nationalconference - ijeee 201… · nationalconference on recent trends in renewable energy...

National Conferenceon

Recent Trends in Renewable Energy Sources& Electronics (RES 2016)

25th-26th February,2016

Organised By:

Publishedin

Special Issueof

International Journal of Electrical & Electronics EngineeringApril, 2016

Volume 3 Special Issue 2 IJEEE: e-ISSN: 1694-2310 | p-ISSN: 1694-2426

FromEditor's DeskDear Reader,

I take the privilege to welcome all of you to this specialissue of IJEEE. The intention of our journals is to createan atmosphere that stimulates creativity, research anddevelopment in the respective area of study.

The development and growth of the mankind is theconsequence of brilliant researches done by scientistsand Engineers in every field. Our journals provides anoutlet for research findings and reviews in areas ofengineering, technology, science, management, law,literature & philosophy found to be relevant for nationaland international development. The aim and scope ofthe journals is to provide an academic medium and animportant reference for the advancement anddissemination of research results that support high-level learning, teaching and research in the concerneddomains and to become a virtual knowledge grid forresearchers, scholars, postgraduate students, developersand innovators in the respective fields.

Finally, I express my sincere gratitude to the EditorialBoard Members for their continuous support to makethe release of this special issue of the journals possibleand bring it to its present form. Undoubtedly we couldnot issue an outstanding journal without the amazingefforts of our reviewers. The motivation and expertise ofthe professionals who serve as reviewers are necessaryto maintaining technical and editorial standards. So it iswith great pleasure and thankfulness that I offer mythanks to each of the peer reviewers.

I hope you will enjoy reading this issue and wewelcome your feedback on any aspect of the Journal.

NaveenKumarEditor-in-Chief,Elixir Publications (Journals)[email protected]

IJEEE is indexed in:

:

DisclaimerThe responsibility of the contents and the opinions expressed in this journal is exclusively of the author(s) concerned. Thepublisher/editor of thejournals and organizers are not responsible for errors in the contents or any consequences arising from the use of information contained in it. Theopinions expressed in the researchpapers/articles in this journaldonotnecessarily represent theviews of the publisher/editor.

IJEEE, Volume 3, Spl. Issue 2 (2016) National Conference on Recent Trends in Renewable Energy Sources & Electronics

RES -2016 1

Single-Core Symmetrical Phase ShiftingTransformer Protection Using Multi-Resolution

AnalysisMeenakshi Sahu1 , Mr. Rahul Rahangdale2

1,2Department of ECE, School of Engineering And I.T. , Mats University, Raipur, (C.G.) , [email protected], [email protected]

Abstract: This paper presents performance evaluation ofconventional current differential protection of single-coresymmetrical phase shifting transformer. The first part ofthis paper describes the modeling of a single coresymmetrical phase shifting transformer. The second partof the paper discusses the problem associated with theconventional protection scheme by evaluating thesimulation results. A phase shifting transformer (PST) isusually used for varying the voltage phase angle betweenthe two systems, which provides controlling of activepower transfer. Some types of PSTs allow controllingphase shift in a certain defined range, which means thatphase shift between the two systems, can be varied on-line. Therefore, this type of phase shifting transformersmust be protected by relays, which are able to on-linecompensate for additional phase shift introduced by PST.The main attention is paid to testing of differentialprotections for transformer energization and externalfaults for different phase shift values. It has beenobserved that for such situations, standard differentialrelays may sometimes mal-operate. MATLAB Simulink,is used for the simulation of delta-hexagonal PhaseShifting Transformer (PST) using On Load TapChangers (OLTC). Relaying algorithm has alsodeveloped using MATLAB Simulink.

Key words: Current differential protection, Phase shiftingtransformers, FACTS devices, Fault identification andclassification, Relaying technique.

I. INTRODUCTIONIncreased energy demand, deregulation, and privatization ofthe power supply industry often cause utilities to operate andstress transmission systems to, and occasionally beyond,their original design capabilities. Maintaining reliable,secure, and economical operation of interconnected networksunder these conditions requires that transmission operator’sbetter control and manages network power flows. Speciallydesigned power system equipment can control the flow ofactive or reactive power in inter connected power systems byaffecting one or more parameters. Traditionally, the onlydevice available to power system operators that controlledboth the magnitude and direction of power flow was the PST.Today other devices are also available to control networkpower flows, which are categorized under different power

electronics based FACTS devices, Further based on thecontrolling parameter and working principle they areclassified as Series and Shunt FACTs devices such as FixedSeries Capacitor, Thyristor Controlled Series Capacitor,Unified Power Flow Controllers, Thyristor-ControlledPhase-Angle regulators, and Inter-phase Power FlowControllers [1]. This paper focuses especially on the PST.A phase shifting transformer (PST) is one of several devicesthat can be classified to series elements of Flexible ACTransmission Systems (FACTS) technology. When the PSTis series-connected to line reactance between two systemsthen the equation for calculating the active power flowthrough transmission line can be rearranged and is as follows[2], [3], [4]:

…(1)It is clearly visible form (1) that due to use of a PST it ispossible to provide controlling of active power transfercapability by adding (or subtracting) an additional phase shiftφ to the existing phase angle δ between the line sending Vsand receiving VR end voltages. In the other words, higher orlower amount of active power can be transferred through atransmission line for smaller value of δ. Another advantage isa possibility of improving power system stability andproviding flexible power flow control. Generally, PSTs areusually installed between two systems (or cross borderinterconnections) being connected with one or more paralleltransmission lines (paths) – examples of using of PST arepresented in details in [5], [6], [7], [8].A differential relay is always used as a unit protection. Theoperating principle of the differential relay is to monitor thecurrent entering and leaving the zone of protection. Duringthe normal or external fault conditions, the current measuredinto the zone is always equal to the current leaving the zone,and therefore the vector sum of both currents is always zero(ideally). However, in the event of internal fault conditions itmeasures a large differential current. Differential protectionis commonly used for the protection of transformers, motors,generators and short-transmission lines. The three mainobjectives of transformer protection are:Detection of internal faults with high sensitivity


RES -2016 2

High speed isolation of the transformer in the event of a faultSecurity/stability against the external fault, or no-faulted,system conditions for which tripping of the transformer arenot required.Proper protection minimizes the cost of repair, productionloss, adverse effect on the balance of the system, damage tothe adjacent equipment and the period of unavailability of thedamaged equipment [9].The conventional differentialprotection principles for the two commonly used PST types,two-core symmetrical and delta-hexagonal PST, areproposed by [10]. However, due to PST unique design andconstruction, it brings new challenges in addition to theaforementioned traditional challenges associated with thestandard transformer differential protection. Commonlyknown challenges are non-standard phase shift between twoends, saturation of the winding exposed to high voltages,turn-turn fault protection. Various new solutions, also basedon the differential philosophy, have been proposed by [11]-[12] to cope with these challenges. However, no newtechnique has reached the practical level yet. The provisionof various new and old PST differential based approaches isto detect and discriminate the internal/external faults.However, they are different in differential current measuringprinciple. The standard transformer design is solely based onthe concept of magnetic-coupling therefore, differentialcurrent measuring principle reflects the ampere-turn relationof the magnetically coupled transformer windings. However,design and construction of the PST is such that it representsboth magnetic-coupling and electrically connected circuits.Therefore, PST differential current measuring principle canbe the representation of magnetically coupled or electricallyconnected windings or simply the vector sum of thecompensated currents entering and leaving the two ends ofPST without considering the PST design and construction.Differential current measuring principles proposed byvarious approaches are reviewed in previous sectionfollowed by the detail comparative analysis and discussinabout PST in next section, further the differential protectionscheme is discussed then simulation study and results willshow us the problems associated with conventional PSTdifferential protection.

II. PHASE SHIFTING TRANSFORMERSPhase shifting transformers are widely used for the control ofpower flow over parallel transmission lines. Power flowcontrol becomes necessary in today’s deregulated powersystem market, when parallel transmission paths are ownedor operated by different operators. PST offers a complete,reliable and more economical solution for the control ofpower flow as compared to FACTS devices. PSTs areavailable in unique designs and constructions whencompared to the standard power transformers. Moreover,they are among the most expensive transformer kinds in theirfamily. The advantage of the symmetrical design overasymmetrical is that the phase shift angle is the onlyparameter that influences the power flow.A Phase-ShiftingTransformer is a device for controlling the power flowthrough specific lines in a complex power transmissionnetwork. The basic function of a Phase-Shifting

Transformer is to change the effective phase displacementbetween the input voltage and the output voltage of atransmission line, thus controlling the amount of activepower that can flow in the line as shown in equation (1) foractive power control. Phase-shifting transformer using on-load tap changers (OLTC) for introducing a phase shiftbetween three-phase voltages at two buses in a transmissionsystem. Controlling phase-shift on a transmission system willaffect primarily flow of active power. Although the phase-shifting transformer does not provide as much flexibility andspeed as power-electronics based FACTS, it can beconsidered as a basic power flow controller. The dynamicperformance of the phase-shifting transformer can beenhanced by using a thyristor-based tap changer instead of amechanical tap changer. The delta hexagonal connectionconsists of three pairs of windings interconnected in ahexagonal configuration.Simulated model in MATLABSimulink [13] has been utilized here to test differentialprotection algorithm, the operation of delta-hexagonal PhaseShifting Transformer (PST) using On Load Tap Changers(OLTC) is also tested with differential protection algorithm.One 120 kV 1000 MVA networks are interconnected througha phase shifting transformer (PST). The phase shift can bevaried on load by means of On Load Tap Changers(OLTC).In order to observe impact of phase shift on powertransfer, the phase shift is increased from zero to 32.2degrees lagging (tap +5), then phase shift is reduced to zeroand increased again up to 32.2 degrees leading. This isperformed by sending 5 pulses to the "Up" input, and then,10 pulses to the "Down" input ". As the tap selection is arelatively slow mechanical process (3 sec per tap as specifiedin the "Tap selection time" parameter of the block menus),the simulation Stop time is set to 50 s

As the tap of the transformer changes for requirement ofactive or reactive power demand, protection algorithm needsto be reconfigured. The OLTC is used to change tap onlinebut relay algorithm needs to reconfigure off-line. Therefore areliable and sensitive relaying algorithm required for thispurpose. The simulated model is shown in Fig. 1.

III. CURRENT DIFFERENTIAL PROTECTIONOF PHASE SHIFTING TRANSFORMERS

From many years, current differential protection has beenwidely used for the protection of bus-bars, generators,transformers. Speed and selectivity are the two mainadvantages of differential protection, and it therefore onlyresponds to internal faults. The differential protectiontechnique is based on the comparison of the differentialcurrent with the restraining current [14]. Differential currentis measured using Kirchoff’s law by taking the vector sum ofthe current entering and leaving the zone of protection.

Fig. 1 Simulated model in MATLAB


RES -2016 3

The zone of differential protection is normally defined by thepositioning of the current transformers (CTs). However, thezone of protection can be the representation of either amagnetically coupled circuit (e.g. transformer) or electricallyconnected circuit (e.g. bus-bar or lines). Any change in thedefined zone leads to an imbalance between the currententering and leaving the zone and, therefore, results in thedifferential current. Therefore, current differential protectionis sensitive to most of the internal faults. However, thesensitivity of differential protection is dependent on the fault-current level. There are various old and new differentialprotection-based approaches [15]−[17] could bedistinguished based on the differential current measuringprinciple that reflects the type of circuit it is representing.

IV. SIMULATION STUDY AND RESULTSThe current differential algorithm is tested using simulatedfault events in MATLAB/SIMULINK® software. Thesampling frequency of three phase current signals isconsidered as 1.2 kHz. Both primary and secondary windingsfaults are considered for validation of differential algorithm.Three phase current signal from both primary and secondaryis obtained from bus 3 and bus 4 respectively throughinstrument transformers. Here instrument transformer isassumed ideal. The Digital signal processing toolbox inMATLAB software has been used for data acquisitionprocess, fundamental component obtained through DFT areused for relaying algorithm.The behaviour of current signals during different position ofTap and there second harmonics behaviour on normaloperating condition are shown in Fig. 2 and Fig. 3.

Fig. 2Current signal of PST secondary side (in per unit) for all thethree phase (no fault)

Fig. 3Second harmonic component of current signal of PST duringtap change

Fig. 4 Differential current signal of PST during single turn fault(AG)

Fig. 5 Differential current signal of PST during single turn fault(BG)

Fig. 6 Differential current signal of PST during single turn fault(CG)

Fig. 7 Differential current signal of PST during inter turn fault(ABG)

0 1 2 3 4 5 6 7

x 104

0

2

4

6Phase A

0 1 2 3 4 5 6 7

x 104

0

2

4

6Phase B

0 1 2 3 4 5 6 7

x 104

0

2

4

6Phase C

0 1 2 3 4 5 6 7

x 104

0

0.05

0.1

0.15

0.2Phase A

0 1 2 3 4 5 6 7

x 104

0

0.05

0.1

0.15

0.2Phase B

0 1 2 3 4 5 6 7

x 104

0

0.05

0.1

0.15

0.2Phase C

0 1 2 3 4 5 6

x 104

0

5

10

15Phase A

diff

curre

nt

0 1 2 3 4 5 6

x 104

0

2

4

6Phase B

diff

curre

nt

0 1 2 3 4 5 6

x 104

0

2

4

6Phase C

diff

curre

nt

0 1 2 3 4 5 6

x 104

0

2

4

6Phase A

diff

curre

nt

0 1 2 3 4 5 6

x 104

0

5

10

15Phase B

diff

curre

nt0 1 2 3 4 5 6

x 104

0

2

4

6Phase C

diff

curre

nt

0 1 2 3 4 5 6

x 104

0

2

4

6Phase A

diff

curre

nt

0 1 2 3 4 5 6

x 104

0

2

4

6Phase B

diff

curre

nt

0 1 2 3 4 5 6

x 104

0

5

10

15Phase C

diff

curre

nt

0 1 2 3 4 5 6

x 104

0

5

10

15

20Phase A

diff

curre

nt

0 1 2 3 4 5 6

x 104

0

5

10

15

20Phase B

diff

curre

nt

0 1 2 3 4 5 6

x 104

0

2

4

6

8Phase C

diff

curre

nt


RES -2016 4

Fig. 8 Differential current signal of PST during inter turn fault(BCG)

Fig. 9 Differential current signal of PST during inter turn fault(CAG)

Fig. 10 Differential current signal of PST during inter turn fault(ABCG)

Fig. 4 to Fig. 6 shows that differential current algorithmcorrectly identified any single turn short circuit faultoccurred in the PST. For simulating different types of singleturn fault on secondary side of the PST three phase switch isused in MATLAB. During single turn to ground faultconditions it is clear that from Fig. 4 to Fig. 6 before no faultcondition differential current is near about zero value butwhen fault occurred in PST it will larger than the thresholdvalue set. Here threshold value is set based on the differentfault analysis done including all internal and external faultcondition. Whenever differential quantity is greater thanthreshold value it indicate internal fault presence in PST. Insimilar manner for inter turn fault condition has evaluatedand shown in Fig. 7 to Fig. 10. Previously mentioned result

for differential algorithm is for secondary side inter-turn faultcondition. This differential algorithm is also successfullyidentifying primary side fault of PST.It is here by noted that this algorithm is working on some ofthe common case only. This algorithm fails where phase shiftof the PST is larger than 90 degree. It has also observed fromthe above case study that threshold value is set based ondifferent fault scenario, which should be set manually duringtime to time according to tap position of the PST. Duringexternal fault condition near by the transformer connectionthis algorithm mal-operate and gives false tripping signal tocircuit breaker.

