Microgrid-“A Smart Grid for Community Users” · PDF fileProceedings of National Seminar on “Dispersed Generation & Smart Grid” 2 ... the power grid, and if a microgrid

Authors are solely responsible for Plagiarism Copy right reserved DGSG-2013@NIST

1Proceedings of National Seminar on “Dispersed Generation & Smart Grid”

Microgrid-“A Smart Grid for Community Users”Bikash Narayan Panda#1

#Department of Electronics & Communication EngineeringNational Institute of Science and Technology

[email protected]

Abstract---Micro grids are modern, small-scaleversions of the centralized electricity system,having specific local goals, such as reliability,carbon emission reduction, diversification ofenergy sources, and cost reduction, established bythe community of users being served. This paperpresents an intelligent micro grid that is installablein small localities and consists of sources like gridpower, solar power etc. By efficient use of sourcesthe overall cost for the users can be reduced. Smartswitching at the user end as per the requirementsso as to restrict the user for overloading. Localpower generation and storage allow portions of thegrid and critical facilities to operate independent ofthe larger grid, Smart switches and sensorsautomatically fix and even anticipate powerdisturbances, unlike today’s system where switcheshave to be reset manually in case of any outage.

I. INTRODUCTIONA microgrid is a localized grouping of

electricity generation, energy storage, andloads that normally operate connected totraditional centralized grid .The connectivitybetween the microgrid and the centralizedgrid can be established or dislodgedaccording to the requirement, hence enablesthe microgrid to function standalone. Theloads and the generation points or sources areusually of low voltages. A connectedmicrogrid can be controlled and monitored asa single entity.

Microgrids allow consumers to procurepower in real-time at significantly lowercosts, while using local generation tocompensate power costs. The mostsignificant environmental benefit of a smartmicro grid is its ability to use localgeneration and the resulting “waste” heat todisplace coal-fired generation. A local powergenerator can be renewable- or natural gas-fuelled. The smart micro grid can reuse theenergy that is produced during electricity

generation for heating buildings, hot water,sterilization, cooling etc.

Smart microgrids reliability significantlyreduces costs. Smart meters that allow for thetwo-way exchange of pricing, usage data andelectricity. Programmable smart appliancesand devices that come on when the price ofpower reaches consumer’s desired price point.User-friendly home energy control systemsthat allow customers to interface with thesmart micro grid to automatically controlevery aspect of a home’s power usage.

Energy efficiency improvements that helpconsumers use less energy and ultimatelysave money on monthly electricity bills.Computerized controls that constantly scanfor, and even anticipate, potential instabilitiesto correct problems before users experienceany disruption in service. One more thing isthe locality can also sell the excess unutilizedenergy to other grids and earn revenue.

One interesting concept is vehicle to gridtechnology. According to which a vehiclecan be used to store some energy to the smartgrid after a long driving.

Microgrids can aggregatecomplementary distributed energy resources.Through the matching of supply and demandresources within a given microgrid, it ispossible to tailor the performance of thatnetwork to provide specific operating orenvironmental performance characteristics.For instance, individual microgrid can bedesigned for interruptibility, or efficiency ofgenerating sources and loads, or a specificlevel of reliability and power quality, or an



environmental emissions profile, or even tomaximize economic value by selling servicesto the centralized grids. In this way,microgrid makes a fairly commoditizedelectricity system customizable to the qualityneeds of an individual customer.

Microgrids can ease the pressure onutilities and rates while simultaneouslymodernizing the grid.

II. ARCHITECTURE

The basic architecture of a microgridsystem is presented in Figure 1, which showsthat a microgrids system generally consists ofdistributed generation (DG) resource, storagesystems, distribution systems, andcommunication and control systems

Fig.1.Microgrid Architecture.

Sources:Distributed generation technologies

applicable for microgrids may includeemerging technologies such as—windturbine, solar PV, micro-hydropower, dieselgenerators.

Solar PV generation involves thegeneration of electricity from solar energy.Due to enormous improvement in invertertechnologies, PV generation is now preferredworldwide as Distributed Energy Resources(DERs). The major advantages of a PVsystem are

(i)The sustainable nature of solar energy,(ii)Positive environmental impact,(iii)Longer lifetime and noiseless operation.

Wind turbine converts wind energy intoelectrical energy using the wind energyconversion systems (WECSs). Wind energyhas been popular for decades. Usuallyinduction generators are used in WECSs. Themain part of the wind turbine is the tower,the rotor, and the nacelle. The nacelleaccommodates the mechanical transmissionand the generator. Wind turbine captures thekinetic energy of wind flow through rotorblades and transfers the energy to theinduction generator through the gearbox. Thegenerator shaft is driven by the wind turbineto generate electric power. Wind turbinesmay have horizontal axis or vertical axisconfiguration. The average commercialturbine size is 300 kW. Micro-hydropowersystem uses the energy of flowing water toproduce mechanical or electrical energy. Thisenergy generation system depends on thetopography and annual precipitation of thearea.

Fig.2.Micro Hydropower System.

The system suffers from large variationof water flow due to uneven rainfall andresults in a variation in generation. Run-of-river system is often used in micro-hydropower systems which do not requirelarge storage reservoir. A portion of the riverwater is diverted to a water-conveyancechannel to rotate a turbine or a water wheelthat spins a shaft. The motion of the shaft canbe used for mechanical power such aspumping water or can be used to power agenerator to generate electricity.

Energy Storage:One of the main criteria of successful

operation of microgrid is the inclusion ofenergy storage devices, which balances theshort-term power and energy demand with



generation. Generally the microgrid powersystems have storage through the generatorinertia. When a new load comes online, it canresult in a slight change in system frequencydepending on its size. Storage devices arevery important to balance the powerfollowing system disturbance and/orsignificant load changes .In the case ofsudden system changes, these devices can actas an AC voltage source. Because of theirphysical limitations, they have limited energystorage capacity. The backup energy storagedevices should be included in microgridsystems to ensure uninterrupted powersupply. Suitable storage devices formicrogrid system include batteries, flywheelsand supercapacitors.

Control System:The users of the community being served

and the sources are connected to a centralizedcontrol system using a Local Area Network.This enables a centralized monitoring of eachand every users activity and also the sources.At the user end smart meter enablesmonitoring the usage whereas the sources aremonitored using sensors like temperature,pressure and others as per the requirementalong with the smart meters. Standard usagemonitoring and scheduling up to some extendare provided at the user end using LCDdisplays and keys to enhance the experience,and also leads to tracking the usage.

III.IMPLEMENTATION OF MICROGRID

The microgrid is implemented using 3sources solar, wind and grid where the gridsupply is used as a backup to the othersources. The control system is implementedusing the Arduino open source device. Whereeach sources and users are monitored usingsmart meters, whose data like voltage,current, units consumed are retrieved by theArduino device and are fed to a computer.The controller has a software panel by whichhe can control and monitor the sources.

Fig.3.implemented Hardware

The software is implemented usingJAVA. Statistics like variations in power,energy usage etc can be obtained from thesoftware.

Arduino communicates with the metersusing the standard RS-232 protocol with theModbus one. Where the device acts as themaster and the meters acts as slaves.

A computer collects the data from thehardware and stores in a database for futurereferences. Controlling of user usage is alsodone by the computer. The computer acts asa server to the Local Area Network to whicheach user access as per the requirement.Fig.2& Fig.3 are the screenshots of the developedapplications.

Fig.4. Source Usage Graph

Fig.5 Contolpanel



IV. ADVANTAGESIn a microgrid system the sources and

loads are situated at a close proximity so longdistance transmission lines are not requiredunlike in the centralized grid systems. Thisexclusion gives prevention to the losses thoseoccurs in the long running transmission linesand are accounted up to 20%.

One of the key advantages to the microgridapproach is that it allows local users to makesmarter choices regarding their use of power,turning them empowered consumers in aflexible energy economy.

Recovered waste heat could also cool anddehumidify buildings, using thermallyactivated processes. This is doublyadvantageous. Cooling buildings placestremendous strain on the power grid, and if amicrogrid shares some of this load, it willhelp both the microgrid customer andeveryone using the larger grid.

A major advantage of a Microgrid is itsability, during a utility grid disturbance, toseparate and isolate itself from the utilityseamlessly with little or no disruption to theloads within the Microgrid.

In peak load periods it prevents utility gridfailure by reducing the load on the grid.

Significant environmental benefits madepossible by the use of low or zero emissiongenerators.

V. PROBLEMS &CONSTRAINTS

The set up cost for the microgrids are hugecurrently as the energy sources fetches largeamount. But after a complete setup the gridsupplies power to the customers at no costexcept the maintenance as mostly renewablesources are used with the microgrids.

