Back to the programme printer.gif Print




Delegates are invited to meet and discuss with the poster presenters in this topic directly after the session 'Advanced control concepts' taking place on Wednesday, 12 March 2014 at 11:15-12:45. The meet-the-authors will take place in the poster area.

Vinko Lesic University of Zagreb Faculty of Electrical Engineering and Computing, Croatia
Co-authors:
Vinko Lesic (1) F P Mario Vasak (1) Goran Stojcic (2) Thomas Wolbank (2)
(1) University of Zagreb Faculty of Electrical Engineering and Computing, Zagreb, Croatia (2) Vienna University of Technology, Faculty of Electrical Engineering and Information Technology, Vienna, Austria

Printer friendly version: printer.gif Print

Presenter's biography

Biographies are supplied directly by presenters at EWEA 2014 and are published here unedited

Vinko Lesic received his master degree in electrical engineering and information technology from University of Zagreb Faculty of Electrical Engineering and Computing (UNIZG-FER) in 2010. Currently, he is with the Department of Control and Computer Engineering at UNIZG-FER where he is pursuing his PhD degree. His research interests are in the area of control theory with application to advanced control of electrical drives and wind turbine generators.

Abstract

Fault-tolerant control of permanent magnet synchronous generator in wind turbines

Introduction

Increase by 18% in 2010 of direct-drive wind turbine (WT) concepts and announcements of new 7-8 MW WTs [1] makes it evident that permanent magnet synchronous generator (PMSG) is the future trend in wind energy. However, some of the previous gearbox problems will be transferred to the generator, which already has a fault frequency about 10 times greater than equivalent-power industrial machines [2]. Raising concerns for generator availability are to be expected and a lot of effort is currently put into development of new diagnostic methods. Motivated by this we propose a fault-tolerant control (FTC) for generator electromechanical faults.

Approach

With the goal of increasing the total efficiency of wind energy extraction and market competence, FTC for increase of WT availability emerged as a promising area of research. Different strategies are proposed, mostly based on redundancies of sensors and electronic components, and control algorithm is used to evaluate the trustworthiness of each [3]. Focus here is on generator electromechanical faults, which are besides gearbox and power converters faults the most common in wind turbine systems [2].
About 30% to 40% of electric machine faults are related to stator insulation [2,4,5,6,7]. Some of the most common causes are moisture in the insulation, winding overheating or vibrations. Modern voltage-source inverters also introduce additional voltage stress on the inter-turn insulation caused by the steep-fronted voltage surge.
Built on available diagnostic methods, we isolate the fault and reallocate the generator stress such that the fault propagation is stopped and the generator is operating safely in the fault presence. Specifically, we consider stator winding faults such as insulation degradation that will evidently cause inter-turn (within the same phase), phase-to-phase (between two different phases) or phase-to-ground short circuits. Researches in public domain [4,5,6,7] show that emergence of this kind of faults occurs gradually and can be detected on-line before they are fully developed to catastrophic proportions. This gives an opportunity for an autonomous reaction of the control algorithm to protect the generator and enable safe operation. The FTC is used to modulate generator electromechanical variables in a way that main cause of rapid fault spreading is removed but also that maximum possible power production in the faulty conditions is maintained.
In our recent work, we proposed FTC of squirrel-cage induction generator for rotor-cage defect [8] and stator inter-turn short circuit [9]. Here we make a contribution to the PMSG with specific distinctions of synchronous operation and permanent magnetic flux but also extend the FTC possibilities for early insulation degradation stage of fault development. The method can be applied to any variable-speed variable-pitch wind turbine.

Main body of abstract

The new and intact insulation in healthy generator conditions is negatively affected by the voltage derivative that arises from the pulse-width modulation. However, once degraded, the insulation has pronounced resistive character and is rapidly damaged further with the stator voltage amplitude [5], i.e., the voltage amplitude becomes dominant to voltage derivative contribution in insulation degradation and is proportional to inter-turn currents through degraded insulation that result in local over-heating. Therefore, in order to stop the fault development, induced voltage in the generator stator windings needs to be restricted and kept under some safe value K obtained from the machine diagnostics. To this aim, the K restriction is to be imposed on stator flux time-derivative, which is the main contribution to the induced stator voltage.
The magnetic flux is generated with permanent magnets on the rotor and can be considered constant. However, the magnetic flux perceived by stator windings can be reduced with adequate stator currents such that permanent magnets flux is dissipated in the air-gap instead of closing through stator coils. This feature is commonly used in the above-rated-speed operation, known as the flux-weakening method, which is controlled by proper negative value of d-current component in the field-oriented control (FOC). We utilize this possibility to form a FTC restrict of the flux derivative.
Approach with globally weakened flux in the below-rated-speed operation can be used to avoid the fault propagation, but it reduces the power production unnecessarily (Fig. ). Theoretical maximum of power production in faulty operation and boundary condition for fault suppressing is the case when flux derivative reaches the exact value of K. Arising from this condition the flux is modulated to obtain a triangular waveform with slope value of K such that its derivative is equal to fault restriction coefficient (Fig. ).
The FTC is used to calculate the stator currents that modulate the stator flux amplitude, based on the instantaneous flux position, as shown in Fig. . Note that targeted flux amplitude is in fact a time-sequence of Sinc functions, which ultimately results in a triangular waveform when multiplied with sine magnetic flux distribution generated by permanent magnets rotation. With modulated stator flux amplitude, the stator flux rate of change and consequently the induced voltage in stator windings are restricted and the fault development is stopped or greatly delayed. The method works both for the insulation degradation and inter-turn short circuit cases and can be applied to any stage of the fault development.
Since FTC deals only with d-current component of FOC, mechanical variables (torque and speed) are kept on the same values as in the healthy case, making the generator behave identically from the outer WT control loops perspective, only with reduced operating range. For salient magnets machine, the q-current can be slightly modulated as well to compensate the influence of d-current changes caused by the proposed FTC. Therefore, the generator electromechanical fault is only reflected in stator voltages and currents, and thereby as periodical changes in power production, which are further on rectified to DC-link, then inverted back and filtered by grid-side converter. The DC-link capacitor also acts as a power buffer that smoothes the power production oscillations.
Usually WTs operate in two regions: (i) low wind speed region where the generator torque is used for tracking the optimum power production characteristic and (ii) high wind speed region where blade pitching is used for keeping the WT at rated values of speed and torque. Possible operating area of a megawatt-class generator is shown in Fig. . The flux modulation is applied between ‘A’ and ‘B’ points. In the faulty state, optimum power production can be followed to the edge of possible generator operating area and blade pitching is used to keep the WT at new rated point denoted with 'B'. The flux can be reduced to about 50% of rated value for megawatt-class direct-drive generators and this represents the possible FTC application area.
Elaborated and detailed mathematical proof and controller design will be given in the full paper, together with brief description of conventional control with prominent FTC extension. Step-by-step FTC algorithm will be stated. Simulations will be performed in Matlab on a megawatt-class direct-drive WT model, both for steady-state operation and dynamical behaviour with realistic turbulent wind profile.