V. CONCLUSIONIn this paper, the effects of the PST on the differential relayalgorithm were investigated. Research results revealed thatthe during PST presence in the system, differential algorithmmal-operates. It was noted that the PST affects the currententering and leaving to the relaying point, which is calculatedby the differential relay. It is also observed that differentialrelay mal-operated during external fault conditions which isnot acceptable in high voltage application. Therefore, there isa need for an improved algorithm for protecting phaseshifting transformers with better stabilization for externaldisturbances and more sensitivity to internal fault. Inaddition, further simulation testing should also be done forthe newly developed protection criteria. In the scope ofinterest are the solutions based on intelligent techniques andadaptation concepts.

REFERENCES[1] N. G. Hingorani and L. Gyugyi, “Understanding FACTS Concept and

Technology of Flexible AC Transmission System”, IEEE Press, 2000.[2] D. A. Tziouvaras, R. Jimenez, "138 kV phase shifting transformer

protection: EMTP modeling and model power system testing", in Proc.Eighth IEE International Conference on Developments in Power SystemProtection, 5-8 April 2004, Vol. 1, pp. 343 – 347.

[3] L. Sevov, C. Wester, "Phase angle regulating transformer protectionusing digital relays", in Proc. Eighth IEE International Conference onDevelopments in Power System Protection, 5-8 April 2004, Vol. 1, pp.376 - 379.

[4] J. Verboomen, D. Van Hertem, P.H. Schavemaker, W.L. Kling,R.Belmans, "Phase shifting transformers: principles and applications",in Proc. International Conference on Future Power Systems, 18 Nov.2005, DOI: 10.1109/FPS.2005.204302.

[5] P. Bresesti, M. Sforna, V. Allegranza, D. Canever, R.Vailati1,"Application of Phase Shifting Transformers for a secure andefficient operation of the interconnection corridors" in Proc. IEEEPower Engineering Society General Meeting, 2004, pp. 1192–1197.

[6] B. K. Patel, H. S. Smith, T. S. Hewes, W. J. Marsh, "Application ofphase shifting transformers for Daniel-Mcknight 500kVinterconnection", IEEE Transactions on Power Delivery, Vol. 1, No. 3,July 1986, pp. 167-173.

[7] A. S. Siddiqui, S. Khan, S. Ahsan, M.I. Khan, Annamalai, "Applicationof Phase Shifting Transformer in Indian Network", in Proc. InternationalConference on Green Technologies (ICGT), 2012, pp. 186 - 191.

[8] IEEE PSRC Working Group K1, "Protection of Phase Angle RegulatingTransformers", IEEE Special Publication, Oct. 1999.

[9] IEEE Std C37.91, “IEEE Guide for Protecting Power Transformers”,2008.

0 1 2 3 4 5 6

x 104

0

2

4

6

8Phase A

diff

curre

nt

0 1 2 3 4 5 6

x 104

0

5

10

15

20Phase B

diff

curre

nt

0 1 2 3 4 5 6

x 104

0

5

10

15

20Phase C

diff

curre

nt

0 1 2 3 4 5 6

x 104

0

10

20Phase A

diff

curre

nt

0 1 2 3 4 5 6

x 104

0

5

10Phase B

diff

curre

nt

0 1 2 3 4 5 6

x 104

0

10

20Phase C

diff

curre

nt

0 1 2 3 4 5 6

x 104

0

10

20

30Phase A

diff

curre

nt

0 1 2 3 4 5 6

x 104

0

10

20

30Phase B

diff

curre

nt

0 1 2 3 4 5 6

x 104

0

10

20

30Phase C

diff

curre

nt


RES -2016 5

[10]Protection of Phase Angle Regulating Transformers IEEE PowerSystem Relaying Committee prepared by Working Group K1, 1999.

[11]TammamHayder, Ulrich Schaerli, Kurt Feser, Fellow, IEEE, LudwigSchiel, “Universal Adaptive Differential Protection for RegulatingTransformers” IEEE Transactions on Power Delivery, vol. 23, no. 2, pp568-575, April 2008.

[12]B. Kasztenny, E. Rosolowski, “Modeling and Protection of HexagonalPhase-Shifting Transformers-Part II: Protection” IEEE Transaction onPower Delivery, vol. 23, no. 3, pp 1351-1358, July 2008.

[13]SimPowerSystems. User Guide, TheMathWorks, Inc. Natick, MA.[14]IEEE Std C37.91, “IEEE Guide for Protecting Power Transformers”,

2008.[15]“Protection of Phase Angle Regulating Transformers”, IEEE Power

System Relaying Committee prepared by Working Group K1, 1999.[16]Michael J. Thompson, Hank Miller, John Burger, “AEP Experience

With Protection of Three Delta/Hex Phase Angle RegulatingTransformers”, in Proc.Advanced Metering, Protection, Control,Communication, and DistributedResources, 2007, pp. 96-105.

[17]Z. Gajic, “Use of Standard 87T Differential Protection for SpecialThree-Phase Power Transformers-Part 1: Theory”, IEEE Transaction onPower Delivery, vol. 27, no. 3, pp 1035-1040, July 2012.

[18]Solak, K.; Rebizant, W.; Schiel, L., "Differential protection of single-core symmetrical phase shifting transformers," in Electric PowerEngineering (EPE), 2015 16th International Scientific Conference on ,vol., no., pp.221-226, 20-22 May 2015.


RES -2016 6

Inter Area Power Oscillation Damping For MultiMachine Interconnected Power System

Shubi Sharda1, Mr. Geoffrey Eappen2

1,2Department of ECE, School of Engineering And I.T. , Mats University, Raipur, (C.G.) , [email protected] , [email protected]

Abstract: Power systems are steadily growing with everlarger capacity. Formerly separated systems areinterconnected to each other. Modern power systemshave evolved into systems of very large size. Withgrowing generation capacity, different areas in a powersystem are added with even large inertia. As aconsequence in large interconnected power systems, lowfrequency oscillations have an increasing importance.Low frequency oscillations include local area modes andinter-area modes. Inter-area modes of oscillations may becaused by the either high gain exciters or heavy powertransfer across weak tie line. The occurrence of the inter-area power oscillations depends on various reasons suchas weak ties between interconnected areas, voltage level,transmitted power and load. At time, the oscillations maycontinue to grow causing the instability of the powersystem. lots of power system stabilizers have beendeveloped by the researchers in the past few years, butthe area is still open for the efficient power stabilizerdevelopment which can efficiently able to handle thepower oscillations without increasing the systemcontroller system complexity. This paper presents recentdevelopments in power system stabilizers for two areafour machine system transmission lines powerstabilization.

Keywords: Multi Machine Power System, Power SystemStabilizer (PSS), Inter area Power Oscillation, Poweroscillation damping.

1. INTRODUCTIONToday, the electrical power systems are no longer operated asisolated systems, but as interconnected system s which mayinclude thousands of electric elements an d be spread overvast geographical areas. The advantages of interconnectedpower systems are that they[l] [2]:

Provide large blocks of power and increasereliability of the system.

Reduce the number of machines which are requiredboth for operation a t peak load, and required asspinning reserve to take care of a sudden change ofload.

Provide economical sources of power to consumers.

On the other hand, interconnection of systems also bringsabout new problems. The interconnecting ties betweenneighboring power systems are relatively weak whencompared to the connections within the system. It easily

leads to low frequency Inter area oscillations. Many of theearly instances of oscillatory instability occurred at lowfrequencies when interconnections were made.

The study of power system stability is an interesting topic inelectrical engineering research. Power system stability maybe defined as[l]:

The property of the system that enables synchronousmachines of a system to respond to a disturbance from anorm al operating condition so as to return to a conditionwhere their operation is again normal. Depending on thenature and order of magnitude of the disturbance, stabilitystudies are usually classified into three types, namely[l] [2][3]:

Steady State Stability - It refers to the behavior of a powersystem around a fixed operating point: the system issubjected to small and gradual changes in the operatingconditions.

Dynamic Stability - It refers to the long time response of apower system to relatively small disturbances. It differs fromthe steady state stability because it assumes that the system issteady state stable, and the system is subjected to smalldisturbances.

Transient Stability - It is aimed at determining if a systemwill remain in synchronism following major disturbancessuch as transmission system faults, sudden load changes, lossof generating units, or line switching. Transient stabilityproblems can be subdivided into first-swing problems wherethe time period under study is the first second following thedisturbance, and multi swing stability problems where theperiod under study may be extended to over 10 seconds.

For the steady state and dynamic stability problems, a powersystem can be modeled by linear differential equations. Forthe transient stability problem, a power system must berepresented by nonlinear differential equations. In allstability studies, the objective is to determine whether or notthe rotors of the machines being perturbed return to constantspeed operation.

2. STABILITY ANALYSIS OF POWER SYSTEMSThe purpose of this section is to introduce and explain thephenomenon of stability in power networks. In staticsecurity, we consider the electrical system at steady-state andignore the transient state. This case it is based on the fact that


RES -2016 7

in order to ensure a good operation of a power grid, it isessential that the voltage in different parts of the network andthe power (or current) flow is maintained within theacceptable limits. In dynamic security, we consider thedynamic evolution of the electrical system in response to anydisturbance (such as short-circuit, circuit tripping, loss ofload, etc.). Dynamic security is thus defined as the system'sability to "survive" the transient (not electromagnetic) andthe dynamic that can occur after disturbance. Dynamicsecurity is mainly based on assessing the stability of thesystem. Thus, decisions and actions are taken based on theresults of the evaluation of the stability and dynamicbehavior of the system.

2.1 Categories and classes of stability problemsThere are several forms of stability. While historicallyangular stability (with reference to the timing generators) hasbeen the dominant problem of stability of electricity grids.With the evolution of electrical networks and the tendency tooperate them near their physical limits, other forms ofstability have emerged. These include voltage stability andfrequency stability, or oscillatory stability.

2.2.1 Angular Stability.As angular stability is related to the synchronous machinewhich plays a fundamental role in understanding this type ofphenomenon. Indeed, most of the electric power generated isprovided by synchronous type machines at the synchronousfrequency of 50 Hz or 60 Hz. Therefore, the synchronousmachine representation and modeling plays a vital role in theanalysis of stability.

2.2.2 The Equation of MotionThe basic equations describing the reaction of rotatingmasses of synchronous machines to various disturbances arerelated to the inertia of the synchronous machine, anddescribe the resulting imbalance between the electromagnetictorque and mechanical torque of these machines.This imbalance may be expressed by the followingrelationship:

ema ...( 2.1)

WhereΓa: acceleration torque Nm;Γm: mechanical torque Nm;Γe: electromagnetic torque N.m.

This equation applies for both generators and motors.However, for generators, Γm and Γe are positive while in thecase of motors they are negative. In the case of generators,Γm is the torque that mechanically produced by the turbine inthe direction of rotation on the shaft of the machine. Thisallows the rotor to accelerate in the positive direction ofrotation.

The electromagnetic torque Γe created by theinteraction of the magnetic flux of the rotor and stator,opposes the mechanical torque and corresponds to theelectric power across the air gap of the machine. In the caseof motors, Γe corresponds to the air gap power provided by

the network and Γm is the opposite torque provided by themechanical load and mechanical losses due to friction. Inwhat follows, we focus exclusively on the synchronousgenerator.Under steady-state conditions, the accelerating torque is zerobecause Γm = Γe. In this case, there is no acceleration ordeceleration of moving masses. Thus, the speed is constantand corresponds to synchronous speed.

The equation of motion of masses in the case ofsynchronous machines is based on equation [2.1]. When animbalance is product between Γe and Γm, there is anacceleration (or deceleration) of the rotating masses. Thelatter is expressed as the product of the moment of inertia ofthe masses by its angular acceleration, i.e,

emm

a dtdJ

......( 2.2)Where:J: moment of inertia of all rotating masses includinggenerator and turbine shafts (kg/m2);ωm: angular velocity of rotor (mech rad/ s).2.2.3 Transient stabilityAs indicated earlier, the network suffers disturbances whichare more or less severe on a regular basis. From the point ofview of stability, low amplitude perturbations can be treatedin the studies of static stability. However, severedisturbances (such as a fault, a sudden loss of a significantload or generation) are treated in the context of transientstability studies. Indeed, this type of disturbance often resultsin strong reactions in the network particularly in alternators.This leads to large variations of different variables (voltage,power, internal angle, etc.). In extreme cases, part of thenetwork, if not all, becomes unstable.

Transient stability is defined in relation to thenetwork capacity to keep the machines synchronized as aresult of severe disturbance. The system then settles to a newpoint of stable operation once the disturbance hasdisappeared. The linearization in this case is not appropriate.The study of transient stability is relatively complex andrequires various levels of more or less complex modeling.This complexity increases with the accuracy of the modelsand size of the network. One will have to use a software toolthat is dedicated to perform stability studies. However, thereis a simple method that allows a rapid prediction of transientstability with simplifying assumptions. This method isknown as the “method of equal areas”.The equal area method is used for the case of a machineconnected to an infinite bus, or a network with twomachines. The method is not applicable for complex multi-machine systems that require high levels of modeling andhigh precision.

2.2.4Case of a multi-machine systemThe stability analysis described above is based on the case ofa machine connected to a network of infinite power. Thecriterion of equal areas applies only to this case and to thecase of two machines. In addition, we have neglected theeffects regulators which have a significant impact in the


RES -2016 8

process of stability especially in the dynamic post-fault phaseand oscillations that follow. However, in larger networkselectrical, several hundreds of machines may be connected tothe system and these machines are of various types andtechnologies.