VI. CONCLUSION

Market-linked microgrids providecustomers options that can help drive downcosts and risks. The traditional electriccustomer is often a price-taker at the mercyof the utility’s pricing structure. Customerinvestment in efficiency and distributedresources like solar PV provide the customer

with some power, but the microgridrepresents the ability to completely flip thescript. The microgrid puts power in the handsof the customer and opens up choices forhow to manage energy risks and optimizecosts.

References:

[i] R. H. Lasseter and P. Paigi, “Microgrid: a conceptualsolution,” in Proceedings of the IEEE 35th Annual PowerElectronics Specialists Conference (PESC 04), pp. 4285–4290, June 2004.[ii] R. H. Lasseter, “Microgrids,” in Proceedings of theIEEE Power Engineering Society Winter Meeting, vol. 1, pp.305–308, 2002.[iii] S. Chowdhury, S. P. Chowdhury, and P. Crossley,

Microgrids and Active Distribution, Networks, London, UK,2009.[iv] F. Katiraei, R. Iravani, N. Hatziargyriou, and A. Dimeas,“Microgrids management,” IEEE Power and EnergyMagazine, vol. 6, no. 3, pp. 54–65, 2008.[v] DTI, “Micro-generation strategy and low carbonbuildings programme-consultation,” 2005,[vii] S. K. Khadem, M. Basu, and M. Conlon, “UPQC for

power quality improvement in DG integrated smart gridnetwork—a review,” International Journal of EmergingElectric Power Systems, vol. 13, no. 1, 2012.[viii] N. W. A. Lidula and A. D. Rajapakse, “Microgridsresearch: a review of experimental microgrids and testsystems,” Renewable and Sustainable Energy Reviews, vol.15, no. 1, pp. 186–202, 2011.[ix] C. Bossi, T. Degner, and S. Tselepis, “Distributedgeneration with high penetration of renewable energysources,” Dispower, Final Public Report, Laboratory TestsCase Studies and Field Experience, Kessel, Germany, 2006.[x]. D. Infield and F. Li, “Integrating micro-generation intodistribution systems—a review of recent research,” inProceedings of the IEEE Power and Energy Society GeneralMeeting (PES '08), pp. 1–4, July 2008.



Effect of wind speed on daily yield of a Solarstill operated at indoor and outdoor conditions

1Hemant Kumar Nayak, 2Md.Saifullah KhalidDepartment of Mechanical Engineering,

1National Institute Of Science & Technology, Palur Hills, Berhampur2Anand Engineering College, Kheetam, Agra1E-mail: [email protected]

Abstract - The effect of wind speed on the daily productivity of a solar still operated atindoor and outdoor conditions is studied. Numerical calculations have been carried out ondifferent days in a year in order to see the effect of speed of wind for same masses of basinwater mw for the basin type still. It is found that for the basin type still, Productivity increaseswith the increase of wind speed up to a critical velocity Vc beyond which the increase inProductivity becomes insignificant for both the Indoor and Outdoor conditions.

Keywords: Solar energy; passive cooling; Wind speed; Productivity.

I. INTRODUCTION

One major problem for the wholeworld and, in particular, in the third worldis the availability of pure clean and healthywater especially in remote areas.Desalination systems using solar energyhave been used in many countries toproduce fresh water. The working of solardistillation units has been broadlyclassified into active and passive modes ofoperation. It is reported that the overallthermal efficiency of a passive distiller ishigher than of an active one due to thelower operating temperature range [1].Several researchers [2, 3, 4, 5,] haveinvestigated the effects of climatic,operational and design parameters on theperformance of single, double and multi-effect active and passive solar stills. It hasbeen concluded that the productivity of thesolar stills increases with the increase ofsolar radiation and ambient temperature [6and7]. However, it should be pointed outthat there are conflicting results about theeffect of wind speed on solar still

productivity. Garg and Mann [6], Cooper[8], Rajvanshi [9], Soliman [10] and Malikand Tran [11] have concluded that theincrease in wind speed causes an increasein productivity; while Eibling et al. [7],Hollands [12] and Yeh and Chen [13 and14] indicated that an increase in windspeed causes a decrease in productivity. Ithas also been reported by Morse and Read[15] that the wind speed has no significanteffect on productivity.

Nomenclature

Ac collector area,

S solar radiation absorbed by

absorbing plate

ta ambient temperature

hfg latent heat of vaporization

of water, J/kg

Pci partial pressure of water

vapour at the glass

temperature, N/m2

Pd Yield or productivity



ηs efficiency of solar still

Ig total (global) solar

radiation, W/m2

Isc extra-terrestrial radiation on

a horizontal surface, W/m2

Ib beam component of the

total (global) solar

radiation,

W/m2

Id diffuse beam component of

the total (global) solar

radiation, W/m2

Cp specific heat at constant

pressure, J/kg oC

S solar flux reaching the

absorber plate, W/m2

I(t) solar radiation on

horizontal

surface, W/m2

k thermal conductivity, W/moC

Itb beam components of the

total solar radiation on

tilted surface, W/m2 .

Itd diffused components of the

total solar radiation on

tilted surface, W/m2.

t time, s

v wind velocity (S), m/s

Vc critical velocity, m/s

kt hourly clearness index

n day of the year

U heat loss coefficient, W/m2

oC

n2 / n1 refractive index of air and

glass cover, respectively.

R reflectivity

Rb beam radiation on the tilted

surface to that on the

horizontal surface.

Greek lettersα absorptivity

β surface tilted from the

horizontal

δ declination angle

ε emissivity

ρ reflectance

ω hour angle

τ transmittivity of the cover system of a

collector

τr transmittivity obtained by considering

only reflection and refraction

τα transmittivity obtained by considering

only absorption

θ1 angle of incident

θ2 angle of refraction

Φ latitude angle

δe glass cover thickness

Subscriptsa ambient air

Glass inr glass inner surface

Glass otr glass outer surface

II. EXPERIMENTAL SET-UP

There is a solar still which is made-up ofcopper sheet covered insulation, so thatthere is no flow of heat from still to



outside then minimize the heat loss. Thesheet of copper has black painted forincreasing the absorbing property of heatof copper metal. In this stil we employedtwo numbers of U-shape heater, whoselongitudinal length was 1.5 feet andheating element capacity 1500 watt eachfor proper heating of water. The basin typestill is closed with the help of glass sheetof thickness 0.005m which is used forcondensation of vapour of water at the topof still and this glass have a slope of 11.5ºfor installation of condensed liquid waterinto drain of still from where the distilledwater out from still. On above the drain weuse a sharp edge of glass strip to stop thedroplets of condensed water and fall downinto the drain. For measuring innertemperature of basin, we employed 14thermocouples (copper+constanton)mounted on fiber beeding. First 9thermocouples starting from basin plateare situated 1 cm apart from each otherand remaining five thermocouples 2cmapart from each other and last one is 5cmapart from second last one which give theglass inner temperature. For measuringglass outer and ambient temperature weused a thermocouple attached outsidesheet, for reading the temperature ofthermocouples we used a digitaltemperature indicator which convert heatsignal into electrical signal and displaydigitally on the screen. This solar still iskept on a iron made stand so that we usedspace under the still as our requirementsand also this stands prevents the heat lossfrom solar still to the ground. Theschematic diagram is shown below in Fig.(a) & the photograph of solar still shownbelow in Fig. (b). Now the heater removedfrom the solar still and for heating water,solar energy used at outdoor condition.

Fig. (a)

Fig. (b)

III. COMPUTATIONALPROCEDURES

A. Solar still efficiency (Outdoor -Conditions):

The efficiency of solar still is calculated onthe basis of solar heat used for distillationand the incident solar radiation. It isdefined as ratio of the two given below:

Efficiency of solar still (ƞs)= Distillate output (kg/s) * Latent heat ofvaporization of water (J/kg) Solarradiation (W/m2) *Area of basin of still(m2).

B. Solar still efficiency (Indoor-Conditions):

The efficiency of solar still is calculated onthe basis of solar heat used for distillationand the incident solar radiation. It isdefined as ratio of the two given below:



Efficiency of solar still (ƞ s)= Distillate output (kg/s) * Latent heat ofvaporization of water (J/kg) Solarradiation (W/m2) *Area of basin of still(m2).

