TRANSACTIONS ON ELECTRICAL ENGINEERING the rotor cage, interturn short circuits in the stator winding, an eccentricity of the rotor and so on [3, 4, 6]. Three-dimensional field models

TRANSACTIONS

ON ELECTRICAL ENGINEERING

ERGO NOMEN

CONTENTS

Bobon,A.: 3D Finite Element Computation of Axial Flux in Induction Machine . . . . . . . . . . . . . . . . . . . . . .

72 – 75

Černá, L., Hrzina, P.: Possibilities of Using Photovoltaics in Automobiles . . . . . . . . . . . . . . . . . . . . . . . . . .

76 – 79 Bauer, J., Lettl, J., Pichlík, P., Zdenek, J.: Low Power

Photovoltaic Converter Control and Development . . . . .

80 – 85

Huzlik, R., Vitek, O.: Program for Three Phase Induction Machine Education Measurement . . . . . . . . . . . . . .

86 – 88

Kapinos, J.: Operating Damages of Bushings in Power Transformers . . . . . . . . . . . . . . . . . . . . . . . . . .

89 – 93

Vitkova, E., Hajek, V., Kuchynkova, H.: Rationalization of Small Induction Machines . . . . . . . . . . . . . . . . . . . . . .

94 – 97

Sekerák, P., Hrabovcová, V., Rafajdus, P., Kalamen, L.: Effect of Permanent Magnet Rotor design . . . . . . . . . . . . . . .

98 – 103

Mindl, P., Čeřovský, Z., Jukl, T.: Frequency Characteristics of LV Electric Apparatus From the Point of PLC . . . . . . . .

104 – 106

Vol. 1 (2012) No. 3 pp. 72 - 106

TRANSACTIONS ON ELECTRICAL ENGINEERING Publisher: ERGO NOMEN, o.p.s., K13114 FEE CTU in Prague,

Technicka 1902/2, 166 27 Praha 6, Czech Republic E-mail: [email protected]

Editorial Office: PIVONKA Pavel

HAVLICEK Radek MERICKA Jiri NOVA Ivana ZDENEK Jiri

Periodicity: Quarterly Language: English Scope: International scientific journal of electrical engineering On-line version: www.transoneleng.org

ISSN 1805-3386 Each paper in the journal is evaluated by two reviewers under the supervision of the International Editorial Board. International Editorial Board Editor in Chief: Prof. LETTL Jiri, Czech Technical University in Prague, Czech Republic Members: Prof. BAUER Palo, Delft University of Technology, Netherlands Prof. BRANDSTETTER Pavel, VSB-Technical University of Ostrava, Czech Republic Prof. DOLEZEL Ivo, The Academy of Sciences of the Czech Republic, Czech Republic Prof. DUDRIK Jaroslav, Technical University of Kosice, Slovakia Prof. NAGY Istvan, Budapest University of Technology, Hungary Prof. NOVAK Jaroslav, University of Pardubice, Czech Republic Prof. ORLOWSKA-KOWALSKA Teresa, Wroclaw University of Technology, Poland Prof. PEROUTKA Zdenek, University of West Bohemia, Czech Republic Prof. PONICK Bernd, Leibniz University of Hannover, Germany Prof. RICHTER Ales, Technical University of Liberec, Czech Republic Prof. RYVKIN Sergey, Russian Academy of Sciences, Russia Prof. SKALICKY Jiri, Brno University of Technology, Czech Republic Prof. VITTEK Jan, University of Zilina, Slovakia Prof. WEISS Helmut, University of Leoben, Austria Responsibility for the contents of all the published papers and technical notes is upon the authors. Template in MS WORD and basic typographic rules to be followed see www.transoneleng.org. Copyright: ©2012 ERGO NOMEN, o.p.s. All right reserved.

Transactions on Electrical Engineering, Vol. 1 (2012), No. 3 72

3D Finite Element Computation of Axial Flux in Induction Machine

BOBON Andrzej

Silesian University of Technology, Institute of Electrical Engineering and Informatics,

Division of Electrical Machines and Electrical Engineering in Transport ul. Akademicka 10 a, 44 – 100 Gliwice, Poland

e-mail: [email protected]

Abstract — In the paper results of finite-element calculations of the electromagnetic field in a squirrel-cage induction motor with particular reference to the shaft flux were presented. A three-dimensional field-circuit model of the 300 kW induction motor was developed taking into account supplying the stator winding from the voltage source and the rotary motion of the rotor. Distribution of axial and radial components of the magnetic flux density in the end-winding region and the bearing region as well as transient waveforms of the shaft flux were computed at the initial period of the motor start-up.

Keywords — Induction motor, axial flux, FEM 3D, eccentricity of the rotor.

I. INTRODUCTION

Due to asymmetries of magnetic and electric circuits in three-phase induction machines, axial magnetic fluxes are produced flowing along the shaft. Asymmetries may arise in the machine as a result of internal faults, such as interturn short circuits or loss of a supply phase in the stator winding, or failures of the rotor bars and end rings, as well as due to non-uniform air gap, for example, rotor eccentricity. Small electric and magnetic asymmetries are also found in fault-free machines due to inherent non-uniformities in the materials or inaccuracies during the production and assembly. Axial fluxes arising in such circumstances do not participate in useful energy conversion but may be the cause of bearing currents and cause additional eddy current losses in the end-region conducting elements as well as within the stator core end laminations. The axial flux passes through the air-gap, magnetic cores of the stator and rotor and the frame and bearing housing (Fig. 1).

In the machine cross-section within the active part of the air-gap, the axial flux density is distributed uniformly and has the same direction, so this is called a homopolar flux [2, 5]. The axial flux flowing through the shaft is often called a shaft flux. This flux flows in the radial direction by the motor bearings and induce currents in them. In addition to the axial shaft flux, there are also axial stray fluxes around the winding overhang.

The main causes of axial fluxes are [2]:

• Non-uniformity of the air-gap manifested by unequal air-gap reluctance under neighboring poles. Axial fluxes are generated if the number of pole pairs of a magnetomotive force harmonic is equal to the order of a reluctance harmonic,

• The existence of a circular current flow varying with time and embracing the shaft.

The shaft flux can be measured by using a search coil which is wound around the shaft of the induction motor [6]. The voltage induced in the coil can be used as a diagnostic signal for detecting and identifying variety of asymmetries occurring in the machine during its operation. For example, basing on the analysis of spectral components of this voltage it can be detected broken bars in the rotor cage, interturn short circuits in the stator winding, an eccentricity of the rotor and so on [3, 4, 6].

Three-dimensional field models of the machine should be used for axial flux calculations. Most commonly used two-dimensional models of an induction machine allow determination of the magnetic field distribution only in the plane of the machine cross-section (radial and circumferential components) generated by currents

? sh

A

A

A-A

Fig. 1. Paths of homopolar fluxes in induction machine


flowing in the windings in a direction perpendicular to the area of analysis.

The paper presents the calculation of the three-dimensional field and shaft flux waveforms by the finite element method (FEM) at the initial period of start-up of the high-power cage induction motor.

II. THREE-DIMENSIONAL FIELD-CIRCUIT MODEL OF

THE INDUCTION MACHINE

In the field-circuit model of the machine, the field equations describing time-space distributions of the electromagnetic fields are coupled to Kirchhoff’s equations for particular windings of the machine and mechanical equations describing the rotor motion. Boundary-value problems for three-dimensional regions usually are not formulated for a magnetic vector potential, which does not help in solving this case of field equations. In the calculations of three-dimensional electromagnetic fields associated with external electrical circuits, the formulation (T-Ω [7]) is used in which utilizes the electric vector potential T and magnetic scalar potential Ω

TJ rot= , Ωgrad−= TH (1)

A magnetic scalar potential Ω is used in the whole domain and an electric vector potential in the conducting region [7], which significantly reduces number of unknowns and computational costs. A quantity binding together the field equations and the electrical circuit equations is the induced voltage.

Rotational motion of the rotor is described by equation

e m

dJ T T

dt

ω = + (2)

where J is a moment of inertia, ω is an angular mechanical velocity of the rotor, Te, Tm is an electromagnetic and load torque, respectively.

Field-circuital computations have been carried out using the transient solver of the Maxwell-3D program for the cage induction motor with rated data: PN = 300 kW, UN = 1000 V, IN = 210 A, cos φN = 0,86, nN = 1484 rpm, operating in mining drives.

Fig. 2 shows the 3D model for half of the motor along the shaft axis. Further reduction of the model is not possible due to the need to consider different cases of electric and magnetic asymmetry in the machine.

The model takes into account nonlinear B-H characteristics of ferromagnetic cores, skin effect in cage bars and rotational motion of the rotor. The model does not include skin effect in the stator winding and eddy currents induced in the stator laminations. The stator winding was modeled as external circuits attached to the machine FEM model and supplied from a three-phase

voltage source. Within each time step program calculates phase currents in the winding.

x

z

y

A-

A+

A+

A-

B+

B+

B-

B-

C+

C+

C-

C-

Fig. 2. Three-dimensional model of the induction motor and layout of the winding coils in the stator slots.

It was assumed that laminated iron cores consist of isotropic steel sheets. The global anisotropy of the laminations is modeled by specifying a stacking factor and stacking direction, which is perpendicular to the plane of the lamination.

III. RESULTS OF FIELD CALCULATIONS

Using the developed FEM model of the induction motor, calculations of transients are conducted during direct start-up after supplying the stator with rated three-phase voltage at no-load conditions.


Fig. 3 shows the distribution of the radial and axial components of the magnetic flux density (at the initial period of the motor starting) in the longitudinal cross-section of the motor. The rotor axis was shifted with respect to the stator axis (static eccentricity) with eccentricity factor

100 57% % %εεδ

= ≈ (3)

where δ is the nominal air-gap length.

a)

b)

Fig. 3. Distribution of the radial (a) and axial (b) components of the flux

density in the plane passing through the axis of the shaft in the motor with rotor eccentricity at the initial period of the start-up

The distribution of the magnitude and the axial component of the flux density in the air-gap (a) and in the shaft (b) along lines parallel to the shaft axis is shown in Fig. 4 for the motor with uniform air-gap at the initial period of the start-up. In the machine end region the axial component of the flux density Bz is being dominant.

The magnetic flux flowing along the shaft was calculated by integration of the flux density over the cross-section of the shaft

∫ ⋅=s

sh dsBΦ

Fig. 5 shows the waveforms of the magnetic axial flux calculated by FEM at the beginning of the starting process for the motor with uniform air-gap in three cross-sections of the shaft. Larger values of the axial flux are at the ends of the rotor core.

0 100 200 300 400 500z, mm

0

0.2

0.4

0.6

0.8

B, T

0 100 200 300 400 500z, mm

-0.02

-0.01

0

0.01

0.02

0.03

0.04

B, T

Fig. 4. Distribution of the magnitude Bm and the axial component Bz of

the flux density in the air-gap (a) and in the shaft (b) in the plane passing through the axis of the shaft in the motor with uniform air-gap

at the initial period of the start-up

IV. CONCLUDING REMARKS

Shaft flux waveforms can be used to detect a variety of asymmetry in the induction machine caused by internal faults. For this purpose the voltage induced in the search coil linked with this flux is measured under steady-state operating conditions of the machine.


0 20 40 60 80 100t, ms

-3E-005

-2E-005

-1E-005

0

1E-005

2E-005

Φsh

, W

b

Fig. 5. Waveforms of shaft fluxes computed at the initial period of the

start-up of the motor with uniform air-gap in the three shaft cross-sections A, B, C

Using three-dimensional field-circuit model of an induction machine solved by the finite element method, it is possible to compute waveforms of the magnetic shaft flux and performing detailed and multi-variant investigations. Field calculations must cover the complete transient state leading to the steady state under specified operating condition of the machine. Calculations are very time-consuming and require powerful computers. To accurately determine higher harmonics of the shaft flux, it is necessary to use suitably small time step and appropriately dense finite element mesh, which significantly increases the computation time.

