OMAR SH. ALYOZBAKY et al : THE BEHAVIOUR OF THREE PHASE THREE- LEG 11KV TRANSFORMER CORE.

Transcription

1 The Behaviour of Three Phase Three- Leg 11KV Transformer Core Type Design Under Sinusoidal and Non-Sinusoidal Operating Conditions for Different Core Materials Omar Sh. Alyozbaky 1,2 *, Mohd Zainal A. A.Ab Kadir 2, Mahdi Izadi 2, Chandima Gomes 2, Norhafiz B. Azis 2 1 Electrical Department, Collages of Engineering, University of Mosul, Iraq. 2 Centre for Electromagnetic and Lightning Protection Research (CELP) Faculty of Engineering, University Putra Malaysia UPM Serdang, Selangor, Malaysia. *Corresponding author: O.SH.Alyozbaky@gmail.com Abstract - Ansys Maxwell software, which is FEM-based software, was utilized to run the simulations for various operation conditions. The FEM technique enabled software to simulate and analyze various electromagnetic parameters including flux lines, flux density, losses, etc., under varying solutions very accurately. This paper investigates the behavior of three-phase three-leg in power transformer core type design under sinusoidal and non-sinusoidal operating conditions. Three types of cold-rolled grain-oriented core materials grades(m4,m5,m6)have been chosen to compare the core losses in steady-state and transient operation condition. The influence of magnetic hysteresis, flux couplings in the core-structure and material performance in three-phase transformers has been simulated. Simulation results of core loss behaviour based on the different materials are presented and discussed. Keywords - Power transformer, core losses, magnetic flux, core materials 1. INTRODUCTION A transformer is a static device with the capability of transferring electrical energy from a certain voltage and current level to another by electromagnetic induction and with no power change. No-load and load losses are two primary groups of transformer losses:.no-load loss involves the energy needed to maintain the continuity of different magnetic flux in the core, and independently of the loading of the transformer. On the other hand, Load loss originates from losses of resistance from windings, and depends on the power of the load current. Power transformer considering one of the most important component of power system and their efficiency is reach to 98% [1]. In EU-25, core loss accounts for almost 70% of total transformer losses while the operating (or energy) efficiency) is 93.38%. Thus, there is worldwide concern about the core losses which should be reduced [2]. Different methods have been proposed in the literature for the reduction of these losses [3-8]. The magnetic field is three dimensional (3-D) in nature and surrounds the transformer core, and solving 3D models poses many problems such as a protracted simulation time, and requires full computer equipment (hard disk, ram and cpu). Applying 2-D models significantly shortens the computation time and complexity of the problem [10]. Achieving precise 3-D modelling of the core with its laminations is not possible. In light of these problems, engineers and transformer designers have attempted to use 2- D simulations [7,8, 9,10] instead of the 3-D core model. However, most of the investigations have been performed for the three-phase transformer three-limb core. For instance, Valkovic [11, 12] has investigated the effect of core design, joint geometry, core material, and induction on core losses. E.G tenyenhuis et al. [13] applied 2-D FDM (finite difference method) to examine magnetic loss and flux distribution in cores with this structure. The core losses depend on the physical properties of the core material, the instant value of the magnetic field in the laminations, the construction parameters of the joint zone, and how often it is operated, all of which are shown in [14-17] The techniques of the core model are presented in [18-19]. The working hours for the distribution transformer are 24 h, which shows normal core loss, regardless of the load delivered. Consequently, the unbroken loss of energy enlarges the magnetic core beyond the energy loss from the transformer s conductive material [20]. When the performance characteristics of the steel improve, reduction of the core size can be achieved. By using higher quality material, the result will be enhanced operating flux density. As such, for a given flux in core, it is possible to minimize the core area with improved flux density. If flux density is maintained at a consistent level, the reduced core size will decrease the mean turn length of copper, which could provide compensation for the additional copper content due to the additional turns required. Consequently, the coil will require less copper and insulation. With a reduced core and coil, the tank size can be reduced, resulting in a reduction of the required insulating oil as show in Fig. 1. The aim of this paper is to highlight the issue of understanding the behaviour of core materials in different operation conditions. DOI /IJSSST.a ISSN: x online, print