Since distillate output in ml/h, to convert itinto kg/s, Experimental results are

shown in Table Aml/h = (10-6 m3*103 kg/m3) / 3600 s = 10-3/3600 kg/sPartial Pressure of vapour calculate by theequation given by Dunkle’s [36]

Pci = exp

ciT273

5144317.25 (pascal)

Experimental results are shown in Table A

Table ATable A

Water depth in linar = 0.04 mIndoor Operated Still Outdoor Operated StillDistillate output Distillate output

S.No. Time S = 0 m/s S = 36 m/s S = 39 m/s S = 7 m/s S = 8 m/s S = 10 m/s1 11:00 0 0 0 0 0 02 11:10 0 0 0 0 0 03 11:20 0 0 0 0 0 04 11:30 0 0 0 0 0 05 11:40 0 0 0 0 0 06 11:50 0 2 4 0 0 07 12:00 0 12 10 0 0 08 12:10 5 23 25 0 0 09 12:20 12 34 37 5 7 410 12:30 25 48 48 10 12 811 12:40 40 68 68 20 23 1712 12:50 56 88 88 30 34 2813 1:00 73 117 108 45 50 4014 1:10 95 143 132 65 70 6015 1:20 120 164 150 85 90 8016 1:30 150 195 180 105 110 10017 1:40 182 230 207 125 132 12018 1:50 217 270 240 145 155 14019 2:00 255 320 280 165 185 160

IV. RESULT AND DISCUSSION

The distillate outputs measured atdifferent times from the start-up periodare plotted in Figs. 3.1 & 3.2 for differentconditions. There is no output of distillatefor nearly 30 to 50 minutes from the start-up period. The effect of cooling rate ofthe glass cover is shown in Fig. 3.1. Thedistillate output, at 775 W (constant heatinput) of heat input for 3 hours of heating,

are 255 ml, 320 ml and 280 ml withnatural cooling, wind speed of 36m/minute and wind speed of 39 m/minute

respectively, showing more output at awind speed of 36 m/minute. This was alsohappen in the similar fashion for out dooroperated still. The distillate output for 3hours of heating, are 165 ml, 185 ml and160 ml with wind speed of 7 m/minute,wind speed of 8 m/minute and wind speedof 10 m/minute respectively, showingmore output at a wind speed of 8m/minute.

Figures 3.3 to 3.4 show efficiency of thestill for different conditions. Higher valuesof efficiency are achieved at moderatecooling rate of the glass cover. Theefficiency of the still after three hours ofheating at the wind speeds of natural, 36



m/minute and 39 m/minute, the efficiencybecomes 21.17%, 27.57% and 24.19%.This was also happen in the similar fashionfor out door operated still. The efficiencyof the still after three hours of heating atthe wind speeds of 7 m/minute, 8m/minute and 10 m/minute, the efficiencybecomes 12.15%, 12.43% and 11.94%.

V.CONCLUSION

On the basis of the results obtained for thevarious operative conditions of the solarstills, the following conclusions may bedrawn: (i) The daily productivities Pd ofthe basin type solar stills increase as windspeed increases up to the crtical velocityVc. (ii) After the typical velocity Vc, thechange in Pd becomes insignificant. (iii)The value of Vc is independent on the stillshape, mode of operation (active orpassive) and the heat capacity of brine, butit shows some seasonal dependence. Vc isfound to be 36 m/s for indoor and 8 m/sfor outdoor conditions respectively. (iv)The results which have been achieved inthe present study may be used to explainthe conflicting results which are reportedin the previous studies [10, 12, 13, 14]about the effect of wind speed on the dailyproductivity of the solar stills.

0

50

100

150

200

250

300

350

Dis

tilla

te o

utpu

t (m

l)

Natural

TimeFig. 3.1 Variation in the distillate output due to wind speed blowing past the glasscover of the Indoor distillation plant (Heat input=775 W and water depth=0.04 m).




V.CONCLUSION


Fig. 3.1 Variation in the distillate output due to wind speed blowing past the glasscover of the Indoor distillation plant (Heat input=775 W and water depth=0.04 m).




V.CONCLUSION


Fig. 3.1 Variation in the distillate output due to wind speed blowing past the glasscover of the Indoor distillation plant (Heat input=775 W and water depth=0.04 m).



REFERENCES

1. G.N. Tiwari, Recent advances in solardistillation. In: Raj Kamal Solar Energyand Energy Conversion, Wiley Eastern,New Delhi (1992), pp. 32–149 (ChapterII).

2. S. Kumar and G.N. Tiwari,Optimization of daily yield from an activedouble effect distillation with water flow.Energy Convers. Mgmt. 40 (1999), pp.703–715.

3. B.A. Jubran, M.I. Ahmed, A.F. Ismailand Y.A. Abakar, Numerical modeling ofa multi-stage solar still. Energy Convers.Mgmt. 41 (2000), pp. 1107–1121.

4. S. Suneja and G.N. Tiwari, Effect ofwater depth on the performance of aninverted absorber double basin solar still.Energy Convers. Mgmt. 40 (1999), pp.1885–1897.

5. H.M. Yeh, C.D. Ho and K.J. Hwang,Energy and mass balance in open-typemulti-effect solar distiller with air flowthrough the last effect. Energy 24 (1999),pp. 103–115.

6. H.P. Garg and H.S. Mann, Effect ofclimatic, operational and designparameters on the year-round performanceof single-sloped and double-sloped solarstills under Indian arid zone conditions.Solar Energy 18 (1976), pp. 159–164.

7. J.A. Elbling, S.G. Talbert and G.O.G.Lof, Solar stills for community used-digest

of technology. Solar Energy 13 (1971), pp.263–276.

8. P.I. Cooper, Digital simulation oftransient solar still performance. SolarEnergy 12 (1969), pp. 313–331.

9. A.K. Rajvanshi, Effect of various dyeson solar distillation. Solar Energy 27(1981), pp. 51–65.

10. S.H. Soliman, Effect of wind on solardistillation. Solar Energy 13 (1972),pp403–415.

11. M.A.S. Malik and V. Tran, A simplemathematical model for predicting thenocturnal out put of a solar still. SolarEnergy 14 (1973), pp. 371–385.

12. K.G.T. Hollands, The regeneration oflithium chloride brine in a solar still foruse in solar air conditioning. Solar Energy7 (1963), p. 39.

13. H.M. Yeh and L.C. Chen, Basin-typesolar distillation with air flow through thestill. Energy 10 (1985), pp. 1237–1241.

14. H.M. Yeh and L.C. Chen, The effectof climatic, design and operationalparameters on the performance of wick-type solar distiller. Energy ConversionMgmt. 26 (1986), pp. 175–180.

15. R.N. Morse and W.R.W. Read, Arational basis for the engineering dev-elopment of a solar still. Solar Energy 12(1968), p. 5.



Power Quality Issues in Dispersed GenerationAnita Shial

HI-TECH COLLEGE OF ENGINEERING, BPUT,ODISHAEmail:[email protected]

Abstract - Power quality refers to thedegree to which power characteristics alignwith the ideal sinusoidal voltage & currentwaveform, with current & voltage inbalance. Dispersed generation contributes tothe improvement of power quality. In theareas where voltage support is difficult,distributed generation offers sufficientbenefits for voltage profile & power factor

corrections. Distributed generation couldserve as a substitute for investments intransmission & distribution capacity. Thispaper describes various power quality issueslike Voltage fluctuation, Voltage Sag,flicker etc. & their possible solutions. Theirapplication is not limited to present utilitiesbut can also be applicable for future smartgrid.

I. INTRODUCTION

Power quality concerns the electricalinteraction between the network and itscustomers. It consists of two parts: thevoltage quality concerns the way in whichthe supply voltage impacts equipment; thecurrent quality on the other hand concernsthe way in which the equipment currentimpacts the system [1].

Dispersed generation has beenrecommended as one of the environmentallyfriendly solutions for improving the energysystem, decreasing the losses and increasingeffectiveness [2].

Connecting new producers and generators tothe distribution network can drasticallychange the working parameters of the grid.Dispersed units affect the current quality andthrough the grid also the voltage quality asexperienced by other customers [2].

One of the key aspects of electricityproduction and distribution is the powerquality and supply reliability for thecustomers.

As it can be expected, a great number ofsmall generation units will be connected todistribution grids in quite near future. Asdue to the rise of harmonics in the grid, themachines would become more vulnerable,their faults become even more difficult todetect, so one could expect a growingnumber of unexpected downtimes due todifferent faults of the generators. This is theissue why real time condition monitoring isof utmost importance in the dispersedgeneration situation.

II. DISPERSED GENERATION

Dispersed generation is the production ofelectricity at or near the point of use. Mostor part of consumed energy is produced atpoint of use and rest of the electricity goesinto the distribution grid [1].

The working group of CIGRE [3] devoted todispersed generation defines dispersedgeneration as all generation units with amaximum capacity of 50MW to 100MW,that are usually connected to the distributionnetwork and that are neither centrallyplanned nor dispatched.



The IEEE [4], on the other hand, definesdispersed generation as the generation ofelectricity by facilities that are sufficientlysmaller than central generating plants so asto allow interconnection at nearly any pointin a power system.

In most cases it is assumed that the electricalcurrent and voltage have a sinusoidal waveshape. But if hundreds or thousands of smallpower production plants are connected to agrid, it could mean that the sinusoidalcurrent and voltage waveforms are distortedand the waveform is no longer sinusoidal.Also, all small generators themselvesproduce harmonics. So the large-scale use ofrenewable energy sources for the productionof electricity will bring major challenges forthe electricity network.

