyzing the responses of the HVDC model for different transmission types, network strengths and control parameters. Results will be evaluated, and presented so that it is possible to define performance and what needs to be improved. Evaluation will be made of the responses obtained using the existing parameters and those obtained using the set of rules proposed in the thesis. 5 Acknowledgement This thesis work was conducted at ABB AB Power systems, Ludvika, Sweden. I would like to thank Mr. David Shearer, for providing me this big opportunity and helpful discussions during this period. I would like to express my sincere gratitude to my leading supervisor Mr. Per-Erik Björklund for his guidance and encouragement. Also, thank him again for helping me with writing and correcting this report. I would like to thank Mr. Wenkan Huang for his assistance regarding modeling and support during the whole period of this work. Further, I wish to thank Mr. Per Holmberg and Mr. Boris Nordström for their good suggestions and comments. I wish to thank also Mr. Massimo Bongiorno, my examiner at Chalmers University of Technology for reading carefully and correcting the manuscript even though he is in parental leave. Eventually, I would like to thank my family for their love and unconditional support. Arunkumar Muthusamy 6 List of abbreviation AC Alternating Current HVDC High Voltage Direct Current CCA Current Control Amplifier VDCOL Voltage Dpendent Current Order Limiter VCA Voltage Control Amplifier VSC Voltage Source converter CFC Converter Firing Control CEA Constant Extinction Angle CC Constant Current RAML Rectifier Alpha Minimum Limiter PSS/E Power System Simulator for Engineers List of Symbols Valve ignition delay angle Overlapping angle Commutation margin angle Udio No load direct voltage dXN Commutation reactance dRN Commutation resistance Udc Direct voltage Iorder DC current order max Alpha maximum minRec Alpha minimum rectifier minINV Alpha minimum Inverter Imargin Current Margin Udref DC voltage reference IorderLim Current order limited (VDCOL output) Porder Power order ref Gamma reference Gammao Commutation margin angle UD Low Direct voltage lower threshold UD High Direct voltage higher threshold AMAX Alpha maximum (rectifier) AMIN Alpha minimum (rectifier) Io low Current order lower limit Io abs max Current order maximum limit order Alpha order CFprev_angle Commutation failure prevention angle Idmeas Measured direct current Ud_Tc_Dn Direct voltage down time constant Ud_Tc_Up Direct voltage UP time constant 7 Contents Abstract:.................................................................................................................................4 Acknowledgement..................................................................................................................5 List of abbreviation.................................................................................................................6 List of Symbols ......................................................................................................................6 Contents .................................................................................................................................7 Scope of thesis work:............................................................................................................10 Outline of thesis work: .........................................................................................................10 1.0 Introduction: ...................................................................................................................11 1.1 HVDC technology: .................................................................................................11 1.2 History:...................................................................................................................11 1.3 Types of HVDC configuration: ...............................................................................11 1.4 Line commutated converter. 6 Pulse Bridge ...............................................................14 1.5 Case with no overlapping period .................................................................................14 1.6 AC and DC current relationship: .................................................................................15 1.7 Case with overlapping period less than 60 degrees ......................................................16 2.0 HVDC Classic control characteristics: ............................................................................20 2.1 HVDC control overview .........................................................................................20 2.2 Alpha-minimum characteristics at rectifier: .............................................................21 2.3 Constant current characteristics at rectifier: .............................................................21 2.4 Constant extinction angle characteristics: ................................................................22 2.5 Alpha minimum at inverter: ....................................................................................22 2.6 Current margin:.......................................................................................................22 2.7 Modified inverter characteristics: ............................................................................23 2.8 Regulator Overview ................................................................................................24 2.9 Active power control:..............................................................................................24 2.10 Voltage dependent current order limiter (VDCOL):...............................................25 2.11 Current control Amplifier (CCA): .........................................................................26 2.12 Voltage controller: ................................................................................................28 2.13 Alfamax inverter control: ......................................................................................31 2.14 Rectifier alpha min limiter (RAML) ......................................................................32 2.15 Commutation failure prevention control: ...............................................................32 2.16 Tap changer control: .............................................................................................33 2.17 Inverter gamma0 start function:.............................................................................33 2.18 Equidistant firing control:......................................................................................34 2.17 Recovery of AC and DC system faults: .........................................................................34 2.17.1 AC system faults: ...............................................................................................34 2.17.2 DC system faults: ...............................................................................................34 2.17.3 Commutation failure: .........................................................................................35 2.18 Conclusion: ...........................................................................................................36 3.0 Performances scan of HVDC classic models:..................................................................38 3.1HVDC Modelling in PSS/E: ........................................................................................38 3.1.1Introduction to PSS/E:...........................................................................................38 3.2 Power Flow data: ........................................................................................................38 3.2.1Bus data: ...............................................................................................................38 3.2.2 Load data: ............................................................................................................38 3.2.3 Generator data:.....................................................................................................39 3.2.4 Branch data: .........................................................................................................39 8 3.2.5 Fixed shunt data: ..................................................................................................39 3.2.6 Two-terminal DC line data:..................................................................................39 3.3 Benchmark cases: .......................................................................................................41 3.4 Test cases................................................................................................................43 3.4.1 3-phase to ground AC fault: .................................................................................43 3.4.2 Commutation failure: ...........................................................................................43 3.4.3 DC line fault: .......................................................................................................43 3.4.4 Dc line fault and recovered with reduced voltage: ................................................43 3.4.5 Power step: ..........................................................................................................44 3.4.6 D.C. voltage step:.................................................................................................44 3.4.7 Gamma reference step:.........................................................................................44 3.5 AC fault at inverter: ....................................................................................................44 3.5.1 The basic control operation (fault at inverter ac system) .......................................44 3.6 Ac fault at rectifier station:..........................................................................................45 3.6.1 Basic control operation (rectifier side ac fault) .....................................................45 3.7 DC line fault at rectifier end:.......................................................................................45 3.8 Control sequence during step change ..............................................................................45 3.8.1 Gamma reference step:.........................................................................................45 3.8.2 Power step: ..........................................................................................................46 3.8.3 Dc voltage step:....................................................................................................46 3.9 Control Parameters (default): ..........................................................................................47 3.9.1VDCOL (Voltage dependent current order limiter) .......................................................47 3.9.2 CCA (current control Amplifier) ..................................................................................47 3.10 Conclusion of performance scanning: ...........................................................................50 4.0 Control parameter