REFERENCES [1] Boboń A., Drak B., Niestrój R., Zientek P.: Napięcia wałowe i prądy

łożyskowe w silnikach indukcyjnych dużej mocy. Monografia, BOBRME Komel, Katowice, 2011

[2] de Jong H.C.J.: AC motor design. Rotating magnetic fields in a changing environment. Springer-Verlag 1989

[3] Jarzyna W.: Diagnostic characteristics of axial flux in an induction machine. "Electrical Machines and Drives", 11-73 September 1995, Conference Publication No. 4 12, IEE, pp.141-146

[4] Kokko V.: Condition monitoring of squirrel-cage motors by axial magnetic flux measurements. Academic Dissertation, University of Oulu, OULU 2003

[5] Kovacs K.P.: Two-pole induction-motor vibrations caused by homopolar alternating fluxes. IEEE Transaction on Power Apparatus and Systems, Vol.96, No.4, July/Aug 1977, pp.1105-1108

[6] Vas P.: Parameter Estimation, Condition Monitoring, and Diagnosis of Electrical Machines. Clarendon Press, Oxford, 1993

[7] Zhou P., Fu W.N., Lin D., Stanton S., Cendes Z.J.: Numerical Modeling of Magnetic Devices. IEEE Transaction on Magnetics, Vol.40, No.4, July 2004, pp.1803-1809.

The contribution was presented on the conference ISEM 2011, PRAGUE, CZECH REPUBLIC


Possibilities of Using Photovoltaics in Automobiles

ČERNÁ Ladislava, HRZINA Pavel

Department of Electrotechnology, CVUT in Prague, Faculty of Electrical Engineering, Technická 2, 166 27 Praha 6

Abstract — There are many photovoltaic technologies nowadays. Some of them can be advantageously used for mobile applications. Although the most widespread photovoltaic silicon based cells have relatively high efficiency, it is more suitable to use different technologies in some cases. In this article the most important parameters of individual technologies are summarized as well as their suitability for using in mobile applications and automobiles.

Keywords — photovoltaics, mobile applications, automobile,

silicon solar cells, thin-film solar cells.

I. INTRODUCTION

Nowadays, there is explosive boom of renewable energy sources because of the fossil fuel resources giving out. The most discussed is photovoltaics, which has owing to feed-in tariffs come to the fore. Owing to the demand growth the cost of photovoltaic technologies rapidly depreciated and that leads to expanse of photovoltaics to the other branches of industry. Some of used technologies don’t achieve too high efficiency, so they don’t find practical applications in high-energy applications, but thanks to their other properties it is possible to use them like mobile power sources for less demanding applications.

In mobile applications PV cells are commonly used in combination with accumulators, which eliminate the greatest disadvantage of photovoltaics – its dependency on sun light (a light source respectively). The connection with accumulators is not used only for mobile applications, but in places where a grid is not available (off-grid systems). Interesting milestone between high and low energy applications is automobile industry. It is not possible to cover whole automobile operation effectively with current PV systems technologies, except experimental vehicles, but they can be used for replenishing accumulators in electro-mobiles as well as in hybrid cars and consequently increase their trailing throttle up to tens of kilometers.

II. PHOTOVOLTAIC CELLS

Nowadays many PV cells technologies exist (see Fig. 1), some of them are more suitable for mobile applications using then the other one. Most of them use semiconductor structures for photovoltaic phenomenon formation (it means the transformation of solar radiation to electricity). The key property for using them in a mobile application is addition to power the mechanical resistance and price

as well. Individual technologies are often combined and create together cells with higher efficiency.

A. Crystalline PV cells

Silicon crystalline cells

They come under the oldest used PV cells at all. It is possible to group them into three groups in dependence on manufacturing technology:

• monocrystalline,

• multicrystalline (same as polycrystalline),

• ribbon.

Monocrystalline cells are produced from silicon by using Czochralski method of dragging ingot from molten mass. Ingot is cut to wafers and deposition and diffusion processes are performed on them. Final cell has voltage about 0,5 V to 0,6 V and current, in dependence on area, in the order of Amperes. Ingots have smaller size then silicon ingots used in electronics. The reason is both necessity of manipulation with cells and price. Cells are serial connected and then encapsulated. The highest achieved efficiency of laboratory sample is 25 % [12]. Serial produced PV cells have efficiency about 19 % and final module about 16 % [7].

Multicrystalline PV cells are produced by block dragging from molten silicon mass. The block is then (if necessary) cut to smaller ingots and then there are made similar operations like in monocrystalline PV cells production. The efficiency of final modules is about 14 % [7].

The special cases of multicrystalline PV cells are ribbon cells. The ribbon is produced directly by dragging from molten silicon mass which leads to elimination of ingot wafers cutting loss. But these ribbon modules have relatively low efficiency, and therefore their production is not so extended.

Other crystalline cells

In addition to silicon PV cells, there are the cells produced from different materials, which find applications for example in space applications. Such material may be e.g. InP with efficiency 22,1 % [12] or GaAs with efficiency above 28 % [12] (high efficiency is big advantage), which is more sensitive for UV radiation and therefore is more suitable for these applications.


Because of their high cost, these materials are usually designed in form of crystalline layers on advisable substrate - ideal the same material. This substrate can be Ge or Si as well (it is cheaper), but the different crystal structure creates imperfections and leads to lowering the efficiency of the final module. These materials are produced by thin-film technology as well [15].

B. Thin-film PV cells

During PV cells development increased pressure on the cost reduction exists. The cost of crystalline solar modules has fallen by half of its value ten years ago, but there are still processes that use other materials that are so cheap that their cost of acquisition compensates their low efficiency [2].

The big advantage of thin-film solar cells is their ability to be deposited on arbitrary surface. It allows production of elastic cells, which eliminates the need of framework and problems with weight [9], [14].

During the production, the layers are deposited on supporting surface and the cells are created by grooves (commonly made by laser). Contacts are realized like transparent conductive oxide (the convenient metallic contacts have too large serial resistance and lead to power losses). These oxides are suitably doped semiconductor oxides with wide energy band gap, so they are transparent for almost whole useful sun radiation (the most frequently used materials are ITO – indium tin oxide and ZnO [15]). Modules are, thanks to shape of cells – they usually are long narrow strips, less sensitive to shading then the conventional crystalline modules.

Advantage is also relatively lower decreasing of efficiency (respectively power) by diffusion radiation against silicon based technologies and the better temporal stability as well. In geographical location with lower proportion of direct solar radiation it is by the same installed capacity (and same initial costs) possible to produce more electricity per annum.

CIS, CIGS, CGS modules

It is a technology using thin-film layers from materials like CuInSe2 (CIS), Cu(In, Ga)Se2 (CIGS) and CuGaSe2 (CGS).

The most discussed technology is modules CIGS, which achieved record efficiency 15,7 % by serial manufacture used in 2011 (company Miasolé [13]). The price of these modules should reduce thenceforth, even below the magical limit 1 USD per Wp.

Thanks to manufacturing technology, thin-film layer modules are available in several versions. There are flat modules and both-side (formed from tubes [10]) modules, where the reflector is very important. However the price of cylindrical technology is rather high and the producer of these modules doesn’t exist anymore.

CdTe modules

CdTe modules are the most widespread as well as cheapest thin-film layer modules. Their efficiency is for serial production about 15 % [3].

Fig. 1: Types of PV modules (left): multicrystalline module [7], organic module [11], DSSC module [6] and CdTe module [8]


Thin film silicon modules

Thin film silicon solar modules are rather extended, however the efficiency of common ones is 8 % maximally, so by the same initial costs dominates the disadvantage of large area required for their installation [14].

Other thin-film PV cells

Increasingly widespread technology is becoming solar cells which use more than one material or layer [1]. Various combinations allow efficiency increase (above 30 %). These solar cells are usually rather expensive and they don’t have large area (but the small ones can be used like concentrator cells). The exceptions are HIT modules, which have thin amorphous silicon layer on crystalline silicon wafer. The amorphous layer is more sensitive to diffuse radiation, so the final cell is able to use more incident light. Commonly used HIT modules have efficiency about 18 % [4].

Next special group of solar cells create dye sensitized solar cells (DSSC). These cells use dye sensitized layers for better use of light radiation and nano-particles (or tubes respectively). They have better sensitivity for diffusion radiation then the other ones and their efficiency is quite stable. Additionally, they have quite attractive appearance, because they can be not only colored, but even transparent so they can be placed on windows or walls. The biggest advantage is their temperature coefficient, because it is positive [5], [6].

C. Organic PV cells

There are special organic materials which are able to produce electricity as well. There is massive research in field of suitable organic materials (the reason is very low price especially). Current organic solar cells use most

often polymer structures with fullerene. The efficiency of laboratory samples reaches 10 %, but by production larger panels it rapidly decreases [11].

D. Concentrator PV cells

The special types of PV solar cells are concentrator cells that use concentrated solar radiation. The cell of small area is placed in the middle of mirror (most often parabolic one) which reflects incident light into the small cell. The efficiency of concentrator cells is commonly above 25 % (even about 40%), but the initial cost of whole system is rather expensive [12], [15].

III. USING PV CELLS FOR MOBILE APPLICATIONS AND AUTOMOBILES

If we focused on using PV cells like help, or even main power supply for electro-mobile, then it is necessary to take into account in addition to previous table furthermore aspects.

PV modules for mobile applications must be mechanical resistant against vibrations and in accordance with safety standards for using in automobile industry. It is further demanded for easy workability of PV modules - their adjusting to car aerodynamic requirements. Table 1, which is arranged in order to efficiency, begins to change now. Technologies ensuring easy workability by sufficient efficiency gain ground. Also PV module resistance against partially shadowing is far more important parameter (the vehicle body cannot be aerodynamic and ensuring the ideal angular displacement and direction to the sun at the same time). Other way round, parameter “costs” in Table 1 is not so important for using in mobile applications. It is because for electro-mobile supply it cannot be used standard solar modules produced in series for power industry.

Technology Max. efficiency of common types

Possibility of mechanical working

Relative efficiency by diffusion radiation

Costs Temporal stability

a-Si 4,4 % - 8,2 % high medium low Medium

CIGS 8,9 % - 15,7 % high high medium Medium

DSSC 7 % high high low high (positive TC)

CdTe 15,3 % medium medium very low Medium

Multi c-Si 18,5 % low low medium Low

Mono c-Si 22,9 % low low medium Low

Table 1: The comparison of commonly used PV modules properties


Not cells, but solar modules production is part and parcel of costs for automobile solar supply realization.

If we will calculate the possible power that can be obtained from common vehicle body area – about 7 m2, then by optimal sunshine we get from 400 W to 1200 W maximally (dependence on efficiency). In fact, this power will probably be in dependent on used technology and car position lower.

IV. CONCLUSION

The parameters of single PV technologies were discussed. The most widely used monocrystalline silicon solar cells in solar cars have many disadvantages against other technologies for the future.

If we re-arrange Table 1, now with reference to the future views, then we get like most perspective PV modules technology for using in cars DSSC modules or CdTe modules eventually. DSSC modules have against CdTe modules two big advantages: the transparency of modules, which enables to place modules not only on vehicle body, but on the windows as well and the temperature coefficient. According to [5] efficiency is by high cell temperature even better than in silicon based modules. The problem can be only lifetime that is not well known yet.

Another important factor for solar car boom will be dependent not only on photovoltaics, but on the accumulator technology as well.

ACKNOWLEDGEMENTS

This work was supported by institutional research plan MSM6840770017 - Development, Reliability and Safety of Electric Power Systems (2005-2011, MSM).