As dispersed generation means also agrowing number of small power plants suchas small hydro and wind applications, thisharmonic problem can become a seriousissue for the power quality and supplyreliability in smart grid or dispersedgeneration situation.

III. POWER QUALITY

Connection of dispersed resources andchanging dispersed generation to thedistribution grid can affect the power qualityin a great amount [2].

Power quality can be controlled andimproved in whatever point of the electricsystem beginning from the means in thesystem or the grid and ending with singledevices at the consumer level. Connection ofthe dispersed generation of renewableenergy to distribution grid can have bothpositive and negative effects to the powerquality. It depends on possibilities ofinformation and communication systems to

control and maintain voltage in the feeders,turn the loads in or out and replace lostpower with the reserves.

IV. HARMONIC DISTORTION DUETO DIFFERENT LOADS

Electrical devices, which are coming ontomarket, are becoming more and morecomplex. They may help to reduce energyconsumption, but their performanceregarding power quality is still ratherimproper.

The problem is that their current curve is nota perfect sinusoid. The widespread use ofnonlinear loads may implicate significantreactive power and problems with higherharmonics in a grid [5].

Harmonic currents produced by nonlinearloads are injected back into the supplysystems. These currents can interactadversely with a wide range of powersystems equipment causing additionallosses, overheating and overloading. Theseharmonic currents can also causeinterference with telecommunication linesand errors in power metering [6]–[8]. Thatproblem may come more important whensmart grid solutions where communicationis very important are adapted.

While the applied voltage is almost perfectlysinusoidal, the resulting current is heavilydistorted.



Fig.1. Currents of typical nonlinear loads versusvoltage.

Harmonics generated by consumer’sappliances must not cause voltage rise in theconnection point [5]. Fixing limits maybecome important before using numerousharmonics emitting devices together. Insome papers [7] measurements withnonlinear loads are done when 5% current´stotal harmonic distortion value at connectionpoint is followed. For example most of thecommon compact fluorescence lamps havethe total harmonic distortion over 100% [8].

Harmonic currents injected from individualend users on the system should be limited.These currents propagate toward the supplysource through the system impedance,creating voltage distortion. Thus by limitingthe amount of injected harmonic currents,the voltage distortion can be limited as well.This is the basic method of controlling theoverall distortion levels proposed by IEEEstandard 519-1992. Example for illustratingnonlinear loads influence on distributiongrid a study with compact fluorescencelamps (CFLs) is made [7], [8].

V. POSSIBLE SOLUTIONS FORPOWER QUALITY

5.1 Increasing Voltage Quality and GridCapacity by Reactive Power

While the grid capacity and grid voltagequality have primarily been provided bynetwork expansion so far, this project aimsto effectively use installations which aredistributed in the grids. This is done by theuse of distributed measurement technology,intelligent control of power electronics, newinformation and communication technologyand the possibilities of the grid control.

The concept is developed and tested on theexample of distributed PV systems.However the use is not restricted to thisapplication. Efficiency generally can beincreased by distributed network services.The operational status of the grid has to bemeasured continuously at connection pointsof large loads and decentralized generation.Solar inverters are equipped with dataacquisition capabilities because they need tosynchronize their voltage and frequency tothe grid voltage. For load connection pointsmeasuring Technology is to be installed.

As shown in Fig. 2 a main computer isnetworked to a number of data acquisitiondevices and solar inverters.Data acquisitiondevices and solar inverters monitor

Fig. 2.Data acquisition and control structure



voltage, current and power flow at theirlocations in the grid. Data acquisitiondevices are located at large loads(e.g. industrial plants) and grid nodes.

The reactive power control structureconsists of three different controls:• The first part is the limitation of the gridvoltage by reactive power absorption of theinverters. To avoid unnecessary losses onlyas many inverters as needed have to absorbonly as much reactive power as needed tolimit the grid voltage. Thus the maincomputer only activates the inverters withthe highest voltage levels in the grid.• Additionally, voltage fluctuations due tofast load and generation changes e.g.moving clouds can be compensated andsmoothed by injecting and absorbingreactive power through the solar inverters.• The inverters can also be used for localcompensation of reactive power required byother loads in order to minimize powerlosses in the grid.

5.1.1 Reactive power compensation

Reactive power compensation to this daterequires additional equipment and associatedinstallation and commissioning costs whichshould be compensated by greaterefficiencies. So far, compensation is mainlyused in large industrial plants. Therefore,generating decentralized reactive power forcompensation significantly lowers the powerlosses due to short transmission distances ofthe reactive power. For generating reactivepower short term energy storage is required.This can be done with capacitors orinductors. Voltage source solar invertersusually have capacitors, so the alreadyinstalled capacity can be used for reactivepower. The existing reactive power reserveswhich are inherently present by thedistributed inverters can be used to providereactive power to the overlaid medium

voltage grid or to reduce the reactive powerconsumption of the low voltage grid tominimize losses.

VI. CONCLUSION

If more and more dispersed generation isgoing to be installed all over the powernetworks like it seems to go then it is mostimportant to find measures for guarantyingquality and security of supply.

In the situation where generation as well asconsumption produces decrease of powerquality in the grid, it is essential to analyzeboth generation and consumption in a verythorough way. If it proves to be necessary itmight make sense to limit the usage of newplants and appliances or use some othermethods to decrease their negative effect topower quality.

Usage of nonlinear loads like compactfluorescent lamps has risen rapidly in thelast decade, but their harmonic emission,reactive power consumption and otherdrawbacks have been ignored.

Beside the problem that harmonics areextremely dangerous to electrical motors,distorted supply makes the diagnostics ofthem more difficult. A growing number ofmachines are driven through frequencyconverters. This means that also diagnosticfor appropriate setups with frequencyconverters should be investigated.Frequency converters add additional noiseand harmonics to the traditional currentspectrum of the machines and thus suchdrives need a slightly different approach indiagnostics than traditional grid suppliedmachines. Nevertheless, appropriate on-linediagnostics of dispersed generation unitsmust be applied to guarantee sufficientpower quality, supply reliability and overall



safety of customers and different facilitiesconnected to the grid.

REFERENCE

[1] M. Bollen, Understanding power qualityproblems: Voltage Sags and Interruptions,1st ed., Wiley-IEEE Press, 2000, p. 543.

[2] K.Purchala, R. Belmans, L. Exarchakos,A. D. Hawkes, “Distributed generation andthe grid integration issue”, KULeuven,Imperial College London.

[3]CIGRE, International Council on LargeElectricity Systems, http://www.cigre.org

[4] IEEE, Institute of Electrical andElectronics Engineers, http://www.ieee.org

[5] T. Ackermann, G. Andersson, L. Söder,“Distributed generation: a definition”,Electric Power Systems Research, vol. 57,pp. 195–204, 2001. [Online]. Available:

http://dx.doi.org/10.1016/S0378-7796(01)00101-8

[6] T. Vaimann, J. Niitsoo, T. Kivipõld,“Dispersed generation accommodation intosmart grid”, in Proc. of the 52ndInternational Scientific Conference of RigaTechnical University. Section of PowerandElectrical Engineering, Riga TechnicalUniversity Press, 2011, ID-42.

[7] European Standard, EN 50160:2011Voltage characteristics of electricity, 2011.

[8] M. Bollen, Adapting electricity networksto a sustainable– smart metering and smartgrids, Swedish Energy Markets Inspectorate,2011, pp. 1–115.

[9] Kerber, G.; Witzmann, R.: StatistischeAnalyse von NS-Verteilungsnetzen undModellierung von Referenznetzen. ew -



Abstract - This paper presents an overview of the smart griddevelopments, analyzed from a power-quality viewpoint;describes the main relations between integration ofdispersed generations to the smart grids and power qualityissues. Power-quality monitoring in the smart grid isdiscussed in detail. Although the smart grid brings manynew power-quality challenges, these should not result inunnecessary barriers against the introduction of newtechnology. One of the most important issues in future gridsare the power quality and supply reliability issues. Thispaper describes how the change to dispersed generation andsmart grids should look like and what are the mainproblems, that need quick and active solutions, so thatfuture grids would be fully functional and reliable.

Key words – Dispersed Generation, DistributionAutomation, Harmonics, Power Quality, Power-QualityMonitoring, Smart Grids.

I. INTRODUCTION

With recent developments in power system, the smart gridallows seamless integration of renewable energy sourcesinto the conventional grid. In the ongoing discussions aboutdispersed generation & smart grids, power quality is animportant aspect and should not be neglected. An adequatepower quality guarantees the necessary compatibilitybetween all equipment connected to the grid. It is thereforean important issue for the successful and efficient operationof existing as well as future grids.