optimization:......................................................................................52 4.1 Current controller amplifier:....................................................................................52 4.2 Voltage Dependent Current Order Limiter (VDCOL):.............................................54 4.3 Optimized VDCOL parameters: ..............................................................................55 4.4 Optimized CCA parameters: ...................................................................................55 4.5 Parameters of Plot:......................................................................................................55 4.6 Results:...........................................................................................................................56 4.6.1 Cable transmission ...............................................................................................56 4.6.2 Overhead line transmission ..................................................................................65 4.6.3 Back to Back........................................................................................................69 4.7 Conclusion: ....................................................................................................................72 5.0 Nordic-32 network..........................................................................................................75 5.1 Modified Nordic-32 network with HVDC links.......................................................79 5.2 3-phase to ground fault at node 4072.......................................................................80 5.3 3-phase to ground fault at node 4042.......................................................................82 5.4 3-phase to ground fault at node 4045.......................................................................84 5.5 3-Phase to ground fault at node 4040.......................................................................85 5.6 Conclusion:.............................................................................................................86 6.0 Future work: ...................................................................................................................87 7.0 Reference: ......................................................................................................................88 8.0 Appendix:.......................................................................................................................89 8.1 The case study simulation results (version 3.0.14.beta) ...................................................89 8.2 Nordic-32 network..........................................................................................................91 8.3 Frequency dependent representation of DC system .........................................................91 8.4 PSS/E raw file ................................................................................................................96 8.5 PSS/E DYR file ..............................................................................................................96 9 10 Scope of thesis work: ABB have different types of HVDC simulation models. One type is called rms-models or ac quantities represented as phasor. This type of model is suitable for planning studies and stability studies, typically performed by utilities. The rms-model used in this thesis work was Power System Simulator for Engineers (PSSE). It is mainly used for steady state analysis like load flow analysis, optimal power flow, switching studies etc. and dynamic simulations like transient, dynamic and long-term stability analysis. The PSSE HVDC classic model has to be robust and user friendly to the customer, so the control parameters should be optimized to get good dynamic performance. Mainly, this document describes about dynamic performance studies of HVDC classic system connected to weak ac networks. Outline of thesis work: Performance analysis of HVDC classic models The dynamic performance of the HVDC system could be analysed by applying faults at ac or dc bus nearer to the converter, it could be inverter or rectifier. There are certain factors which shall be verified, like recovery time of dc system after fault clearing, non-repetitive commutation failure during recovery, avoid rectifier inrush current at inverter fault or dc fault in order to prevent commutation failure at rectifier, oscillatory power recovery after fault clearing due to improper damping. Basically, dynamic performance study is to investigate the performance of the HVDC control system. Also, some specific tests could be done to check the performance of the each controller i.e. current control, voltage controller. Control parameter optimization The overall good dynamic and steady state performance of HVDC classic system is dominated by the CCA and VDCOL parameters. Verification of optimized parameters The optimized parameters have been verified by simulating with these parameters in different cases which has defined in the performance analysis. The results are plotted. Modelling of HVDC links in Nordic-32 system The classic HVDC links and Light links are implemented and applied shunt faults at different nodes which nearer to rectifier and inverter in order to analyse the control system performance during transients. Conclusion and future work Frequency dependent representation of dc system 11 1.0 Introduction: 1.1 HVDC technology: The tradition HVDC classic technology is used to transmit power for long distances via overhead lines or submarine cables with reduced losses. There is a breakpoint between ac and dc transmission distance, where after this point dc transmission is smarter and effiecient.It also reduces the synchronous constraints between the two ac systems. It enhances steady state and dynamic stability of the ac system. The recent 15 years, also a new technology is used, HVDC Light, based on VSC (Voltage Source Converters). HVDC Light has considerably higher dynamic performance compared to HVDC Classic, but still HVDC Classic is dominating for low cost bulk transmissions. 1.2 History: The first HVDC transmission in the world was begun in 1954. It was 150KV, 20MW DC link between Swedish main land and island of Gotland by ASEA. Until 1970, mercury valves are used for conversion of direct current. After a powerful invention of high power electronic device so called thyristors for static power transfer has been encouraged by power industries and its substantial increase of rating and reliability over the years. 1.3 Types of HVDC configuration: There are different configurations of DC links used for transmitting power based on power capacity. Monopole, ground return The power is transmitted from one converter station to another station through one conductor (positive or negative polarity) and return is grounded at both stations (figure 1). Figure 1 Monopole, ground return Monopole, metallic return. The power is transmitted from one converter station to another converter station through one conductor and metallic conductor is used as return and grounded at one end (figure 2). 12 Figure 2 Monopole, metallic return Bipolar Bipole (figure 3) has two conductors, upper pole is operating in positive current and positive voltage and lower pole is operating in negative voltage and negative current. Both poles transmit a power in same direction. It is grounded at both stations. Both poles are operating at equal currents during steady state, therefore zero current through the ground. It can be operating as a single pole during fault at another pole. Figure 3 Bipole Monopole, midpoint grounded. Monopole, midpoint grounded (figure 4) has two conductors as same as monopole metallic ground, but instead of using the return conductor as power transmission conductor, grounded at one end. It reduces the transmission loss since cables can have full voltage, the total transmission voltage is doubled, and consequently the current needed is half for same power and transmission capacity has been increased. 13 Figure 4 Monopole, mid-point grounded Back to back. Both rectifier and inverter stations are located at same place (figure 5). The normal configuration is to use monopolar blocks, but several converter blocks can be installed in parallel, each with separated dc circuit. The purpose of this kind of configuration is to connect two asynchronous systems. It reduces the total system cost, due to absence of lines/cables; current rating of the system shall be increased with reduced voltage. Thus, transformer size could be reduced. Figure 5 Back to Back 14 1.4 Line commutated converter. 6 Pulse Bridge Figure 6 6-pulse GRAETZ bridge Since the operation of 6 pulses GRAETZ circuit is well documented in many books, even though some basics are needed to understand the below chapters. Line commutated converter is nothing but current source converter, in order consider the theoretical analyse of 6-pulse bridge ( figure 6) we have to assume infinite 3-phase ac source connected to the converter transformer with finite leakage inductance (L) , ideal valves and infinite smoothing reactor. Because of finite leakage inductance in converter transformer (protect the valves during valve short circuit) commutation from one valve to another valve is not instantaneous. During commutation, three valves will be conducting. Such a period is called overlapping period ( ) which is normally less than 60 degrees. 1.5 Case with no overlapping period Instantaneous line to neutral voltages of three phases is 60)tcos(EmUa (1.1) 60)tcos(EmUb (1.2) 180)tcos(EmUc (1.3) Corresponding line to line voltages are )30tcos(Em732.1UcUaUac (1.4) )90tcos(Em732.1UaUbUba (1.5) )150tcos(Em732.1UbUcUcb (1.6) The instantaneous direct voltage( shown in figure 7) across the 6-pulse bridge is composed of 60 degrees of line to line voltage across the upper and lower conducting valves. Therefore, the average direct voltage can be found by integrating the 60 degree area with no ignition delay. 