Panel 17 8 Taking the Pulse

Slide1: 

“Taking the pulse” of Power Systems: Monitoring Oscillations by Wavelet Analysis and Wide Area Measurement System DEE - Politecnico di Bari - Italy S. Bruno, M. De Benedictis, M. La Scala

Slide2: 

Introduction Market & Security: more uncertainties in restructured power systems Deregulation dragged interests and investments on generation and distribution (“core of the business”) Transmission networks (capacity limits, coordination between neighboring SOs, interconnection systems, limited control resources: centralized and corrective control) Not-sufficient investments in Transmission Systems (authorization problems, environmental constraints) Blackouts re-introduced the issue whether competition and security are mutually exclusive New technologies can help:

Slide3: 

Introduction Methodologies for automation and control – Dynamic Security Assessment (DSA); Communication systems for real-time data exchange; Adaptive relays; FACTS and HVDC; Wide Area Measurement and Control Systems; Enhancement of system security There are several countermeasures that can be adopted on different levels (soft and hard investments): Technology and knowledge are at hand: need of decisions at political level

Slide4: 

WAMS WAMS: a grown-up technology All basic technologies for Wide Area Measurement Systems are quite established (PMUs, GPS, fiber optics, broadband communication systems) PMUs can be installed with minor additional costs when relays on transmission lines are substituted with newer digital relays (average estimation of 100 distance protections substituted per year on the Italian HV power grid) The system observability issue is overcome since the cost of a PMU is negligible compared to the cost of a substation Delays due to the communication system in the range of 100-700 ms with the present technology Fast data processing can be obtained with suitable routines and distributed computing environment

Slide5: 

WAMS Some possible applications of WAMS Improvement of dynamic power system modeling and implementation of on-line dynamic database Implementation of WACS architecture for enhancing dynamic performances and preventing cascade events Integration with dynamic security constrained optimization methodologies (DSA approach) Response-based approach to dynamic security Monitoring dynamic behavior and diagnosis of on-going degraded dynamic performances

Slide6: 

WAMS 30 PMU installed Monitoring through PMUs TSO control centers should be able to monitor the system dynamic behavior, recognize threats to the integrity of the grid …

Slide7: 

WAMS … evaluate and implement suitable control actions 30 PMU installed

Slide8: 

Monitoring the dynamic system behavior Monitoring through PMUs The availability of real-time measurements allows to keep track of oscillation modes during system operation every time a perturbation occurs On-set of poorly damped oscillations can be detected; Need of showing only significant results and having a suitable Man-Machine Interface By processing trajectories with Wavelets Transform, it is possible to obtain information on amplitude and damping of the modes, have a good MMI with a color scale presentation of results

Slide9: 

Monitoring the dynamic system behavior WT-based signal processing The approach allows to identify dominating modes and evaluate their damping (it can be proved that damping is preserved in the time-frequency space) This case is referring to small disturbance response, results have been compared with the eigenvalue analysis results

Slide10: 

Monitoring the dynamic system behavior In Swiss territory the first fault: the 380kV line Mettlen-Lavorgo is tripped (and the reclosure fails) The blackout on September 28th 2003 1° event at 03:01 a.m.

Slide11: 

Monitoring the dynamic system behavior Simulation of transients at 3:01 a.m. Simulations of the transient subsequent to the disconnection of the Mettlen-Lavorgo transmission line. Monitoring of voltage magnitude at the Brindisi bus (located in the Apulia Region, almost 1000 km far from Switzerland). Simulation interval [0, 7s] and disconnection of the Swiss EHV transmission line at t0 = 1s. The simulated voltage behavior is supposed to be acquired by a PMU. A random uniform noise has been added to the simulated signal in order to take into account the effects of measurement system noise.

Slide12: 

Monitoring the dynamic system behavior Wavelet spectral analysis Identification of the main modes: around 1.0 Hz and 0.5 Hz The modes are damped. The 0.5 Hz mode shows a worrying tendency to be poorly damped. The detection of such condition could have alerted the system operators in order to investigate on the nature of the disturbance. Morlet wavelet (ω0=8)

Slide13: 

Monitoring the dynamic system behavior The blackout on September 28th 2003 2° event at 03:25 a.m. Still in Switzerland, a second fault causes the tripping of the 380 kV line Sils-Soazza

Slide14: 

Monitoring the dynamic system behavior Wavelet spectral analysis Identification of the same oscillatory modes. The 0.5 Hz mode appears to be less damped than the other mode. Morlet wavelet (ω0=8)

Slide15: 

The full-scale experiment in the Russian Far-East Some objectives of the experiment Installation and testing of PMU and DFR Preliminary results on the feasibility of implementation of WAMS Keep track of the modes of oscillation by means of PMU and WT-based spectral analysis in presence of noise Developing international cooperation in the area of large power system studies with the participation of experts from different countries

Slide16: 