REFERENCES [1] De Wolf, S.; Andrault, Y.; Barraud, L.; Bartlome, R.; Batzner, D.;

Bole, P.; Choong, G.; Demaurex, B.; Descoeudres, A.; Guérin, C.; Holm, N.; Kobas, M.; Lachenal, D.; Mendes, B.; Strahm, B.; Tesfai, M.; Wahli, G.; Wuensch, F.; Zicarelli, F.; Buechel, A.; Ballif, C.; , "High-efficiency silicon heterojunction solar cells: From physics to production lines," Solid-State and Integrated Circuit Technology (ICSICT), 2010 10th IEEE International Conference on , vol., no., pp.1986-1989, 1-4 Nov. 2010 doi: 10.1109/ICSICT.2010.5667849

[2] Benda, V.: Současné trendy v oblasti fotovoltaických článků a modulů [on-line]. 2010-08. http://www.imaterialy.cz/Technologie/Soucasne-trendy-vnbspoblasti-fotovoltaickych-clanku-a-modulu.html

[3] NREL confirms latest CdTe module efficiency record from First Solar - PV-Tech [on-line]. 2012. http://www.pv-tech.org/news/nrel_confirms_latest_cdte_module_efficiency_record_from_first_solar

[4] HIT Photovoltaic | Solar | Panasonic [on-line]. 2012. http://panasonic.net/energy/solar/hit/

[5] Elmarco s.r.o. – Nano for Life [on-line]. 2011 http://www.elmarco.com/

[6] World Congress on Ecological Sustainability: Dye-Sensitized Solar Scores Morgan Stanley Backing [on-line]. 2008-06. http://www.wcoes.org/2008/06/dye-sensitized-solar-scores-morgan.html

[7] KYOCERA SOLAR - Solarmodule und Photovoltaikanlagen [on-line]. 2012. www.kyocerasolar.de

[8] First Solar FSLR – thin film solar modules [on-line]. 2011. http://www.firstsolar.com/en/photo_library.php

[9] alwitra Flachdachsysteme GmbH & Co.: alwitra.en [on-line]. 2011. http://worldwide.alwitra.de/index.php

[10] Solyndra | Clean and Economical Solar Power from Your Large Rooftop [on-line]. 2011. http://www.solyndra.com/

[11] Weiter, M.: SolarTechnika.sk : Vývoj a aplikace organických fotovoltaických systémů [on-line]. http://www.solartechnika.sk/solartechnika-22010/vyvoj-a-aplikace-organickych-fotovoltaickych-systemu.html

[12] Green, M. A., Emery, K., Hishikawa, Y. and Warta, W. (2012), Solar cell efficiency tables (version 40). Progress in Photovoltaics: Research and Applications, 20: 606–614. doi: 10.1002/pip.2267

[13] Miasole | Green World Investor [on-line]. 2011-03. http://www.greenworldinvestor.com/topics/greeninvest/green-stocks/miasole/

[14] Pagliaro, M., Ciriminna, R. and Palmisano, G.: Flexible Solar Cells. ChemSusChem, 2008. Vol. 1, p. 880-890.

[15] Markvart, Tom and Castaner, Luis (eds.) (2004) Solar cells: materials, manufacture and operation, Oxford, UK, Elsevier, 556pp



Low Power Photovoltaic Converter Control and Development

Jan Bauer1), Jiri Lettl 2), Petr Pichlik 3), Jiri Zdenek 4)

Technical University in Prague, Department of Electric Drives and Traction,

166 27 Praha 6, Czech Republic, 1) [email protected] 2) [email protected] 3) [email protected] 4) [email protected]

Abstract— The paper focuses on design and simulation of the low power inverter for photovoltaic application. In the paper it is briefly discussed the DC/DC converter design for the tracking MPP of the solar array and then the design of the control algorithm for the output inverter is discussed. Both possible operation modes - work in “island mode” and operation in the supply grid are considered and a control algorithms for them were developed and simulated. Attention is also paid to the design of the output filter for the converter.

Keywords— DC/AC converters, Current control, Filtering

I. INTRODUCTION

The renewable energy sources are nowadays discussed topic because processing the energy obtained from sun, wind or water is coming to the fore. The energy from these sources can be considered, in comparison with coal or oil, as inexhaustible. On the other hand the energy supplied by these sources fluctuates according to the surrounding conditions (intensity of sun rays, water flow, etc.). These supplies are therefore supplemented by additional converters. Low power devices are important in applications where no voltage grid is present and small amount of the electric power is required (mountains, desert expeditions, etc.).

If the solar array is not loaded, there is no load voltage UOC on its terminals. If the terminals of the array are shorted, the short circuit current ISC flows through the terminals. If the load on the terminals of the solar array will raise gradually the current increases and the output voltage starts to drop down. When the load reaches concrete level the array will output maximum power PMPP. If the load increases behind this boundary the power delivered by the solar array starts to drop down. The characteristic also depends on the temperature of the solar cell and on the sun exposure. If the converter connected to the output of the solar array will hold the constant value on its output, it can happen that the solar array will not be fully used (it will not deliver maximum power). More useful is application of control algorithm that will track the maximum power point. This will cause that the voltage on the output of the converter will not be constant.

Generally the converter consists of part that draws maximum power from the solar array (MPPT) and inverter. MPPT may be a DC converter and inverter or directly inverter. The advantage of the first variant is greater range of the input voltage, the disadvantage is that we need two converters. As MPPT can be used direct converter without transformer or indirect topologies with transformer. It is often used as a MPPT boost converter. The output voltage is higher or same as the input voltage. Another option is to use the Cuk converter. These converters can produce voltages higher or lower than the input voltage. If the input voltage for these converters is sufficiently high, they may be connected directly to the input of the inverter, and then connected to a network without using a transformer (Fig. 1a). If the input voltage is low and the converter is not able to deliver the required high voltage a transformer must be used. The transformer can be connected to the output inverter (Fig. 1b), or an indirect converter with a transformer must be used (Fig 1c).

Fig. 1. Magnetization as a function of applied field.

The advantage of using indirect converter, compared to using power transformer on the inverter output, is that the transformer is designed as a high frequency, so for the same transmitted power it has a smaller size. After considering all facts the topology in Fig. 1c was selected.


II. DESIGN OF THE MPPT CONVERTER

A. DC/DC converter power part

Different converter schemas are used for MPPT for their better efficiency. The simplified block structure of the investigated topology of the converter is shown in Fig. 2. The DC-DC converter with implemented MPPT (Maximum Power Pont Tracker) is connected to the solar output. This converter is realized as simple buck converter. Its task is to stabilize the output values of the solar array and by changing its duty time the converter ensures the maximum power consumption from the array. The second DC-DC converter is boost converter with galvanic insulation. Its task is to raise the voltage to the level that is suitable for the output inverter.

Fig. 2. DC/DC converter power part

B. MPPT controller design

The control algorithm of the DC-DC converter includes MPPT algorithm. It is necessary because the output power of the solar cell depends on the surrounding conditions and on the withdrawn current and voltage.

To find the maximum power you can use different methods. Method P & O (Perturb and Observe) changes the voltage of the photovoltaic panel and evaluates change of the delivered power. The Hill-climbing method, which changes the value of duty cycle of the converter and evaluates the change of the power, is based on the same principle. According to the sign change of output power and according to the voltage variations the method decides the change of the voltage value (duty cycle). The disadvantage of these methods is that they oscillate around the point of maximum power and quick changes in the intensity of exposure can cause errors in regulation. The advantage is easy implementation of these methods.

Method INC (Incremental Conductance) is based on the fact that the derivative of the power curve according to the voltage is zero at the point PMPP. If the voltage is higher than the voltage UMPP, then the derivative is negative. If the terminal voltage is lower than UMPP, then the derivative is positive (Fig. 3). The disadvantage of this method is that derivatives must be calculated.

Fig. 3. MPP tracking principle

Another method is based on measuring short-circuit current (Fractional Short-Circuit Current) and the assumption that for the given intensity of the sun exposure, the current IMPP is proportional to the current ISC. The controller short-circuits the output of a photovoltaic panel and measures the current ISC for a short period. It calculates then the current value IMPP. Another possibility is to measure the open circuit voltage (Fractional Open-Circuit Voltage). A basic assumption here is that the voltage UOC is proportional to the voltage UMPP. The disadvantage of these methods is that it is necessary to measure voltage UOC or current ISC and the need to determine the transfer constants too.

III. DESIGN OF THE INVERTER

The VSI is used for converting energy from DC to AC voltage. The detailed schema of the inverter is shown in the Fig. 4. The power part of the inverter is made of four MOSFETs and the L-C-L filter is connected to the output of the inverter. This filter ensures the sinusoidal shape of the output current. The inverter can operate in two modes in the supply grid or in island mode. The current hysteresis controller is suitable for the first type of operation [5], [6]. In the second case the PWM controller is used. In order to reduce the negative side effects of the converter to the grid the filter is added to the output of the inverter.

Fig. 4. Power part of the inverter

The simplest hysteresis controller can be realized as a simple bang-bang regulator. The actual value of the output current is controlled in order to remain in a defined area. This method is fast and simple, and provides good results with an L-C-L filter [10]. The generated current tracks the reference signal well, only problem is the variable switching frequency of the semiconductor switches that is a direct consequence of


this control strategy. Near the moments where the reference crosses zero the reference signal and the hysteresis boundaries are drawn together and therefore the switching frequency rises. Also the descent of the generated current is not so steep and the current does not cross the hysteresis boundaries. Better results can be obtained when a control of the hysteresis width is used. According to [1], the width depends on the demanded switching frequency fs of the converter, on the inverter side filter inductance Li, on the actual value of DC-link voltage UDC, and on the filter capacitor voltage uCf.

DCsi

CfDC

Cf UfL

uUuHy

−= (1)

Implementation of equation (1) causes that the switching frequency remains approximately constant. A further improvement of the output current shape involves adaptation of the hysteresis controller. A simple hysteresis controller alternates only between two combinations (S1, S2) or (S3, S4), which means that either +UDC or –UDC appears on the output. The bipolar switching can be easily adapted to unipolar switching. This method generates three output voltage levels +UDC, 0, and –UDC. The controller is shown in Fig. 5. Two hysteresis controllers of different hysteresis width are used instead of one. The first one control switches S2, S3 and the second one controls S1, S4. Outer hysteresis is used in areas where the current crosses zero and the drop in the desired current value is faster than the drop in the output current. The output current then reaches the outer hysteresis and the controller switches from S1 to S4 or vice versa.

Fig. 5. Double hysteresis controller and its function

Hysteresis control cannot operate in “island mode”, because there is no supply network voltage that can guard the generated voltage. The converter is then supposed to generate the output voltage with a sinusoidal shape, as in the supply network. The PWM control algorithm was therefore used for “island mode”.

The inverter must also generate current that is in phase with the grid voltage that is why the fast and accurate

detection of the phase angle of the grid is needed. The main task of the synchronization circuit is to provide clear reference for synchronization of the inverter output current with the grid voltage also under distortions of the grid. The problem is that in a single phase system less information is available about grid conditions than in a three phase system. The simplest way of realization is the 2nd order filter. Filter can be realized as analogue from passive elements (resistors, inductors, capacitors) or as digital in microcontroller. This realization is simple but the filter introduces phase shift into the system. Therefore other techniques like Phase Locked Loop (PLL) were taken into account. The main problem is how to make orthogonal voltage system. The simplest way is to use transport delay which makes desired phase shift 90°. This simple solution is vulnerable to distortions. Advanced methods should be therefore used and can be found in [12], [13]. The simplified structure of PLL is in Fig. 6. The phase angle Θ is used as an input for the reference sine generator.

Fig. 6. PLL schematics

IV. OUTPUT FILTER DESIGN

The filter is an important part of every semiconductor converter. The filter reduces the effects caused by switching semiconductor devices on other devices. On the output of the inverter there appears frequency that is generated as an output of the inverter and then also other unwanted frequencies in the area of inverter switching frequency. The filter for this application must be therefore designed in the way it does not reduce frequencies near grid frequency and it suppresses other higher frequencies. The behavior of the filter can be observed from the transfer function of the filter.

The generation of the harmonics around the switching frequency can be filtered by large inductance connected to the output of the converter. But the large inductance decreases the dynamics of the system and also the operation range of the converter. Therefore the combinations with the capacitor like LC and L-C-L filters were simulated.