0 60 0 60 d)30cos(Em3dUac3Udo 15 Figure 7 DC voltage waveform without delay angle Em33Udo (1.7) Where, Em is the peak value of the line to neutral voltage. Udo is the ideal no load direct voltage with =0 LLE23Udo (1.8) Where, ELL is the rms Line to Line voltage. With ignition delay angle (shown in figure 8), the integration limit simply increased by and therefore Figure 8 DC voltage waveform with delay angle cosE23cosEm33Udo LL (1.9) The effect of delayed angle will reduce the average direct voltage by cos factor. Since the practical alpha can be varied from 5 degrees to 142 degrees, dc voltage polarity can be changed for power reversal and current can not be reversed since unidirectional device has been used. 1.6 AC and DC current relationship: Assuming no losses in the converter bridge, Power sending from the AC network is equal to the DC power cos.IU3IdUdPd 1LLN (1.10) Where, Pd is direct current power ULN is the line to neutral voltage 16 IL1 is the fundamental frequency line current is the phase angle between line voltage and line current shown in figure The line current is in rectangular shape (figure 9) since Id is constant (smoothing reactor prevents current from changing). Each valve in the 6 pulse bridge conducts 120 degree per cycle. With no delay angle, line current is in phase with voltage. With delay angle, the line current lags voltage. This is the reason why the converter consumes reactive power. By fourier analysis, rms value of line current is given by 60 60 1L dcosId2I Id6I 1L (1.11) Figure 9 AC and DC current relationship with and without delay angle 1.7 Case with overlapping period less than 60 degrees As same reason as explained before, due to converter transformer inductance, valve takes some finite time to transfer a current from one valve to another valve. Such as called overlapping time. 17 Figure 10 converter bridge circuit with valves 1,2 and 3 Figure 11 commutation of current in valve 1 and 3 If suppose the valve 1 and valve 2 is conducting at an instant (figure 10), and valve 3 is fired and starts conducting. But still valve 1 is conducting until the stored magnetic energy in the converter transformer goes to zero. During the commutation Id is summation of i1 and i3, i1 is decreasing from Id to zero and i3 is increasing from 0 to Id (figure 11). Figure 12 direct voltage with effect of delay angle and overlapping angle 18 Voltage drop due to overlapping angle (figure 12) can be achieved by IdL 2 6IdLf6Ud IdL3Ud (1.12) Since the current changes through the inductance is 6 times for 6-pulse converter Figure 13 rectifier direct voltage Therefore, the total rectifier DC voltage will be IdLUUd LL 3cos35.1 (1.13) Figure 14 Inverter direct voltage 15 The inverter DC voltage will be found by 180 )3cos35.1( IdLULLUd (1.14) 19 20 2.0 HVDC Classic control characteristics: 2.1 HVDC control overview Normally, HVDC system operates in constant power control mode. Power order is given by the user. Current order (Iorder) derived from the power controller, which is send to the VDCOL (voltage dependent current order limiter) and into the current control amplifier (CCA). The alpha order from the CCA is send to the converter firing control which determines the firing instant of valves (shown in figure 16). The function of VDCOL and CCA will be explained later in this report. Figure 16 HVDC control overview The primary function of HVDC controls are: Fast and flexible power control between the terminals under steady state and transient operation. Better stability of ac system. Fast protection of ac and dc system faults. i) it minimizes over voltage across the valves ii) it reduces the short circuit current through the valves and lines/cables iii) it reduces the reactive power consumption iv) avoids repetitive commutation failures These above advantages are achieved by varying exact firing instant of valves. The converter firing control which determines the firing instants for each valve to determines the rated DC voltage. The input for the firing control system could be the output of current control, voltage control, minimum alpha control, and minimum commutation margin control mode or alphamax control. Usually, dc transmission controlling and co-operation between rectifier and inverter has been explained based on Ud/Id characteristics (figure 17). Traditionally rectifier controls the current and inverter operates with constant commutation margin under normal operation. 21 Figure 17 Ud-Id characteristics Under steady state, typically rectifier would be act as constant current source i.e. constant current control and inverter will operate as constant counter voltage source i.e. constant extinction angle. The current order at the rectifier is determined by the manipulation of power order and inverter dc voltage. To maintain stability at rectifier, it is necessary to have less (Idref – Id) deviation in dc current and also ( meas- ref) deviations should be keep as low as possible for inverter stability. The intersection of two modes gives normal operation point. 2.2 Alpha-minimum characteristics at rectifier: This characteristics is determined by the equation shown below, U U d d U I Idi xN rN di N dcN dcdc 0 0cos (1.15) The above equation determines the dc voltage across the converter. If we assume practical minimum alpha of 5 degrees in order to have certain voltage across the valve before firing and transformer reactance (dxN + drN) Udi0N/IdcN are also always constant. Hence, increasing dc current reduces the dc voltage i.e. negative slope determined by the transformer reactance and dc current (reduced voltage due to overlapping of valve currents). 2.3 Constant current characteristics at rectifier: This characteristic could also be explained by the same equation (1.15), by assuming current as constant and alpha as variable. It can be seen from the figure 18 that higher dc voltage at minimum alpha and increasing of alpha decreases the dc voltage. The direct current is determined based on the current order, which could be selected between minimum current capability and the rated current of valves. The maximum current carrying capacity of valves would be determined for a transient time period to limit valve stress. 22 Figure 18 Ud-Id characteristics 2.4 Constant extinction angle characteristics: Inverter is normally operating as alpha-max or constant commutation margin mode in order to have certain extinction angle to commutate the valves without fail. Under normal operation, inverter operates at =17 at 50Hz, it is not recommended to increase or decrease to limit reactive power consumption and avoid commutation failure. At steady state, inverter operates normally as constant dc voltage control mode. Assuming gamma constant and Idc as variable gives negative slope characteristics. This slope would be even more negative if the ac system is weaker. dc dcN N0di rNxN0didc I I U ddcosUU (1.16) 2.5 Alpha minimum at inverter: The power reversal could be obtained by increase the current order of the inverter higher than rectifier. In case of dc line fault, it is recommended that both converters should operate as inverter to make the fault current in dc line to zero as fast as possible. If there is no minimum alpha limit at inverter, it could also operate as rectifier by reduced alpha cause feeding of dc fault. Therefore, always minimum alpha (figure 19) at the inverter is limited to 1100. However, rectifier could be operating as inverter for reason explained above. Also because of one more reason, inverter should have minimum counter voltage to start current flow after the fault clearance. 2.6 Current margin: To de activate the inverter current controller at normal operation, current order at inverter is subtracted from rectifier by 10% .such as called current margin. The solution with a current controller also at the inverter, but normally deactivated by the current margin, avoids the current to become zero during disturbances at rectifier. When there is sudden voltage drop at the rectifier ac system cause hit the minimum alpha limit, if there is no current controller at inverter, reversed potential difference between inverter and rectifier will force the current to 23 zero since the unidirectional current device has been used. Alpha at inverter could increase up to when it hits minimum extinction angle. 2.7 Modified inverter characteristics: Figure 19 Ud-Id characteristics with VDCOL Figure 20 constant Beta and constant voltage Constant beta control (figure 20) is nothing but Alphamax control which is explained in the previous page. Constant voltage control is normally used under reduced voltage operation; the Udref is set to higher value at normal voltage operation to avoid hunting between Alphamax and Voltage controller. But in reduced voltage operation Udref shall be reduced to nominal reduced voltage. 24 2.8 Regulator Overview Figure 21 Overview of HVDC regulator 2.9 Active power control: Power control is normally a closed loop control required for stable operation of HVDC system. The power order is determined by the user at master station (rectifier). The current order to the current control amplifier can be derived from the equation Iord= Pord/Ud Current controller is normally a PI regulator, input for the PI regulator will be error of measured value and reference value of current, the output from the regulator determines the firing instant of valves hence the dc voltage to maintain constant potential difference between converters. Ud is determined by the inverter under normal operation (constant extinction angle). The current order is sent to slave station (inverter) by telecommunication link. In order to de activate the current controller at inverter, current order is reduced by current margin, typically 10% of nominal current. 25 Under normal condition, system operates in constant power control mode; it can be switched to constant current control mode when telecommunication failure happens. 2.10 Voltage dependent current order limiter (VDCOL): Figure 22 Voltage dependent current order limiter The main function of VDCOL (figure 22) control is to reduce the current order to lower value when there is a reduction in dc voltage due to contingencies, in order to prevent the higher consumption of reactive power and valve voltage stress. If the system is operating in constant power control mode, current would be increased to maintain the constant power at low voltage causes higher consumption of reactive power will decrease the ac voltage further and induce commutation failure during the recovery of dc system. The break points (figure 23) for dc voltages are typically between 70% and 30%. These breakpoints could vary relative to the strength of the ac system. Normally, the long lines have same Udhigh breakpoints for both rectifier and inverter but in contrast long cables need different Udhigh break points for charging the cable i.e. 50% for rectifier and 90% for inverter. If there is fault in the inverter end, voltage would decrease greatly, if the VDCOL is not activated, the power control mode will increase the current to keep the power constant. Increased current will increase the reactive power consumption of converter; it will increase the risk of subsequent commutation failures if the ac system is relatively weak. The low pass filter time constants (figure 23) Ud_Tc_Dn and Ud_Tc_Up are down time and up time delay of break point limits. The down time for the inverter should be fast to prevent commutation failures but for rectifier it need not be too fast during inverter fault there is a threshold limit for VDCOL (0.8), if the dc voltage reaches threshold value; current order is decreased down rapidly to predefined lower value to prevent the consecutive commutation failure during inverter side ac fault. The down time constant could be same for both stations. But the up time constant could not be, rectifier should restart before the inverter. So the up time constant for inverter should be higher than rectifier (mostly depends upon how strong the ac system is). Contrary, if the inverter restart first, it start build the counter voltage thus the dc current, might loose the current margin causes power reversal. This function prevents the commutation failures during the recovery. Obviously the valve stress is reduced. 