Electrical Power quality plays an important role in smoothoperation of sensitive loads, appliances & sophisticatedplant processes. Most of the manufacturing industries facedifficulty due to the distortion of the power supply whichresults to both manufacturing & production losses. Becauseall kinds of transducers, PLC apparatuses and automaticproduction processor based on computers are sensitive tosmall distortion of supply voltage. More attention has beengiven in the power quality problems [1]. Since powerquality distortion has harmful effects on the electric powersystem, and everyone is concern about quality of products &services as those are paid for. Hence there is a need tomonitor the power quality disturbances by proper instrumentfor accurate detection and localization so as to take properpreventive measures to assure qualitative power to theconsumers.

1. .Associate Professor & HOD, Department of EEE, VITAM,Berhampur, India, e-mail: [email protected]

2. Asst. Professor & Asst. HOD, Department of EEE, VITAM,Berhampur, India, e-mail: [email protected]

3. Asst. Professor, Department of EEE, VITAM, Berhampur, India ,e-mail: [email protected]

If centralized generation is characterizing factor fornowadays energy systems, then smart grids of the futuremean also the spreading of dispersed generation. The causeof these changes is the strict environmental norms andliberalization of electricity markets [2].

Dispersed generation has been recommended as one of theenvironmentally friendly solutions for improving the energysystem, decreasing the losses and increasing effectiveness[3]. In addition increasing the ratio of small producers inelectricity generation has been proposed.

Connecting new producers and generators to the distributionnetwork can drastically change the working parameters ofthe grid. This situation is extremely important when the newconnected power plant is equal to or even greater than theload in this particular area. In such case the new power plantaffects the voltage adjustment and power flux. It isimportant to evaluate the existing grid, capacity and loads inthis certain area of the distribution network [3].

Dispersed units affect the current quality and through thegrid also the voltage quality as experienced by othercustomers [3]. Power quality concerns the electricalinteraction between the network and its customers. Itconsists of two parts: the voltage quality concerns the wayin which the supply voltage impacts equipment; the currentquality on the other hand concerns the way in which theequipment current impacts the system [4].

For these reasons it is important changing of the loads haveto be observed in smart grids as well as the growth ofdispersed generation. In addition loads are getting more andmore nonlinear which means that the cooperation ofuntraditional generation and loads affect the grid inunpredictable ways.

One of the key aspects of electricity production anddistribution is the power quality and supply reliability forthe customers. Traditionally the problems have been solvedbut as the world is moving towards smart grids anddispersed generation, those problems need more active andprecise control.

As it can be expected, a great number of small generationunits will be connected to distribution grids in quite nearfuture. Most probably it would require certain onlinediagnostic systems to secure the full functionality andreliability of those units.

As due to the rise of harmonics in the grid, the machineswould become more vulnerable, their faults become even

Monitoring and Mitigation of Power Quality Issuesin Dispersed Generation & Smart Grids

Debasis Tripathy1, Manoj Pattnaik2, Amar Kumar Barik3



more difficult to detect, so one could expect a growingnumber of unexpected downtimes due to different faults ofthe generators. This is the issue why real time conditionmonitoring is of utmost importance in the dispersedgeneration situation.

II. DISPERSED GENERATION

Dispersed generation is the production of electricity at ornear the point of use. Most or part of consumed energy isproduced at point of use and rest of the electricity goes intothe distribution grid [5].

In most cases it is assumed that the electrical current andvoltage have a sinusoidal wave shape. But if hundreds orthousands of small power production plants are connected toa grid, it could mean that the sinusoidal current and voltagewaveform are distorted and the waveform is no longersinusoidal. Also, all small generators themselves produceharmonics. So the large-scale use of renewable energysources for the production of electricity will bring majorchallenges for the electricity network.

Generators are typical electrical devices that are usuallysetup together with frequency converters to drive them anddifferent inverters to synchronize their work with the grid.Not only generators themselves but also frequencyconverters and other electronic devices produce a vastnumber of harmonics that can be a problem to electricalmachines they are set up with and also the grid they areworking in. Due to financial benefits usually no additionalfilters are used to lessen the amount of induced harmonics.A typical harmonic distortion of a frequency converter isshown on figure 1.

As dispersed generation means also a growing number ofsmall power plants such as small hydro and windapplications along with renewable energy resource plants asshown in figure 2, this harmonic problem can become aserious issue for the power quality and supply reliability insmart grid or dispersed generation situation.

Figure. 1. Typical harmonic distortion of frequency converter.

For showing the probability of large extent of distributedgeneration in near future the potential is pointed out.

III. SMART GRIDS

Smart Grid technologies are advanced electrical net-worksthat support new generation interactive energy andcommunication services for the final customer. Theelectrical networks must be live, available, interconnectedand coupled with real-time communications [6]. Smart gridtechnologies are the vision of the future delivering variousbenefits to society in order to accomplish sustainable energydevelopment.

The main advantage of smart grid implementation in termsof utility benefits include reduced perturbations and outages;minimal power losses and blackout prospects; lowermaintenance and operational cost; lower greenhouse gasemissions; increased energy efficiency; increased large scalerenewable energy and distributed generation integration;enabled micro-grid applications and Energy ManagementSystems (EMS); environmental benefits and economicgrowth through clean power markets. Similarly, the ad-vantages of smart grid implementation in terms of customerbenefits include enabled electricity consumption, enabledconsumer services through distributed generation and costsavings through EMS. These benefits have a positive impacton society stimulating improvement in new power systemnetworks as well as encouraging multiple engineers frompursuing research developments in smart grid technologies.

Smart grid technologies offer numerous characteristics toimprove this state such as power quality enhancement, selfhealing, generation sources, optimization of asset utilizationand empowerment of users[6]. Figure 3, illustrates thearchitecture of an integrated smart grid communicationsplatform

IV. POWER QUALITY

Power quality disturbance is a general term used todesignate a number of electromagnetic phenomena thatcause voltage supply to deviate from its constant magnitudeand frequency of ideal sinusoidal wave shape.

Two main groups of power quality disturbances can bedefined: stationary (or quasi-stationary) and transientdisturbances. Harmonic and interharmonic distortion,voltage fluctuation, voltage flicker and voltage unbalancemake up the first group, whereas voltage transients, voltagedips, voltage swells, short interruptions in voltage supplyand other high-frequency disturbances constitute the lattergroup. Categories and Characteristics of ElectromagneticPhenomena responsible for Power Quality disturbances aregiven in Table-1.

Connection of dispersed resources and changing dispersedgeneration to the distribution grid can affect the powerquality in a great amount [7]. Smart distribution grid mustsecure the end users with power that has the demandedquality [8-10]. This is why the modern control systems, thatare monitoring the important components of the distributiongrid, must react precisely to the changes in power quality.

Power quality can be controlled and improved in whateverpoint of the electric system beginning from the means in thesystem or the grid and ending with single devices at theconsumer level.



Figure 2. Dispersed Generation integrated to Smart Grid

Figure 3. Architecture of integrated smart grid communications platform



Connection of the dispersed generation of renewable energyto distribution grid can have both positive and negativeeffects to the power quality. It depends on possibilities ofinformation and communication systems to control andmaintain voltage in the feeders, turn the loads in or out andreplace lost power with the reserves.

For example small amounts of wind power have negligibleeffects on electricity networks, but when electricitygeneration from wind power exceeds a certain thresholdlevel, investments in the power system will be required. Thisthreshold level is known as the hosting capacity [5].

The principle of hosting capacity is explained at Figure 4.Hosting capacity does not say anything about how muchgeneration from renewable energy sources that is connectedto the grid, only how much can be connected without havingto invest in measures to strengthen the grid.

Figure 4. The principle underlining hosting capacity.

V. POWER-QUALITY ISSUES

5.1 Emission by new devicesWhen smart grids are introduced, we expect growth both inproduction at lower voltage levels (distributed generation)and in new types of consumption (for example, chargingstations for electric vehicles, expanded high-speed railways,etc.). . Some of these new types of consumption will emitpower-quality disturbances, for example harmonic emission.Preliminary studies have shown that harmonic emission dueto distributed generation is rather limited. Most existing end-user equipment (computer, television, lamps, etc) emitalmost exclusively at the lower odd integer harmonics (4,10, 13 etc), but there are indications that modern devicesincluding certain types of distributed generators emit abroadband spectrum [11,12,13]. Using the standard methodsof grouping into harmonic and interharmonic groups andsubgroups below 2 kHz will result in high levels for evenharmonics and interharmonics. For frequencies above 2 kHzhigh levels have been observed for the 200-Hz groups. Anexample is shown in Fig. 3: the spectrum of the emission bya group of three full power converter wind turbines, where 1A is about 1% of the emission is low over the wholespectrum, being at most 0.5% of the nominal current. Thecombination of a number of discrete components at thecharacteristic harmonics (10-14] together with a broadbandspectrum over a wide frequency range, is also being emittedby other equipment like energy-efficient drives, micro-generators, and photo-voltaic installations. The levels arenot always as low as for the example shown here. The

existing compatibility levels are very low for somefrequencies, as low as 0.2%.