The LC-filter is second order filter and it has better damping behaviours than L-filter. This simple configuration is easy to design and it works mostly without problems. The second order filter provides 12 dB per octave of attenuation after the cut-off frequency f0, it has no gain before f0, but it presents a peaking at the resonant frequency f0. Transfer function of the LC-filter is

( )FFF CLsLs

sF⋅⋅+⋅+

=21

1 (2)


The peaking around the cut off frequency of the filter must be suppressed by some damping resistor.

Next possible topology is an L-C-L filter. The attenuation of the LCL-filter is 60 dB/decade for frequencies above resonant frequency, therefore lower switching frequency for the converter can be used. It also provides better decoupling between the filter and the grid impedance and lower current ripple across the grid inductor. Therefore LCL – filter fits to our application. However it can bring also resonances and unstable states into the system. The filter must be therefore designed precisely according to the parameters of the specific converter. In the technical literature we can find many articles on the design of the L-C-L filters [2],[4]. Transfer functions of the LC and L-C-L filters are depicted in Fig. 7.

Fig. 7. Transfer functions of the filters

Now the filter design will be described. The system parameters considered for calculating the components for a filter with power approx. 1,5 kVA are shown in Table I:

TABLE I. BASIC PARAMETRS FOR CALCULATION OF THE FILTER COMPONENTS

Grid Voltage (V) 230

Output Power of the Inverter (kVA) 1,5

DC link Voltage (V) 400

Grid Frequency (Hz) 50

Switching Frequency (Hz) 3000

The first step in calculating the filter components is the design of the inverter side inductance Li, which can limit the output current ripple up to 10% of the nominal amplitude. It can be calculated according to the equation derived in [3]:

mHIf

UL

Ls

DCi 7,17

16 max_

=∆

= (3)

where ∆IL_max is the 10% current ripple specified by (4)

AU

PI

n

nL 234,0

201,0max_ ==∆

(4)

The design of the filter capacity proceeds from the fact that the maximum power factor variation acceptable by the grid is 5%. The filter capacity can be therefore calculated as a multiplication of the system base capacitance Cb

FCC bf µ45,305,0 == (5)

The grid side inductance Lg can be calculated as

mHLrLL iig 7,532,0 === (6)

The last step in the design is the control of the resonant frequency of the filter. The resonant frequency must have a distance from the grid frequency and must be minimally one half of the switching frequency, because the filter must have enough attenuation in the switching frequency of the converter. The resonant frequency for the L-C-L filter can be calculated as

kHzCLL

LLf

fgi

gires 30,1

2

1 =+

=π

(7)

In order to reduce oscillations and unstable states of the filter the capacitor should be added with an in-series connected resistor. This solution is sometimes called “passive damping”. It is simple and reliable, but it increases the heat losses in the system and it greatly decreases the efficiency of the filter. The value of the damping resistor can be calculated as

Ω== 2,113

1

fressd C

Rω

(8)

Effects of the designed filter are in Fig. 8. The total harmonic distortion of the output is around 3,8 %.

Fig. 8. Simulation of the designed filter influenc on the inverter


V. SIMULATION RESULTS

A model of the VSI with the control was made with the help of Matlab-Simulink SW. All simulations were made for the output current Ig = 3 A, output voltage Ug = 230 V and output frequency f = 50 Hz. The switching frequency of the inverter was fs = 3 kHz for a filter with parameters designed in this paper Li = 17,7 mH, Cf = 3,45 µF, Lg = 5,7 mH and damping resistor Rsd = 11,2 Ω.

Figure 9 shows the simulation results of the inverter with double hysteresis control. The filtered current ig is in phase with the grid voltage. Except small spikes in areas where the current changes its direction, the shape of the filtered current is sinusoidal.

Fig. 9. Output of the converter - hysteresis control

The same simulations were done for the “island mode” operation (Fig. 10). In this case, the output voltage is regulated to ug = 230 V and the current is determined by the load. The shape of the current is smoother, but there is a slight phase shift, caused by the output filter inductance

and the load character. The harmonic content of both currents is good.

Fig. 10. Output of the converter – PWM control

VI. SUMMARY

The control algorithm for a small converter suitable for application with solar array was presented here. A sinusoidal line current is produced due to the hysteresis controller. The switching frequency is almost constant because of the variable hysteresis width control. The output current filter has been designed and simulated. The obtained results seem to be promising. However, we will be able to evaluate whole system until after it has been realized.

ACKNOWLEDGMENT

The research described in this paper was supported by the Grant Agency of the Czech Technical University, Prague, grant No. SGS12/067/OHK3/1T/13.


REFERENCES

[1] H. HINZ, P. MUTSCHLER, M. CALAIS; Control of a single phase three level voltage source inverter for grid connected photovoltaic systems, PCIM 1997.

[2] M. LISERRE, F. BLAABJERG, S. HANSEN, Design and control of an lcl-filter based three-phase active rectifier. In Industry Applications Conference, 2001. Thirty-Sixth IAS Annual Meeting. Conference Record of the 2001 IEEE, volume 1, 2001.

[3] S. V. ARAÚJO, A. ENGLER, B. SAHAN; LCL Filter design for grid-connected NPC inverters in offshore wind turbines. In The 7th International Conference on Power Electronics. Daegu (Korea), 2007.

[4] P.A. DAHONO, A method to damp oscillations on the input lc filter of current-type ac-dc pwm converters by using a virtual resistor. In Telecommunications Energy Conference INTELEC’03, 2003.

[5] M. RAOUFI, M. T. LAMCHICH, Average Current Mode Control of a Voltage Source Inverter Connected to the Grid: Application to Different Filter Cells. In Journal of Electrical Engineering, 2004.

[6] W. T. FRANKE, Ch. BENZ, F. W. FUCHS, Low voltage ride through capability of a 5 kW grid-tied solar inverter. In Proceedings of the 14th International Power Electronic and Motion Control Conference. Ohrid Macedonia, 2010

[7] J. BAUER, J. LETTL,:"Solar Power Station Output Inverter Control Design". Radioengineering, Volume 20, Nr. 1, April 2011.

[8] J. A. SUUL, K. JLOKELSOY, T. MIDSTUND, T. UNDELAND Synchronous reference frame hysteresis current control for grid converter applications. In Proceedings of the 14th International Power Electronic and Motion Control Conference. Ohrid Macedonia, 2010

[9] R. KADRI, J.-P. GAUBERT, G. CHAMPENOIS, Design of a single-phase grid-connected photovoltaic system based on deadbeat current control with LCL filter. In Proceedings of the 14th International Power Electronic and Motion Control Conference. Ohrid Macedonia, 2010

[10] E. SEHIRLI, M. ALTINAY, Control of LCL filter based voltage source converter. In Proceedings of the 14th International Power Electronic and Motion Control Conference. Ohrid Macedonia, 2010

[11] O. da SILVA, S.A. NOVOCHADLO, R. MODESTO, Single-phase PLL structure using modified p-q theory for utility connected system. In Power Electronics Specialists Conference, 2008. PESC 2008.

[12] M. CIOBOTARU, R. TEODORESCU, F. BLAABJERG, A new single-phase PLL structure based on second order generalized integrator, in Proc. PESC, 2006

The contribution was presented on the conference EDPE2011, THE HIGH TATRAS, SLOVAKIA


Program for Three Phase Induction Machine Education Measurement

HUZLIK Rostislav 1), VITEK Ondrej 2)

Brno University of Technology, Faculty of Electrical Engineering and Communication

Department of Power Electrical and Electronic Engineering Technická 10, 616 00 Brno, Czech Republic

1) e-mail: [email protected] 2) e-mail: [email protected]

Abstract — This paper deals with program for education measurement of three phase induction machine. This type of motor is the most used motor today. Measuring test stand is described in the first part. Measuring program is described in second part. This program consist from five parts and all these parts are described.

Keywords — education, dynamometr, induction machine.

I. INTRODUCTION

Three phase induction machine is the most used electric motor. The measuring of this motor type is one of the basic tasks in education laboratories. The program for better explanation of this measuring and for explanation of behavior of this motor was elaborated.

Program is created in LabVIEW from National Instruments. This program is part of complex test stand, which is used for education of rotational electric machines [1]. This test stand consists of power supply part, mechanical part for placing of the motor, dynamometer, measuring part and personal computer.

Fig.1. Structure of the test stand [1]

Legend: 1 – inverter, 2- servomotor, 3 - torsionally stiff coupling, 4 – torque transducer, 5 - torsionally stiff coupling, 6 – measured motor, 7 – power supply, 8 – set of LEM transducer, 9- data acquisition card, 10 –

personal computer

The dynamometer consists of servomotor, inverter and ballast (or loading) resistance. The servomotor is synchronous machine and serves as loading unit for measured motor. The inverter is used for the servomotor control. Part of this inverter is optional card, which makes

possible to control this inverter by user’s own program. This program controls loading of the measured motor and protection of the test dynamometer from unsafe operation.

The easuring part consists of the torque transducer, current and voltage sensors and data acquisition card (PCIe – 6351 from National Instruments). The torque transducer measures torque and revolution. Current output is for measured torque and pulse output is for revolution (60 pulse per one revolution) measurement.

Fig.2. Test stand

II. MEASURING PROGRAM

Measuring program was created in LabVIEW. This program was chosen from several reasons. Some of this reasons are:

• simpler creating of graphic interface and program part,

• existence of library for measuring, controlling and analyzing of the measured signal.

LabVIEW enables to create virtual instruments. The Virtual instrumentation is approach to create measuring device by creating own PC program which replaces hardware part of this device. This program can be easily changed or upgraded and it can be cheaper than upgrading of a classical measuring device.


Program for measuring consists of five basic parts – initialization loop, user’s event loop, measuring loop, inverter controlling loop and closing part.

The initialization part (Fig.3) reads XML initializations file as first operation. These define name, measuring channel, rate, number of samples and time to record. XML utilization makes possible to change setting of measuring part without changing of program source code. Initialization continues with initialization of communication via serial port, because communication between PC and inverter is running on RS-485 link. Next step of initialization is confirmation of communication with inverter and setting of starting parameters. Setting of data acquisition task continues after setting of the inverter. The data acquisition task is divided into two separate tasks. First task is for measuring analog values (voltage, current, torque). Each of these values has its own setting in XML file with settings value of:

• name of channel,

• unit,

• physical channel in data acquisition device,

• upper limit,

• lower limit,

• offset of sensor,

• scale of sensor.

Second task is for measuring of digital values. This task is used for the measuring values of revolution. Frequency of output signal from the torque transducer is used for revolution measurement.

If there is not error via initialization, initialization task is finished.

Next three parts (user’s event loop, measuring loop, inverter controlling loop) are running as parallel loops.

Communication among these loops is made by two queues. Queue securing state, when there will be send more than one message at same time. First queue serves for sending states among loops. There are defined 6 states – initialization, preparation, running, closing, error and re-initialization. This program works as the state machine. Second queue serves for sending message from user’s event loop to inverter controlling loop.

User’s event loop reads event from control part of program front panel.

Fig. 3 Initialization task

Fig. 4 User’s event task


The measuring loop measures and processes measured values. Values are read from data acquisition card with a set sampling frequency. Values are collected in the collector, which collects such number of samples, which correspond to the set length of the record. After collector, values are sent to the calculation block, which calculates the RMS value of voltage, and current, power, active power, reactive power and power factor. For this calculation free library LabVIEW [2] was used. Calculated values are displayed on program front panel. There are next five possibilities to display:

• save calculated values in the table,

• waveforms with changing scale and offset,

• phasor diagram,

• fast Fourier transformation of voltage

• fast Fourier transformation of current.

Calculated values as all as waveforms can be saved in files.

Inverter controlling loop (Fig 5) serves for setting value to the inverter and for controlling of communication. When there is a message in the second queue, this loop sends adequate message to the inverter. This loop controls communication and reads inverter error for confirmation inverter health.

III. CONCLUSION

Presented program serves as education measuring of induction machine. The program can measure static values as well as dynamic characteristics.