26 2.10.1 Current limits of current order: Maximum current is depends upon the thermal overload of valves. Minimum current should be around 0.3 p.u. in order to avoid valve extinction and high valve voltage stress since the overlapping angle is directly proportional to the dc current. So very low dc current gives small overlapping angle causes voltage spikes at starting and ending of commutation combines together creates twice the voltage spike. Figure 23 VDCOL characteristics with voltage and current limits The power reversal could be obtained by increase the current order of the inverter higher than rectifier. In case of dc line fault, it is recommended that both converters should operate as inverter to make the fault current in dc line to zero as fast as possible. If there is no minimum alpha limit at inverter, it could also operate as rectifier by reduced alpha cause feeding of dc fault. Therefore, always minimum alpha at the inverter is limited to 1100. However, rectifier could be operating as inverter for reason explained above. Inverter should have at least minimum counter voltage (i.e. cos(110)) to have current flow during start-up after fault with zero dc current. 2.11 Current control Amplifier (CCA): The current control amplifier (figure 24) is used as the main function which used to control the firing angle of the converter under steady state and dynamics of HVDC system. The current controller is basically a Proportional and Integral regulator. As we all know the proportional part helps to give fast response with respect to the feedback and integral part is a slower part which used to make steady state error zero. The current error (Iorder-Idc) is send as input to the PI regulator. It gives out alpha order as output to the converter firing control. Traditionally, rectifier will operate as current controller in order to have optimal operation point with reduced consumption of reactive power. Direct current is indirectly regulated by controlling the firing angle of the thyristor. The firing angle at rectifier station kept within a steady state range of ±2.5º by tap changer control. The ac voltage can be maintained constant by switching in and off of shunt capacitor and filter banks at nominal frequency. During steady state, current order I0LIM from voltage dependent current order limiter (VDCOL) and measured dc current are same; hence the error of zero would be send to the saturated PI regulator in current control amplifier. In contrary, during transients IoLIM would be varied as a function of dc voltage when the dc voltage hits the breakpoint of VDCOL, the error of IoLIM and measured current will be sent to PI regulator with maximum limit (AMAX =164) and minimum limit (AMIN = 5). The max and min limit of CCA is determined by voltage control 27 amplifier. The output from the PI regulator would be change in alpha order which is directly proportional to the change in current. There is one more important function called alpha retard, it will be activated only when the dc line fault occurs. The current controller gives output of AMAX when dc fault happens. Figure 24 current control amplifier In order to eliminate the non-linearization of the current controller, since the dc voltage dependent on change in firing angle i.e. cos causes decrease in loop gain as increase in alpha; thus leads to instability of current controller. Therefore, linearization factor 1/sin is added to improve the stability. The current margin would be added based on controller (rect/inv). The maximum limits of CCA integrator are output from voltage controller(VCA order) and max as well as minimum limits are RAML order, Umin, overvoltage order, VCA order and mininv, these limits are selected depends upon the contingency of the system. If it is inverter, lower limit of alpha should be defined as 110 ( mininv) and it could be reduced even 10 degrees further by commutation failure prevention function. CCA set value would be activated at higher remaining voltage inverter ac fault by the function called Gamma0 explained below. CCA set value is nothing but max or VCA order. Once it is activated by gamma0 function, both limits max and min limit of integrator would be CCA set value. During inverter side 3 phase to ground ac fault, the inverter voltage would collapse consequently abrupt rise of dc current. CCA maximum and minimum limits of PI regulator would be selected as output of voltage regulator and ( mininv – Cfprev_angle). When 3 phase ac fault occur at rectifier side, voltage would collapse consequently current decreased to zero. The maximum and minimum limits of CCA would be Alphamax and output of voltage regulator ( min or RAML order). In order to have a good dynamic operation of current controller, proper value of KP and Ki should be selected. The proportional part is a faster part which executes the instantaneous change in feedback and integral part which reduces the steady state error. The proportional 28 and integral gain shall be chosen to give fast response and well damped oscillation with minimum overshoot. 2.12 Voltage controller: The input for the voltage regulator (figure 25) is measured direct voltage and reference voltage error. The maximum and minimum limits of PI regulator are determined by the direct current (ID low). If the direct current is less than Idlowref value, maximum and minimum limits are manipulated by normal dc voltage equation =cos-1(Ud/Udio). On the other hand, Alphamax as maximum limit and lower limits are RAML alpha order, Umin, Alphamin inverter and overvoltage alpha order, these limits could be chosen depends upon the converter (rectifier/inverter) and contingency of the system. Figure 25 voltage control amplifier The output of voltage regulator (figure 26) is used as maximum and minimum limits of current control amplifier with respect to converter (rectifier/inverter). In normal voltage operation the reference voltage of the voltage regulator is set above the operating voltage in order to avoid the hunting between normal tap changer controller and voltage controller. The output from the voltage and angle reference calculation has multiplied by factor of one tap changer step for inverter voltage. This reference voltage is sent as input for voltage regulator. Even higher reference value shall be used in rectifier voltage controller in order to de-activate if it is active at inverter station. 29 Figure 26 combined CCA and VCA At reduced dc voltage operation, reference voltage could be set lower than operating voltage in order to reduce the dc voltage. It can be see that voltage regulator gives output of higher limit of integrator for inverter operation and lower limit for rectifier operation. When 80% remaining voltage fault at inverter occurs, commutation failure happens due to the increase of dc current cause’s higher overlapping angle. If it is high remaining voltage fault, commutation failure will be detected by the comparison of dc short circuit current and ac current, thus increase the by 15 degrees. Hence, it advances from alphamax position by 15 degrees. The dc current in the inverter would be reduced to zero by decreasing the voltage at the rectifier faster by current controller since voltage difference is reversed. The Alphaminref_inv will become 100 degrees from 110 degrees. It will be used as a minimum limit of integrator of voltage controller until the Idlow is activated and also Alphamax will be used as maximum limit of integrator until the Idlow is detected. The Idlow gives an output signal of 1 after 40ms delay from the instant it detects the Idmeas is less than 0.015p.u. Once the Idlow gives output of 1, maximum and minimum limits of the integrator will be based on the equation Ud= Udio*cos . Otherwise, maximum alpha will be increased further since there is no overlapping angle effects increase of dc voltage above rated at pole until the voltage error becomes zero. max =180- cos (1.17) Consequently, dc system recovery time will be increased. Gamma0 function will be activated after 50ms time delay from when the inverter fault is detected gives Udlowinv output equal to 1, consequently CCA set output will be 1. So the CCA set value output will be voltage regulator output or Alphamax with respect to the Udabovezero signal. Just before the CCA set is activated, maximum limit of current controller integrator is output of voltage regulator and minimum limit is (alphamin_inv-Cfprev_angle). Hence the output of inverter current controller gives advanced alpha order in order to prevent the consecutive commutation failure. But it is not good to have higher commutation margin during the recovery, since it consumes more reactive power as current increases causes ac voltage reduction leads to voltage collapse or could take longer time to recover from the fault if the ac system is weaker. Therefore, the CCA set value will be used as both max and min limits of current controller integrator until the voltage error becomes zero to reduce the recovery time, CCA set value is nothing but Alphamax or output of voltage regulator as explained above. During inverter side 3-phase ac fault, voltage at inverter would collapse hence sudden rise of dc current Idlow signal will select the maximum and minimum limits of Alphamax and 30 (Alphamin_inv-Cfprev). Cfprev is an output of commutation prevention failure function; this value could be maximum of 10 degrees. Same sequence happens as 80% rem. Voltage fault. 2.12.1 Control sequence in dynamics When ~80% remaining voltage 3 phase ac fault occurs at rectifier, dc pole voltage decreased to ~80% of rated causes decrease in dc current since potential difference becomes reversed. Current controller will advance the firing angle in rectifier in order to increase the voltage to set the current equal to reference value. Once the alpha at rectifier hits the minimum limit (Umin, minimum voltage required to turn on the valves), no further possible to increase the voltage in order to have constant potential difference between two converters. Therefore, converter will be recommended to operate in reduced voltage or reduced power. This forces the rectifier into constant voltage controller and inverter into constant current controller operation by increasing the gamma. As soon as fault cleared, ac voltage recovered to nominal value causes transient increase of current since the alpha is at minimum value. Hence the alpha increased transiently to a value higher than nominal to decrease the current then the alpha settles in nominal value. In this case RAML function is not activated because of higher impedance fault. Figure 27 Maximum and minimum limits of VCA When 3 phase ac ~10% remaining voltage fault occur at rectifier side, ac system voltage would collapse consequently current decreased to zero, alpha at rectifier hits the minimum limit. Collapsed commutation voltage would also leads to commutation failure at the rectifier. RAML function will be activated with a delay of 20ms from the instant measured dc voltage is less than RAML reference value. It increases the alpha to predefined value (ex. 60 for cables since higher charging current) and keeps it constant until the measured voltage is greater than reference value. When the fault clears, voltage would increase to nominal value results RAML alpha order ramp down to nominal value (ramp step depends upon the type of transmission, normally the value in the range of 0.2 deg/ms to 0.8deg/ms). It can be seen from the result, the alpha order will be increased transiently from 60 to 70 deg at fault clearing due to the abrupt increase of dc current. The current controller at inverter would try to advance the alphamax to set the reference value equal to measured value at starting. But the decreasing current due to reversed voltage difference retarding the alphamax thus the ac voltage at the inverter will increase. Once the Ud below reference value is detected by gamma0, activates the Udlow_inv after 50ms time delay. Thus the CCA set value gives alphamax output for both max and min limits of integrator in order to reduce the recovery time as same explained 31 in inverter ac fault. The ac voltage at inverter hits the maximum limit when the dc current goes to zero due to the switching capacitance. We can conclude that during the inverter fault, max output as same as VCA order until Idlow is detected. The VCA order will be sent as input for CCA as max limit and min limit would be obviously minINV. In the rectifier side, output of VCA order is minREC; it will be used as CCA min limit and max limit will be obviously max in order to increase alpha above 90 transiently to decrease the fault current. During the rectifier side ac fault, the output of VCA order is minimum limit of VCA i.e. RAML order or minREC (Umin). VCA order would be min limit and max as max limit of rectifier CCA. In the inverter side, max would be the output of VCA order, it probably a max limit of inverter CCA and minINV as min limit until CCA set is not activated. 