2 In addition, to investigation which is the important core parameters that affected by sinusoidal and non-sinusoidal power supply. Fig. 1. Relation between a smaller core and the transformer price, when the performance characteristics of the steel improve. Fig KV Transformer model II. ANSYS MAXWELL ANSYS Maxwell is a high-performance interactive software package that utilizes finite element analysis (FEA) to provide solutions for electric, magneto-static, eddy current and transient problems. Maxwell s solutions for the electromagnetic field problems are achieved by solving Maxwell's equations in a finite region of space with suitable boundary conditions and, when required, with user-specified initial conditions so as to obtain a guaranteed and unique solution [21]. Maxwell equations used to solve the electromagnetic field problems are: Faraday's induction law: 1 Gauss' law of magnetism:. 0 2 Ampere's circuital law: μ 3 Gauss' law of electricity:. 4 A. Transformer Model III. CASE STUDY The main apparatus consists of a core model for threephase transformer which rating 1000KVA, 11 KV / KV,50Hz, assembled with three limbs as shown in Fig 2. The real dimension of core is shown in Fig.3. Core stack in T-joint section as butt lap 90ᵒ is assembled from grain oriented silicon steel (CRGO) with 0.3 mm thickness laminations. B. Materials Use Fig. 3. Dimension (mm) of the core Transformer The materials of HV, LV windings and core laminations were copper and cold rolled grain oriented silicon steel (CRGO) grades M4, M5 and M6, respectively. Oil was used to fill the space between windings and core. On the other hand, the magnetizing curves or B-H curves of magnetic core materials to software is presented. For examples Fig. 4 shows the B-H curve for grade M4. The calculation of core loss, coefficients related to hysteresis and eddy current losses for M4 laminations were calculated in relation to the B-P (flux density versus core loss) curve it was introduced. The B-H and P-B curves were obtained from the manufacturer s data. A typical B-P curve for M4 core is presented in Fig. 5 [22]. DOI /IJSSST.a ISSN: x online, print

3 important for many core transformer designs. Fig..6 shows a diagrammatic representation of these results. Fig. 4. The B-H curve [22] Fig.6. Diagram Results in Transient solution FEM has been proven to be a beneficial method of obtaining a precise characterization of the electromagnetic behaviour in magnetic components such as transformers. Fig. 7 is a flowchart that illustrates the steps of the transformer core design used in this study and provided in Transformer modelling. The core design simulation can be attained with Ansys Maxwell software together with the core lamination material. Fig.5. B-P curve [22] IV. METHODOLOGY FOR SIMULATION A. ANSYS Maxwell Solver Software In Ansys Maxwell solver software, there are three types of solutions: magneto static (Dc), eddy current (Ac), and Transient (a very comprehensive solution) that can be sinusoidal and non-sinusoidal and any kind of time variation can be represented. This study focus on the one type of simulation solution to calculate some parameters which affect the loss profile. The transient solver type of solution can be used in more comprehensive study in relation to actuator, transformer, inductance, motor, permanent magnetic and energy driven applications. In addition, it can be used in low and high frequency applications and to calculate the magnetic field in any each step time. Many sources such as voltage and current can be used in transient solution and applied to windings and permanent magnetic fields. The materials here can have linear or non-linear properties. Magnetic field (H) and current density (J) are in the time domain, and some of the results in transient solution can be obtained, such as Magnetic flux density (B), forces, Torques, Core loss and flux linkage, the results of which are very Fig.7. A flowchart of the transformer design in the simulation. The simulation s output data namely: flux density, flux flow, energy density and permeability are required for determining the losses of the transformer and the inductance design of the windings as well as the processes of the Ansys Maxwell simulation are shown in the flowchart presented in Fig. 8. DOI /IJSSST.a ISSN: x online, print