A stability study for a multi-machine systemrequires taking into account all the components of thesystem. For a better accuracy, the components and thegenerators in particular must be modeled in detail includingregulations. The study of stability becomes complex in suchreal cases. Several computer software tools are available forthese studies, and allow various levels of details in modeling.However, if we want to focus only on the very first momentsafter the occurrence of fault, it is always possible to simplifythe modeling and thus the complexity of calculation bytaking the same simplifying assumptions as those used ongenerators during the presentation of transient stabilityabove.

3. POWER SYSTEM STABILIZATION REVIEWThis section presents recent developments in power systemstabilizers for two area four machine system transmissionlines power stabilization.

Junbo Zhang; Chung, C.Y.; Shuqing Zhang;Yingduo Han [2], combined Frequency response and residuemethods to design a wide area damping controller whichobviates the need for identifying the open loop dynamicmodel. The method includes feedback loop selection, residuephase calculation and power system stabilizer (PSS) tuningusing modal decomposition control (MDC). The core issue,residue phase calculation, has been validated in a two-areafour-machine system and the whole procedure for wide areadamping controller design was performed in a real powersystem, China Southern Grid. Their results demonstrate thatthe method is feasible and effective in real power systems.

Hussein, T.; Shamekh, A [3], this work presents aDirect adaptive Fuzzy technique as a Power SystemStabilizer (DFPSS). The technique utilizes fuzzy systemsdirectly as a controller that incorporates fuzzy control rulesin the design of the control law. In direct adaptive control,the parameters of the controller are directly adjusted toreduce the norm of the output error that represents mismatchbetween the plant and the reference model. The mainobjective of the designed controller was to damp local andinter-area oscillations that normally occur following powersystem disturbance. The performance of thetechnique wasillustrated through simulation studies of a two-area four-machine system. They have also shown a comparativeanalysis between DFPSS and a well-tuned conventionalpower system stabilizer (CPSS). After comparative analysisthey have proven effectiveness of their technique forpractical power oscillation damping.

Qudaih, Y.S.; Mitani, Yasunori; Mohamed, T.H.[4], This work presents a new approach to deal with theproblem of robust tuning of power system stabilizer (PSS)and automatic voltage regulator (AVR) in multi-machinepower systems. The proposed method was based on a modelpredictive control (MPC) technique, to improve the stabilityof the wide-area power system with multiple generators and

distribution systems including dispersed generations. Thismethod provides better damping of power system oscillationsunder small and large disturbances even with the inclusion oflocal PSSs. The effectiveness of their approach wasdemonstrated through a two areas, four machines powersystem. A performance comparison between the proposedcontroller and some of other controllers was also carried outconfirming the superiority of the proposed technique. Thedeveloped algorithm can be successfully applied to largermulti-area power systems and do not suffer fromcomputational difficulties. The algorithm carried out usingMATLAB /SIMULINK software package.

Babaei, E.; Golestaneh, F.; Shafiei, M.; Galvani, S[5], Damping of electro-mechanical oscillations ininterconnected power systems is important to guaranteesecure and stable operation of the system. Power systemstabilizers (PSSs) are applied to damp these oscillations. Toovercome flows associated with conventional algorithm ofPSS design, a modified version of non-dominated sortinggenetic algorithm was used in this work, to regulate the PSSparameters. Additionally, fuzzy logic principle wasemployed to develop PSS in order to improve the stabilityand reliability of the power systems subjected todisturbances. Two-area (four-machine) power system wasconsidered as the case study. PSS parameters were obtainedfor four PSS connected to four generators. The effectivenessof their algorithm in damping the system oscillations duringoverall disturbances was evaluated on simulationbackground.

Alsafih, H. A.; Dunn, R. W [6], This workillustrates the robust performance of a new wide-area basedglobal power system stabilizer (GPSS) using fuzzy logiccontrol design approach. The superiority of fuzzy logic basedpower system stabilizers is illustrated first by comparing theperformance of local conventional stabilizers CPSS withlocal fuzzy logic based stabilizers (LFPSS). Then anadditional control loop is added to include a wide-area basedfuzzy logic global stabilizer (GFPSS). The LFPSS uses thegenerator speed deviation and the accelerating power as inputsignals to provide the stabilizing signals through theexcitation systems of the generators to which they areconnected to. On the other hand, the GFPSS is designed toact between coherent areas within a power system It uses thedifference in the average speed deviation between any twoconnected coherent areas and the deviation in the tie-lineactive power transfer between these coherent areas as inputcontrol signals. It provides additional stabilizing signals tothe generators within these coherent areas. The proposedcontrol scheme was tested using the four machines two-areaKundur test system. The obtained results show the robustnessand effectiveness of their scheme in enhancing systemstability.

Daryabeigi, E.; Moazzami, M.; Khodabakhshian,A.; Mazidi, M.H. [7], this work presents a new method todesign a Power System Stabilizer (PSS) using Smart BacteriaForaging Algorithm (SBFA). The proposed algorithm canconduct bacteria at a smart direction to decrease the costfunction better than conventional Bacteria Foraging (BFA)method. This algorithm not only considers social


RES -2016 9

intelligence, but also emphasizes the individual intelligenceof bacteria for finding a better nutrition path. This approachleads to have a higher convergence speed and also a betterperformance than BFA. Simulation results in Two-area four-machine power system show that SBFA improves thedynamic performance of the power system.

Babaei, E.; Galvani, S.; AhmadiJirdehi, M [8],Damping local mode and inter-area oscillations are twocritical issues for optimal power flow and stability in a powersystem. Power system stabilizers (PSSs) are applied to dampthese oscillations. Conventional algorithms of PSS design arebased on classical control theory and eigen values basedcalculation and involve numerous mathematical calculationswhich may not present an accurate solution. In this paper, aparticle swarm optimization (PSO) based PSS design issuggested which is based on time domain response of powersystem and also is robust for the overall credible disturbancesin power system. Two-area (four-machine 11-bus) powersystem is considered as the case study in this paper. PSSparameters are obtained for four PSS connected to fourgenerators. Finally, the effectiveness of the proposedalgorithm is evaluated for damping the system oscillationsduring overall disturbances. The simulation results illustratethis issue.

Hadidi, R.; Jeyasurya, B [9], in this paper, aframework for applying reinforcement learning (RL) to thedesign and control of power system stabilizers (PSSs) isproposed. A near-optimal coordinated design for severalPSSs is achieved using reinforcement learning. The objectiveof the control policy is to enhance the stability of a multi-machine power system by increasing the damping ratio of theleast-damped modes. An RL method called Q-learning isapplied to find a near-optimal control policy for controllingPSSs. With this control policy, the agent can change the gainof PSSs automatically in such a way that a predefined goal isnearly always satisfied. A modified Q-learning algorithm isproposed to enhance the convergence speed of theconventional algorithm toward a near-optimal policy. This isachieved by using selective initial state criteria instead ofchoosing the initial state randomly in each episode. Thevalidity of the proposed method has been tested on a two-area, four-machine power system using nonlinear time-domain simulation under severe disturbances.

Huaren Wu; Qi Wang; Xiaohui Li [10], A rotorspeed of remote generator is used as wide area measurementsignal in this paper. The signal is detected and transmitted bythe phasor measurement unit (PMU). The wide area signalinputs to the local power system stabilizer (PSS) to regulatethe generator excitation and enhancing the power systemdamping. The gains of the wide area controller aredetermined by solving linear matrix inequalities (LMI). TheLMI solving method is introduced for the wide area controlof power systems. A proportional plus derivative network isused to compensate the communication delay. Kundur's four-machine two-area system is used to test the performance ofthe wide area damping control. The simulation results showthat the wide area control can improve power systemstability.

Athanasius, G.X.; Pota, H.R.; Subramanyam,P.V.B.; Ugrinovskii, V [11], This paper addresses theproblem of designing decentralized power system stabilizersfor interconnected power systems using robust outputfeedback minimax control techniques. We have consideredthe problem of designing a robust controller for a generatorconnected to a multi-machine grid system. To make thecontroller robust against parameter variations around theoperating point, system parameter variations, due to loadchanges, are represented using integral quadratic constraints(IQCs). The interconnection effects from other machines inthe system and local uncertainties are also included in thecontroller design using IQCs. Generator speed is used as theoutput feedback variable for the controller, due to the factthat other generator states are difficult to measure. Thecontroller design is validated using a real-time digitalsimulation (RTDS) facility for a four-machine two-area testsystem.

Hunjan, M.; Venayagamoorthy, G.K. [12], Powersystem stabilizers (PSSs) are used to damp intra-area andinter-area oscillations in a power network. They provideeffective supplementary control by supplying auxiliarycontrol signals to the excitation system of the generators. Theproper tuning of PSSs has a significant influence on itseffectiveness in providing the required damping underdifferent operating conditions and disturbances. Variousalgorithms have been successfully implemented tosimultaneously design multiple optimal PSSs in powersystems. As the power network's operating conditionschange, the performance of PSSs degrade. Optimal PSSparameters obtained using bacteria foraging algorithm (BFA)have shown to successfully damp out system oscillationsduring disturbances for various operating conditions. Thispaper presents an artificial immune system based PSS designto adapt the optimal parameters of the PSSs. The innateimmunity to system oscillations is provided by the optimalPSS parameters while the adaptive immunity is provided byadapting the PSS parameters during transients. Theeffectiveness of the 'adaptive' optimal PSSs (APSSs) isevaluated on the two-area four-machine benchmark powersystem.

Dobrescu, M.; Kamwa, I [13], this paper describesthe performances of a new power system stabilizer, the fuzzylogic PSS (FLPSS). It is basically a PID (proportional-integral-derivative) type FLPSS with adjustable gains addedoutside in order to keep a simple structure. The FLPSS usesthe generator speed deviation as primary input from whichthe accelerating power is derived as a secondary input. Inorder to validate the FLPSS, it has been compared with tworeference stabilizers, the IEEE PSS4B and IEEE PSS2B fromthe IEEE Std 421.5. Conclusions are supported by a range ofsmall and large signal analyses, performed on a four machinetwo areas test system (with two configurations).

4. CONCLUSIONIn the present era, with growing power generation systems,security and stability of power generation systems becomesvery crucial part. In past so many blackout have beenobserved due to lack of power generation system stability.


RES -2016 10

As a consequence in large interconnected power systems,low frequency oscillations have an increasing importance.Low frequency oscillations include local area modes andinter-area modes. Inter-area modes of oscillations may becaused by the either high gain exciters or heavy powertransfer across weak tie line. The occurrence of the inter-areapower oscillations depends on various reasons such as weakties between interconnected areas, voltage level, transmittedpower and load. At time, the oscillations may continue togrow causing the instability of the power system. lots ofpower system stabilizers have been developed by theresearchers in the past few years, but the area is still open forthe efficient power stabilizer development which canefficiently able to handle the power oscillations withoutincreasing the system controller system complexity.Presently researchers are relying on the use of softcomputing techniques like fuzzy, neural network andadaptive neuro fuzzy etc. based advance controllers toefficiently handle power oscillation damping problem. Evenmore optimized controllers are now become popular which isa fusion of optimization techniques with soft computingbased controllers.

REFERENCES

1) PrabhaKundur, Power System Stability and Control, McGraw-Hill, Inc,1994.

2) Junbo Zhang; Chung, C.Y.; Shuqing Zhang; Yingduo Han, "PracticalWide Area Damping Controller Design Based on Ambient SignalAnalysis," Power Systems, IEEE Transactions on , vol.28, no.2,pp.1687,1696, May 2013.

3) Hussein, T.; Shamekh, A., "Direct Adaptive Fuzzy Power SystemStabilizer for a Multi-machine System," Computer Modeling andSimulation (UKSim), 2013 UKSim 15th International Conference on ,vol., no., pp.33,38, 10-12 April 2013.

4) Qudaih, Y.S.; Mitani, Yasunori; Mohamed, T.H., "Wide-Area PowerSystem Oscillation Damping Using Robust Control Technique," Powerand Energy Engineering Conference (APPEEC), 2012 Asia-Pacific ,vol., no., pp.1,4, 27-29 March 2012.

5) Babaei, E.; Golestaneh, F.; Shafiei, M.; Galvani, S., "Design anoptimized power system stabilizer using NSGA-II based on fuzzy logicprinciple," Electrical and Computer Engineering (CCECE), 2011 24thCanadian Conference on, vol., no., pp.000683, 000686, 8-11 May 2011.

6) Alsafih, H. A.; Dunn, R. W., "Performance of Wide-Area based FuzzyLogic Power System Stabilizer," Universities' Power EngineeringConference (UPEC), Proceedings of 2011 46th International, vol., no.,pp.1, 6, 5-8 Sept. 2011.

7) Daryabeigi, E.; Moazzami, M.; Khodabakhshian, A.; Mazidi, M.H., "Anew power system stabilizer design by using Smart Bacteria ForagingAlgorithm," Electrical and Computer Engineering (CCECE), 2011 24thCanadian Conference on, vol., no., pp.000713, 000716, 8-11 May 2011.

8) Babaei, E.; Galvani, S.; AhmadiJirdehi, M., "Design of robust powersystem stabilizer based on PSO," Industrial Electronics & Applications,2009. ISIEA 2009. IEEE Symposium on, vol.1, no., pp.325, 330, 4-6Oct. 2009.

9) Hadidi, R.; Jeyasurya, B., "Reinforcement learning approach forcontrolling power system stabilizers," Electrical and ComputerEngineering, Canadian Journal of, vol.34, no.3, pp.99, 103, summer2009.

10) Huaren Wu; Qi Wang; Xiaohui Li, "PMU-Based Wide Area DampingControl of Power Systems," Power System Technology and IEEE PowerIndia Conference, 2008. POWERCON 2008. Joint InternationalConference on, vol., no., pp.1, 4, 12-15 Oct. 2008.

11) Athanasius, G.X.; Pota, H.R.; Subramanyam, P.V.B.; Ugrinovskii, V.,"Robust power system stabilizer design using minimax controlapproach: Validation using Real-time Digital Simulation," Decision andControl, 2007 46th IEEE Conference on, vol., no., pp.2427, 2432, 12-14Dec. 2007.

12) Hunjan, M.; Venayagamoorthy, G.K., "Adaptive Power SystemStabilizers Using Artificial Immune System," Artificial Life, 2007.ALIFE '07. IEEE Symposium on, vol., no., pp.440, 447, 1-5 April 2007.

13) Dobrescu, M.; Kamwa, I., "A new fuzzy logic power system stabilizerperformances," Power Systems Conference and Exposition, 2004. IEEEPES, vol., no., pp.1056, 1061 vol.2, 10-13 Oct. 2004.

14) V. Vittal, “Consequence and Impact of Electric Utility IndustryRestructuring onTransient Stability and Small-Signal StabilityAnalysis”, Proceedings of the IEEE,Vol. 88, No. 2, Feb. 2000.