2.13 Alfamax inverter control: This control is nothing but constant beta control in order to have positive slope in the current margin region to increase the stability of the current controller. If there is no positive slope, during the mode shift sudden change in current causes power oscillations in system. Therefore, constant gamma operation gives stability problem, if the inverter ac network is weak. Since constant extinction angle characteristics could be written as )refcos Udio Id*Xtarccos( (1.18) Increase in current will give negative slope; especially receiving end network is weaker leads to more number of crossing point’s results instability. The key is constant alpha or beta operation gives positive slope. The direct voltage can be calculated by using the equation Ud = 0.5 Udio (cos + cos ( )) (1.19) = . (1.20) Therefore, ) U U*2arccos(coscos dio d (1.21) ) Udio UdioN* IdN Io*)drdx((cosUdioUd (1.22) Insert equation (1.7) in (1.6), will get ) U U IdN Io*dx*2arccos(cos dio d (1.23) Where, dr is neglected since resistance will not play any role in commutation of valves. The stabilization could be introduced if a contribution derived from the difference between current order and measured current is added when calculating beta. The gain and time constant of this added regulator shall be varied to get better stability. ))IdIo(K U U IdN Io*dx*2arccos(cos dio d (1.24) Alphamax = 180- (1.25) 32 During the transients will be increased by commutation failure prevention function based on how severe the fault is. If it is high remaining voltage fault, commutation failure will be detected by the comparison of dc short circuit current and ac current, thus increase the by 15 degrees. Otherwise, if it is solid fault gamma will be increased based on the comparison of fault voltage and pre-fault voltage. It is explained briefly in commutation failure prevention function. 2.14 Rectifier alpha min limiter (RAML) When there is fault in ac system at rectifier station, ac voltage will drop down, thus the dc voltage at pole decreases. But if it is cable, voltage after the smoothing reactor will decrease slowly (large capacitance). This reduces the dc current to zero faster. Therefore the current controller will decrease the alpha to minimum value (5 degree). When the fault clears, voltage will bring back to normal value instantaneously; it leads to high dc current if the alpha is still in minimum value at fault clearing time. There is two different function Uac_min_hold and Uac _max _hold. This function detects the maximum and minimum ac voltage and compared with nominal reference value (0.6) and triggers the alpha to predefined minimum value (ex. 60 degrees). Uac_min_hold function is used to detect the unbalanced fault and Uac_max_hold function is used to detect the balanced fault. After the fault, alpha is decreased to nominal value by 0.2deg/ms. This function forces the inverter into current controller during rectifier shunt fault and recovery to avoid consecutive commutation failures. 2.15 Commutation failure prevention control: It is not possible at all to avoid commutation failure at inverter when sudden increase of dc current and decrease of commutating voltage. However, it could be possible to avoid subsequent commutation failures during the high remaining voltage fault at inverter ac network by advancing the firing angle in order to increase the commutation margin. This makes the system could transfer power during disturbance. There are two possibilities for commutation failure such as commutation voltage reduction or distortion and phase angle jump at unbalanced fault. Therefore, two different sets of control function used to avoid consecutive commutation failure caused by balanced and unbalanced fault. Since the reduction of voltage at single phase fault is not severe as three phase fault, but creates phase shift. So the unsymmetrical fault is detected by zero sequence voltage detection, then it compared with pre-defined voltage level, if zero sequence voltage is lower than predefined level, it advances the firing angle and keeps it for whole fault duration. Symmetrical fault will not give zero-sequence, therefore it directly compared with predefined value and voltage (prefault voltage – fault voltage) difference, it will decrease the firing angle if the difference is higher than predefined value. In brief, the three phase voltages are transformed to dc quantity by alpha-beta transformation; it gives dc output only when the three phases are symmetrical. However, three phase fault will not decrease the voltage instantaneously in all the phases, it decrease the voltage phase by phase hence creates negative sequence the same happens when clearing the fault. Therefore, oscillations will occur in alpha-beta output. It is not possible to compare the oscillating waveform with pre-fault voltage, so max-hold function is used here to keep the maximum value for half cycle. 33 Commutation failure is more often detected when there is higher dc current compared to ac current, since the short-circuited dc grid effects open circuited ac grid for 2/6th time period of one cycle. 2.16 Tap changer control: The purpose of Tap changer control is to maintain the delay angle alpha in a certain range at rectifier station. If the alpha reduces below the predefined value, it starts increase the voltage by varying the number of turns in the primary side to keep the alpha in certain range for fast controlling of current. Also, if alpha increased above a certain predefined maximum value then the tap changer would decrease the voltage to keep the alpha within the limit. In inverter station, tap changer control maintains the dc voltage equal to reference voltage. Normally gamma is kept as constant for proper commutation. Any increase or decrease in dc voltage will be controlled by tap changer. Since tap changers would take more time to control the converter voltage than current or voltage controller, dynamic interaction between both controllers can be avoided. 2.17 Inverter gamma0 start function: If the measured voltage is lower than the predefined minimum value of 0.8p.u. This function indicates an output of 1 after a certain delay. It will be reset to 0 after the voltage recovered to nominal value with certain time delay. Gamma0 function is used only under ac faults, if the measured voltage is less than reference value for a certain time period would give udlowInv output as 1. It activates the CCAset value. CCA set value could be an output of voltage regulator alpha order and Alphamax with respect to measured voltage whether it is above or below zero. Gamma0 function is mainly used to recover the dc system as soon as possible with reduced consumption of reactive power. As soon as inverter side ac fault occurs, Udlowinv value will be activated after a 50ms time delay with delay off time of 100ms. It will be reset to zero after 150ms time delay from the instant voltage recovered. If gamma0 function is not activated, the lower limit of PI regulator of current controller would be alphamininv; hence it consumes more reactive power causes ac voltage reduction leads to collapse especially if the ac system is weaker. To avoid this, during the recovery it recommended to operate the inverter in Alphamax position. 34 2.18 Equidistant firing control: Figure 28 Equidistant firing control Equidistant firing control (figure 28) method of firing the valves will determine the equal distance between two firing. The converter firing control (CFC) is normally open loop control. It provides firing of each valves of 12 Pulse Bridge with delay of 30 degrees between each firing. It is always synchronized with ac voltage before firing using phase locked oscillator. Equidistant firing control automatically adds the change in angle to 30 degrees (30+ ), 60 degrees for 6 pulse bridge circuit. Typically, the change in firing angle is determined by current control amplifier. 2.17 Recovery of AC and DC system faults: 2.17.1 AC system faults: The ac system fault could occur both rectifier and inverter station. The fault at the rectifier station would suddenly collapse the voltage at bus bar, since the thyristor is unidirectional device the dc current goes to zero immediately. Therefore, no power would be transferred to sending end unless the fault is cleared. Also fault at the inverter station cause counter voltage collapse at inverter ac bus, resulting sudden rise of dc current causes increased overlapping angle reduce the commutation margin. Hence commutation failure occurs .Commutation failure is nothing but short-circuited dc grid and open circuited ac grid. Therefore, no power transfers by the valves into ac network. These energy losses due to these faults should be recovered as fast as possible to prevent transient instability of synchronous machines and energy loss to customers. The recovery of the dc system is mainly depends upon how strong the ac network is. If the SCR of the receiving end ac system is very low, more difficult in system recovery without commutation failure. 