Harmonic resonances are more common at these higherfrequencies so that any reference impedance for linkingemission limits to compatibility levels should be set ratherhigh. Keeping strict to existing compatibility limits andexisting methods of setting emission limits could putexcessive demands on new equipment.Table 1: Classification of Power Quality Disturbances [7]

Sl.No. Categories

Typicalspectralcontent

Typicalduration

TypicalVoltage

magnitude1.0 Transients1.1 Impulsive1.1.1 Nanosecond 5 ns rise <50 ns1.1.2 Microsecond 1s rise 50 ns–1 ms1.1.3 Millisecond 0.1ms rise >1 ms1.2 Oscillatory1.2.1 Low frequency <5 kHz 0.3–50 ms 0–4 pu1.2.2 Medium frequency 5-500 kHz 20s 0–8 pu1.2.3 High frequency 0.5-5 MHz 5s 0–4 pu

2.0 Short-durationvariations

2.1 Instantaneous

2.1.1 Interruption 0.5–30cycles <0.1 pu

2.1.2 Sag (dip ) 0.5–30cycles 0.1–0.9 pu

2.1.3 Swell 0.5–30cycles 1.1–1.8 pu

2.2 Momentary

2.2.1 Interruption 30 cycles –3 s <0.1 pu

2.2.2 Sag (dip ) 30 cycles –3 s 0.1–0.9 pu

2.2.3 Swell 30 cycles –3 s 1.1–1.4 pu

2.3 Temporary2.3.1 Interruption 3 s – 1 min <0.1 pu2.3.2 Sag (dip ) 3 s – 1 min 0.1–0.9 pu2.3.3 Swell 3 s – 1 min 1.1–1.2 pu

3.0 Long-durationvariations

3.1 Interruption,sustained >1 min 0.0 pu

3.2 Undervoltages >1 min 0.8–0.9 pu3.3 Overvoltages >1 min 1.1–1.2 pu4.0 Voltage unbalance Steady state 0.5 – 2%5.0 Waveform distortion5.1 DC offset Steady state 0 – 0.1%

5.2 Harmonics 0 – 100th

harmonic Steady state 0 – 20%

5.3 Interharminics 0 – 6 kHz Steady state 0 – 2%5.4 Notching Steady state5.5 Noise Broadband Steady state 0 – 1%6.0 Voltage fluctuations <25 Hz Intermittent 0.1 – 7%

7.0 Power frequencyvariations <10 s

The measurement of these low levels of harmonics at higherfrequencies will be more difficult than for the existingsituation with higher levels and lower frequencies. Thismight require the development of new measurementtechniques including a closer look at the frequency responseof existing instrument transformers.

The presence of emission at higher frequencies than beforealso calls for better insight in the source impedance at thesefrequencies: at the point of connection with the grid as wellas at the terminals of the emitting equipment.



5.2 Interference between devices and powerline-communicationSmart grids will depend to a large extent on the ability tocommunicate between devices, customers, distributedgenerators, and the grid operator. Many types ofcommunication channels are possible. Power-linecommunication might seem an obvious choice due to itseasy availability, but choosing power-line communicationcould introduce new disturbances in the power system,resulting in a further reduction in power quality. Dependingon the frequency chosen for powerline communication, itmay also result in radiated disturbances, possibly interferingwith radio broadcasting and communication.

It is also true that modern devices can interfere with power-line-communication, either by creating a high disturbancelevel at the frequency chosen for power-linecommunication, or by creating a low-impedance path,effectively shorting out the power-line communicationsignal. The latter seems to be the primary challenge topower-line communication today [11].

So far, there have been no reports of widespreadinterference with sensitive equipment caused by powerline-communication, but its increased use calls for a detailedstudy.

5.3 Allocation of Emission LimitsWhen connecting a new customer to the power system, anassessment is typically made of the amount of emission thatwould be acceptable from this customer without resulting inunacceptable levels of voltage disturbance for othercustomers. For each new customer a so-called emission limitis allocated. The total amount of acceptable voltagedistortion is divided over all existing and future customers.This assumes however that it is known how many customerswill be connected in the future [17].

With smart grids, the amount of consumption will have nolimit provided it is matched by a similar growth inproduction. This continued growth in both production andconsumption could lead to the harmonic voltage distortionbecoming unacceptably high. Also the number of switchingactions will keep on increasing and might reachunacceptable values. One may say that production andconsumption are in balance at the power-system frequency,but not at harmonic frequencies.

Another way of looking at this is that the system strength isno longer determined by the maximum amount ofconsumption and/or production connected downstream, butby the total amount of harmonic emission coming fromdownstream equipment. This will require a different way ofplanning the distribution network.

5.4 Improving Voltage QualityOne aim of smart grids is to improve the performance of thepower system (or to prevent deterioration) without the needfor large investments in lines, cables, transformers, etc.

From a customer viewpoint, the improvements can be interms of reliability, voltage quality or price. All otherimprovements (e.g. in loading of cables or transformers,

protection coordination, operational security, efficiency) aresecondary to the customer.

These networks should therefore be addressed first. Thesame balance between “production” and “consumption” canin theory also be used for the control of harmonic voltages.When the harmonic voltage becomes too large, either anemitting source could be turned off, or a harmonic filtercould be turned on, or a device could be turned on that emitsin opposite phase (the difference between these solutions isactually not always easy to see). Smart grid communicationand control techniques, similar to those used to balanceconsumption and production (including market rules), couldbe set up to reduce harmonic emissions. This could be asolution for the growing harmonic emission with growingamounts of production and consumption.

Microgrids with islanding capability can, in theory, mitigatevoltage dips by going very quickly from gridconnectedoperation to island operation. The presence of generatorunits close to the loads allows the use of these units inmaintaining the voltage during a fault in the grid.

5.5 Immunity of devicesSimultaneous tripping of many distributed generators due toa voltage-quality disturbance develops a voltage swag. As asmart grid attempts to maintain a balance betweenproduction and consumption, mass tripping of consumptioncould have similar adverse consequences. This should befurther investigated.

5.6 Weakening of the transmission gridThe increased use of distributed generation and of largewind parks will result in a reduction of the amount ofconventional generation connected to the transmissionsystem. The fault level will consequently be reduced, andpower-quality disturbances will spread further. This willworsen voltage dips, fast voltage fluctuations (flicker) andharmonics. The severity of this has been studied for voltagedips. The conclusion from the study is that even with 20%wind power there is no significant increase in the number ofvoltage dips due to faults in the transmission system [17].

VI. POWER QUALITY MONITORING

Growing service quality expectations and reducedpossibilities for grid enforcements make advanceddistribution automation (ADA) an increasingly necessarydevelopment for network operators and the next large step inthe evolution of the power systems to smart grids.

The management of the distribution system is mainly basedon the information collected from the power flows by anintegrated monitoring system. This enables realtimemonitoring of grid conditions for the power systemoperators. It also enables automatic reconfiguration of thenetwork to optimize the power delivery efficiency and toreduce the extent and duration of interruptions. The basicpart of the monitoring system infrastructure is based onsensors, transducers, intelligent electronic devices (IED) and(revenue) meters collecting information throughout thedistribution system.



A number of network operators have already proposed thatthe smart grid of the future should include: Network monitoring to improve reliability, Equipment monitoring to improve maintenance, Product (power) monitoring to improve PQ.

In order to achieve these goals, the actual distributionsystem infrastructure (especially meters and remotelycontrolled IEDs) should be used to gather as muchinformation as possible related to network, equipment andproduct (i.e. power quality and reliability) to improve thedistribution system overall performance. Among the mostimportant ADA operating systems, that a smart grid willinclude, it can be mentioned: Voltage & VAR control (VVC), Fault location (FL), Network reconfiguration or self-healing. Network operators with an ambitious energy efficiency

program have focused on two targets: Capacitor banks installation, Voltage control.

There is also another important goal: to reduce the durationof interruptions. The VVC system requires a permanentmonitoring of the voltage magnitude (averaged over 1 to 5min) at the end of the distribution feeder and the installationof switched capacitor banks. Besides that, the monitoringallows the detection of power quality disturbances such aslong duration under voltages and over voltages, and voltageand current unbalance.

Basically, the voltage regulation system at the substation isreplaced with an intelligent system that uses networkmeasurements to maintain a voltage magnitude for allcustomers within the acceptable upper and lower limits. TheVVC system also analyzes the reactive-power requirementsof the network and orders the switching of capacitor bankswhen required.

An important goal is to prevent potential power qualityproblems due to the switching operations of capacitor banks(with rating up to 1.2 Mvar). [18].