RES -2016 11

Mathematical Modeling of Thin Layer DryingProcess of Red Chili in Solar Dryer and Under

Open SunSitesh Kumar Pandey1, Richa Soni2, Abhishek Shrivastav3

1,3Mechanical Engineering Deptt., C.I.E.T, Raipur, IndiaMechanical Engineering, R.I.T, Raipur, India

[email protected], [email protected]

Abstract: Drying is an important technique to remove themoisture of the product. Various drying techniques areemployed to dry different agricultural products. Alltechniques have their own advantages and limitations. So,choosing the right drying techniques is very important in theprocess of drying. The raw agricultural products with 80-90% moisture content are brought down to equilibriummoisture content for keeping in the short or long termstorages. In this present work, I will experimentallyinvestigate the drying characteristics of red chili, one of themain ingredients used in almost every cooking in the world. Itis mainly cultivated in the Indian states of Andhra Pradesh,Karnataka and Maharashtra, Madhya Pradesh using thinlayer solar drying technique. In this study, direct-type naturalconvection solar dryer was developed and used in theexperiments. An experimental study was performed todetermine the thin layer drying process in a solar dryer andunder open sun with natural convection of red chili. Thedrying data were fitted into four different thin layer dryingmodels. The performance of these models is investigated bycomparing the coefficient of determination (R2), reduced chi-square(X2) and root mean square error (RMSE) between theobserved and predicted moisture ratio. Among these models,Wang and Singh model was found best to explain the thinlayer behavior of red chili.

Keywords- convection; coefficient of determination (R2);reduced chi-square (X2); root mean square error (RMSE)

I. INTRODUCTION

Quality of product depends on preservation of fruits,vegetables, and certain food items for which it is essentialto keep them for a long time without further deterioration.Up to 50% of the fruits and vegetable and 25% of theharvested food grain gets rotten in the tropical countries.During the harvest season either the high quantity of foodcannot be sold or the price is not good. Hence, drying is theoldest and simplest method of preserving fruits, vegetablesand food. It is defined as the process of moisture removaldue to simultaneous heat and mass transfer. But, drying inthe direct sun has the risk of damaging the completeharvest by rainfall. Also, risk of animals decreases thequality of the product.

By drying the farmer can have more income and for alonger time. Dried food generates a much better benefit onthe market than the unprocessed fruit. By drying, the tasteof the fruits is intensified and the nutrients are containedmore intensely and they taste more sweet and richer. Driedfruit and vegetables are free from preservatives. Dried foodhas a very good nutritive quality (vitamins etc).

The requirement of energy for drying can be supplied fromvarious resources namely electricity, fossil fuel, naturalgas, wood, forest residual and solar. According to thesource of energy used, there are different types of dryingprocess. Amongst all, the sun is the biggest carbon-freeenergy resource for human being. It is used to extendagricultural product life, improves quality, minimize lossesduring storage and lower transportation cost. As per thesolar energy used for drying, it can be categorized as opensun drying and solar drying.

In India, red chili is one of the important spices of meal,and is compulsory course in the kitchen. There is mainingredient in all cooking, while has high nutritional value.For red chili, it was found to contain very nutritious, withhigh vitamin C calcium, protein. The use of chili is not justfor adding food palatability, because there are a number ofstudies have shown that is very beneficial for human health.Furthermore, chili is a good source of antioxidants, beingrich in vitamins A and C, minerals and otherphotochemical, which an important source of nutrients inthe human diet. Traditionally, chili was dried in the opensun drying. The farmers expose their chilies to the open sunon a mat, earthen floor, and cemented floor or on a tin shed.In this method, drying cannot be controlled and a relativelylow quality dried product is obtained. Open sun dryingrequires large open space area, long drying times and verymuch dependent on the availability of sunshine, susceptibleto contamination with foreign materials. Otherwise insectsand fungi, which thrive in moist conditions, render themunusable. Moreover, in this method, a low quality driedproducts was obtained. Most agricultural commoditiesrequire drying process in an effort to preserve the quality ofthe final product including the drying temperature andduration of drying time. As an alternative approach to opensun drying, solar drying system is one of the most attractiveand promising applications of solar energy systems. It isrenewable and environmentally friendly technology, alsoeconomically available in most developing countries.

II. LITERATURE REVIEWAkpinar et al. [5] performed an experimental study andmathematical modeling of thin layer drying process of redpepper. The thin layer drying behavior of red pepper sliceswas experimentally investigated in a forced convectivedryer consisting of fan, heaters, drying chamber andinstrument for measurement. Experiments regarding dryingwere conducted at inlet temperatures of drying air of 55, 60and 70 C and at a drying air velocity of 1.5 m/s. Initial


RES -2016 12

moisture of the red peppers was 87.25% (wb), drying ofthe pepper continued until no further changes in their masswere observed. Final moisture content of the red pepperobtained was about 10% (wb). The time taken to reach10% moisture content from the initial moisture content atthe various drying air temperatures were found to bebetween 160 and 300 min. According to the results of themultiple linear regression analysis, among the 11 thinlayer-drying models, the approximation of the diffusionmodel could adequately describe the thin layer dryingbehavior of red peppers. Constant drying rate period wasnot observed under any of the test conditions of thisinvestigation, the red pepper drying process occurring inthe falling rate.

Madamba et al. [6] studied the thin layer dryingcharacteristics of slices of garlic. The thin-layer dryingcharacteristics of garlic slices (2-4 mm) were investigatedfor a temperature range of 50-90 ºC, a relative humidityrange 8-24%, and an airflow range (0.5-l) m/s. Initialmoisture content by garlic slice varied from 60-64% wetbasis and was determined in a vacuum oven under certaincondition. The effects of temperature, thickness, and RHand air velocity were analyzed using analysis of variance(ANOVA), and this clearly showed that temperature andslice thickness are significant factors in drying. In thisexperiment, it was concluded that most of the drying ofgarlic slices takes place in the falling rate period. Fourmathematical models available in the literature were fittedto the experimental data, among which the Page and thetwo-compartment models gave better predictions incomparison to single-term exponential and Thompson’smodel.

Rafiee et al. [7] studied thin layer drying properties ofsoybean. The experiments were conducted at inlet dryingair temperatures of 30, 40, 50, 60 and 70 ºC and at a fixeddrying air velocity of 1 m/s. The thin layer drying behaviorof soybean was experimentally investigated and themathematical modeling performed by using thin layerdrying models provided in the literature. The experimentalmoisture ratio was fitted to thirteen thin layer dryingmodels. Midilli [4] model gave the lowest RMSE and thehighest regression coefficient values. Hence, the Midillimodel was chosen to represent the thin layer drying ofsoybeans.

Akpinar [8] investigated suitable thin layer drying curvemodel for some vegetables and fruits. It is a study ofmathematical modeling of thin layer drying of potato,apple and pumpkin slices in a convective cyclone dryer. Heperformed experiment in convective cyclone dryer. Thedryer consists of fan, resistance and heating control system,air duct, drying chamber in cyclone type and measurementinstruments. In which Potato slices of 83% (wb), appleslices of 87% (wb), pumpkin slices of 93% (wb) averageinitial moisture content were dried to 10% (wb), 13% (wb),6% (wb), respectively, at temperatures of 60, 70 and 80 ºCin the velocities of drying air of 1 and 1.5 m/s. It wasnoticed that the Midilli– Kucuk model [9] gave the highestcorrelation coefficient ‘r’ and the lowest chi-square (X2) forall drying conditions.

III. EXPERIMENTAL PROCEDURE AND SETUP

A. EXPERIMENTAL SETUPFig. 3.1 shows the arrangement of the experimental setupof the solar dryer. The solar dryer consist of wooden block,insulating foam, aluminum sheet and 4mm glass plate. Thedryer consists of aluminum plate of thickness 0.5 mm.Aluminum plates is painted by black color because blackcolor enhances the absorbtivity of a material. Sheet ofaluminum is used as an absorber plate. Insulating materialfoam is provided at all side of the dryer to minimize theheat loss.

Fig. 3.1Pictorial view of the experimental setup of solardryer

B. EXPERIMENTAL PROCEDURESolar drying experiments were performed during periodsof Nov.-Dec 2015 in RIT Rewa, India.

The initial moisture content of freshly harvested red chiliwas determined by hot air oven method before carrying outthe drying experiment. 100 g of chili was dried in a hot airoven. On wet basis average initial moisture contentobtained from the experiment was 80.62%.

Fig.3.2. Flow diagram of thin layer natural convection solar

drying process

100 g fully ripe chili sample was taken and placed on thedrying tray of the solar dryer in a single layer and another100 g was taken for the open-sun drying. Experiment wasconducted at a regular interval of time of 1 hr. From fig.3.6 it is clear that fresh air enters into the solar dryer and ittakes heat available inside the dryer due to intense solarradiation .Air also extracts moisture from the chili and gets


RES -2016 13

humidified. Weight loss of the chili was measured at aregular interval of time of 1 hr.

Reduction in weight of the chili provided the amountmoisture removed during the drying process. Dryingprocess was continued till the moisture content of the chiliwas reduced up to 10% on wet basis. During experiment itwas clear that chili drying inside the drying chamber driedfaster comparison to chili drying under open sun.

C. MATHEMATICAL MODELING PROCEDURE OFDRYING CURVES

Mathematical modeling of the chili was done on the basisof drying characteristics obtained from the experimentalinvestigation. Thin layer drying requires some of theimportant parameters to get the drying characteristics ofthe product.

D. IMPORTANT PARAMETERSa) Moisture content

The quantity of moisture present in a material can beexpressed either on the wet basis or dry basis and also canbe expressed either as decimal or percentage. The moisturecontent on the wet basis is the weight of moisture presentin a product per unit weight of the undried material,represented as,

Mwb = (w0-wd)/w0 (1)

While the moisture content on the dry basis is the weightof moisture present in the product per unit weight of drymatter in the product and represented as,

Mdb = (w0-wd)/wd (2)

The moisture contents on the wet and dry basis are inter-related according to the following equation,

Mwb = 1- [1/Mdb+1] (3)

b) Equilibrium moisture content (Me)A crop has a characteristic water vapor pressure at aparticular temperature and moisture content. Theequilibrium moisture content is the moisture content atwhich the product is neither gaining nor losing moisture. Itis a dynamic equilibrium which changes with relativehumidity and temperature.

c) Moisture ratio (MR)Moisture ratio is one of the important criteria to determinethe drying characteristics of agricultural product. MR canbe determined according to external conditions. If therelative humidity of the drying air is constant during thedrying process, then the moisture equilibrium will beconstant too. In this respect, MR is determined as in Eq.(4).

MR= (Mt - Me) / (M0 – Me) (4)

If the relative humidity of the drying air fluctuatescontinuously, then the moisture equilibrium continuouslyvaries so MR is determined as in Eq.(5) given by Diamanteand Munro[26]

MR=Mt / M0 (5)

d) Drying ratesDrying rate is expressed as the amount of evaporatedmoisture over time.

Drying rate = (Mt+dt – Mt)/ dt (6)

Where Mt+dt and Mt (g water /g dry matter.h) are themoisture content at the moment t and the moisture contentat the moment t+dt, respectively.

E. DETERMINATION OF THE APPROPRIATE MODELMathematical modeling of the drying of chili often requiresthe statistical methods of regression and correlationanalysis. Thin layer drying equations require MR variationversus time‘t’. Therefore, MR data plotted with time t andcurve fitting was done for the models shown in Tab. 3.1.The validation of models can be checked with differentstatistical methods.

Table 3.1 Thin layer drying curve modelsModelno.

Modelname

Model References

1 Newton MR=exp(-kt) Mujumdar[30]

2 Page MR=exp(-ktn) Diamanteand Munro[31]

3 Hendersonand Pabis

MR=a exp(-kt) Zhang andLitchfield[32]

4 Wang andSingh

MR=1+at+bt2 Wang andSingh [33]

The coefficient of determination (R2) is one of the primarycriteria to select the best model compare with theexperimental data. In addition to R2, reduced chi-square(X2) and root mean square error(RMSE) are also beused to compare the goodness of the fit.

Reduced chi-square (X2)

=∑ ( , , ) (7)

Root mean square error (RMSE)-

RMSE = [∑ ( MR , − MR , ) ] (8)

Correlation coefficient (r2)-


RES -2016 14

r2= ∑ , , ∑ , ∑ ,∑ ( , ) ∑ , ∑ , ∑ ,(9)

IV. RESULT AND DISCUSSIONa) DRYING CHARACTERISTICS OF RED CHILI

The variation of moisture content of 300 g of red chili withdrying time is shown in Fig. 5.1 and Fig. 5.2. The finalmoisture content (10% wb) of the chili was achieved in 27h in solar dryer, while it was achieved in 38 h in open-sundrying. The open-sun drying took more time to reach thefinal moisture content due to slow drying rate. It isobserved that the drying of chili occurred only in thefalling rate period and the constant rate period wascompletely absent in the drying period.

Table 4.1 Experimental data of solar drying

Fig.4.1 Moisture content variation with drying time on wetbasis.

Fig.4.2. Moisture content variation with drying time ondry basis

The variation of drying rates with drying time is shownin Fig. 5.3. It is obvious that the drying rate in solardrying is faster than the open-sun drying, and thedrying rates decrease continuously with drying time.The drying rate is found to be fluctuating in nature bothin the solar drying and the open-sun drying, which ismainly caused by variation of solar radiation during theexperiment. During summer, solar radiation changesfrequently due to clouds in the sky. The drying rate in

Fig 4.3.Solar dryer with T1 and T2 hole

solar drying is faster than the open-sun drying till the first18 h of the drying hour. However, the drying rate in thesolar drying is almost same as the open sun drying in thelater part of the drying hour.

0

10

20

30

40

50

60

70

80

90

0 10 20 30 40

Moi

stur

e co

nten

t,wb(

%)

Drying time (hr)

solar drying

00.511.522.533.544.5

0 10 20 30 40Moi

stur

e co

nten

t,db(

g/g

of d

m)

Drying time (hr)

Open sun drying

TimeWeight of the

sampleWeight

reductionMoisturecontent

Moistureratio

(hr) (gm) (MR)

0 100 80.62 1.000

1 96.07 3.93 76.14 0.944

2 93.3 2.77 73.37 0.910

3 89.26 4.04 69.33 0.860

4 83.34 5.92 63.41 0.787

5 78.16 5.18 58.23 0.722


RES -2016 15

Table 4.2 Experimental data of open sun drying

b) TEMPERATURE MEASUREMENT OF INLETAND OUTLET AIR DURING EXPERIMENT

T1 – Inlet air temperatureT2 – Outlet air temperature

During experiment temperature at position 1 and position 2was obtained by placing the thermometer at that position.Temperature at the inlet and outlet depends upon the solarradiation intensity, higher the solar radiation intensityhigher the temperature of the air and vice versa.