2.17.2 DC system faults: Dc line faults are most probably dc line to ground faults, cable fault, station equipment faults and insulation breakdowns. For some special reasons the dc breakers are not used as main 35 equipment to clear the fault, even though the dc circuit breaker has resonance circuit to achieve zero crossing for arc absence opening of switch. The HVDC control system plays main role in clearing of dc fault. The function called retard, as soon as the dc fault is sensed the retard function will be activated. The retard is nothing but operating the rectifier with high alpha i.e. inverter to send energy back to the ac system to make zero current at dc system. If there is a fault in pole one dc line of bipolar system, it will be operated as a monopolar system with reduced power. During the de-energization the of rectifier station, it is acting as pure inductance. If the rectifier station ac system is weak, it could be possible for commutation failure at rectifier. In some cases, the recovery of pre-fault voltage may not be recovered successfully due to some reason like insulator failure. The dc faults are detected by derivatives of the dc voltage and current if it is still negative after fault clearing; it would be good to restart with reduced voltage. 2.17.3 Commutation failure: Commutation failure is inevitable at inverter during disturbance at ac system. Mostly the inverter would always operate at constant beta; let’s say gamma reference value at normal operation is around 18 degrees in order to have certain negative valve voltage to deionize the outgoing valve. The minimum extinction angle need to turn off the valves is ~8 degrees. This could not be higher than 18 degrees for reducing the consumption of reactive power and valve voltage stress. Most probably the commutation failures will take place at the inverter side during ac faults since low commutation margin at inverter compared to rectifier. The commutation failure could be expected at inverter side at 90% to 85% remaining voltage fault at this margin. If the margin is increased bit, it could withstand up to even some lower remaining voltage fault. The commutation voltage of the converter valves nothing but phase to phase ac voltage at converter transformer. Commutation failure would take place not only due to voltage reduction and also depends on phase shift of voltage. The leading (advanced zero crossing) voltage phase shift will reduce the commutation margin leads to failure. The phase shift will occur not only for unbalanced fault, it also depend on the ac system active power injection reduction during fault. In general, weaker receiving ac system would cause more commutation failures when remote ac system fault happens due to higher reduction of commutation voltage. The recovery time after the fault clearing will be longer for low SCR than high SCR system. Since fast recovery of dc system with weak ac system causes voltage instability resulting consecutive commutation failures during recovery. Hence, very weak systems may be provided switching capacitors or SVC to avoid voltage instability. Most probably, commutation failure happens at the star winding transformer connected valves of 6 Pulse Bridge at high remaining voltage ac fault. However, the valves in the lower bridge commutating normally causes short circuit of dc grid for 2/6th time period since the valve fails to commutate once, it will get next chance to commutate after a cycle or 360 el degrees. Most often the commutation failure happens in star winding connected bridge but not in delta connected bridge. It mainly depends upon the type of transmission, ac system strength and how fast the commutation failure prediction is. The well defined control strategies shall be included to have good recovery of dc system without following fault consecutive commutation failures. The main control used for proper recovery of dc system after fault clearing is voltage dependent current order limit (VDCOL). The purpose of this control is to limit the current order as a function of the reduction of direct 36 voltage. This reduces reactive power consumption during recovery hence reduce the risk of commutation failures. 2.18 Conclusion: This chapter gives some introduction of traditional control system of HVDC classic. The consequences of ac and dc faults of HVDC system connected to weak ac network. Moreover, the basics of commutation failure and prediction of commutation failure was covered. In the next chapter, ac/dc system interaction of HVDC link connected to weak network will be analyzed with existing control parameters. 37 38 3.0 Performances scan of HVDC classic models: 3.1HVDC Modelling in PSS/E: 3.1.1Introduction to PSS/E: Power System Simulator for Engineers (PSS/E), which is kind of software using for studies on power system transmission network and generation in steady state and dynamic conditions. 3.2 Power Flow data: In this paragraph, data required for power flow in steady state analysis is described. Figure 29 HVDC model diagram in PSS/E 3.2.1Bus data: Each bus data should be recorded in PSS/E. bus data includes not only basic bus properties but also shunt capacitance or reactor connected to the bus. Bus data format: I, ‘NAME’ , BaseKV, IDE,GL,BL,AREA,Zone,VM,VA,owner I , Bus number NAME Name of the bus BaseKV Base voltage IDE Bus type, whether load bus, generator bus or swing bus. GL Active component of shunt admittance to ground BL Reactive component of shunt admittance to ground Area Area of the bus Zone Zone number of the bus VM Bus voltage magnitude in p.u. VA Bus voltage phase angle Owner Owner number 3.2.2 Load data: Each network bus at which load should be represented. Load data format: I, ID, STATUS, AREA, ZONE, PL, QL, IP, IQ, YP, YQ, Owner I Bus number ID Number of loads connected in bus STATUS whether it is in service or out of service AREA Area number 39 ZONE Zone number PL Active power component of constant MVA load QL Reactive power component of constant MVA load IP Active power component of constant current load IQ Reactive power component of constant current load YP Active power component of constant admittance load YQ Reactive power component of constant admittance load Onwer Owner number 3.2.3 Generator data: The generator data should be recorded in PSS/E Generator data format: I, ID, PG, QG, QT, QB, VS, IREG, MBASE, ZR, ZX, RT, XT I Bus number ID Number of generators connected PG Active power generation QG Reactive power generation QT,QB Maximum and minimum generator reactive power output VS Regulated voltage in p.u. IREG Extended bus number MBASE Total MVA Base ZR,ZX Complex machine impedance in p.u. RT,XT Transformer impedance in p.u. 3.2.4 Branch data: Branch data format: I, J, CKT, R, X, B I Branch from bus number J Branch to bus number CKT number of branches R branch resistance in p.u. X branch reactance in p.u. B Total branch charging suceptance in p.u. 3.2.5 Fixed shunt data: Fixed shunt data format: I, Bus name, ID, Status, G-shunt, B-shunt I Bus number ID number of shunt element Status In-service or out of service G-shunt Active component in MW B-shunt Reactive component in MVAr 3.2.6 Two-terminal DC line data: Record 1: I, MDC, RDC, SETVL, VSCHD, VCMOD, RCOMP, METER, I DC line number MDC Control mode, whether blocked(0), power(1) or current(2) 40 RDC DC line resistance in ohms SETVL Current or power demand VSCHD Scheduled compounded DC voltage VCMOD Mode switches dc voltage, if the inverter voltage falls below certain limit, mode switch to current control from power control mode. RCOMP Compounding resistance in ohms, 0% resistance holds the dc voltage in inverter, 50% resistance holds the constant dc voltage in centre of dc line, 100% to control the rectifier end dc voltage. METER Rectifier or inverter Record 2: IPR, NBR, ALFMX, ALFMN, RCR,XCR, EBASR, TRR,TAPR, TMXR, TMNR, IPR rectifier converter bus number NBR Number of bridges in series ALFMX,ALFMN Alpha Max limit and Alpha Min limit RCR Rectifier commutating transformer resistance per bridge XCR Rectifier commutating transformer reactance per bridge EBASR Rectifier primary base ac voltage TRR Rectifier transformer ratio TAPR Rectifier tap setting TMXR Maximum rectifier taps setting TMNR Minimum rectifier taps setting Same format for inverter except alpha instead gamma. The power flow data are loaded in PSS/E in a raw file format. Raw file is nothing but power flow input data file see appendix (8.4). Solve the raw file using Newton rapson method. The HVDC model representation in PSSE is shown in figure (29). The generator at both station are represented as infinite source behind the short-circuit impedance with PSSE dynamic generator model GENCLS. The short circuit ratio at both converter terminals are 3, representing as weak ac network. The fixed shunt compensators are connected at converter bus which produce reactive power = 0.5 active power. The load flow data are called from raw file, such as bus data, branch data, generator data, fixed shunt data, and two-terminal dc line data. The dynamic data are called from dyr file shown in appendix (8.5), which has dc line data such as rated dc voltage, current, inductance of the line/cable, capacitance of the line/cable and smoothing reactor value. The main circuit, dc system and control systems are called by(software independent) dll file, which contains program written in fortran language. Ac/dc system interaction studies are mainly about the recovery of the system after the disturbance, voltage stability and over voltages, especially HVDC system connected to weak network. However, the rms models are not used for dynamic performance studies, though selection of control parameters should be verified by dynamic performance studies. The good dynamic performance of HVDC system could be achieved by tuning of CCA and VDCOL control parameters. HVDC system stability is measured by how fast the dc system recovered from the ac or dc faults. Recovery time is defined as the time from fault clearing to the instant 90% of the pre fault power is restored. In general, recovery time depends on the characteristics of the dc and ac system 41 DC control strategy The characteristics of ac/dc system are nothing but strength of the network, dc line parameters i.e. line inductance, capacitance (especially if its cable), smoothing reactor and transformer reactance. The rate of power recovery affects system stability, since after a long period of ac system disturbance, fast recovery of power may cause voltage instability if the inverter ac system is weaker; effects subsequent commutation failures. In particularly, low SCR system requires slow recovery of power to ensure the ac system voltage maintained at satisfactory level. Short circuit ratio (SCR): The ratio between three phase short circuit MVA to dc power. Stronger the system lower the ac system impedance and weaker the system higher the impedance. Effective short-circuit ratio (ESCR): The shunt capacitors including ac filters connected at ac terminal of dc link can increase the effective ac system impedance considerably. Effective short circuit ratio is defined as PdN QcSCRESCR (2.1) Where, Qc is the three phase fundamental MVar PdN is nominal rated dc power. Voltage stability: Voltage stability of the power system is, if any fault occurs; it is required to maintain the voltage within acceptable limits in all the connected buses except faulted bus. Otherwise, significant effects on the local ac network leads to voltage instability for some reasons like higher impedance line, or increase in load demand. In particular, the higher line impedance builds higher inductive voltage drop across the line, consequently voltage instability would occur if there is not enough capacitive reactive power to compensate. Thus, Voltage stability also mainly depends upon system strength. The voltage stability is measured by factor called voltage stability factor. The voltage factor is defined as incremental change in voltage due to small reactive power injection in to commutation voltage bus bar at rated dc power. dQ dUVSF (2.2) The control parameters of current control amplifier and VDCOL have to be optimized to have good response during transient operation. There are some specific time of dc system recovery based on the transmission types (line/cable). All benchmark models are designed with SCR of 3 for both sending and receiving station. The damping angle used for both station ac systems is 80 degrees. In order to optimize, following tests have to be investigated. 