Conclusion

We show that gradually developing faults in stator insulation can be greatly delayed or completely stopped by proper modulation of generator electrical variables and safe operation can be achieved regardless of the fault presence. We further show how power production can be safely maximized in the faulty conditions.
Depending on the diagnostic method sensitivity, the FTC can achieve wide scope of generator operation, from inter-turn short circuit and from about 30% to almost intact power production, i.e., it can be applied permanently or only temporarily for avoiding the wind turbine shut-down and power production opportunity costs until the next scheduled repair.
The FTC is designed as an extension of conventional WT control. Its main task is the calculation of adequate stator d and q-current components that are further on passed to classical FOC as reference values. Algorithm inputs are stator-flux-derivative constraint imposed by the machine monitoring subsystem, speed and currents measurements, i.e., only already available sensors are required for the implementation. It is important to point out that FTC keeps the operation below rated values of generator variables – they are just shaped and reallocated properly to achieve fault-tolerant operation. This relaxes increased generator iron losses due to non-sine waveforms and increased harmonic composition caused by FTC.

Added market value of proposed control algorithms lays in the fact that they are conceived as cheap, efficient and modular software upgrades to the existing classical generator control algorithms and whole wind turbine control strategy, without any mechanical interventions. The nature of upgrades allows the FTC to be easily incorporated in new concepts, but also in already available and working WTs and for any type of generator. Moreover, the method can be applied for any inverter-fed AC electric machine, regardless of size and application.


Learning objectives
Generator insulation faults can be stopped from spreading with proper modulation of magnetic flux perceived by stator and safe operation in the presence of the fault can be achieved. Generator electrical stress can be reallocated from faulty to healthy part and power production in the faulty conditions can be significantly improved.


References
[1] Renewable Energy Policy Network for the 21st Century (REN21), “Renewables 2013,” Global status report, 2013.
[2] Z. Daneshi-Far, G. A. Capolino and H. Henao, “Review of Failures and Condition Monitoring in Wind Turbine Generators,” 19th International Conference on Electrical Machines ICEM, pp. 6, Sept. 2010.
[3] P. F. Odgaard, J. Stoustrup and M. Kinnaert: “Fault-Tolerant Control of Wind Turbines: A Benchmark Model,” IEEE Transactions on Control Systems Technology, vol. 21, no. 4, pp. 1168-1182, 2013.
[4] G. Stojicic, G. Joksimovic, M. Vasak, N. Peric and T. M. Wolbank, “Increasing Sensitivity of Stator Winding Short Circuit Fault Indicator in Inverter Fed Induction Machines,” 15th International Power Electronics and Motion Control Conference, EPE-PEMC ECCE Europe 2012, pp. DS2a.10-1–6, Sept. 2012.
[5] J. Yang, J. Cho, S. B. Lee, J. Yoo and H. D. Kim: “An Advanced Stator Winding Insulation Quality Assessment Technique for Inverter-Fed Machines,” IEEE Transactions on Industry Applications, vol. 44, no. 2, pp. 555-564, 2008.
[6] P. Nussbaumer, A. Mitteregger and T. M. Wolbank: “Online Detection of Insulation Degradation in Inverter Fed Drive Systems Based on High Frequency Current Sampling,“ 37th Annual Conference on IEEE Industrial Electronics Society, IECON 2011., pp. 1954-1959, 2011.
[7] T. Boileau, N. Leboeuf, B. Nahid-Mobarakeh and F. Meibody-Tabar: “Synchronous Demodulation of Control Voltages for Stator Interturn Fault Detection in PMSM,” IEEE Transactions on Power Electronics, vol. 28, no. 12, pp. 5647-5654, Dec. 2013.
[8] V. Lesic, M. Vasak, N. Peric, T. M. Wolbank and G. Joksimovic, “Fault-tolerant Control of a Wind Turbine with a Squirrel-cage Induction Generator and Rotor Bar Defects,” Automatika, vol. 54, no. 3, pp. 316-328, 2013.
[9] V. Lesic, M. Vasak, N. Peric, G. Joksimovic and T. M. Wolbank, “Fault-tolerant Control of a Wind Turbine with Generator Stator Inter-turn Faults,” Automatika, vol. 54, no. 1, pp. 89-102, 2013.