Another goal was to evaluate the joint impact of the VVCsystem and voltage dips occurring on the grid. The results ofthe study indicate that the impact can be quantified by twoeffects: Increasing number of shallow voltage dips is expected.

Voltage reduction from 2 to 4% is obtained due to VVCsystem. Added to this is the voltage drop due to thefault: drops of 6 to 10% (not counted as dips) becomedrops of 10 to 12% (which are counted as dips).

Equipment malfunctioning or tripping: the jointcontribution of the VVC system and the disturbancebrings the residual voltage level below a criticalthreshold, around 70% of the nominal voltage for manydevices.

Fault location is based either on a voltage drop fault locationtechnique that uses waveforms from distributed powerquality measurements along the feeder or on a fault current

technique based on the measurement of the fault current atthe substation. The average error in locating the fault withthe first technique was less than 2%, in terms of the averagemain feeder length. An accurate fault-location techniqueresults in a significant reduction in the duration of(especially) the longer interruptions. The informationcollected by the fault-location system can also be used forcalculating dip related statistics and help to betterunderstand the grid behaviour [19].

The third application, network reconfiguration or selfhealing, is based either on local intelligence (belonging tomajor distribution equipment controllers) or on decisionstaken at the power system control centre, which remotelycontrols and operates the equipment used for networkreconfiguration (reclosers and switches).

The impact of these applications on the distribution networkand its customers is permanently evaluated. Theinfrastructure belonging to ADA systems can be shared by apower-quality monitoring system capable of real timemonitoring. Depending on the type of ADA application orsystem, the monitoring can be done either at low-voltage orat medium-voltage level. In the first case monitoring devicesmay belong to an Advanced Metering Infrastructure (AMI)and in the second case they may belong to the distributionmajor equipment itself (i.e. controllers) (see Fig. 5) [20].

Figure 5. Distribution of automation equipments in a Smart grid

The smart grid will allow a continuous power-qualitymonitoring that will not improve directly the voltage qualitybut will detect quality problems helping to mitigate them.

VII. POSSIBLE SOLUTIONS FOR POWER QUALITYFALL IN SMART GRID

Equipment responds very differently to harmonicdistortions, depending on their method of operation. Forexample incandescent lights and different household heatersare not affected by them. On the other hand, induction motorwindings are overheated by harmonics, causing accelerateddegradation of insulation and so the lifetime of the machinecan shorten in an abrupt way. The problem is that harmonicvoltages can give correspondingly higher currents than do50 Hz voltages and one can easily underestimate the degreeof additional heating in the motor [20].

It is a widely known fact that faults such as the broken rotorbars induce sideband harmonic components to the stator



current spectrum of the induction machine. Those harmonicscan be used for detecting the faults. As most of electricalmachines today are used in hand with frequency converters,then those converters add additional variables to theproblem. Frequency converter causes supply frequency tovary slightly in time and, as a result, some additionalharmonics in the current spectrum are induced andsidebands are reduced or even hindered. This phenomenonalso raises the amount of noise in the test signals, whichmakes the faults more difficult to detect.

In that sense it could prove to be useful to use a certain on-line diagnostic system in the grids with dispersed generationand the wind generators that are integrated to this system.This could be a helpful tool to detect the faults at an earlystage where the repairing of the machines is still possibleand reasonable. Also it would help to differentiate thedeviations and harmonic distortions in the grid from thefaulty cases of the machines.

VIII. CONCLUSION

The new technology associated with smart grids &integration of dispersed generation, offers the opportunity toimprove the quality and reliability as experience by thecustomers. However it also results in the increase ofdisturbance levels in several cases and thereby introduces anumber of new challenges, which should not be used asarguments against the development of smart grids; rather weshould develop appropriate monitoring & mitigationtechniques [14] for power quality.

Irrespective of how the term smart grid is defined, one cansafely state that electricity networks will face newchallenges in the future, and that current and futurechallenges can be solved by a set of technologies that eitherexist today, or are being actively developed. If more andmore dispersed generation is going to be installed all overthe power networks like it seems to go then it is mostimportant to find measures for power quality and security ofsupply.

In the situation where generation as well as consumptionproduces decrease of power quality in the grid, it is essentialto analyze both generation and consumption in a verythorough way. If it proves to be necessary it might makesense to limit the usage of new plants and appliances or usesome other methods to decrease their negative effect topower quality. Nevertheless, appropriate on-line diagnosticsof dispersed generation units must be applied to guaranteesufficient power quality, supply reliability and overall safetyof customers and different.

However they should attract attention to the importance ofpower quality for the successful and reliable operation ofsmart grids. New developments need new approaches andperspectives from all parties involved (network operators,equipment manufacturers, customers, regulators,standardization bodies, and researchers).

REFERENCES

[1] J. Douglas, "Solving problems of power quality", inEPRI journal, vol. 18, no. 8, pp 6-15 Dec. 1993.

[2] Estonian Ministry of the Environment. Euroopa Liidukliimapoliitika, 2012.

[3] K.Purchala, R. Belmans, & et al, “Distributedgeneration and the grid integration issue”, KULeuven,Imperial College London.

[4] M. Bollen, Understanding power quality problems:Voltage Sags and Interruptions, 1st ed., Wiley-IEEEPress, 2000, p. 543.

[5] M.Bollen, M. Häger, “Power quality: interactionsbetween distributed energy resources, the grid, andother customers”, Electric Power Quality andUtilisation Magazine, vol. 1, no. 1, pp. 51–61, 2005

[6] M. Tremblay, R. Pater, F. Zavoda, G. Simard,Accurate fault-location technique based on distributedpower-quality measurements, CIRED Conf- 2007.

[7] R. C. Dugan, M. F. Mcgranaghan, S. Santoso & H.W.Beaty, Electrical Power Systems Quality, 2nd Edition,2004, TMH Education Private Ltd., New Delhi.

[8] International Electro technical Commission, IEC61000-4-7.Electromagnetic compatibility (EMC).

[9] Institute of Electrical and Electronics Engineers, IEEEStd 1159-1995, IEEE Recommended Practice forMonitoring Electric Power Quality, 1995, pp.11 – 25.

[10] T. Vaimann, J. Niitsoo, T. Kivipõld, “Dispersedgeneration accommodation into smart grid”, in Proc. ofthe 52nd International Scientific Conf. of RTU. Sectionof Power and Electrical Engineering, 2011.

[11] F. C. De La Rosa, Harmonics and power systems, CRCPress, 2006, pp. 1–184

[12] . Ackermann, G. Andersson, L. Söder, “Distributedgeneration: a definition”, Electric Power SystemsResearch, vol. 57, pp. 195–204, 2001

[13] M. Bollen, Adapting electricity networks to asustainable– smart metering and smart grids, SwedishEnergy Markets Inspectorate, 2011, pp. 1–115.

[14] A. K. Barik, D. Tripathy & et al, “Detection &Mitigation of Power Quality Disturbances using WPT& FACTS Technology”, International Journal ofScience, Engineering & Technology Research, Vol. 2,Issue 2, pp.336-343, Feb. 2013.

[15] T. Senjyu, Y. Miyazato& et al. Optimal distributionvoltage control and coordination with distributedgeneration. IEEE Transactions on Power Delivery,23(2):1236–1242, April 2008.

[16] I. Wasiak, M.C. Thoma& et al. A power-qualitymanagement algorithm for low-voltage grids withdistributed resources. IEEE Transactions on PowerDelivery, 23(2):1055–1062, April 2008.

[17] M.H.J. Bollen, M. Speychal, K. Lindén, Impact ofincreasing amounts of wind power on the dipfrequency in transmission systems, Nordic WindPower Conference, Helsinki, Finland, May 2006.

[18] F. Zavoda, C. Perreault, A. Lemire, The impact of avolt & var control system (VVC) on PQ andcustomer’s equipment, IEEE PES T&D Conf- 2010

[19] F. Zavoda, M.H.J. Bollen, M. Tremblay, Thebehaviour of power distribution feeder dips, CIREDConference 2009.

[20] N. Watson, S. Hirsch, “The impact of compactflourescent lamps on power quality”, in Proc. ofConference of the Electric Power Supply Industry,1994, pp. 416–428.

***



Power Factor, its Impact on net Energy savingsand other Tangible Benefits

Mallikarjun A Kambalyal. Certified Energy Auditor-EA 3485.SUNSHUBH RENEWABLES & RESEARCH CENTRE,

31, Pagadi Street, Hubli-580020. Karnataka.

The significance of the article is to evaluateand table the findings of my energy audit. Ithas been the responsibility of everyelectrical energy user to fix his PF nearer tounity. In the interest of those who are nontechnical, please be known that the PF inelectrical terms is very much similar toblood pressure of a human being. Just as thedeviation from the normal range isdangerous, so is the case of PF in theelectrical system. A low PF or a Very high(leading) PF will cause heavy loss; in caseof the human system it is the life loss.