3.3 Benchmark cases: Overhead Line transmission Cable Transmission Back to Back configuration DC Power DC DC Power DC DC Power DC 42 (MW) Voltage (KV) (MW) Voltage (KV) (MW) Voltage (KV) 3600 800 3000 800 2500 800 2000 660 1575 660 1500 500 Monopolar, Ground return 1000 500 800 500 500 500 3600 800 3000 800 2500 800 2000 660 1575 660 1500 500 Monopolar, Metallic return 1000 500 800 500 500 500 7200 800 6000 800 5000 800 4000 660 3150 660 3000 500 Bipolar 2000 500 1600 1000 500 500 1000 900 Monopole,Midpoint grounded - - 600 900 7200 800 6000 800 5000 800 4000 660 3150 660 3000 500 2XMonopole, ground return 2000 500 1600 1000 500 500 600 500 150 167 139 42 Tabel 1 Benchmark models in PSS/E 43 3.4 Test cases 1) 100ms AC fault at the inverter side with 10% remaining voltage 2) 100ms AC fault at the rectifier side with 10% remaining voltage 3) 100ms AC fault at the rectifier side with 70% remaining voltage 4) 100ms AC fault at the inverter side with 70% remaining voltage 5) 0.92 p.u. DC power step for 100ms 6) DC fault for 100ms at node1 nearer to rectifier. 7) DC fault for 100ms at node1 and restart with reduced (80%) voltage. 8) Inverter side AC fault with range of %remaining voltage (95% to 10%) 9) 3 degree gamma step at the inverter. 3.4.1 3-phase to ground AC fault: 10% remaining voltage ac fault is used to investigate the transient stability of the ac system and how fast the dc system recovers the power to consumer after the fault clearance. The recovery should be smoother without any subsequent commutation failure. Also, one more important thing to note that fault current should be limited within permissible limits and within certain time. ~70% remaining voltage fault is to investigate the performance under reduced voltage operation i.e. current controller at inverter and Umin controller at rectifier. Recovery time Fault type Transmission type Rectifier Inverter 1 3-phase to ground-10% remaining voltage Overhead line 100ms to 120ms 120ms to 150ms 2 3-phase to ground-10% remaining voltage Cable 100ms to 150ms 120ms to 250ms 3 3-phase to ground-10% remaining voltage Back to back 90ms to 120ms 120ms to 150ms 4 Dc line fault nearer to rectifier Overhead line 120ms to 150ms - 5 Dc line fault nearer to rectifier cable 120ms to 200ms - Tabel 2 Power recovery time of different transmission types 3.4.2 Commutation failure: Commutation failure could be expected at 90% to 80% remaining voltage ac fault at inverter for overhead lines and for cable, it happens at 95% to 90% remaining voltage ac fault. 3.4.3 DC line fault: The dc line fault applied for 100ms, the order retard will be activated after 5ms from the fault applied instant and shall be de-activated after 50ms from fault clearing instant. The fault is applied nearer to the rectifier station. 3.4.4 Dc line fault and recovered with reduced voltage: In some cases, faults are caused by insulator failures so recovered with full voltage in not possible. This is same as dc line fault, but recovered with 80% of nominal voltage. 44 3.4.5 Power step: The negative power step is applied for 500ms, power decreased to 0.92p.u. at 0.1s and back to nominal at 0.6s This is to investigate the performance of the active power controller. The expected response time is between 70ms to 100ms. 3.4.6 D.C. voltage step: The voltage step applied for 300ms, voltage decreased to 80% of nominal value at 0.1s and step back to nominal value at 0.4s. This is to investigate the performance of the voltage controller. The response time is typically in the range of 50ms to 100ms. 3.4.7 Gamma reference step: The gamma step is applied for 500ms, positive 3 degree gamma step applied at 0.1s and step back to nominal at 0.6s. This is to investigate the performance of the Alphamax controller. The normal response time is in the range of 50ms to 100ms. 3.5 AC fault at inverter: 3.5.1 The basic control operation (fault at inverter ac system) When 70% remaining voltage 3-phase ac fault to ground applied at the inverter end, suddenly dc voltage at the inverter end decreased but the dc voltage at the line would not be decreased as fast as pole voltage due to line capacitance. This potential difference builds a immediate rise in dc current. As a consequence, overlapping angle increases, alpha at the inverter could not change immediately to prevent the commutation failure hence shrink the margin of negative voltage for commutation of valves. At certain current, commutation failure occurs at inverter results abrupt increase of direct current due to short-circuited dc grid for 2/6th period of 6 pulse bridge. This surge current will be limited by smoothing reactor. If the dc voltage at inverter reaches the VDCOL break point voltage, current order will be reduced to predefined value after down time constant of low pass filter as well as current order will be recovered to nominal value with different time constant as increased dc voltage hits breakpoint after clearing of fault. After a travelling time, dc line voltage at the rectifier decreases, but the internal dc voltage is constant. This potential difference builds an abrupt rise of rectifier current. Consequence, current controller at rectifier starts increase of alpha in order to decrease the current by reducing the pole voltage at rectifier to maintain equal voltage difference between both converters. VDCOL will start its action when the dc voltage at rectifier hits the threshold value. Increase in current makes a converter to consume more reactive power as a consequence ac voltage at rectifier decreases. Then the internal dc voltage at the rectifier decreases, current decreases and consumption of reactive power decreases .Thus, the ac voltage at the bus increases. But if the system is very weak, the decrease in voltage and increase in dc current make commutation failure at the rectifier. But the commutation failure could be prevented at rectifier proper value selection of CCA gain and VDCOL time constants. If it is solid fault, collapse of commutating voltage leads to commutation failure. The transient current due to commutation failure will increase to 2 p.u. if it is line transmission; cable discharge current reaches maximum value of 8 p.u. at inverter. Commutation failure prevention function advances the firing angle alpha to its minimum limit of 110o in order to increase the extinction angle to prevent consecutive commutation failure during the recovery. The alpha at rectifier also transiently increases above 90 to decrease the increase of dc 45 current. Alpha order at rectifier increased to maximum value of around 140 to decrease the increase of dc current. Rectifier VDCOL helps to decrease the current order to predefined value when the threshold dc voltage value reaches. The inverter will control the current to current order minus current margin. 3.6 Ac fault at rectifier station: 3.6.1 Basic control operation (rectifier side ac fault) When there is voltage collapse at rectifier station ac system, suddenly voltage at the pole would decrease but not in after the smoothing reactor(inverter voltage), this voltage difference decrease the current rapidly to zero (since the current cannot reverse).As soon as the current start decreasing, current feed back sent to the current controller and it reduces the alpha to minimum value to increase the current to match the predefined reference .Then dc voltage reaches the break point ,VDCOL action forced to decrease the Iorder to predefined lower value. RAML function would be activated to increase the alpha to predefined maximum value when it hits the reference value; this prevents dc surge current at fault clearing. After the travelling time, voltage at the inverter node start decrease but not the pole counter voltage since the ac voltage is constant, results decrease in current .This decrease in current shall reduce the reactive power consumption hence voltage at the ac bus increase. If dip in ac voltage is 70%, then the dc current would decrease as described above, results current controller at rectifier hits the minimum alpha limit to increase the dc voltage hence dc current to provide constant voltage difference. Current controller at the rectifier is valid until the alpha >alpha min limit. Once the alpha hits the minimum limit, no further increase of voltage is possible. Current control switch to inverter since we have large range of voltage at inverter and rectifier operate at alpha min or constant voltage. 3.7 DC line fault at rectifier end: When dc line to ground fault occurs at the rectifier end, will collapse the dc voltage results abrupt increase of current effects higher consumption of reactive power. The short circuit shall be limited by smoothing reactor. As soon as the fault is detected, alpha retard will be activated to operate the rectifier at higher alpha of 1640 to reverse the pole voltage hence dc current. Inverter also advances the alpha to minimum alpha limit of inverter 1100. Technically, both converters operating as inverter to discharge the line or cable as fast as possible to extinguish the dc current. The retard will be released to normal value by ramp step after clearing the fault. The VDCOL increase the current to nominal current after the removal of fault. 3.8 Control sequence during step change 3.8.1 Gamma reference step: Gamma reference step is to investigate the performance of the Alphamax controller at inverter. The proper value of gain should be selected to have good dynamic behavior. ))II(K U U I I*d*2arccos(cos180 do dio dioN dN o xrefmax (2.3) When apply 3 deg positive gamma step from 17 deg, alphamax will be suddenly advanced by 3 deg consequently voltage decreases hence current increases. Transiently gamma would decrease back to 16 deg by the term )II(K do in the alpha-max controller. The intention with 46 this term is add a positive slope at the operating point in Ud-Id diagram. The applied 3 degrees in gamma step will provoke a temporary current rise which activates this term. After a travelling time, increased rectifier current would increase the alpha in order to reduce the dc voltage to decrease the current consequently gamma starts increasing to the reference value. After 100ms gamma settle down to reference value of 20 deg and alpha increased to 18 deg from 15 deg. And at 0.6s the gamma reference step back to nominal value of 17 deg results undershoot in the dc current because of the overshoot in dc voltage. It back to steady state with response time 100ms (see section.4.6.1.5). 