4 After running the whole simulation in a time duration of 200ms, the results comprising core loss, flux line, flux density and flux leakage were collected. The value of running simulation (200ms) has been chosen to check the all simulation it is become steady state response. The second set of results in this simulation study was obtained by using nonsinusoidal input voltage supply as shown below: cos 2 π 50 time 8 cos 2π 50 time 2/3 time 9 cos 2π 50 time 4/3 time 10 V. RESULTS AND DISCUSSION Fig. 8. A flowchart of the magnetic analysis. The development model of the objective design has four stages: the geometry design, field sources, boundary conditions, and mesh generation. In the simulation window, the geometric model is utilized for the collection of all the geometric shapes of given case study in Transformer modelling, linking the contained objects with material properties that were related, field sources, and boundary conditions and a 3D design of the transformer was developed as shown in Fig. 10. B. Input Voltage As we know the phase shift between three phase power supply is 120. For this simulation the power supply was obtained by using sinusoidal input voltage supply: sin 2 π f time (5) sin 2 π f time 120 (6) sin 2 π f time 120 (7) Where V m is input voltage, and f is frequency The input voltage in primary winding for three-phase transformer are show in Fig.9. Fig D design of transformer Every vertex signifies points and point coordinates which the user is able to clearly state or Ansys Maxwell can automatically have computed at the intersection point of two edges. Each vertex gives a definition of its mesh spacing value and its label, while the value of the mesh spacing provides a definition of the estimated distance that separates the mesh nodes in the vertex neighbourhood. As mentioned in the literature [7,8,9,10], the 2D design created from 3D design is presented in Fig. 11, to obtain the results as fast as possible. Fig. 9. input voltage. DOI /IJSSST.a ISSN: x online, print

5 On the other hand, Fig. 14 shows the flux lines between the core and the low voltage windings in the core transformer using M4 grade material. Fig D design of transformer There is an association between the property of the mesh spacing and vertices which is calculated in the current units of length. Fig.12, presents the mesh in the transformer model using a fully automatic method Fig. 14. flux lines between the core and the low voltage windings One of the most important factors which many researchers have been paying attention to is a way to reduce the core loss. The aim in this study is to compare the core loss between different core materials in a three-phase transformer and so notes the behaviour of core under different operation voltages. The flux density (B) distribution on the core flux alternates between the Yoke and limb. Fig. 15 presents the sinusoidal voltage supply M4 grade, and the flux density distributed on the core. Fig. 12. Mesh diagram analysis Magnetic fields play an essential role in transformers and the flux line distribution presented in Fig.13, shows the field lines are concentrated around the limbs of the core transformer and especially around the conductors, where increasing density of current and voltage is marked. Fig D design of transformer Fig. 13. Flux line distribution Most of the manufactories of power transformer are focusing generally to reduce the losses in transformer especially, for the loss which always be appeared when the transformer work. Therefore, core loss considers one of the important factor for transformer designers who want to reduce it. The value of core loss in sinusoidal operation voltage about 1220 watt for material grade M4 as shown in Fig. 16. DOI /IJSSST.a ISSN: x online, print