Table 4.3 Temperature during one day experiment of solar

drying

During solar drying of red chili in our experimental setupinlet and outlet temperature obtained is shown in the abovetable. From the table it is clear that temperature at theoutlet reaches at its maximum level during mid day.Temperature depends upon the solar radiation intensity.During sunny days solar radiation intensity is maximumwhen sun is overhead.

Table 4.4 Temperature during one day experiment of opensun drying

Thermometer is used to measure ambient temperature.Ambient temperature is almost same during theexperiment.

c) MODELING OF MOISTURE RATIOACCORDING TO DRYING TIME FOR THIN LAYERDRYING

Open sun drying of Red chili

Fig. 4.4 Newton model fit

Time Inlettemperature (T1)

In

C

Outlettemperature (T2)

In

C

07:3027 30

8:3027 32

09:3028 34

10:3028 33

11:3029.5 39.2

12:3030 43.5

01:3030 37.5

02:3029.5 35.5

03:3029 34

04:3029 33

Time Ambient temperature (Tamb)

In

C

07:3027

8:3027.5

09:3028

10:3028

11:3029.5

12:3030

01:3030

02:3029.5

03:3029

04:3029

TimeWeight of the

sampleWeight

reductionMoisturecontent

Moistureratio

(hr) (gm) (MR)

0 100 2.36 80.62 11 97.64 1.74 78.26 0.9674742 95.9 2.94 76.52 0.9434933 92.96 3.06 73.58 0.9029744 89.9 3.09 70.52 0.8608015 86.81 1.36 67.43 0.818214


RES -2016 16

Fig. 4.5 Page model fit

Fig. 4.6 Wang and singh model fit

Fig. 4.7 Henderson and pabis model fit

Solar drying of red chili

Fig. 4.8 Newton model fit

Fig. 4.9 Page model fit

Table 4.5 Results of the fitting statistics of various thinlayer models for open sun drying

Modelno.

Modelname

Coefficientsand constants

RMSE

1 Newton k=0.0496 0.93831

0.00576

0.07589

2 Page K=0.01161,n=1.4785

0.98207

0.00167

0.04091

3 Henderson andPabis

a=1.09552,k=0.05474

0.94991

0.00468

0.0684

4 WanghandSingh

a=-0.03287,b=0.000171

0.99876

0.000115

0.01072


RES -2016 17

Fig. 4.11 Henderson and pabis model fit

Fitting of the four thin layer drying models has been donewith the experimental data of chili drying using open sunand solar drying. Models of drying which were fitted withthe experimental data of drying were the Newton, Page,Henderson and Pabis and Wang and Singh model.Experimental data for drying were fitted to the model ofdrying in the form of changes in moisture ratio versusdrying time. These drying models were fitted intoexperimentally obtained moisture ratio versus drying timecurve with the help of origin pro 8.5 software. The resultsof the statistical analysis undertaken on these models foropen sun drying and solar drying are given in table 4.3 and4.4. The models were evaluated based on R2, X2 andRMSE .The model which was having the higher value ofR2 and the lower values of X2 and RMSE was selected tobetter estimate the drying curve.

For the thin layer open sun drying of red chili Wang andSingh model gave R2 =0.99876, X2 =0.000115 andRMSE=0.01072. For the thin layer solar drying of red chiliWang and Singh model gave R2 =0.99646, X2 =0.00032and RMSE=0.01802.Wang and Singh model is consideredas best model for thin layer solar drying and open sundrying of red chili.

Fig. 4.10 Wang and singh model fit

Thin layer drying modelling equation for open sun dryingof red chilli is given by

MR=1+(-0.0327)t+0.000171t2

Thin layer drying modelling equation for solar drying ofred chilli is given by

MR=1+(-0.05515)t+0.0007045t2

Fig.4.12. A view of freshly harvested red chilli

Table 4.6 Results of the fitting statistics of various thin

layer models for solar drying.

Modelno.

Modelname

Coefficientsand

constants

RMSE

1 Newton

k=0.07696 0.96307

0.00339

0.05822

2 Page K=0.03325,n=1.31756

0.98626

0.00126

0.03551

3 HendersonandPabis

a=1.07169,k=0.08277

0.96916

0.00283

0.05321

4 WanghandSingh

a= -0.05515,b=0.0007045

0.99646

0.00032

0.01802


RES -2016 18

Fig. 4.13. A view of red chili after open sun drying

Fig.4.14 A view of red chili after solar drying

In solar drying good quality red chili was found and colorwas not degraded after drying. In open sun drying red chiliquality was not as good and color quality was alsodegraded.

REFERENCES

[1] Sharma, A., Chen, C.R., Lan, N. V, Solar-energy drying systems: Areview, Renewable and Sustainable Energy Reviews, 13 (2009) pp.1185-1210.

[2] Singh, S., Singh, P.P., Dhaliwal, S.S. Multi-shelf portable solar dryer.Renewable Energy, 29 (2004) 753–65

[3] Sreekumar, A, Manikantan, P.E., Vijayakumar, K.P. Performance ofindirect solar cabinet dryer’ Energy Convers Manage, 49(2008)1388–1395.

[4] Seveda M S,JhajhariaD.Design and performance evaluation of solardryer for drying of large cardamom (amomumsubulatum).Journal ofrenewable and sustainable energy 4,063129 (2012).

[5] Akpinar, E.K., Bicer, Y., Yildiz, C., .Thin layer drying of red pepper.Journal of Food Engineering 59(2003), 99–104.

[6] Madamba, P.S., Driscoll, R.H., and Buckle, K.A. Thin-layer dryingcharacteristics of garlic slices. Journal of Food Engineering.29(1996)75–97

[7]Rafiee Sh., Sharifi M, JafariA ,Mobli H., and Tabatabaeefar A..ThinLayer Drying Properties of Soyabean (ViliamzCultivar).Journal of J.Agric. Sci. Technol. (2009) Vol. 11: 289-300.

[8] Midilli, A. and Kucuk, H. Mathematical Modelling of Thin LayerDrying of Pistachio by Using Solar Energy. Energy ConversionManage.44(7) (2003): 1111–1122

[9] Akpinar, E.K. Determination of suitable thin layer drying curvemodel for some vegetables and fruits. Journal of Food Engineering73,(2006) 75–84.

[10]Rao G.V,MandeS,Kishorev V.V.N. Study of drying characteristics oflarge-cardamom. Journal of Biomass and Bioenergy 20 (2001)37-43.

[11]Aghbashlo M, Kianmehr M.H., Arabhosseini A. Modeling of thin-layer drying of potato slices in length of continuous band dryer.Journal of Energy Conversion and Management 50 (2009) 1348–1355.

[12]Garavand A.T.,Rafiee s., Keyhania A, Mathematical Modeling ofThin Layer Drying Kinetics of Tomato Influence of Air DryerConditions. International Transaction Journal of Engineering,Management, & Applied Sciences & Technologies (2011).

[13]Panchariya, P.C., Popovic, D., Sharma, A.L. Thin layer modelling ofblack tea drying process. Journal of Food Engineering (52) (2002),349-357

[14]Meisamiasl E., Rafiee S. “Mathematical Modeling of Kinetics ofThin-layer Drying of Apple.Agricultural Engineering International:the CIGR. journal.Manuscript 1185.Vol. XI.(2009).

[15]Ertekin, C., and Yaldiz, O. Drying of eggplant and selection of asuitable thin layer drying model.Journal of FoodEngineering.63(2004) 349–359. Madamba, P.S.Thin layer dryingmodels for osmotically pre-dried young coconut. DryingTechnology.21(2003).1759–1780.


RES -2016 19

Big Data Analysis and Renewable EnergyExtraction: A New Path For Research

Chitra Sharma1, C. K. Dwivedi2

1Dept. of EEE, Kruti Institute of Technology and Engineering, Raipur, CG, India2Dept. of EEE, Columbia Institute of Engineering Technology, Raipur, CG, India


Abstract— Big data analysis is future for the upcominginnovation and invention. In this paper a renewableapplication of big data analysis have been presented usingsolar PV power. United Nation has now considered big dataanalytics for various healthcare applications applied withhealth care analysis. In this paper a survey of application ofbig data analysis has been presented. Open source softwareRETScreen is used to monitor the data taken from NASA.This data will help to design the solar power system withMPPT for stand alone and grid connected system.

Keywords-Big Data analysis, solar power, renewableenergy, Hadoop

I. INTRODUCTION

Big Data bring new opportunities to modern society andchallenges to data scientists. In the age of digital world, allthe information is being stored in the cloud. Also withdigitization of various power sector has increased the flowof sensor data and the cloud computing. The first definitionof Big Data comes from Merv Adrian, “Big data exceedsthe reach of commonly used hardware environments andsoftware tools to capture, manage, and process it within atolerable elapsed time for its user population”[1]. In year2014 IBM has introduced the HyRef (Hybrid RenewableEnergy Forecasting) system for the wind and solar power.By utilizing local weather forecasts, HyRef can predict theperformance of each individual wind turbine and estimatethe amount of generated renewable energy. This level ofinsight will enable utilities to better manage the variablenature of wind and solar, and more accurately forecast theamount of power that can be redirected into the power gridor stored. It will also allow energy organizations to easilyintegrate other conventional sources such as coal andnatural gas.

However, this area of research is very new andpromising and carries lot of financial potential. Manyresearch communities are trying to develop a sophisticatednetwork comprised with big data analytics. In addition,solar and wind power forecasting is developed according tothe soft-modeling of the system with the sensor data withvarious level of weather analysis. Several researches haveproven the effectiveness of big data in health care andfinancial forecasting. International Data Corporation havebeen paid by Xerox, Univac and Burrough companies forpredicting future of computer technology in year 1999.[2]Various survey and analysis has been done by the IDC from1964 to till date.

In addition, Min Chen in year 2014 [3] presented adetailed survey on the big data. In this work state-of-the-art

of big data is presented with its pros and cons. Now a days,big data related to the service of Internet companies growrapidly. For example, Google processes data of hundreds ofPetabyte (PB), Face book generates log data of over 10 PBper month, Baidu, a Chinese company, processes data oftens of PB, and Taobao, a subsidiary of Alibaba, generatesdata of tens of Terabyte (TB) for online trading per day.

Figure 1 Stand alone Photo voltaic System

In 2010, Apache Hadoop defined big data as “datasetswhich could not be captured, managed, and processed bygeneral computers within an acceptable scope.” On thebasis of this definition, in May 2011, McKinsey &Company, a global consulting agency announced Big Dataas the next frontier for innovation, competition, andproductivity.

II. DEVELOPMENT OF BIG DATA

In the late 1970s, the concept of “database machine”emerged, which is a technology specially used for storingand analyzing data. With the increase of data volume, thestorage and processing capacity of a single mainframecomputer system became inadequate. However, manychallenges on big data arose. With the development ofInternet services, indexes and queried contents were rapidlygrowing. Therefore, search engine companies had to facethe challenges of handling such big data. Google createdGFS [4] and Map Reduce [5] programming models to copewith the challenges brought about by data management andanalysis at the Internet scale.


RES -2016 19







I. INTRODUCTION










RES -2016 19







I. INTRODUCTION










RES -2016 20

Now big data is being applied for the future location ofcharging station and the IoT (Internet of Thing), V2G(Vehicle to Grid) communication etc. Image processing,voice processing and signal analysis tool has shown thebenefit of big data from various sources.

III. APPLICATION OF BIG DATA IN RENEWABLEENERGY

Continuous data have been generated from the weatherforecasting network for upcoming 24hr system. In this caseone system applied for distributed generation for solar orwind is needed to capture and process these data. Severalassessments are always required before utilizing it at fieldlevel. In figure 1 fundamental arrangement of solar PV cellis given for a standalone system with dc dc converter to getregulated power supply. In fig 2 a collected data has beenshown from the NASA for Raipur region. The above data isone day forecasting for complete year.

Figure 3 photovoltaic system (a) Basic Circuit and (b)Characteristic Curve

ph

s

exp 1 (1)

Where ;

I = Current to the load

I = Photo current

I = Reverse saturation current of the diode

q = Electron charge

V= Voltage across the diode

K= Boltzmann con

S Sph S

sh

q V R I V R II I I

KT R

s

sh

stant

T= Junction temperature

=Ideality factor of the diode

R =Series resistors

R = Shunt resistor

Equation mentioned above is fundamental equation ofcurrent. Since solar power behaves as the current sourcefor different irradiance capacity and different irradiancelevel. In figure 1 operation of DC DC converter is based onthe MPPT algorithm from the various sensor data asvoltage or current. If the solar irradiance is available fromthe data forecasting and big data assessment, algorithmbecomes more effective. Various topologies of DC-DCconverter have been proposed till date to provide the bestassessment of power and to get the high efficiency powerconversion. It is found that the DC-DC conversionefficiency is very high for the application other than the PVpanel. Since the effect of partial shading and variation intemperature does show the variation on the actual voltagelevel and the DC DC converter does show its optimal

Figure 2 RETScreen data collected for the Raipur, Chhattisgarh


RES -2016 20





ph

s

exp 1 (1)

Where ;


I = Photo current


q = Electron charge


K= Boltzmann con

S Sph S

sh

q V R I V R II I I

KT R

s

sh

stant



R =Series resistors

R = Shunt resistor




RES -2016 20





ph

s

exp 1 (1)

Where ;


I = Photo current


q = Electron charge


K= Boltzmann con

S Sph S

sh

q V R I V R II I I

KT R

s

sh

stant



R =Series resistors

R = Shunt resistor




RES -2016 21

behavior for such variation. Fig-4 shows a setup for 3kWPV array with partial cloud shading and the partialvariation of the parameter.

Figure 4 A common setup for the 3kW PV system

In figure 4 effect of cloud makes a great impact on theconversion efficiency of converter. To achieve good levelof power conversion one need to consider other variable ofsolar system.

Figure 5 P-V characteristic for different irradiance

Fig-5 shows the PV characteristic of a system for differentirradiance level during a day. In a day irradiance variesaccordingly for the various position of sun. There arevarious mathematical expressions for position of sun forthe day and night.

Figure 6 P-V characteristic for different Temperature.

Figure 7 Log characteristic of Voltage at different battery current.

In fig 6 power vs output voltage is presented for thedifferent temperature level. In this curve it is very muchclear that the temperature variation greatly affect the outputvoltage of solar PV panel.