A full-scale experiment in the Russian Far-East Description of the experiment The experiment took place in the Russian Far East Interconnected Power System and consisted in the separation of the 500 kV transmission grid

Slide17: 

Results of the full-scale experiment WT of voltage trajectories in Test #1 The WT-based spectral analysis of voltage trajectories is able to filter out the noise (introduced by the network and by the measurement system) The dynamic behavior exhibited two damped modes of oscillation (the one with higher frequency is related to local machine oscillations and the other one to the regulating system) Voltage trajectory and related wavelet chart for the bus in Primoskaya (S-E area)

Slide18: 

Conclusions Lesson learned (up to now…) WAMS are complex projects and need on-the-field and full-scale experiments; “Taking the pulse” of power systems monitoring oscillations by WAMS and spectral analysis is helpful in diagnosis and ex-post analyses; Spectral analysis by wavelets is immune to noise originated by actual power system operation; Suitable MMI can be developed in order to fully exploit the proposed approach and to provide an additional tool for monitoring the dynamic security of power systems.

Slide20: 

WAMS-based pilot projects Italian WAMS project 30 PMUs are installed on the Italian grid One sample for cycle (20 ms) and a maximum delay of 10 ms due to the PMU computing time Dynamics database requires maximum 20ms for each cycle Possible applications of the architecture: state estimation evaluation of security margins real-time TTC monitoring of inter-area oscillation dynamic parameter identification model validation ex-post analysis of contingencies validation of protection devices tuning

Slide21: 

Conclusions Some numbers The detection of the event needed 8ms, the transfer 25 ms; Time delay to evaluate the control action by the SPS was 7-9 ms; every retransmission of the wide area detection system added 10 ms to the time delay of the data transfer; the distribution of the control action needs 25 ms and retranmsission 10 ms. Extension of the system 2000 km. Power flow crossing the separation section was limited to 34-36 MW. SEL 421 Relay SEL. Inc. , RES521 1.0 ABB, BEN 6000 LEM. IEEE C37.118-2005 (SEL, ABB).

Slide22: 

Results of the full-scale experiment Secondary regulation Operating conditions in the two tests, were chosen in such way that, as a consequence of a separation of the system, the two parts experienced a power unbalance of 35 MW. The Test #1 (line tripping with power flowing towards the South-East area) encountered problems with secondary regulation in the North-West area, that kept on accelerating after the system separation This was due to an error in the secondary regulation scheme since the frequency feedback was taken in the S-E area even for the generators belonging the N-W area The system kept on accelerating since the regulating system was receiving a decelerating frequency signal Frequency regulation was operated manually for Test #2

Slide23: 

Results of the full-scale experiment WT of voltage trajectories in Test #1 The WT-based spectral analysis of the voltage trajectory clearly shows that modes of oscillation with higher frequency are all damped out Other modes, with lower frequencies, probably related to the secondary regulation that was malfunctioning, are not damped out revealing that something was going wrong Voltage trajectory and related wavelet chart for the bus in Zeya (N-W area)

Slide24: 

Results of the full-scale experiment SPS time response The very first part of the experiment regarded the testing of the communication system Adopted communication systems were both optic fibers and wires, depending on the transmission line voltage levels (220kV / 500kV) The overall time delay was in the range 70 -100 ms (25 ms for each one-way data transmission) These timings are compatible with the time requirements of the proposed extended-real time approach or with on-line DSA This result is also consistent with the estimation of around 100ms for transferring data measured by the PMUs to the control center in the Terna-WAMS project

Slide25: 

Introduction Some consequences of restructuring Transmission grids and interconnections are means for economical exchanges and surplus maximization Request of more transfer capacity (also due to the load demand increase) Investments mostly focused on the generation side Transmission systems are exploited at their limits Few new transmission lines, TC is increased by relaxing security constraints Lacks of investments on transmission Scarcity of investments in the physical infrastructure, new technology, reinforcement of defense plans Security must be enhanced!

Slide26: 

Introduction Weak points of power systems Security must be enhanced by reinforcing some weak points that had been indicated as main general causes of past blackouts: The infrastructure suffers of aging and lacks of investments Reliable real-time data were not available on the fast-dynamic time scale and operators had not enough time to take suitable remedial actions Automated and coordinated controls were scarce or failed the chance to take immediate remedial actions

Slide27: 

Integrating the optimization methodology with WAMS Response-based approach WACS might allow to achieve the passage from a control strategy based on the forecasting of possible system states (event-based) to a strategy where the system might react to the actual system dynamic behavior (response-based)

Slide28: 

Integrating the optimization methodology with WAMS Response-based simulation Simulation of such approach were performed with our dynamic optimization tool Control actions were evaluated by processing system trajectories in a specific time window and implemented with a delay that takes into account the time necessary for the data communication (from/to) and computing In our simulations, this approach proved to be effective only with a 0.1s maximum delay