6 Fig. 16. Core loss in M4 grade (sinusoidal voltage) We can note the curve response of core loss it was become steady state after 75 ms., but we proposed the time is 200ms to check the core loss still steady state value in this type of operation condition. Figure 17 illustrated the core loss in material grade M5. The figure also explained the duration until the value is 1325 watt, because it is running in transient simulation type and the running time is 200ms. Fig. 18. core loss in M6(sinusoidal voltage) On the other hand, Fig.19 shows the core loss for different grade materials (M4, M5 and M6) in non-sinusoidal input voltage. Moreover, the percentage increase in core losses as a compare when operation in non-sinusoidal power supply between material grade M4 and grade M5 ranging 11.25% while 27.6 % as a compare between grade M4 with grade M6. Fig. 19. Core loss for different materials in non-sinusoidal input voltage Fig. 17. core loss in M5(sinusoidal voltage) The different results of core loss in different materials were obtained. The percentage increase in core losses as a compare between material grade M4 and grade M5 ranging 8.6% when operation in sinusoidal power supply. On the other hand, the core loss in material grade M6 was presented in Fig. 18. Gradually increasing from 0 until 1547 watt, and the percentage increase in core losses as a compare with material grade M4 it is about 26.8 % while % as a comparator with material grade M5. It is noted that there is an increase in the proportion of the core loss when change the material which used in core design. The reason for this is due to the nature of the behaviour of this material under different operating conditions and how to interact with them. The nature of the manufactured material and the circumstances surrounding them during the manufacturing stages play these circumstances is an important factor to improve the quality of this material. Therefore, the response it will be different from material to another material in spite of working in the same operation conditions After running the simulation in three types of core materials, under different operation conditions, the resulted showed the increase of core loss it is about 50% in material grade M4 as a compared in different operation conditions as shown in Table I. TABLE I: CORE LOSS FOR DIFFERENT MATERIALS IN SINUSOIDAL AND NON-SINUSOIDAL INPUT VOLTAGE Core materials Core loss in sinusoidal supply Core loss in Non- sinusoidal supply The percentage % of core loss increased Grade M watt 1840 watt % Grade M watt 2047 watt % Grade M watt 2348 watt % The value of the field density (B) is 2.9 Tesla in core transformer grade M4 under sinusoidal voltage, while in nonsinusoidal voltage is increased to 4.14 Tesla as illustrated in fig 20, which means an increase of about 42.75%. The scenarios of other materials were used in this study but with different percentages, for example, the maximum value of the magnetic field density (B) happens in grade M6, while DOI /IJSSST.a ISSN: x online, print

7 the minimum value is found in M4 grade in different operation voltages. Fig. 20. Flux density (B) in grade M4 material in sinusoidal and nonsinusoidal supplies In the case of leakage flux, energy is stored alternately in and discharged from the magnetic fields with each power supply cycle. It is not a direct loss of power but causes poor voltage regulation. This prevents the secondary voltage from achieving a direct proportion of the primary voltage, especially in cases of heavy load. Typically, by design, transformers have very low leakage inductance. Fig.21. illustrates the flux linkage between the phase windings for different materials of the core transformer. Furthermore, the percentage increase in flux linkage as a compare when operation in sinusoidal power supply between material grade M4 and grade M5 ranging 1.79% while 4.42 % as a compare between grade M4 with grade M6. While the percentage increase in flux linkage as a compare between material grade M4 and grade M5 ranging 3.46% and 4% as a compare between grade M4 with grade M6 when operation in non-sinusoidal power supply. VI. CONCLUSION The full details about flux distribution in the core transformer was presented with the following outcomes: 1) A precise model for the whole magnetic circuit of the transformer was obtained in different materials. 2) The exact contribution of parameters which affect to the core losses was determined. 3) The analysis was employed for the simulation of nonsinusoidal power supply effects on no-load loss of distribution transformers. Core design parameters such as the materials were taken into account. The values of flux linkage, flux density and core losses derived from grade M6 were greater than those of grade M4. The Findings 1. The materials in three-phase transformer have an effect on the losses inside the core of the transformer. 2. Studies have been made by researchers to find a way to reduce the no load losses. 3. Research on this trend is already actively ongoing, although shape and outlines are still not understood. An insight into what is actually going on in this emerging line is in order. 4. The behaviour of the core transformer under different conditions can affect the loss profile. This typically reflects the types of available apps in the power distribution system but gives a clear indication of where the gaps are in apps development and/or evaluation. 5. Researchers have expressed their concerns in the literature, and many have suggested resolving the seen and anticipated challenges, the list of which opens many opportunities for research in this field. Finally, it was determined that reduction of the core losses could be achieved by controlling the flux distribution using good materials. ACKNOWLEDGMENT The authors wish to thank the Center of Electromagnetic and Lightning Protection Research (CELP), Electrical & Electronic Engineering Dept., University Putra Malaysia, Malaysia, for the support given to this work. REFERENCES Fig. 21. Flux linkage in different materials with (sinusoidal and nonsinusoidal voltage) All the simulation was done using the latest version 16.0 of Ansys Maxwell software in the laboratory of the Centre for Electromagnetic and Lightning Protection Research (CELP), Electrical & Electronic Engineering Dept., University Putra Malaysia, Malaysia [1] Ilo, A., Pfutzner, H., Nakata, T., Critical induction-a key quantity for the optimisation of transformer core operation, Journal of Magnetism and Magnetic Materials, [2] Targosz, R., and Topalis, F.V., Energy efficiency of distribution transformers in Europe, 9th int. IEEE conf. Electrical and Power Quality and Utilization (EPQU 2007), Oct. 9-11, [3] Olivares, J.C., Liu, Y., Canedo, J., Escarela-Perez, R., Driesen, J., and Moreno, P.,Reducing losses in distribution transformers, IEEE Transaction on Power Delivery, 18 (3) Jul [4] Amoiralis, E.I., Tsili, M.A., Georgilakis, P.S., Kladas, A.G., and Souflaris, A.T., A parallel mixed integer programming-finite element method technique for global design optimization of power transformers, IEEE Transaction on Magnetics, 44 (6) Jun [5] Kefalas, T.D., Georgilakis, P.S., Kladas, A.G., Souflaris, A.T., and Paparigas, D.G.,Multiple grade lamination wound core: a novel DOI /IJSSST.a ISSN: x online, print