Figure 8 Optimal application of PV using MPPT.

It is noteworthy that effectiveness of any power trackingalgorithm is function of different variable as voltagecurrent and power. However, for better control andextraction of power from the solar panel requires the nearto exact estimation of temperature, irradiance and the windspeed.

IV. CONCLUSION

This paper ideate the future of renewable energy sourcesfor the stand alone and large scale system where the


RES -2016 22

analysis for control action is performed using the Big dataanalysis. Using big data analysis one can estimate the windspeed accordance with the irradiance and temperature level.Big data can be assessed at various levels to get bettercontrol and the fast forecasting for power scheduling.

REFERENCES

[1] Gantz J, Reinsel D (2011) Extracting value from chaos. IDC iView,pp 1–12

[2] Fact sheet: Big data across the federal government (2012).http://www.whitehouse.gov/sites/default/files/microsites/ostp/bigdata fact sheet 3 29 2012.pdf

[3] Chen, Min, Shiwen Mao, and Yunhao Liu. "Big data: A survey."Mobile Networks and Applications 19.2 (2014): 171-209.

[4] Ghemawat S, Gobioff H, Leung S-T (2003) The google file system.In: ACM SIGOPS Operating Systems Review, vol 37. ACM, pp 29–43

[5] Dean J, Ghemawat S (2008) Mapreduce: simplified data processingon large clusters. Commun ACM 51(1):107–113

[6] I. S. Jacobs and C. P. Bean, “Fine particles, thin films and exchangeanisotropy,” in Magnetism, vol. III, G. T. Rado and H. Suhl, Eds.New York: Academic, 1963, pp. 271–350.

[7] S. R. Wenham, M. A. Green, and M. Watt, Applied Photovoltaics.Sydney, Australia: Univ. New South Wales, 1994.

[8] J. Larminie and A. Dicks, Fuel Cell Systems Explained. New York:Wiley, 2000.

[9] D. A. J. Rand, R. Woods, and R. M. Dell, Batteries for ElectricVehicles. New York: Wiley, 1998.

[10] M. Calais, J. M. A. Myrzik, and V. G. Agelidis, “Inverters for singlephase grid connected photovoltaic systems—overview andprospects,” in Proc. 17th PV Solar Energy Conf. and Exhibition,Münich, Germany, Oct. 2001.

[11] T. Noguchi, S. Togashi, and R. Nakamoto, “Short-circuit pulsebased maximum-power-point tracking method for multiple

[12] photovoltaic-and-converter module system,” IEEE Trans. Ind.Electron., vol. 49, pp. 217–223, Feb. 2002.

[13] L. M. Tolbert and F. Z. Peng, “Multilevel converters as a utilityinterface for renewable energy systems,” in IEEE PES SummerMeeting, Seattle, WA, 2000.

[14] M. Calais and V. G. Agelidis, “Multilevel converters for singlephase grid connected photovoltaic systems-an overview,” in Proc.IEEE International Symp. Industrial Electronics, Pretoria, SouthAfrica, 1998.

[15] J. H. R. Enslin, M. S. Wolf, D. B. Snyman, and W. Swiegers,“Integrated photovoltaic maximum power point tracking converter,”IEEE Trans. Ind. Electron., vol. 44, pp. 769–773, Dec. 1997.

[16] R. Giri, R. Ayyana, and N. Mohan, “Common duty ratio control ofinput series connected modular dc–dc converters with active inputvoltage and load current sharing,” in Proc. 18th Annu. IEEE AppliedPower Electronics Conf. Expo. (APEC’03), vol. 1, Feb. 2003, pp.322–326.

[17] G. Walker, “Evaluating MPPT converter topologies using aMATLAB PV model,” J. Elect. Electron. Eng. Australia, vol. 21,pp. 49–56, 2001.

[18] F. Hamma, T. Meynard, F. Tourkhani,and P. Viarouge, “Characteristics and designof multilevel choppers,” in Power ElectronicsSpecialists Conf. (PESC’95), vol. 2, 1995, pp.1208–1214.

[19] D. A. J. Rand, R. Woods, and R. M. Dell, Batteries for Electric[20] Vehicles. New York: Wiley, 1998.

BIBLOGRAPHYChitra Sharma, Completed her bachelorof Engineering in Electrical & ElectronicsEngineering, in year 2012, fromChhattisgarh Swami Vivekananda

Technical University. She is pursuing her Masters ofTechnology in Power System and Control from RKDFCollege of Engineering, affiliated to RGPV. Currently, sheis with Dept. of Electrical & Electronics Engineering,KITE-Raipur. Her area of interest is Testing-Commissioning of Electrical Equipments, High VoltageEngineering, Power System and Apparatus, and RenewableEnergy sources.

Chandra Kant Dwivedi received the B.Sc (Engg.) Degreein Electrical Engineering from B.I.T. Sindri, RanchiUniversity, India in 1971 & M.E. High VoltageEngineering from RTM University, Nagpur. He hastelescopic experience of more than 32 years in the field ofElectrical testing & commissioning of electricalequipments, H.T. & L.T.switch gears at floor level insteel industries as well as 9 years of teaching experienceto UG Students. He guided the team of electricalengineers for testing and commissioning of electrics atBokaro, Bhilai, Rourkela and Visage steel plant. SinceFeb, 2008 to October 2011, he was Sr.Lecturer inElectrical & Electronics Engineering at DIMAT, RAIPUR(C.G). From November 2011 to March 2013 he wasassociated with Columbia Istitute of Engineering &Technology, Raipur as Head of Department, EEE. FromJuly 2013 to June’14 he was HOD of ElectricalEngineering at MGEC Jaipur, Rajastan. From July’14 toJune’15 he was Associate Professor cum HOD of EEE atSBCET Jaipur, Rajasatan. At present, he is working asHOD of EEE in Columbia Institute of Engineering &Technology, Raipur. His main areas of interest are Highvoltage engineering, Testing & commissioning ofelectrical installations. Several research papers have beenpublished in International Journals. He is a member ofInstitute of Engineers, India & IEEE.


RES -2016 23

Optimal Location and Sizing of DistributedGeneration using Sensitivity Indices

S. K. Das1, G. Manikanta2, Srivatsa M3

CIET, RAIPUR, Chattisgarh, [email protected]

Abstract: In this paper optimal location of distributedgeneration in a radial distribution network using voltagestability and loss Sensitivity indices is done. Also, theoptimal size of the DG for all the buses are obtained inthe IEEE 28 bus radial distribution network. The resultsare obtained using MATLAB programming.

Keywords: Distributed generation (DG), optimal location,loss sensitivity index, voltage sensitivity index, optimalsize.

1 INTRODUCTIONThe power demand in the present day is rapidly

increasing. This is resulting in increasing pressure on thepower utilities to supply more and more power to the everincreasing load demand. This has resulted in looking fornew methods to generate power and supply the power whilereducing power losses in distribution network.

One such method is to use distributed generation.In distributed generation, the power is generated at thelocation of load by either using renewable energy sourceslike solar panel, wind mill, etc. or non-renewable sourceslike diesel generators. As the power is being generated nearthe load, the distribution losses reduces considerably andthe stress on the power utilities also reduces. DG alsoreduces environmental impact of power generation andimproves system reliability. These are the reasons why DGis gaining so much of importance now. The inclusion of DGin distribution network has encouraged consumers to userenewable energy technologies like solarcell panel asdistributed generation. Though DG increase reliability,reduce losses, saves energy and is cost effective, it suffersfrom some disadvantages of isolated power qualityfunctioning and voltage control problems.

Here the optimum location for the DG is obtainedby sensitivity analysis and loss sensitivity indices afterwhich the optimal sizes of the DG at each bus is obtained.This is done by using load flow calculation for a 28 busIEEE distribution network. The resulting correspondinglosses after placing optimal DG sizes at each bus aretabulated.

The objective function is the average power at thegenerator. In [4], exact loss formula is used to find theoptimum sizing of DG in each bus. This method givesaccurate losses for corresponding DG sizes and takes lesstime for computation.In this paper, the optimal sizes are obtained and optimallocation is obtained based on those DG sizes, through themethod in [4].

Here in section 1, load flow for a 28 bus radialdistribution network is carried out. In section 2,3 & 4, theoptimal location and sizing of DG is carried out. In section5, the results are tabulated.

2. LOAD FLOW FOR DISTRIBUTION NETWORKHere, for the 28 bus radial network BIBC, BCBV techniqueis used for load flow[6]. A brief explanation of thistechnique is as follows.

For a bus i, the complex power is given bySi = Pi + j Qi = Vi Ii

* (1)Thus the corresponding equivalent current injection is givenby

Ii = (( Pi + j Qi) / Vi)* (2)

Here Pi and Qi are active and reactive powerspecified in ith bus. Vi is the voltage at the ith bus. Thebranch current B is calculated with the help of BIBCmatrix. The BIBC matrix is obtained from relation betweenbus current injections and the branch currents. The elementsof BIBC matrix consists of ‘1’s and ‘0’s.

[B]nbΧ1 = [BIBC]nbΧ(n-1) [I](n-1)Χ1 (3)

Here nb is the number of branches in the systemand n is the number of buses.

The bus voltages at different buses are obtainedfrom the voltage at substation bus, branch currents and thebranch impedence. Thus another matrix representingrelation between drops in branch connecting substation busto different buses and the impedences of the correspondingbranches is formed and is given by

[ΔV] nbΧ1 = [BCBV] (n-1)Χnb [B] nbΧ1 (4)

This expression can be written in terms of bothBIBC and BCBV as

[ΔV] nbΧ1 = [BCBV] [BIBC] [I] (5)

These voltage drop values are then subtracted fromthe voltage value of the substation bus voltage to get thecorresponding bus voltage value.

This process can be continued iteratively withproper convergence criteria to obtaine the accurate voltagemagnitudes at different buses. The result for this load flowproblem are given in the tabular column.

Distribution networks


RES -2016 24

The data about the systems along the single linediagram are as given below with 100 MVA as base and firstbus as substation bus.

Fig.1 : Single line diagram for IEEE 28-bus radialdistribution network.

Table 1: The bus details for 28 bus radial distributionnetwork

Sending Receiving P QBus Bus R(Ω) X(Ω) (KW) (KW)

1 2 1.8216 0.7580 140 902 3 2.2270 0.9475 80 503 4 1.3662 0.5685 80 504 5 0.9180 0.3790 100 605 6 3.6432 1.5160 80 506 7 2.7324 1.1370 90 407 8 1.4573 0.6064 90 408 9 2.7324 1.1370 80 509 10 3.6432 1.5160 90 5010 11 2.7520 0.7780 80 5011 12 1.3760 0.3890 80 4012 13 4.1280 1.1670 90 5013 14 4.1280 0.8558 70 4014 15 3.0272 0.7780 70 4015 16 2.7520 1.1670 70 4016 17 4.1280 0.7780 60 3017 18 2.7520 0.7780 60 302 19 3.4400 0.9725 70 4019 20 1.3760 0.3890 50 3020 21 2.7520 0.7780 50 3021 22 4.9536 1.4004 40 203 23 3.5776 1.0114 50 3023 24 3.0272 0.8558 50 2024 25 5.5040 1.5560 60 306 26 2.7520 0.7780 40 2026 27 1.3760 0.3890 40 2027 28 1.3760 0.3890 40 20

After finding out the power output of kite windgenerator, the optimal location and Sizing of the kite windgenerator is to be done. This is done using voltage stabilityindex, loss sensitivity factors and the optimal sizing formuladerived from exact loss formula as below.

3. OPTIMAL LOCATION USING VOLTAGESTABILIY INDEX

From [4] , the bus stability index for Distribution networksis given asSI(r) =2V1

2V22-V2

4-2V22(PR+QX)-Z2(P2+Q2)

(6)Here, V1 and V2 are the voltage at sending end and

receiving end of a line respectively. P,Q are the real andreactive power sent through the line. R, X are the resistanceand reactance of the line. This equation is used to find the

stability index of each node. The node at which the value ofstability index is minimum, is considered to be the mostsensitive. The values of the stability indices are tabulated inthe results.

4. OPTIMAL LOCATION USING LOSSSENSITIVITY INDEX

The loss sensitivity index gives the indication that a bus issuitable for placing a DG. It is mainly used in DGallocation. For calculating the loss sensitivity indiex, theexact loss formula for real power loss is considered[4]. Thisis given by

PL = αij(PiPj + Qi Qj) + βij (Qi Pj – Pi Qj)](7)

Where :αij = (rij cos(δi - δj))/ Vi Vj (8)βij = (rij sin(δi - δj))/ Vi Vj (9)

The sensitivity factor of real power loss with respect to realpower injection from DG is given by

αi = 2 αij Pj - βij Qj ) (10)the buses are ranked in descending order of their values ofsensitivity factors. The buses with high values of sensitivityfactor are first studied for DG location.

Table2:Total power loss,Total Real & Reactive Power

5. DG SIZINGThe optimum DG sizes can be determined by using the realpower loss sensitivity factors[4]. At minimum losses, therate of change of losses with respect to injected powerbecomes zero.

αij Pj - βij Qj ) = 0 (11)

PDGi = PDi + [βii Qi - αij Pj - βij Qj )]/ αii

(12)Here PDGi = real power injection from DG placed at node iPDi = load demand at node i.DG size other than the values obtained by above relationwill lead to higher losses.

6. RESULTSAfter finding the voltage stability indices, the most

sensitive bus in the network is obtained as Bus number 18which is having the lowest value. From the plots of thevoltage stability indices it is clear that 18th bus is the mostsensitive bus to voltage instability. Here the total loss andtotal real and reactive powers are as given in table.

From the values of voltage stability indices bus no.18 is having the lowest value. From the loss sensitivityvalues the bus no. 13 is having highest value resulting infirst preference for placing the DG.

Total loss innetwork=P(loss)

Total real Power = P(total)

Total reactive Power =Q(total)

108.03 KW

1.9000 MW

1.0700 MVAR


RES -2016 25

As per the optimal sizing formula, the DG optimalsizes at each bus and the corresponding losses in the systemwhen the optimal sizes of the DG are placed at respectivebuses are tabulated. The bus at which, when the optimumsize of the DG is placed, results in minimum total loss inthe system, is considered to be the optimal location of theDG having optimal size. This optimal location is based onminimisation of system losses.

Fig.2: voltage magnitudes and voltage sensitivity indices ateach bus

Table3:voltage magnitudes with Index valuesBusno.