3.8.2 Power step: To investigate the performance of the active power controller at rectifier, the power step of 0.92p.u. applied at 0.1 seconds, active power controller at rectifier sends the current order to the inverter to maintain the current margin and the rectifier waits for send back signal from inverter to change the current order (the transmission delay of 70ms back and forth). The activated current order at inverter start reduces current cause’s alphamax controller decreases the gamma in order to increase the current as well as current controller at rectifier starts decreasing alpha to increase current transiently. After 70ms of telecommunication delay i.e. send back signal from inverter, dc current reduced to 0.92 p.u. then rectifier alpha increases transiently to 21 degrees and settles down in 17deg. Gamma at inverter would decreased transiently and send back to nominal value by alphamax controller. Once the power step back to nominal value at 0.6s; transiently reduces the alpha to 11 deg from 17deg at 0.92p.u. and send back to nominal value after 100ms. Also gamma at inverter decreased to 13 deg transiently and sends back to nominal value. Especially for the long cable with weak receiving network should have high gamma margin at inverter in order to avoid commutation failure during power step. The power step response time is around 80 to 100ms. The response is well damped with minimum overshoot and controlled (see section 4.6.1.4). 3.8.3 Dc voltage step: In order to investigate the performance of the voltage control amplifier, the negative step in dc voltage has to apply, and also gamma reference shall be lowered to avoid hunting between Alphamax and VCA. The proper value of Kp and KI shall be selected to attain good dynamic performance. Under steady state, Udref in voltage controller is set higher than UdN in order to prevent hunting between alphamax and voltage controller. The output from the voltage and angle reference calculation has multiplied by factor of one tap changer step for inverter voltage. This reference voltage is sent as input for voltage regulator. In order to investigate the voltage controller response, required to set the multiplication factor equal to 1 instead 1.025 to have correct voltage response output. Otherwise, the voltage response output would be 0.025 times higher than the step applied. The response time of voltage controller is 50ms to 100ms. The response is well damped with minimum overshoot and controlled (see section 4.6.1.6). 47 3.9 Control Parameters (default): 3.9.1VDCOL (Voltage dependent current order limiter) 3.9.2 CCA (current control Amplifier) No. Control parameters Common values BTB(strong AC network) OH line (strong AC network) Cable (strong AC network) BTB (weak AC network) Long OH line Cable (weak AC network) 1 CcaLinAlpMax 164 - - - - - - 2 CcaLinAlpMin 5 - - - - - - 3 CcaLinMax 1.0 - - - - - - 4 CcaLinMin 0.3 - - - - - - 5 Sin_15_deg 0.25882 - - - - - - 6 CcaKp 2.8 2.5 2.8 1.3 2.5 2.8 2.5 7 CcaGain 23 30 20 50 23 30 21 8 CcaTi 0.01 - - - - - - Tabel 4 CCA parameters The control parameters shown above are used to simulate the HVDC classic benchmark models which are in appendix. The suggested transient simulations are done to analyse the dynamic performance of the HVDC classic control system. The shunt faults are applied at both rectifier end and inverter end with different voltage dips. The shunt fault impedance with respect to remaining voltage could be calculated by using the formula, ZsZf ZfUf (2.4) Where, Uf is remaining voltage is phase angle jump during fault No. Control Parameters Common values BTB (strong AC network) OH line (strong AC network) Cable (strong AC network) BTB (weak AC network) Long OH line Cable (weak AC network) 1 VdColFiltGain 1.0 2 UdTcUpRe 0.02 0.02 0.04 0.02 0.02 0.04 0.02 3 UdTcDnRe 0.015 0.02 0.015 0.07 0.02 0.015 0.07 4 UdTcUpInv 0.05 0.05 0.04 0.04 0.05 0.04 0.055 5 UdTcDnInv 0.015 0.005 0.015 0.001 0.005 0.015 0.001 6 VdColFiltMax 1.0 - - - - - - 7 VdColFiltMin 0 - - - - - - 8 IoAbsMax 2 - - - - - - 9 IoAbsMin 0.1 - - - - - - 10 IoLim 0.345 - - - - - - 11 UdLowParamRe 0.25 0.25 0.15 0.2 0.25 0.15 0.2 12 UdLowParamInv 0.25 0.25 0.15 0.3 0.25 0.15 0.3 13 UdHigh (rectifier) 0.75 0.8 0.8 0.5 0.8 0.8 0.5 14 UdHigh(Inverter) 0.75 0.8 0.8 0.9 0.8 0.8 0.9 15 ConstLim 0 - - - - - - Tabel 3 VDCOL parameters 48 Zf is fault impedance Zs is ac system impedance. The above mention test cases (see section 3.4) have been done. The overhead line transmission and back to back transmission cases justified almost the PSCAD results but the cable cases did not. The recovery time of the dc system for rectifier ac fault was bit faster(less than 100ms) compared to the PSCAD results for both overhead lines and cable cases, in reality it should be more than 100ms. And, when 3-phases to ground fault applied at inverter for cable cases, dc system does not recover after fault clearing and commutation failure happened at rectifier causes dc current oscillation in dc system shown in figure 31. Also commutation failure occurred for some cases when dc line fault applied near to the rectifier shown in figure 32. Hence, the circulating dc current will be available in dc system causes difficult to clear the fault. The dc line fault will be applied only near to rectifier not to the inverter since the short circuit of dc line close to inverter will decrease the dc current to zero due to reversed voltage difference. 0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 0 5000 10000 ID N C_ 1 0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 -1 0 1 2 x 10 4 Q A C_ 1 0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 0 1 2 U AC _1 File: fault-10rv102_Cable_Tr_Binopol_1600MW_500kV.out 0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 -10000 -5000 0 5000 P CO N V_ 1 0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 -1000 0 1000 2000 U D_ 1 0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 0 100 200 AL P O RD _1 A LP H A_ 1 Time [s] Figure 30 3-phase to ground,10% remaining voltage fault at inverter (rectifier outputs) 49 0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 0 5000 10000 ID N C _2 0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 -5000 0 5000 Q A C _2 0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 0 0.5 1 1.5 U A C _2 File: fault-10rv102_Cable_Tr_Binopol_1600MW_500kV.out 0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 -6000 -4000 -2000 0 P C O N V _2 0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 -1000 0 1000 2000 U D _2 0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 -100 0 100 200 G A M M A _1 G A M M A _2 Time [s] Figure 31 3-phase to ground,10% remaining voltage fault at inverter (rectifier outputs) 0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 0 1000 2000 3000 ID N C _1 0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 -500 0 500 Q A C _1 0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 0 0.5 1 1.5 U A C _1 File: dcfault-node1_Cable_Tr_Monopol_MetRet_500MW_500kV.out 0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 -500 0 500 1000 P C O N V _1 0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 -1000 -500 0 500 U D _1 U D IN TR N _1 0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 0 100 200 A LP H A _1 A LP O R D _1 A LP E X E _1 Time [s] Figure 32 DC line fault near to rectifier 50 3.10 Conclusion of performance scanning: The HVDC models were simulated in PSSE by using default existing control parameters; the obtained results are not justifying the PSCAD dynamic performance results. As an example of one case, it can be seen from the figure 30,31, that rectifier commutation failure creates dc current oscillation due to bad CCA parameters. As well as, dc system didn’t recover at all due to bad control performance of VDCOL. The recovery of dc system after fault clearing instant is very fast for benchmark models with existing control parameters at rectifier and inverter ac faults. It results in consecutive commutation failures during the recovery at inverter ac faults and low CCA gain and slow ramping down of current order during inverter ac fault might cause commutation failure at rectifier if the sending end ac network is weaker. The dc system recovery time of rectifier ac fault would be increase or decrease by tuning the RAML alpha order decreasing time constant. Commutation failure occurs at rectifier during dc line fault close to rectifier. 51 52 4.0 Control parameter optimization: The overall good HVDC classic dynamic performance could be achieved by tuning the Current control amplifier (CCA) and voltage dependent current order limiter (VDCOL). 4.1 Current controller amplifier: Current controller amplifier is traditional Proportional plus Integral controller. The proportional part gives fast response to the change in output and integral part is a slower part which reduces the steady state error to zero. Nyquist diagram for CCA stability has been done and verified in document (10). The product of proportional and integral gain should be in the range of 50 to 90 in order to have well damped and controlled response (3). Time constant of the integrator to be kept constant since tuning of the gains are enough to get good response. If the frequency of current oscillation is within the bandwidth of current controller, current oscillations could be suppressed. The transfer function of PI controller could be written as sT KiKp)s(G (3.1) Including linearization of the current controller, the proportional gain will be sin 15sin2Kp1KpKp and integral gain 2KpKi =CCAgain The dc voltage of the converter can be written as dc dcN N0di rNxN0didc I I U ddcosUU (3.2) At constant current, small change in gives sinUdioUdc (3.3) Therefore, the gain G = -Udio Sin (11) sT1 G)s(1G (3.4) For twelve pulse bridge f*12 1T (11) At 50Hz, T= 1.66ms Transfer function of dc circuit can be written as sC )sLR( 1 1sLR 1 Udc Idc)s(2G (3.5) 53 Figure 33 Dc system representation Figure 34 Block diagram of CCA Traditionally, rectifier controls the current at steady state operation. The proper selection of proportional and integral constant of CCA is required to achieve the current controller stability. Obviously, it depends upon the strength of the network. Normally, choosing control parameters for cable is quite difficult than overhead lines. There is chance of commutation failure at the rectifier when 3-phase to ground fault at inverter if the ac network is weaker at rectifier station. Therefore, proper selection of CCA gain would prevent the commutation failure at rectifier. Method of tuning CCA parameters: The integrator gain or CCA gain shall be increased higher than 30 until the current follows the current order faster i.e. without any steady state error for cable with weak rectifier and proportional gain shall be within the range of 50 and 90 in order to get smooth response, hence based on these Kp value could be chosen, typically it is around 2.5. Since increased gain would give fast response, results decrease of current due faster increase of alpha. In contrast, slower response of CCA, effects larger dip in ac voltag