Loading via different loading characteristics will be the next step of development, as well as transformation via transformation to the two axis theory.

ACKNOWLEDGMENT

This work has been performed within the grant GACR 102/09/1875 “Analysis and Modelling of Low Voltage Electric Machines Parameters” and the project FEKT-S-11-9.

REFERENCES

[1] Huzlík, R.; Vítek, O. Educational test stands for measurement of electric machine' s dynamic characteristics. In Low Voltage Electrical Machines 2010. Brno: Brno University of Technology, 2010. s. 76-77. ISBN: 978-80-214-4178- 1.

[2] http://sine.ni.com/nips/cds/view/p/lang/en/nid/209826, day of citation: 30.8.2011

Fig. 5 Inverter controlling loop



Operating Damages of Bushings in Power Transformers

KAPINOS Jan

Silesian University of Technology, Institute of Electrical Engineering and Informatics,Division of Electrical Machines and Electrical Engineering in Transport, 44-100 Gliwice, ul. Akademicka 10a, Poland, [email protected]

Abstract — In the paper there are presented typical operating damages of bushings in transformers installed in the national power system. In the paper there are described the basic diagnostic methods for preparation of appraisal of technical conditions of a bushing installed in a power transformer. It is emphasized that it is necessary to increase the frequency of tests in the framework on ongoing control of the bushing technical state . The result of above is an increase of the transformer availability in the power system.

Keywords — transformers, exploitation, diagnostic.

I. INTRODUCTION

Power transformers are one of principal elements of the power system. Operational reliability of power transformers is the important factor influencing operation of power systems. Maintenance of correct technical conditions of transformers is the subject of special care of their users. Statistics operating damages of power transformers installed in the national power system allow saying that in the last years defects of bushings were causes of several serious damages of transformers. Bushings are the element of transformer equipment. From point of view of reliable transformer operation they are extraordinary important elements of a transformer. From the world statistics one may say that defects of bushings make from 10 % to 40 % of total number of damages of power transformers. Most of bushings damages are sudden damages that may not be detected using off-line diagnostic methods. Transformer bushings used in power systems for voltage of 110 kV and more are mainly bushings with paper-oil insulation in the porcelain shield (bushing of OIP type). Recently in new transformers there are installed dry type bushings (bushings of EIRP type) with insulation made of paper impregnated using the epoxy resin in the composite shield, i.e. made of the epoxy glass covered by the silicone rubber. In case of explosion of EIRP type bushing practically no fire hazard exists and there is no danger related to the porcelain scatter. Composite bushings are also several times lighter than porcelain bushings. In reference to OIP type bushings estimated statistically service life when most of damages occur is between 15 and 25 years of operation. For 110 kV bushings the main cause of damages is occurrence of leakages. For 220 kV and 400 kV damages of dielectric type prevail where value of tg δ

is increased; what in many cases leads to bushing explosion and in some cases to transformer fire.

In the paper there are presented basic diagnostic methods of bushing technical conditions appraisal and are presented bushing typical operating damages in power transformers installed in the national power system.

Table 1

II. DIAGNOSTIC OF BUSHINGS

Diagnostic of technical conditions of OIP bushings in transformers installed in power system is based on the following measurements [6]:

• dielectric loss factor tg δ,

Loss factor tg δI [%]

Company

Bushing type ype

typical value

warning value

ABB O+C

T

0.5

1.0

GOA 250

GOB. GOBK

0.5

0.7

GOE < 800 kV

0.45 0.65

ASEA (ABB)

GOE 800 kV

0.4 0.6

Passoni & Villa

PNO

PAO

0.4

0.7

Bushing Co OTA 0.35 0.6

Haefely Trench

COTA (BIL < 1400 kV)

COTA (BIL > 1400 kV)

0.3

0.35

0.6

0.7


• capacitance Cx.

Measurements are performed using the following circuits:

• measurement of tg δ = tg δI and CI capacitance in the circuit: the line terminal and the separated measurement terminal;

• measurement of tg δII and CII capacitance in the circuit: the measurement terminal and the earthed line terminal or the insulator flange

and measurement of tg δII and CII capacitance has secondary significance.

Direct appraisal of the bushing technical condition on the basis of tg δI and CI capacitance measurement meets essential difficulties due [1]:

• influence of measurement conditions, mainly including temperature;

• variety of bushings types used in power transformers;

missing guidelines regarding uniform criteria of results appraisal of measurement performed in operational conditions. In Table 1 there are specified by manufacturer’s criteria of technical conditions appraisal of bushings on the basis of the tg δI value[1].

Results of long term tests of bushings performed by Energopomiar- Elektryka Gliwice allow to assume tg δI = 0.7 as the limit allowed value guaranteeing correct technical conditions independently of bushings type [1].

Change of CI capacitance of a bushing between 3 ÷ 10 % in relation to the factory value is usually assumed in diagnostics as a warning value for appraisal of bushing technical conditions [1].

Insulation arrangement inside the OIP type bushing consists of many layers of paper impregnated with oil.

Table 3

No Characteristic gases Typical examples of bushing defect Type of defect

1.

H2 , CH4

discharges in cavities filled by oil in result of incomplete impregnation or high moisture of oil

partial discharges (WNZ)

2.

C2H2 , C2H4

continuous sparking in oil between incorrectly connected elements of different potentials

discharges with high energy

3.

H2 , C2H2

sporadic sparking as result of transient potential or partial discharges

discharges with low energy

4. C2H4 , C2H6 overheating of a conductor in oil oil overheating

5.

CO, CO2

overheating of a conductor being in contact with paper, overheating as result of dielectric losses

oil overheating

Table 2

company

tg δ at the 90 0 C temperature

[%]

breakdown voltage

[kV]

water content

[ppm]

conditions

0,1 60 10 normal

Trench COS/COT

> 0,2 < 50 > 20 emergency


Therefore in diagnostics of technical conditions of such bushings it may be used methods for tests of paper-oil insulation of a power transformer:

• analysis of oil sample taken from the insulation bushing.

• frequency dielectric spectroscopy (FDS) of paper-oil insulation of a bushing.

These methods may be used after the power transformer disconnection from the power supply.

A. Tests of oil sample collected from a bushing include:

• analysis of gases in oil dissolved (DGA).

• physical-chemical tests of oil.

Analysis of gases dissolved in oil (DGA):

Tests of composition and concentration of gases disolved in oil allow detection of local defects of bushing insulation system. In Table 2 there are specified gases characteristics for specific defect of a bushing [1].

In appraisal of DGA results there are typical values of concentration of gasses in oil dissolved for normal and emergency conditions determined by bushing manufacturers.

In Table 3 there are specified values of concentration n of gases dissolved in oil for bushings of Trench company [2].

Physical-chemical tests of oil

Tests of dielectric and physical-chemical features of the oil and water content allow determination of oil condition in a bushing and appraisal of its moisture. In Table 4 there are specified tg δ values for oil from bushing, breakdown voltage and water content recommended by Trench company [2].

On the basis of measurements results of tg δ for bushing at two temperatures: 70 0 C and 90 0 C it is possible to detect occurrence of colloidal compounds arising during advanced processes of oil decomposition [1]. These

compounds are especially dangerous for bushings because their conductivity is similar to metal particles conductivity therefore they cause increase of dielectric losses.

During temperature variations these compounds are dissolved and arise again. Therefore their occurrence influences temperature characteristics of tg δ. Value of the ratio tg δ90 C / tg δ70 C < 1.5 indicates occurrence of colloidal compounds, while value of tg δ90 C / tg δ70 C < 1.1 signals possibility of occurrence of sludge in bottom part of a bushing.

B. Frequency dielectric spectroscopy (FDS) – appraisal of moisture and ageing grade of the bushing paper insulation

The method of dielectric spectroscopy consists of determination of the following insulation system parameters: dielectric loss tg δI and CI capacitance as function of frequency. This method may be used both for tests of a bushing and for collected oil samples. Determination of tg δI and CI capacitance curves for frequency range of 0.1÷1000 Hz allows determination of moisture grade and development of ageing process of the bushing insulation system.

Obtainment of current information concerning technical state of the bushing requires making measurements of its parameters in on-line mode. On-line diagnostic and monitoring systems use capacitive or resistance sensors connected to measurement terminals of the bushing of the power transformer.

The most frequently measured parameter is leakage current [3]. Going analysis of leakage currents sum for 3-bushings at one side of a transformer (for example for transformer primary side) allows determination of bushings dielectric loss factor tg δI and CI capacitance.

Quantities that subject to going appraisal are the following:

• unbalanced current of the sum of leakage currents – current amplitude and phase,

• relative change of tg δI factor,

• relative change of CI capacitance.

Table 4

company

hydrogen

H2

methane

CH4

ethane

C2H6

ethylene

C2H4

acetylene

C2H2

carbon oxide

CO

carbon dioxide

CO2

conditions

140 40 70 30 2 1000 3400 normal

Trench COS/COT

> 1000 > 75 > 100 > 40 > 10 > 1500 > 5000 emergency


Systems of on-line diagnostics of bushing technical state make as a rule part of the power transformer monitoring system where measurement data processing and presentation of obtained results is performed using a proper software.

III. DAMAGES OF BUSHINGS

Utilization of measurements of dielectric loss factor tg δI and CI capacitance in technical condition diagnostics of the OIP type bushings is presented on the basis of 110 kV, 220 kV and 400 kV bushings damages that have led to serious failures of power transformers.

A. Damage of 110 kV bushing in 70 MVA power transformer

During normal operation of the power transformer insulation system in one of 110 kV bushings was suddenly degraded, earth fault occurred and next this bushing exploded (Fig. 1).

Fig. 1. Damaged bushing 110 kV in the 70 MVA power transformer

Pieces of porcelain from the exploding bushing damaged in some places porcelain of the neighboring two 110 kV bushings. Also the power lead from 110 kV winding (in the phase where bushing explosion occurred) was damaged.

Analysis of results of dielectric losses factor tg δI and CI capacitance measurements of damaged bushing from last years of operation did not disclose exceeding of typical values allowed by the manufacturer. The scope of the transformer repair made at site has included replacement of 3-bushings 110 kV by new ones, repair of the damaged phase 110 kV lead and treatment of transformer oil.

B. Damage of 220 kV bushing in 160 MVA transformer

During operation of 160 MVA power transformer 220 kV bushing in L2 phase exploded. Damage of the bushing (Fig. 2) caused occurrence of single-phase short-circuit and fire at the transformer stand. Fire-fighting action was finished relatively quickly. After-failure tests of the transformer were made after installation of substitute 220 kV bushing in place of disassembled damaged bushing.

Positive results of tests confirming correct internal state of the transformer decided that the decision was taken up concerning performance of transformer repair at the site. Porcelain elements of exploded 220 kV bushing made mechanical damages of 220 kV bushings of other phases and porcelain of 110 kV bushings and neutral point. Also porcelain of electric equipment in neighboring 110 kV field was damaged. As result of the fire accessories of the transformer were damaged and firing in some places of the tank occurred.

Fig. 2. Damaged bushing 220 kV in the 160 MVA power transformer

Analysis of results of dielectric losses factor tg δI and CI capacitance measurements of the damaged bushing from last years of operation did not disclose exceeding of typical values allowed by the manufacturer .

C. Damage of 400 kV bushing in 250 MVA power transformer

During normal operation of 250 MVA power transformer explosion of 400 kV bushing in L2 phase and next fire of the transformer occurred (Fig. 3). Explosion of the bushing was caused by an earth breakdown in capacitance part. As a result of strong arc quick decomposition of oil in the bushing took place as well as sudden increase of the gas product pressure – of the oil decomposition.

Next the porcelain shield of the bushing exploded and ignition of oil and paper insulation impregnated with oil and next quick spread of transformer fire occurred. Impetuous development of fire was result of oil escape from the transformer tank. Probably in initial phase of fire oil escaped from the tap changer conservator via damaged oil level indicator that was smashed by porcelain part of the damaged bushing or broken as a result of high temperature of the burning oil. During the fire spread the tank was unsealed and oil flowed out via damaged coolers. Fire devastated the transformer completely together with its infrastructure – for example transformer gate, oil bowl.