Voltagemagnitude

Voltagestabilityindices

Losssensitivityfactor

12345678910111213141516171819202122

1.00000.98920.97830.97270.96910.95610.94800.94410.93750.92970.92480.92270.91750.91350.91110.90940.90800.90750.98720.98660.98590.9854

0.99800.95600.91530.89450.87980.83410.80700.79320.77070.74580.73100.72310.70720.69530.68820.68290.67900.65410.94930.94680.94380.9415

-0.03500.04060.05120.07250.08390.11140.11990.11940.15220.14600.15080.18350.15180.15720.16080.14090.14200.02230.01700.01830.0157

232425262728

0.97670.97580.97490.95520.95480.9547

0.90930.90490.90120.83210.8310

-

0.02810.02970.03730.04370.04430.0446

As per the values of losses, bus no. 11 is havingminimum loss value when its optimum size DG is placed atit. Thus bus no. 11 is the optimum location of DG in termsof loss minimisation.

Table4:Optimal Sizes with lossesBusno.

OptimumSizes(MW)

Correspondinglosses(KW)

2345678910111213141516171819202122232425262728

1.70501.61131.50101.40051.26701.13581.08401.01580.89930.85560.82470.71880.68370.63400.59110.55100.51380.68300.58100.40740.26580.91380.65620.39851.06320.95910.8701

89.230773.773967.481664.138051.962347.046445.049842.230240.472939.439539.591741.740843.016345.208547.626951.045053.947698.8215

100.0112101.8686103.7337

86.750491.619996.806961.159665.007568.4748

The loss in the system, when optimal size of the DG isplaced at each bus is plotted as below.

7. CONCLUSIONFrom the results obtained after running the load

flow with sensitivity analysis and observing the voltagemagnitudes it can be concluded that when distributedgeneration is located at the most sensitive bus, the voltageprofile improves indicating less stress on the powergeneration. Also placement of optimal DG size at Optimallocation in terms of loss minimization results in minimumpower loss in the system.


RES -2016 26

Fig. 3: Plot of Powerloss vs Bus no.

REFERENCES[1] A. W. Manyonge, R. M. Ochieng, F. N. Onyango and J. M.

Shichikha. “Mathematical Modelling Of Wind Turbine in a WindEnergy Conversion System: Power Coefficient Analysis”. AppliedMathematical Sciences, Vol. 6, 2012, no. 91, 4527-4536.

[2] S. Ghosh and D. Das ,”Method for load flow solution for distributionnetwork”. IEE Proc.-Gener. Transm. Distrib.. Vol. 146, No. 6,November 1999.

[3] Jen-Hao Teng, Member, IEEE,” A Direct Approach for DistributionSystem Load Flow Solutions. ” IEEE transactions on powerdelivery, vol, 18, No. 3, July 2013.

[4] A. Parizad, A. Khazali, M. Kalantar,” Optimal Placement ofDistributed Generation with Sensitivity Factors Considering VoltageStability and Losses Indices”. Proceedings of ICEE 2010, May 11-13,2010.


RES -2016 27

Application of Brushless D.C. Motor for SpeedControl, Using Six Pulse Operations

Namrata Gupta1, Mahesh Ahuja2, C.K.Dwivedi3

1,3C.I.E.T Raipur, Chattisgarh, India2MATS University, Raipur, Chattisgarh, India

Abstract :This paper presents speed control of permanentmagnet brushless dc motor using 6 Pulse operations.According to the input command, feedback and thecontrol algorithm, the pulse for each phase are generatedand is given to the IGBT driver. The output of the driveris 6 independent pulses that have to be given to thecorresponding gates of the six IGBTs power switches usedin the three-phase bridge inverter whose output is givento the stator of the Brushless DC Motor. The completesystem model is simulated in Psim environment.

Keywords - Brushless DC Motor, Pulse Operation., DC drive

I. INTRODUCTIONBrushless DC motors are also known as electronicallycommutated motors. Electronic Commutated motors aresynchronous electric powered by direct current (DC)electricity and having electronic commutation systems, ratherthan mechanical commutator and brushes. The brushless DCmotor can be envisioned as a brush DC motor turned insideout, where the permanent magnets are on the rotor, and thewindings are on the stator. As a result, there are no brushesand commutator in this motor, and all of the disadvantagesassociated with the sparking of brush DC motors areeliminated.This motor is referred to as a "DC" motor because its coilsare driven by a DC power source which is applied to the

various However; "BLDC" is really a misnomer, since themotor is effectively an AC motor. The current in each coilalternates from positive to negative during each electricalcycle. The stator is typically a salient pole structure which isdesigned to produce a trapezoidal back-EMF wave shapewhich matches the applied commutated voltage waveform asclosely as possible. However, this is very hard to do inpractice, and the resulting back-EMF waveform often looksmore sinusoidal than trapezoidal. For this reason, many of thecontrol techniques used with a PMSM motor (such as FieldOriented Control) can equally be applied to a BLDC motor.Brushless DC motors use an electronic controller whichregulates and controls the voltage and current to the fieldcoils. The electronic module or controller uses discretedevices and amplifiers to establish and deliver the desiredvoltage/current. Advantages of a brushless motor include,increased reliability, longer life, elimination of sparks fromthe commutator, reduced friction, precision voltage/currentapplied to the field coils, faster rate of voltage and current.

A. Anatomy of a BLDC

Figure 1 is a simplified illustration of BLDC motorconstruction. A brushless motor is constructed with apermanent magnet rotor and wire wound stator poles.Electrical energy is converted to mechanical energy by themagnetic attractive forces between the permanent magnet.

rotor and a rotating magnetic field induced in the woundstator poles. In this example there are three electromagneticcircuits connected at a common point. Each electromagneticcircuit is split in the center, thereby permitting the permanentmagnet rotor to move in the middle of the induced magneticfield. Most BLDC motors have a three-phase windingtopology with star connection. A motor with this topology isdriven by energizing 2 phases at a time.

Figure-1. BLDC motor construction

An interesting property of brushless DC motors is that theywill operate synchronously to a certain extent. This meansthat for a given load, applied voltage, and commutation ratethe motor will maintain open loop lock with the commutationrate provided that these three variables do not deviate fromthe ideal by a significant amount. The ideal is determined bythe motor voltage and torque constants. Consider that whenthe commutation rate is too slow for an applied voltage, theBEMF will be too low resulting in more motor current. Themotor will react by accelerating to the next phase positionthen slow down waiting for the next commutation. In theextreme case the motor will snap to each position like astepper motor until the next commutation occurs. Since themotor is able to accelerate faster than the commutation rate,rates much slower than the ideal can be tolerated withoutlosing lock but at the expense of excessive current.Now consider what happens when commutation is too fast.When commutation occurs early the BEMF has not reachedpeak resulting in more motor current and a greater rate ofacceleration to the next phase but it will arrive there too late.The motor tries to keep up with the commutation but at the


RES -2016 28

expense of excessive current. If the commutation arrives soearly that the motor cannot accelerate fast enough to catch thenext commutation, lock is lost and the motor spins down.This happens abruptly not very far from the ideal rate. Theabrupt loss of lock looks like a discontinuity in the motorresponse which makes closed loop control difficult. Analternative to closed loop control is to adjust the commutationrate until self locking open loop control is achieved. This isthe method we will use in our application. When the load ona motor is constant over its operating range then the responsecurve of motor speed relative to applied voltage is linear. Ifthe supply voltage is well regulated, in addition to a constanttorque load, then the motor can be operated open loop over itsentire speed range

Key characteristics of the Variable Speed Brushless DCMotor Synchronous external commutation No brush noise, Durable & Robust : Flat Speed Vs Torque Characteristics Position/Speed Servo

II. CIRCUIT

Figure-2 (Schematic schema for 6 Pulse Operation)

The circuit consists of a voltage controlled voltage sourceconnected across three phase IGBT Bridge An IGBT consistsof a transistor in anti-parallel with a diode. It is turned onwhen the gating is high and the switch is positively biased. Itis turned off when the gating is low or the current drops tozero. The gate node of the IGBT can be connected to a switchgating block GATING or the output of a switch controlleronly. The on-off switch controller interfaces between thecontrol circuit and the power circuit. The input is a logicsignal (0 or 1) from the control circuit. The output isconnected to the gate (base) node of a switch (or multipleswitches) to control the conduction of the switch. The signallevel of 1 is for switch on and 0 for switch off. This bridge isconnected to Brushless dc motor whose speed feedbacksystem is closed loop with negative feedback, and finally thisfeedback controls the input voltage source .Thus this circuitshows similarity in characteristics between Brushless dc

motor and conventional dc motor. Thus by changing the dcbus voltage speed can be controlled.The BLDC motor also consists of a Hall position sensorembedded in its stator. A Hall Effect position sensor consistsof a set of hall switches and a set of trigger magnets. The hallswitch is a semiconductor switch (e.g. MOSFET or BJT) thatopens or closes when the magnetic field is higher or lowerthan a certain threshold value. It is based on the Hall Effect,which generates an emf proportional to the flux-density whenthe switch is conducting a current supplied by an externalsource. It is common to detect the emf using a signalconditioning circuit integrated with the hall switch ormounted very closely to it. This provides a TTL-compatiblepulse with sharp edges and high noise immunity forconnection to the controller via a screened cable. For a three-phase brushless dc motor, three hall switches are spaced 120electrical deg. apart and are mounted on the stator frame. Theset of trigger magnets can be a separate set of magnets, or itcan use the rotor magnets of the brushless motor. If thetrigger magnets are separate, they should have the matchedpole spacing (with respect to the rotor magnets), should bemounted on the shaft in close proximity to the hall switches.If the trigger magnets use the rotor magnets of the machine,the hall switches must be mounted close enough to the rotormagnets, where they can be energized by the leakage flux atthe appropriate rotor positions.

III. SIMULATION RESULTS

Fig.3 (Ia Vs Tme)

Fig.4 (Ib Vs Time)


RES -2016 29

Fig. 5 (Ic Vs Time)

Fig. 6 (n Vs Time)

Fig. 7 (n-Ref Vs Time)

Fig.8 (torque Vs Time)

A. Brushless DC CharacteristicsVoltage on the motor determines speed and current in themotor determines torque. These relationships are linear andnearly identical to a standard Brush DC drive. Theapplication of the product then is essentially like the more

familiar brush machine. Speed accuracy is very high, in factwith the most widely used Brushless drive, the accuracy issynchronous (0% speed error) due to a digital encoder anddrive controller position regulation. Torque to inertia ratiosare very high providing high accel/decel rates and excellentdynamic response. Controller bandwidth (30 to 40 Hz) is 5 to8 times higher [7] than the Brush DC drive.Motor efficiencies range from 90 to 96 % and controllerefficiency is 97% giving overall efficiencies better than brushDC systems.

B. AdvantagesIn general, the characteristics of Brushless D.C. that are mostadvantageous to the blow molding process are:• Very precise average speed control over a very wide speedrange.• Precise instantaneous speed control due to high dynamicresponse• Constant power factor means lowest possible input current.• Small physical size of motor compared to brush type.• No recurring motor maintenance (brush replacement)• Feedback device (encoder) is inside the motor not outside.• Higher efficiency overall.The blow molding process always involves an extruder ofsome kind and the variable speed drive on the extruder has toprovide an output sufficient to allow the comparison to beformed in time to meet the cycle requirements. Since the finalproduct may require more or less volume of plastic fordifferent shapes and because the cycle time varies, hence therequirement for variable speed. The result of inconsistentspeed control is simply that more (or less!) material than isnecessary to make the part will be extruded. Speed controlconsistency therefore is important to the production of aconsistent product. Speed that either drifts slowly

IV. SUMMARY AND CONCLUSIONSWhile there are several options available to the manufacturerand user regarding the type of variable speed drive to use onthe Extruder and rotating mold applications of the blowmolding machine, there are some distinct advantages in theuse of the Brushless D.C. drive while other types of drivesmay become available in the future that may provide theseadvantages, they are not yet available for general use in theHP sizes necessary for most blow molding machines. TheBrushless DC technology is here and in wide use in variousextrusion and related applications in the plastics industry.

References[1] T.G. Wilson, P.H. Trickey, "D.C. Machine With Solid State

Commutation", AIEE paper # CP62-1372, Oct 7, 1962.[2] NEMA Publication # MG 1, ref MG-1.41.2, table 12.6B[3] Eaton Dynamatic Electric Drive Applications Guide page M-37.[4] Dennis P. Connors, Dennis A. Jarc, Roger H. Daugherty, "Considerations

in Applying Induction Motors with Solid-State Adjustable FrequencyControllers",IEEE Transactions on Industry Applications, Vol 1A-20,no. 1, January/February, 1984.

[5] John B. Mitchell, "Inverter Power Factor and Noise", PowerTransmission Design magazine, page 45, 46.

[6] Derek A. Paice, "Harmonic Issues and Clean Power Controllers",Westinghouse Electric Corp, Presented at PCIM '90, Oct 25, 1990.

[7] Frank J. Bartos, "Reliability, Ease of Use Widen AC Drives' ApplicationHorizons", Control Engineering News, page 55, February 1992.

[8] Robert E. Lordo, "Comparison of the 150 HP Brushless DC andconventional DC/SCR motor/Control", June, 1992, POWERTECIndustrial Corp. Box 2650, Rock Hill, S.C. 29732, phone 803-328-1888.


RES -2016 30

Flexible Manufacturing SystemKomal Prasad Sahu1, Rahul Tamralkar2

1,2Mechanical Engineering Department, S.S.I.T.M. Bhilai, (C.G.)[email protected], [email protected]

Abstract- Flexible Manufacturing System (FMS) is a totallyautomated manufacturing system and a part of ComputerIntegrated Manufacturing (CIM) system. The main objectiveof supervisory control of FMS is to plan & execute the taskat the cell or work processing centre levels for the safe andunattended processing of a work order. It is a concept ofmachining where a set of machine tools can be used toperform a wide range of machining operations to produce avariety of products. This dynamic nature of machine toolscan be achieved by developing a amalgamation of Hardwareand Software component. Flexible Manufacturing System isan extension of programmable automation which is capableof producing variety of products with virtually no time lostfor changeovers from one part style to the next. There is nolost in production time while reprogramming the system andaltering the physical setups (like tooling, fixtures, machinesetting, etc.) “Automation” is the technology by which aprocess or procedure is accomplished without humanassistance. It is implemented using a set of codedinstructions called “Computer programs”. The Paperattempts to put forth the concept of Flexible ManufacturingSystem. It enlightens us about the various Hardware andSoftware components of this system and how each of themworks in co-ordination with each other. The paper alsoexplains F.M.S can be designed to suit the requirements of aProduction company to achieve the better productionactivity. Paper also tells us how to go about implementingthis system right from procuring the required components tostarting production on the system. The paper consist of casestudies of Companies have set-up F.M.S.