8 technique for transformer iron loss minimization using simulated annealing with restarts and an anisotropy model, IEEE Transaction on Magnetics, 44 (6) Jun [6] Hernandez, C., Arjona, M.A., and Dong, S.H.,Object-oriented knowledge-based system for distribution transformer design, IEEE Transaction on Magnetics, 44 (10) Oct [7] tenyenhuis, E.G., Girgis, R., and Mechler, G.F.,Other factors contributing to the core loss performance of power and distribution transformers, IEEE Transaction on Power Delivery, 16 (4) Oct [8] Valkovic, Z., and Rezic, A.,Improvement of transformer core magnetic properties using the step-lap design, Journal of Magnetism and Magnetic Materials, [9] Mechler, G.F., and Girgis, R.S.,Calculation of spatial loss distribution in stacked power and distribution transformer cores, IEEE Transaction on Power Delivery, 13 (2) Apr [10] Prieto, R., Cobos, J.A., Bataller, V., Garcia, O., and Ucedci, J.,Study of toroidal transformers by means of 2D approaches, 28th Annu. IEEE Conf. Power Electronics Specialists (PESC 1997), Jun , [11] Z. Valkovic, Recent problems of transformer core design, Physica Scripta T24 (3) [12] Z. Valkovic, Flux and loss distribution in three-limb core with staggered T-joint, IEEE Transaction on Magnetics,18 (2) (1982) [13] E.G. tenyenhuis, G.F. Mechler, R.S. Girgis, Flux distribution and core loss calculation for single phase and five limb three phase transformer core design, IEEE Transaction Power Delivery 15 (1) (2000) [14] GIRGIS R.S., TENIJENHUIS E.G., GRAMM K., WRETHAG J.E.: Experimental investigations on effect core production attributes on transformers core loss performance, IEEE Trans. Magn., 1998, 13, pp [15] TENYENHUIS E.G., GIRGIS R.S., MECHLER G.F.,Others factors contributing to the core loss performance of power and distribution transformers, IEEE Trans. Magn., 2001, 16, pp [16] MECHLER G.F., GIRGIS R.S.,Magnetic flux distributions in transformers core joints, IEEE Trans. Power Deliv., 2000, 15, pp [17] TENYENHUIS G.F., GIRGIS R.S.,Measured variability of performance parameters of power & distribution transformers. Proc. IEEE PES Transmission and Distribution Conf. on Exposition, 2006, pp [18] LIEW M.C., BODGER P.S.: Partial-core transformer design using reverse modelling techniques, IEE Proc. Electr. Power Appl., 2001, 148, (6), pp [19] GUERRA F.C.F., MOTA W.S.: Magnetic core model, IET Sci., Meas. Technol., 2007, 1, (3), pp [20] Kefalas TD, Kladas AG. Harmonic impact on distribution transformer no-load loss. IEEE Transactions on Industrial Electronics 2010;57(1):193e200 [21] User's guide Maxwe1l 2D, p [22] Grain-oriented electrical steel sheets catalogue. Allegheny Technologies Incorporated, USA; DOI /IJSSST.a ISSN: x online, print