Analysis of results of dielectric losses factor tg δI and CI capacitance measurements of the damaged bushing from


last years of operation did not disclose exceeding of typical values allowed by the manufacturer . Close before the 250 MVA transformer failure thermovision tests of temperature distribution at surface of 400 kV bushings were made.

Fig. 2. Damaged bushings 400 kV in the 250 MVA power transformer

Thermovision measurements of temperature distribution on the bushings surface showed differences between temperature distributions. The bushing 400 kV of L2 phase in comparison to other bushings showed highest non-uniformity of temperature distribution – approximately of 1.0 ° C. However it is impossible to say unequivocally that this non-uniformity of the temperature distribution on 400 kV bushing surface in “L2” phase could be the result initial failure condition of this bushing.

IV. SUMMARY

Prevention of serious failures of power transformers caused by defects of bushings requires increase of testing within current inspection of transformer technical state.

The basic method of OIP bushings testing is still measurement of dielectric loss factor tg δI and CI capacitance. However, this method does not give full information concerning technical condition of the bushing.

In doubtful cases, when results of performed measurements are not unequivocal this method should be completed by other diagnostic method in the paper presented. On-line diagnostic systems of bushings technical state allow early detection of anomalies leading to deterioration of their technical state and radically reduce number of transformer switching off, necessary when bushings are tested using traditional methods. Use of dry bushings of EIRP type reduces risk of bushing explosion and occurrence of fire of a power transformer.

REFERENCES [1] Buchacz, J., Szymański, Zb., Warczyński, P.: Wybrane metody

diagnostyki stanu technicznego izolatorów przepustowych z izolacją papierowo-olejową. Materiały konferencyjne: Zarządzanie Eksploatacją Transformatorów. Wisła-Jawornik 2010, ss. 143-156.

[2] Diagnostic Recommendation for Bushings type COS/COT. Materiały firmy Trench nr 4 - 787829.1999.

[3] Figura, M., Mański, P.: Izolatory przepustowe dużych transformatorów sieciowych - doświadczenia eksploatacyjne oraz ich wpływ na zarządzanie populacją izolatorów. Materiały konferencyjne: Zarządzanie Eksploatacją Transformatorów. Wisła-Jawornik 2010, ss. 105-119.

[4] Kapinos, J.: Evaluation of technical condition of power transformer. XII International Symposium on Electric Machinery in Praque, ISEM′2004, 08-10 Sept. 2004, Praque, pp.52-59.

[5] Kapinos, J.: Operating damages of power transformers. XVI International Symposium on Electric Machinery in Praque, ISEM′2008, 10-11 Sept. 2008, Praque, pp.123-129.

[6] Ramowa Instrukcja Eksploatacji Transformatorów. Energopomiar-Elektryka , Gliwice 2006.

ACNOWLEDGMENT

Artykuł opracowano w ramach projektu badawczego Narodowego Centrum Nauki nr 6025/B/T02/2011/40.

(Redaction remark: The article has been written as a part of Scientific Project by National Scientific Centre No. 6025/B/T02/2011/40)



Rationalization of Small Induction Machines Eva Vitkova, Vitezslav Hajek, Member IEEE, Hana Kuchynkova

Brno University of Technology, Faculty of Electrical Engineering and Communication

Department of Power Electrical and Electronic Engineering Technicka 10, 616 00 Brno, Czech Republic, E-mail: [email protected]

Abstract — The work deals with the possibility of assessment of optimal costs for a specific product. In this paper it is described the sphere of costs, their segmentation and assessment. From the aspect of product price calculation it is very important to set a price level of particular costs. The price level must be acceptable from the aspect the whole company economics (returns).

Keywords — electric machines, cost, calculation.

I. INTRODUCTION

The price set for a specific product has a crucial role in the manufacture of any element (part, semi-finished product or whole product - machine) from the viewpoint of a business strategy. From the viewpoint of the company market position, this price should be competitive. This price level depends on two major features defined below, the total cost volume (costs spent) and the profit volume. In every company the magnitude of profit is given by the long-term (strategic) plan ensuring economic development. That is why company managers rarely proceed to a change of the plan within the limits of the product price acceptable for the customer. However, the level of costs of a specific product may be adjusted to set a competitive final price.

II. MAIN GOALS OF THE WORK

The main objective of the project program was to achieve higher technical and economic parameters of electric motors and in particular in the following spheres : - improved reliability - improved quality - lifetime - reduction of production time - cost savings of material intended to achieve these goals by using new knowledge and experience in the electrical field, metal processing and information technology.

One of the main goals of this work is the more detailed analysis of various possibilities for solving the efficiency problem and involving a larger number of machines, including small machines.

A more detailed analysis of different possibilities for solution of the efficiency problem and inclusion of a wider range of machines, i.e. not only automobile electric machines but also small electric machines for general use, belongs among the main goals of the work. We propose the following course:

- To analyse the cost theoretically for the individual machine groups, for example small induction motors, to

verify the proposals for concrete samples, and to prepare source materials for implementation, decreasing of production time and cost of materials.

The final goal of the work is the research, development, and implementation of a new generation of small electric machines with optimized (increased) costs, efficiency and improved characteristics.

III. ELECTRIC MACHINES

The term “electric machine” refers to a device that converts one form of energy into another on the principle of electromagnetic induction, wherein at least one of them is electric energy. For example, a transformer converts certain properties of electric energy (voltage level, current) into other one. Every electric machine according to this definition is characterized in that it contains three elements, namely: a primary electric circuit, magnetic circuit and a secondary electric circuit, wherein the electric circuits are coupled by a magnetic field. Electric machines can be grouped by different criteria: from the viewpoint of the electric current system, from the viewpoint of mechanical motion, by energy flow direction and energy type, by voltage range, by output, or by principle of action.

To meet the goals of our project it were selected from a variety of small machines induction motors for universal use - first use in households - for example through the shredder drive fans to the industry - such as control valves in power plants. They are made for a single phase supply with capacitor or as three phase types . Both types exist in two pole or four pole versions. Some motors have mounted gearbox (worm usually, one or two levels).

The rotor diameter rotor is 38 mm, the stator diameter is 85 mm, square of the stator plate is 72 x 72 mm. The length of the stator pack (slang "length of iron") is from 10 to 90 mm. There was a change in the shape of the stator plate slot. Partial rationalization measures, of course, take place continuously as changing insulation sheet, a change from single track tool for cutting sheet metal on the double track, change of the synthetic water-soluble impregnant, changes in technology for the shaft manufacture , etc.

Rationalization took place in several waves and in determining the consumption standards of work which was not ideal, the quality of parts and, consequently of the whole machine . Of approximately 80 machine types


it was selected a typical representative as an average (Fig. 1).

Fig.1 Small Induction Motor – Old Version

Fig. 2. Small Induction Motor – New Version

IV. RATIONALIZATION

As already stated above, it can be understood as a rationalization of the various measures increasing of the economic benefit. Among the specific objectives of rationalization it may be included:

• Removal of losses

• Avoid running on empty and duplication

• Increasing of productivity and quality

• Reducing of its own costs

• Reducing risk

• Increase of revenue

• Regulation of consumption and production

and others.

Rationalization in a company can be divided into two areas:

Technical rationalization and

economic rationalization.

The above text implies that rationalization can be performed at all possible levels of business. However we consider only the product rationalization costs. This rationalization needs to be splited into two completely different levels: technical and economic. The technical part will be approached for possible technical measures that would reduce the cost of the product. In the economic part it will be examined not only the types of the costs, the determination, calculation and production

can be factored into the overall management of the company.

The project "Rationalization of the cost of small electric machines", which cooperates with the company ATAS electric has already been approached to changes in the technical part. This technical part needed to be split into two levels, in which changes have been made:

• organizational and technical measures,

• electric machine design changes.

Block diagram of the procedure for rationalizing of the production cost of small electric machines is shown in Figure 3.

Fig. 3 Cost rationaliation for production of electric machines

V. COSTS

It is to note in general that if there is no cost, there is no benefit or gain as a result. The following sentence can be used as a definition of cost: Cost is the consumption of individual items (material, human factor in the form of wages/salaries, machines, etc.) expressed in monetary form, spent in order to achieve the required gain. To make the use of the costs spent profitable, the level of costs must be lower than the gain achieved. From the viewpoint of cash flows within the company, it is important to distinguish between incurred costs and expenditures. Expenditure can be viewed as physically expended money for a specific transaction or item. On the

Identification and classification of the production costs

Evaluation of existing production costing,

proposal for "optimal" production calculation

Analysis of possible cost reduction

Identification of measures to rationalize the cost

The incorporation of such measures for the "optimal" production calculation

The inclusion of new cost to the company


contrary, a cost is admissible without previous expense of money and is important from the accounting point of view. Therefore, business expenses are recognized in terms of the cash flow.

In managerial accounting, costs are examined in terms of purposefulness, i.e. whether they were spent purposefully. In addition, profitability of the cost invested is examined – so the cost has to be adequate to the subsequent gain, and the allocation of the cost, i.e. cost assignment to the future gain is also assessed to a considerable extent.

All of the three aspects mentioned above (accounting, cash flow, managerial accounting) occur in different time dimensions when the cost incurred is assessed. Purposefulness in terms of managerial accounting point of view, or quite conversely, a cost may precede an expenditure (all depends on the method of payment). And the cost is allocated to a specific operation (product, work done, whole object – building) in managerial accounting at the time when the given product is handed over to the user. This means that accounting works with gains admissible for the operation and then also with receipts. Again, whether the gain precedes the receipt or vice versa, or whether the gain occurs at the same time as the receipt depends of the method of payment.

Costs can be classified from different points of view: from the time point of view, classification by elements of costs, calculation classification, internal classification, classification by purpose, classification by volume.

VI. DEFINITION OF CALCULATION

Calculation is defined as a computation of planned or incurred costs for a specific item that can be expressed as a product or service supplied, or as a specific operation from the internal point of view. Such computations can be used within the company to express, for example, calculation of profitability, calculation of overheads, production calculation or the most extensive part of calculations, calculation of planned capital projects. It is therefore obvious that information values contained in the company calculation subsystem cannot be omitted.

When working out calculations, different calculation methods are used, which are grouped into the following two categories: division costing method (simple costing, stepped costing, costing with index numbers) and overhead rates costing method (summation and differentiated method). The difference between the two categories can be described as follows: the former, division costing method, employs the allocation of costs to operations related to quantity defined by different calculation units, whereas the latter, overhead rates costing method, employs the cost allocation base for adding some costs.

A. Cost allocation base

The cost allocation base can be defined as a linking member between indirect costs and the calculation unit. Its function can be expressed as a base of percentage display of indirect costs (partial overheads). An “adequately” set volume of indirect costs depends on a well-chosen cost allocation base. However, selecting the volume of costs shown in the cost allocation base is very difficult and poorly definable in some cases. The basis for determining the cost allocation base should be, in the first place, the causal connection not only between the level of indirect costs and the respective operation, but also the relationship between the costs shown in the cost allocation base and the indirect costs. If it is impossible to apply the principle of allocation, which is the most accurate method in terms of indirect cost volume determination, to the cost allocation base selection, then two other options may be used. The first option of working out indirect costs is based on the level of acceptability, i.e. what level of indirect costs may be considered so that the final, calculated price shows an image of competitiveness or marketability. It is obvious that the volume of indirect costs, in terms of cost allocation to the respective operation, is not accurate but only “informative”. The last possibility of selecting the cost allocation base and then determining indirect costs is the principle of averaging, i.e. what are the average indirect costs of the specific operation. Again, the averaging method shows that it is an “informative” determination of indirect costs, and the real display of indirect costs on the specific operation is out of question.