The Paper can be reallocated to produce different part typesor a different mix of part types without major delays orinvestment. It briefly describe across the field of engineeringand managements to include operations managementmanufacturing engineering, Industrial engineering operationresearch and management science as they relate to F.M.S. Itincludes the future of industrial automation.

I. INTRODUCTIONCompetitive business environment offers new pressures tobe confronted by the manufacturing systems, such asincreasing product variety with delivery on limited timealong with emphasize conventional requirements of qualityand competitive cost. Therefore, Business firms generallychoose to compete within one or two areas of strength.These areas of strength are often referred to as distinctivecompetencies, core competencies, or competitive priorities.Among the options for competition are price (cost), quality,delivery, service, and flexibility. An ever-increasing number

of firms are choosing to compete in the area of flexibility.Generally, this has meant that the firm's major strength isflexibility of product (able to easily make changes in theproduct) or flexibility of volume (able to easily absorb largeshifts in demand). Firms that are able to do this are said tohave flexible capacity, the ability to operate manufacturingequipment at different production rates by varying staffinglevels and operating hours, or starting and stopping at will.A Flexible manufacturing System (FMS) can be defined as acomputer-controlled configuration of semi-dependentworkstations and material-handling systems designed toefficiently manufacture various part types with low tomedium volume. It combines high levels of flexibility withhigh productivity and low level of work- in-processinventory (Jang & Park, 1996). The exquisiteness of FMS isthat it gleaned the ideas both from the flow shop and batchshop manufacturing system and is designed to imitate theflexibility of job shops while maintaining the effectivenessof dedicated production systems. Such FMS should bedesigned to improve productivity while fulfilling thedemand with decreasing manufacturing time. A genericFMS is able to handle a variety of products in small tomedium sized batches simultaneously. The flexibility of aflexible manufacturing system (FMS) has enabled it tobecome one of the most suitable manufacturing systems inthe current manufacturing scenario of customized and variedproducts with shorter life cycles.

II. LITERATURE REVIEWIn the period immediately following World War II, Japanesecompanies used their low labor costs to gain entry to variousindustries. As wage rates rose and technology became moresignificant, the Japanese shifted first to scale-basedstrategies and then to focused factories to achieve advantage.The advent of just-in-time production brought with it a moveto sought both low cost and great variety in the market.Cutting-edge Japanese companies today are capitalizing ontime as a critical source of competitive advantage:shortening the planning loop in the product developmentcycle and trimming process time in the factory-managingtime the way most companies manage costs, quality, orinventory. The search for ways to achieve even higherproductivity and lower costs continued, however. And in themid-1960s, it led top Japanese companies to a new source ofcompetitive advantage-the focused factory. Focusedcompetitors manufactured products either made nowhereelse in the world or located in the high-volume segment of amarket, often in the heart of their Western competitorsProduct lines. Focusing of production allowed the Japaneseto remain smaller than established broad-line producers,


RES -2016 31

while still achieving higher productivity and lower costs-giving them great competitive power.

The framework of flexible manufacturing systems (FMSs)combines high productivity, quality and flexibility neededfor the fast response to changing market demands (Womack,Jones & Roos, 1990). The term flexible manufacturingsystem (FMS) is generally used to represent a wide varietyof automated manufacturing systems. FlexibleManufacturing System (FMS) can be defined as anintegrated system composed of automated workstations suchas computer numerically controlled (CNC) machines withtool changing capability, a hardware handling and storagesystem and a computer control system which controls theoperations of the whole system (Mac Carthy, 1993).Tempelemeier & Kuhn (1993) define FMS as a productionsystem consisting of a set of identical and/or complementarynumerically controlled machines, which are connectedthrough an automated transportation system. Each process inan FMS is controlled by a dedicatedcomputer (FMS cell computer). As per Parrish (1990), aflexible manufacturing system is a collection of productionequipment logically organized under a host computer andphysically connected by a central support system. The mainimpetus to switch from a traditional system to an FMS is tointroduce flexibility in manufacturing operations so that afirm can compete more efficiently in the marketplace.Suresh and Sridharan (2007) described FMS as a growingtechnology mainly suitable for mid-volume, mid-varietyproduction, they also defined FMS as an integratedproduction facility consisting of multifunctional numericallycontrolled machining centers connected with an automatedmaterial handling system, all controlled by a centralizedcomputer system. An FMS is designed to have capability ofconcurrently handling a range of product types in batches(small to medium sized) and at a high efficiency ascompared to that of traditional production systems which aredesigned to deal with low-variety parts in high volume. Thissystem is able to process any part that belongs to specificfamilies within the prescribed capacity according to apredetermined schedule. Generally, the system is designedin such a way that manual interference and change over timeare minimized (Chan & Chan, 2004). One of the objectivesof an FMS is to achieve the flexibility of small volumeproduction while maintaining the effectiveness of large-volume mass production. The flexibility of a flexiblemanufacturing system (FMS) has enabled it to become oneof the most suitable manufacturing systems in the currentmanufacturing scenario of customized and varied productswith shorter life cycles. Ramesh and Jay Kumar (1991)stated that manufacturing flexibility can be of severaldifferent forms e.g. machine, operation, material handling,routing, program, expansion, process, product, volume,labor and material flexibilities. Sethi and Sethi (1990) gavethe concept of eleven flexibility types, Browne et.al. (1984)illustrated only eight types, which are known as; machineflexibility, process flexibility, routing flexibility, operationflexibility, product flexibility, volume flexibility, part mixflexibility and production flexibility. An FMS can provide

one or more of the above flexibilities. The consideration of aparticular type of flexibility to be considered in the design ofan FMS depends upon the system objectives. The increase inflexibility provides the alternative resources/machines to dothe same processing (Shnits et al., 2004).In recent studies pertaining to the FMS, researchers havebeen very keen to improve the performance of flexiblemanufacturing system (Wadhwa et al., 2005; Chan, 2003).To deal with the operational problems of flexiblemanufacturing systems such as routing and scheduling,simulation modeling has proved to be practical. Manyresearchers used simulation to study the scheduling androuting decisions for FMS.

III. FEATURES OF FLEXIBLEMANUFACTURING SYSTEM

Stated formally, the general objectives of an FMS are toapproach the efficiencies and economies of scale normallyassociated with mass production, and to maintain theflexibility required for small- and medium-lot-sizeproduction of a variety of parts.

Conventional manufacturing systems have been marked byOne of two distinct features:

1. The capability of producing a variety of differentProduct types, but at a high cost (e.g., job shops).

2. The capability of producing large volumes of aProduct at a lower cost, but very inflexible in terms of theproduct types which can be produced (e.g., transfer lines).

An FMS is designed to provide both of these features inaddition to these also, which are as follows:

1. Programmable machine tools.

2. Controlled by common computer network.

3. Combines flexibility with efficiency.

4. Reduces setup & queue times.

5. Ability to adapt to engineering changes in parts.

6. Increase in number of similar parts produced on thesystem.

7. Ability to accommodate routing changes.

8. Ability to rapidly change production set up.

9. There are basically two categories of Biometrics.

The FMS is capable of optimizing each step of themanufacturing process, and can involve one or moreoperations or processes (turning, drilling, milling, and


RES -2016 32

surface grinding), inspection and assembly tasks. An FMSprocesses many parts and it adapts to changes in the partdesigns. as a metric, that is, the measured action. In FMSeach machine can perform a variety of operations on a partwhen all the needed fixtures, tools, and pallets are available.For that reason, the set up time is reduced.

Fig. 2. A block diagram showing the features of FlexibleManufacturing System (FMS)

IV. WHY FLEXIBLE MANUFACTURINGSYSTEM?

Flexible Manufacturing is adapted by a firm because of thefollowing benefits:

1. To reduce set up and queue times

2. Improve efficiency

3. Reduce time for product completion

4. Utilize human workers better

5. Improve product routing.

6. Produce a variety of Items under one roof.

7. Improve product quality.

8. Serve a variety of vendors simultaneously.

9. Produce more Products more quickly.

Because of the above listed flexibility in the productionThe system is so called as “Flexible Manufacturing System”.

V. FMS COMPONENTS

1. Numerical Control (NC) machine tools

2. Automated material handling system (AMHS)

3. Automated guided vehicles (AGV) Conveyors

4. Automated storage and retrieval systems (AS/RS)

5. Industrial Robots

Loading and unloadingSpray paintingWeldingMaterial handlingInspectionMachine Assembly

6. Cells and Centers

7. Automated Inspection

8. Control Software

FMS is a reprogrammable manufacturing system capable ofproducing a variety of products automatically. It is anautomated production system that produces one or morefamilies of parts in a flexible manner. Today, this prospectof automation and flexibility presents the possibility ofproducing nonstandard parts to create a competitiveadvantage. The concept of flexible manufacturing systemsevolved during the 1960s when robots, programmablecontrollers, and computerized numerical controls brought acontrolled environment to the factory floor in the form ofnumerically-controlled and direct-numerically-controlledmachines.


RES -2016 33

Fig .2 - A Table Top Flexible Manufacturing System (FMS)- ‘Ready to run’ system integrates electronic sensors, PLC,robotics, and a CNC mill machine providing an excellentautomation system for applying actual industrial PLCcontrol ,as well as developing FMS concepts. Suppliedcomplete with course materials and available with additionalconveyors to add on or individual systems as well.

VI. COMPUTER INTEGRATEDMANUFACTURING (CIM)

The Integration of the total manufacturing enterprise throughthe use of integrated systems and data communicationscoupled with new managerial philosophies that improveorganizational and personnel efficiency.

• Integration of design, manufacture & delivery viacomputer technology

• CAD - uses software to create & modify designs.• CAM - uses programmable automation in manufacturing.• CAE - links functional design to CAD form design.• CAPP - creates processing instructions for CAM.• GT- classifies designs to benefit from prior experience.

CIM is a series of integrated activities and operationsinvolving the design, materials selection, planning,production, quality assurance, management and marketingof discrete consumer and durable goods.

Fig.3 - Benefits of CIM.

Fig.4 - Components of CIM

VII. LEVELS OF MANUFACTURINGFLEXIBILITY

(A) BASIC FLEXIBILITIES

• Machine flexibility - the ease with which a machine canprocess various operations• Material handling flexibility - a measure of the ease withwhich different part types can be transported and properlypositioned at the various machine tools in a system• Operation flexibility - a measure of the ease with whichalternative operation sequences can be used for processing apart type

(B) SYSTEM FLEXIBILITIES


RES -2016 33














RES -2016 33














RES -2016 34

• Volume flexibility - a measure of a system’s capability tobe operated profitably at different volumes of the existingpart types• Expansion flexibility - the ability to build a system andexpand it incrementally• Routing flexibility - a measure of the alternative paths thata part can effectively follow through a system for a givenprocess plan• Process flexibility - a measure of the volume of the set ofpart types that a system can produce without incurring anysetup• Product flexibility - the volume of the set of part types thatcan be manufactured in a system with minor setup.

(C) AGGREGATE FLEXIBILITIES

• Program flexibility - the ability of a system to run forreasonably long periods without external intervention• Production flexibility - the volume of the set of part typesthat a system can produce without major investment incapital equipment• Market flexibility - the ability of a system to efficientlyadapt to changing market conditions

VIII. THE FUTURE OF FMSFMS systems which deliver directly into warehouse, and donot require labor

· The use of robots that have vision, and tactile sensing toreplace human labor

· Technology will make 100% inspection feasible. Thusmaking faster process adjustment possible.

· Computer diagnosis will improve estimation of machinefailure, and guide work crews repairing failures.

· International coordination and control of manufacturingfacilities.

· Customers have completely custom orders madeimmediately, and to exact specifications, and at a lower cost

· Networks will tend to eliminate the barriers caused byinternational borders

· Standards will be developed which make installation of anew machine trivial

· Networking between manufacturers and suppliers willstreamline the inventory problems

· Marketing will be reduced, as customer desires are metindividually, and therefore do not need to be anticipated byresearch.

· Finished goods inventories will fall as individual consumerneeds are met directly.

· Better management software, hardware, and fixturingtechniques will push machine utilization towards 100%

· The task of Design and Process Planning will becomehighly automated, therefore reducing wasted time onrepetitious design, and discovering careless mistakes.

IX. ADVANTAGES AND DISADVANTAGESOF FMS

Advantages

1. Faster, lower- cost changes from one part toanother which will improve capital utilization.

2. Lower direct labor cost, due to the reduction innumber of workers

3. Reduced inventory, due to the planning andprogramming precision

4. Consistent and better quality, due to the automatedcontrol

5. Lower cost/unit of output, due to the greaterproductivity using the same number of workers

6. Savings from the indirect labor, from reducederrors, rework, repairs and rejects

Disadvantages

1. Limited ability to adapt to changes in product orproduct mix (ex. Machines are of limited capacityand the tooling necessary for products, even of thesame family, is not always feasible in a given FMS)

2. Substantial pre-planning activity3. Expensive, costing millions of dollars4. Technological problems of exact component

positioning and precise timing5. necessary to process a component6. Sophisticated manufacturing systems

ACKNOWLEDGEMENTWe desperately search for words to thank almighty

GOD for making us an instrument to prepare this paper.Before we get into thick things we would like to add a fewheartfelt words for people who helped us in making thispaper to go on.

We would like to express our sincere gratitude towardsour professors of Mechanical Department for theirinvaluable guidance and constant encouragementthroughout. Their unconditional support have proved to beindispensable.Finally, thanks to our parents, family members and also ourdear friends for the encouragement we got from them.

REFERENCESA. Websites and Links:http://claymore.engineer.gvsu.eduAutomationWorld.comwww.automationonline.comwww.9engineer.com[1] www.ibia.org[2] www.zyvex.com


RES -2016 35

B. Books and Articles1. Time -The Next Source of Competitive Advantage byGEORGE STALK, JR.2. Hugh Jack 1993-2001, Automated ManufacturingEngineering on disk3. Towards improving the performance of flexiblemanufacturing system:- A. Singholi; D. Chhabra; M. Ali4. Jim Pinto, Fully automated factories approach realityBrian Huse, Finishing Cell Automation Considerations5. Hary Gunarto, An Industrial FMS CommunicationProtocol, UMI (Univ. Microfilms International), Ann Arbor,Michigan, 160 pp, 19886. The Future Of Industrial Automation - FlexibleManufacturing SystemBy: Morrish Kumar and Jitendra Chaure

C. Industrial robot management[1] Leonard Flom, Aran Safir: “Iris recognition”.[2] John Daugman: “How iris recognition works”.[3] Yun, Yau Wei. The ‘123’ of Biometric Technology.

nationalconference - ijeee 201… · nationalconference on recent trends in renewable energy...

Documents