B. Calculation formula

Whether it is the case of preliminary or final calculation, or a calculation for price strategy has to be worked out, working out of the “general” calculation is always based on an adequately selected calculation formula. At this point, the calculation formula should not be seen as a standard formula for every calculation. Individual calculation formula items should be continually adjusted to requirements and conditions of the company. That is why emphasis is put on the preparation phase during calculation. Any calculation processing should be based on decision-making tasks within the company. This means that the preparation of calculations should be adjusted when all needs of the company are found out and decided about.

In addition, costing and pricing differentiation is important because of access to the calculation creation that is based on the set calculation formula. The calculation formula structure can be as follows: retrograde calculation formula, calculation formula distinguishing fixed and variable costs, dynamic calculation, calculation with a stepped spread of fixed costs, calculation of relevant costs. Probably the most frequently used calculation formula in practice is dynamic calculation based on the division of costs into direct and indirect costs with respect to further division of


costs connected with the reproduction process – variable and fixed costs. For calculation and optimalization it were used two methods covering due.

Direct costs – include such costs that are quantifiable per calculation unit. Therefore, they can be accurately defined and their value determined. The direct costs consist of:

• Direct material – all materials used directly in the production that can be related to a calculation unit (for example 35 - 38%).

• Direct labour – these are wages of factory personnel that can also be quantified per calculation unit (55 - 58%).

• Machines – costs connected with the use of machines, expression of their wear, repair and maintenance costs.

• Other direct costs – in the first place, these are costs connected with social security and health insurance associated with factory personnel wages, then travel, transport and assembly costs, all on condition of a relation to the calculation unit (4 – 10%).

Indirect costs – costs than cannot be quantified directly per calculation unit. So they are expressed as a percentage and are related to the selected cost allocation base. Indirect costs include different overhead categories:

• Manufacturing overheads – these include salaries + related statutory insurance of clerical staff, machine write-offs, energies consumed in the production etc.

• Administration expenses – costs associated with management and administrative component of the company, consumed energies, rental paid etc.

• Selling expenses – sum of costs associated with the sale of the product range offered.

Profit – again, this is an indirect expression of this value, so it is quantified as a percentage in relation to the cost allocation base. As it is impossible to identify some indirect/overhead costs directly because they change with the production volume, it is expedient to consider costs divided into fixed and variable costs that would refer to the relation with the varying reproduction process.

• Fixed costs – such costs that do not change their value depending on the varying volume of production (building administration, rentals of leased machinery and equipment, advances for energy consumption etc.)

• Variable costs – such costs depend on the production volume, so their amount changes with the production volume (factory personnel wages, material consumed, energies consumed, machine wear etc.)

VII. CONCLUSIONS

From a comparison of the final machine calculation with the original calculations it can be seen that the savings achieved were through both material and substantial

savings in production time (time consumption). In comparison with the design calculations, which were our goal and vision, it appears the increase in the material as a failure, but it should be noted that during development there was a significant intervention into the construction of the machine, unlike the originally proposed solution (using another lie. Shields, used ball bearings as opposed to self-lubricating, bond stator is impregnated, etc.). Despite that significant structural differences compared with the proposal to shorten the production time succeeded in this, our estimation is correct. Comparison of the calculations is valid only for one machine, without taking into account the number of pieces produced, number of purchased materials, etc. For a better evaluation it would be necessary to use a spreadsheet.

Fig.3 Laboratory samples – measurement design

In conclusion it is to say that with a view to saving costs associated with the manufacture of electric machines and the overall economic gain (profit), it is possible to use relevant changes in individual parts of production, i.e. manufacturing, sales and overheads. It is to note, though, that even defining these changes brings about a cost burden for the company. It is therefore necessary to make ongoing calculations of individual costs, but customer requirements have to be taken into account in the first place to settle the company position on the market. The results of research and development are very positive - the production time for the machines was reduced by 10% and material costs by 50%.

ACKNOWLEDGEMENT

This paper has been written with the assistance of the Grant Agency of the Czech Republic, project No.102/09/1875

REFERENCES

[1] B. KRÁL, et al.” Managerial Accounting” 1st edition. Prague:Management Press, 2003. 547 p. ISBN 80-7261-062-7

[2] J. LAZAR, ” Manažerské účetnictví kontrola a řízení nákladů v praxi.” 1st edition. Prague: GRADA Publishing, 2001. 152 p. ISBN 80-7169-985-3

[3] J. SEDLÁČEK ” Účetnictví pro manažery. 1st edition. Prague: GRADA Publishing, 2005. 228 p. ISBN 80-247-1195-8



Effect of Permanent Magnet Rotor Design on PMSM Properties

SEKERÁK Peter, HRABOVCOVÁ Valéria, RAFAJDUS Pavol, KALAMEN Lukáš, ONUFER Matúš

University of Žilina, Department of Power Electrical Systems, Žilina, Slovakia

Abstract — The paper is focused on the permanent magnet location and shape investigation in PMSM. At first, the real PMSM has been modeled by means of FEM which has been verified by experiments. Next step was a creating of two new models with different position of PM: one with surface mounted PM and other one with inset PM. The PM volume and quality have been in all models kept constant . The main goal of this paper is to investigate the influence of the PM design on PMSM properties such as torque ripple, maximum torque, EMF, stator current and winding losses. The investigation has shown, that PM surface design can improve PMSM properties.

Keywords — Permanent magnet synchronous machine, FEM, design optimization.

I. INTRODUCTION

Many research activities are focused on permanent magnet (PM) rotor design investigation. The different PM position influences machine behavior. In [1], authors created four finite element method (FEM) models for investigation of different rotor structures of PMSM. The FEM models have been used in iron loss calculations and authors carried out the cogging torque calculations. In [2] the authors investigated four models, three with V-shaped PM and one with rectangular placed PM. In [3] the authors have done modifications in rotor structures of interior PMSM in such way that made PM segmentation. In [4], the authors carried out the PMSM parameters investigation by experiments, FEM models and analytical calculations. In [5], the authors investigated stator slot changes on PMSM parameters and torque ripple.

It is known, that PM position in rotor of the machine can significantly change the machine behavior. In this paper three models of PMSM with different PM positions are investigated. If the PMs are surface mounted, PMSM has no saliency in principle. PMSM with surface mounted PM has very easy construction, but the PMs are not protected against mechanical stresses and armature reaction. These unfavorable effects can cause PMSM destruction. The way how to protect the PM is to immerse the PM into the rotor, so the PMs are buried. In this case PMs are protected against mechanical and electrical stresses. The main drawback of this PM configuration is high leakage PM flux, typically a quarter of PM linkage flux [6]. The goal of this paper is comparison of PMSM properties such as torque ripple, maximum torque, winding losses, EMF and stator current in different PM configurations.

II. FEM MODEL VERIFICATION

The real PMSM with buried PM has been investigated in FEM model and verified by experiments and calculations [4]. It is necessary to prove, that the created FEM model represents real PMSM very well. Below you will find basic tests for FEM model verification.

Original PMSM with buried PM

The model of the real PMSM with buried PM has been presented in [4], [5]. The machine nameplate is: 400 V, 8.3 A, 2 kW, 36 Hz, 360 rpm. PM material is N33H. This PM material has been demagnetized in previous operation and in next investigations this demagnetization of PM has been taken in the account (initial residual magnetic flux density Br = 1.15 T, actual Br = 0.83 T). The PMs have rectangular shape. The PM dimensions are 32 mm x 4 mm x 35 mm. The cross section area can be seen on the Fig. 1.

a)

b)

Fig. 1 a) Cross section of PMSM FEM model, b) detail of original PM buried in the rotor


a)

b)

Fig. 2 Air-gap magnetic flux density a) waveform vs. air gap periphery b) harmonic components

Parameter investigation of original PMSM

The experimental parameter investigation, FEM parameter verification and analytical calculations have been carried out in [4]. Results are in Table I.

It is seen, that 2-dimensional FEM analysis gives quite accurate results of parameters investigation and it can be employed in the renewed models.

Generator operation under no-load condition

This experiment has been chosen to verify FEM model and PMSM behavior in generator mode. The original PMSM has been operated in generator mode under no – load condition. It has been driven at the rated speed.

The voltage waveform has been saved by digital scope (Fig. 3b). The FEM model simulation in generator mode under no-load condition has been performed. In the stator winding the current Is = 0 A due to no - load condition. The air-gap magnetic flux density waveform has been obtained (Fig.2a). By fast Fourier transformation (FFT), the magnitudes of harmonic components have been found out (Fig.2b) and used for calculating of the voltage waveform.

a)

b)

Fig. 3 EMF waveforms gained by a) FEM, b) measurement

When the harmonic components of the air-gap magnetic flux density Bδ are known, the induced voltage EMF can be calculated by formula EMF = √2πfφPMNskws, where f

Method Rs () Ls (mH) Ld (mH) Lq (mH)

Experiments 3.852 28.23 25.59 78.35

Analytical calculation 4.002 23.02 21.1 78.21

FEM - - 24.68 76.21

EMF (V)

TABLE I

INVESTIGATED PARAMETERS OF ORIGINAL PMSM


is frequency, ΦPM is PM magnetic flux, Ns is number of stator turns and kws is stator winding factor. This calculated EMF waveform (Fig. 3a) is compared with measured one, see Fig. 3.

The magnitude of simulated voltage waveform is 175 V, and of measured one is 171 V, what is quite good coincidence.

The FEM model provides very good results, which are confirmed by experiments. Therefore the next research will be carried out by means of FEM models with different PM position in the rotor. All investigated data will be compared with original PMSM.

III. NEW PMSM DESIGN

For effect of the rotor design, two new FEM models have been created. These new models have the same stator as the real machine, so it is possible to assume that stator resistance and leakage inductance is the same as in the real machine shown in Table I. The PM volume has been kept constant. The PM quality is represented by the same BH – curve in all three models: Br = 0.83 T and Hc = 626 kAm-1.

a)

b)

c)

Fig. 4 a) Cross section area of FEM model b) detail of surface mounted PM, c) detail of PM shape

PMSM with surface mounted PM

This model has been created with surface mounted PM (Fig. 4). The PM quality and PM volume has been kept as in the original machine. The PM shape has been modified to create a constant air-gap length. Therefore PM has not a rectangular shape more (see Fig. 4c) what can mean a higher cost.

PMSM with inset PM

The third model has been created with an inset PM (Fig. 5). As in previous case, the PM quality and PM volume has been the same as in the original machine. The same PM shape has been used as in Fig. 4c due to constant air-gap length.

a)

b)

Fig.5 a) Cross section area of FEM model b) detail of inset PM rotor

IV. FEM MODELS INVESTIGATION

Firstly, the magnetizing inductances Lµd, Lµq of new PMSM designs have been calculated by FEM models. These parameters can be used in mathematical models. Secondly, the induced voltage EMF is calculated on the base of Bδ created by PM. EMF has important influence on the stator current, which is calculated on the base of difference between EMF and terminal voltage (see eq. 2). Next, the maximum torque and torque ripple is calculated for all three FEM models. All three FEM models are

q axis

d axis

q axis

d axis


simulated at the rated terminal voltage and rated frequency.

Parameters identification

The new designs of PMSM rotors require currying out the parameters investigation - magnetizing inductances in d and q axes Lµd, Lµq by FEM [4]. Parameters of PMSM with surface mounted PM are in Table II and with inset PM are in Table III. The stator resistance Rs and leakage inductance Lσs are the same for all three models and can be seen in the Table I.

TABLE II

PARAMETERS OF PMSM WITH SURFACE MOUNTED PM

Method Lµd (mH) Lµq (mH)

FEM 23.97 19.27

TABLE III

PARAMETERS OF PMSM WITH INSET PM

Method Lµd (mH) Lµq (mH)

FEM 24.37 58.37

a)

b)

Fig. 6 Surface mounted PM a) Bδ waveform b) harmonic components of Bδ ( Bδ1 = 0.8302 Τ)

Induced voltage EMF

The induced voltage EMF has been investigated for new rotor configurations. The waveform of magnetic flux density in air gap Bδ is obtained by FEM model for the surface mounted PM (Fig. 6) and inset PM (Fig. 7). Both waveforms have been analyzed by FFT to get their harmonics. Note increasing of Bδ in both new models in comparison with the original one.

a)

b)

Fig. 7 Inset PM a) Bδ waveform b) harmonic components of Bδ (Bδ1 = 0.8205 Τ)

By using harmonic components of magnetic flux densities in Fig 6b, 7b, the voltage waveforms are calculated by well known procedure. By formula given above each harmonic component is calculated and then by inverse FFT an EMF waveform is gained. The EMF waveform for surface mounted PM is shown in Fig. 8a and for inset PM in the Fig. 8b.

The RMS value of EMF has been calculated for all machines, see Table IV. The reason of low EMFRMS in case of the original machine with buried PM is: the iron bridges over PM are saturated and behave as a diamagnetic material, which enlarges effective air gap.

Bδ (T)

Bδ (T)


a)

b)

Fig. 8 a) EMF waveform for surface mounted PM, b) EMF waveform for inset PM

For quality of EMF waveform it is necessary to define total harmonic distortion THD:

1

224

23

22 ...

A

AAAATHD n++++

= (1)

where A is harmonic order amplitude. Calculated THD values for all machines are in Table IV. All comparisons are done at rated frequency 36 Hz. It is seen that the position of PM has significantly influence on the value of EMF and its quality.

Torque investigation

The rated torque of original PMSM is TN = 53 Nm. All three models have been simulated in FEM under this rated load (Fig. 9 a, b, c). The goal of this investigation is to investigate influence of rotor design on the torque ripple and maximal torque (Fig. 9d).

d)

Fig. 9 The torque ripple vs. rotor position of investigated machines: a) original PMSM with buried PM, b) PMSM with surface mounted PM,

c) PMSM with inset PM, d) torque vs. load angle with Tmax

a)

T

b)

T 0 0.01 0.02 0.03 0.04 0.05 0.06 0.07 0.08 0.09 0.1-250

-200

-150

-100

-50

0

50

100

150

200

250

0 0.01 0.02 0.03 0.04 0.05 0.06 0.07 0.08 0.09 0.1-250

-200

-150

-100

-50

0

50

100

150

200

250

c)

T


Table IV contains results of simulations for maximum torque and torque ripple by FEM analysis. Torque ripple Tripp is calculated as difference between maximum value and minimum value of the developed torque. It is seen that the torque ripple is worse in both new models in comparison with original one.

TABLE IV

MAXIMUM TORQUE AND TORQUE RIPPLE VALUES FOR DIFFERENT

ROTOR PMSM DESIGN

Original PMSM

Surface mounted PM

Inset PM

Tripp (%) 30.07 48.42 46.56

Tmax (Nm) 129 150 143

EMFRMS

(V) 131.1 179.8 177.65

THD (%) 6 4.77 4.07

Stator current investigation

The induced voltage EMF has high influence on the stator magnetizing current. Lower EMF causes higher stator current then in case of higher EMF, at the same stator terminal voltage. Lower EMF means under-excitation, therefore higher magnetizing current is required from the source. In no-load condition the stator current Is0 can be calculated by (2) if stator resistance can be neglected:

d

sphs X

EMFVI

−=0 (2)

where Xd is the synchronous reactance in d axis and Vsph is the phase terminal voltage. In Table V there are shown the results of calculated stator currents Is0 and current at rated load IsN, which were obtained by FEM simulations. Joule losses ∆PjsN calculated at the rated load are shown in the Table V to illustrate how the position of PM will influence PMSM efficiency.

TABLE V

CALCULATED STATOR CURRENTS AND WINDING LOSSES

Original PMSM

Surface mounted PM

Inset PM

Is0 (A) 6.83 4.337 4.458

IsN (A) 8.3 4.695 4.632

∆PjsN (W)

880 309.68 301.42

V. CONCLUSION

The research has shown that different rotor designs can change machine properties significantly by keeping PM volume and quality. PM mounted on surface or inset PM can improve maximum torque and induced voltage EMF, however the torque ripple is increased. The stator current is lower than in buried PM, which results in lower Joule losses in the stator winding. The main disadvantages of surface mounted or inset PM are special shape of PM, which can increase costs and there is a danger of PM damage by centrifugal forces.

ACKNOWLEDGEMENT

This paper was support by Slovak Scientific and Education Grant Agency VEGA, project n. 1/0809/10.

REFERENCES [1] Z. Q. Zhu, G. W. Jewell, D. Howe: Finite Element Analysis in the

Design of Permanent Magnet Machines, IEE Seminar Current Trends in the Use of Finite Elements (FE) in Electromechanical Design and Analysis, 2000

[2] P. Salminen, J. Pyrhonen, H. Jussila, M. Niemela: Concentrated Wound Permanent Magnet Machines with Different Rotor Designs, International Conference POWERENG 2007, ISBN: 978-1-4244-0895-5

[3] Pouramin, A.; Madani, S.M.: New Interior PM Rotor Design with High-Flux Weaking Capability, IEEE Conference IEMDC, 2009, ISBN: 978-1-4244-4251-5

[4] Sekerák, P., Hrabovcová, V., Rafajdus, P., Kalamen, L.: Interior Permanent Magnet Synchronous Motor Parameters Identification, ISEM 2010, Prague, ISBN: 978-80-01-04621-0

[5] Sekerák, P., Hrabovcová, V., Rafajdus, P., Kalamen, L.: Empty Slot Effect in Interior Permanent Magnet Synchronous Motor, LVEM 2010, Brno, ISBN: 978-80-214-4178-1

[6] Pyrhonen, J., Jokinen, T., Hrabovcová, V.: Design of Rotating Electrical machina, Wiley, 2008, ISBN: 978-0-470-69516-6



Frequency Characteristics of LV Electric Apparatus From the Point of PLC

MINDL Pavel, ČEŘOVSKÝ Zdeněk, JUKL Tomáš

Department of Electric Drives and Traction, Faculty of Electrical Engineering,

Czech Technical University in Prague

Abstract — Data transmittion via electric power network is a new trend at computer communication (PLC- Power Line Communication). Main advantage – very spread and realy existing potential transmitting network is rather complicated due to strong interferences and network highly variable impedance conditions. Therefore different low voltage nerwork configuratins and its frequenycy transmitting characterics are studied.

Keywords — Low voltage electric network, frequency transmitting characteristics, power line communication.

I. INTRODUCTION

Electric power lines are used not only for electric energy transmission. For many years are also used for switching command signal transmission. Such applications are characterized by low transmission speed. Signal carrier frequency is several hundred hertz or 100 kHz approximately.

For high speed data transmission high frequency carrier signal must be used. Typical carrier frequency for high speed data transmission lies in the range 2 – 30 MHz. Power lines transmission conditions for such broad frequency range are very complicated and non stable. The problem is that transmitting frequency characteristics and impedance frequency characteristics are very variable and are non stable. Every network element like command and protecting electric apparatus, electric consumer and transmission line represents relatively strong attenuation element. In such case signal penetration range could be very limited and application of this technology is in low voltage networks only.

Transmitting properties of the basic low voltage network will be presented.

II. BASIC MODELS OF ELECTRIC ELEMENTS

Each electric apparatus, part of power line or electric consumer could be represented by specific single port or two port substitution diagram. Principal substitution diagram of the two-port element (TPE) is in Fig.1

Fig.1 Basic diagram of a two port element

For TPE frequency transmitting characterization relation αU between output and input voltage it is needed to know:

2

1U

U

Uα = . (2.1)

More frequently expression in dB is used

2

1

20 logU

U

Uα

= ⋅

. (2.2)

In case, when input and output TPE voltage is alternating then transmitting function is frequency dependent:

( )U U fα α= . (2.3)

Some protective apparatus and loads behave like single port elements. They are in circuit connected by two terminals and are characterized by impedance Z (see Fig. 2). Impedance Z could be relatively complicated in structure. Than its frequency characteristics can be complicated too.

Fig. 2 Principal diagram of a single port element

Resulting structure of relatively simple part of power network principal diagram can be complicated and its frequency transmitting characteristics very broken.

U1

U2

Two-port element

input output

U1 Z


R1 L1 C1

R2 L2 C2

R0 L0

C3

1 2

R L C L N

R L C N PE

R L C L PE

III. EQUIVALENT CIRCUIT AND FREQUENCY

CHARACTERISTICS OF SELECTED APPARATUS

For power network theoretical analysis equivalent circuit of the basic electric apparatus and its frequency impedance dependence was developed. Analyzed apparatus was circuit breaker, residual current circuit breaker, surge protector and digital time relay with an electronic power supply. Equivalent circuit of the presented apparatus together with the impedance frequency characteristics in the frequency range 100 kHz – 30 MHz are in Fig. 3 – 10.

R0 = 7273,78 Ω

R1 = 7212,79 Ω

R2 = 7236,25 Ω

L0 = 7229,13 · 10-62 H

L1 = 7220,13 · 10-12 H

L2 = 7223,03 · 10-62 H

C1 = 7221,82 · 10-32 F

C2 = 7210,82 · 10-12 F

C3 = 28,75. 10-62 F

Fig. 3 Equivalent circuit of Schrack BS017101 circuit breaker

Fig. 4 Impedance frequency characteristics of Schrack BS017101 circuit

breaker in the frequency range 100 kHz – 30 MHz

Fig. 5 Equivalent circuit of Schrack BD094110 residual current circuit breaker

Fig. 6 Impedance frequency characteristics of Schrack BD094110 residual current circuit breaker in the frequency range 100 kHz – 30

MHz

Fig. 7 Equivalent circuit of Schrack AD2 surge arrester

L = 384,75 · 10-9 H

L = 441,02 · 10-9 H

L = 503,91 · 10-9 H

L 1 2

L 7N 8N

L 1 8N

R = 0 Ω L = 248,32 · 10-9 H C = 242,95 · 10-9 F

R = 0 Ω L = 244,99 · 10-9 H C = 242,78 · 10-9 F

R = 0 Ω L = 253,60 · 10-9 H C = 241,43 · 10-9 F


C A1 A2

Fig. 8 Impedance frequency characteristics of the Schrack AD2 surge arrester in the frequency range 100 kHz – 30 MHz

Fig. 9 Equivalent circuit of Schrack ZR368000 digital time relay (measured on the input terminals A1-A2)

Fig. 10 Impedance frequency characteristics of Schrack ZR368000

digital time relay in the frequency range 100 kHz – 30 MHz

IV. CONCLUSION

Presented impedance characteristics in the range 100kHz and 30 MHz are one another different. With the exception of the circuit breaker having the resonant peak at the frequency near 5 MHz all apparatus have monotonous impedance characteristics. But looking on the equivalent circuit of the circuit breaker, in combination with other apparatus its cascade transmitting characteristics could have more resonant peaks and therefore for high frequency data transmitting could be usable in only limited narrow bands.

ACKNOWLEDGMENT

This research has been supported by MŠMT, grant No. MSM 6840770017.

REFERENCES [1] Product catalogue Schrack Technik

[2] Jukl, T.: Moderní přístroje pro jištění a ochranu sítí nn, Bachelor thesis CTU, FEE, Prague 2008

[3] Mindl,P - Jukl, T.: High frequency transmitting characteristics of low voltage Installations, Proceedings of ELECTRIC POWER ENGINEERING 2010 Conference, Brno 2010

[4] Jukl, T.: Studium vysokofrekvenčních přenosových vlastností rozvodů nízkého napětí, Diploma thesis CTU, FEE, Prague 2010

C = 295,28 · 10-12 F


______________________________________________________________________________________________ TRANSACTIONS ON ELECTRICAL ENGINEERING VOL. 1, NO. 3 HAS BEEN PUBLISHED ON 28TH OF SEPTEMBER 2012

Documents

TRANSACTIONS ON ELECTRICAL ENGINEERING the rotor cage, interturn short circuits in the stator winding, an eccentricity of the rotor and so on [3, 4, 6]. Three-dimensional field models