To bring on table the energy scenario, let usconsider the energy consumption pattern.While every urbanite blames the high powerconsumption by the agricultural sector forthe huge power loss, it is more important tolook at the facts and figures before weconsider the main cause for power loss andthe possible scope for savings. The fact that,the power consumed by industry sector,accounts for 50.7% against the 3.5% ofagricultural sector has been neglected by andlarge but the fact is that the huge power lossexists in this sector is very much a fact. Thisis mainly due to the length of thetransmission cable laid for the agriculturalsector. If the industry sector is by & largeeducated the agricultural sector is largelyuneducated. In both the cases we have theproblem of power factor. The industry is notmuch concerned because the energy usepattern is unpredictable and or is advised byunskilled electrical contractors. Theagriculturists are not to be blamed at all forthe fact that they know not about Power

Factor & its purpose, more important theyget the power for free almost. Hence thepoint of concern is for those who wish tosave on distribution loss either within oroutside. The power factor parameter hasbeen discussed widely by eminenttechnicians and everyone has a point tosupport their views.

Let me put my experience w.r.t. to thepower factor, on the energyconservation/saving drive. It has beenwidely discussed that the improvement inpower factor does not reduce active poweror the units consumed. This statement is trueat some point and false at the other. Let mediscuss with few case studies, below. First,all the technocrats know the power iscombination of Active power, Reactivepower and Apparent power. And the PowerFactor is related to all the three powers inthe system. Secondly, what is it thatconstitutes the system? It is the device(motor, resistance heating coil orilluminating device etc.,) alone or it is inaddition to the cable which interlinks thedevice and the power source, contactors,relays, measuring devices etc., and any otherdevice which forms the part of the system asrequired on site. If you consider the powerdrawn by the device alone then the impact ofpower drawn by it certainly is independentof power factor. Is it just the device alonethat constitutes the system? Can this devicework without the other integral components?Certainly not, it is the cable conductor,contactors and all other devices thatconstitute the system which drives the key



device. Hence the total power drawn shouldbe the power drawn by the complete system.Every piece of the system constitutes toactive power drawn. In this contextimportance of the power factor is highlyrelevant and should be a key parameterwhen the power loss is accounted for.

Let us consider a case. The unit underdiscussion manufactures automobilecomponents with good number of CNCmachines and has a contract demand of 1300KVA. The average power factor that waspossible with APFC system was 0.87. Theunit installed new APFC system and

connected large sized capacitors and still thepower factor showed no improvement (?).Our energy audit showed that the operatingmethod of the unit was much below themanufacturers assumed loading on themachines. In fact the load on the motors wasnever felt by it because there was no majorchange in no-load and on-load power. Theother findings were that the current in thecable was too high when compared to theactive power. The table below gives thecurrent drawn at different timings and theactual active power with power factorreadings.

* These readings are during lunch break, the machines are running idle.

The power factor improvement was taken upand the following readings were recorded.The pre & post power factor correctionexercise showed that current in cable

dropped substantially (as predicted), it wasalso found that the current harmonicsincreased five times

.



The current reduction was also associatedwith the KWs (active power) reduction. Thefact that the same job work was operated,before and after correction, a 10% plus,reduction in active power was seen. This

was mainly attributed to the system loss. Asit is seen from the table 1 above the cableresistance and the current flow in the cablewas measured to arrive at the cable &accessories loss

.

The loss due to accessories is reflectingbeyond acceptance. However it is thereading on site and has to be accepted theway it occurs. The point of discussion is justthat the impact of Power Factor on theworking system as a whole. Extending thediscussion further to the total energyconsumption scenario the total energysavings by virtue of power factor

consideration alone works out to be 5% ofthe total energy consumed in the industrysector and .35% in agricultural sector. Thefact that the agricultural sector is spreadover vast area the actual energy loss is muchmore than 0.35%. If we can generalize theenergy loss in the agricultural sector isnearer if not less than industry sector.



Inference: it is to be considered by the largescale industries and the utilities andascertain the actual loss rather thangeneralize the loss only on agriculturalsector.CONCLUSION

i. It is now very prominent that theimprovement in the power factorhelps reduce power consumptionboth in terms of units and thecontract demand.

ii. The relation of power factor withactive power drawn by the deviceshould be extended to the system aswhole and not the device alone.

iii. The current in the cable reducedinversely.

iv. The cable temperature was also less.v. The loss due to drop in temperature

will further add to our benefits.vi. The voltage level across the

terminals improved marginally.vii. The impact of surge loads is not felt

by the smaller motors/machines.

The other tangible benefits are:

i. Under the given cable size availableon the site, more machines can beloaded, which implies reducedinfrastructure cost.

ii. The power consumption is reducedby around 10% which otherwisewould have been lost as I2R losses.

iii. The other benefits of PF correctionwill also prevail over the entireindustrial establishment.

PRECAUTION

The point of concern or a critical factor isthe RYB polarity that the industry has tofollow. If the polarity is not followed thenthe result is that the PF would reduce and

not increase in spite of any of the valueconnected in the network.

The sequencing of Capacitors is equallyimportant when the connections are beingmade. The centre lead should always beconnected to the Yellow phase when RYBpolarity is maintained within the industry.

EXTENSION

The logic when extended to the utilities thatare instrumental in power distribution willfind that their losses can be substantiallyreduced. Considering the length of thedistribution network the utility can achievetheir distribution loss cutting in the firstinstance alone over 10%.

FURTHER STUDY

The impact of current harmonics on thesystem has now to be carried out in depthand rectification measures implementedaccordingly.



Recent Advancements in Protectionof Smart Grid

Rajlaxmi SahaHI-TECH COLLEGE OF ENGINEERING,BPUT,ODISHA

Email:[email protected]

Abstract - The concept of smart gridstarted with the idea of advanced meteringinfrastructure (AMI) and its initial goalwas improving the demand sidemanagement, energy efficiency anddeveloping a self healing electrical grid toimprove the power system reliability.Smart grid also called as future grid orintelligent grid. It uses two way flow ofelectricity and information to createautomated and distributed advancedenergy delivery network. It changes thepower flow and recovers the powerdelivery service. It is also two way cybersecure communication technologies andcomputational intelligence in an integratedfashion across electricity generation,transmission, substation, distribution andconsumption to achieve a systemsustainable. There are also ways toimprove protection as well as security andprivacy issues in the smart grid. Cybersecurity is critical issue due to increasingpotential of cyber attacks. Organisationsare working on the development ofsecurity and protection of smart grid. It isused to provide secure authentication andauthorization and smartly protecting thegrid is key feature in this paper.

I. INTRODUCTION

Smart grid is a method to decreasedependency on energy sources, reduceemission of global warming componentsand create a reliable sources of electricity.

It is two way flow as grid as in smart grid,electricity can also be put back into grid byuser. For example So, it is necessary toprovide security and protection. Securitymeans the cyber attacks and protectionfrom transient. Internet based IPV4andIPV6 developed many years willprovide cost effective transport. One wayis by SCADA i.e. Supervisory control anddata acquisition solution. Which havevarious capabilities and securities andother one is transient environment of smartgrid. The lightning will continue toproduce direct and coupled transient thatpropagate in conductor until it groundedby surge devices. This paper shows theoverall protection of smart grid.[1][6]

II. SMART GRID TECHNOLOGY

Smart grid can be regarded as an electricsystem that uses two way cyber securecommunication technologies and largewind turbines and photovoltaic arrays. Itwill also require higher degrees of networkwhich connected to new features. Cybersecurity is critical is critical issue it canlead to user errors, equipment failure andnatural disaster. Its aim is to improvingdemand side management and energyefficiency and constructing self healingreliable grid protection. Reducing oilconsumption, reducing greenhouse gas,accommodating[7] distributed powersources, automatic maintenance andoperation and increasing consumer choice.



Many organisation uses the wind turbineand photovoltaic technology together ashybrid. Which utilized by individualorganisation or residential consumer andconnect to smart grid. As in fig(1).

Fig:1. Smart grid conceptual model.

1). Security Requirement and Solution.

Smart grid has structural securityvulnerability in power transmissionnetworks. The cascade based attackvulnerabilities in power grid have beenfounded. Attack way is from cyber whichcause traditional topological metrics andother power operation failure. The UScentral intelligence agency says criminalshave hacked into the computer system ofutility companies, cutting the power toseveral internal cities [1]. Among variousstandards bodies the security of griddepend on authentication, authorizationand privacy technology. Federalinformation processing standard (FIPS)approved , Advanced encryption standard(AES) and Triple data encryptionalgorithm (3DES) solution, offering strongsecurity and high performance. Specificexample given as National Institute ofStandards and Technology [2] (NIST)which has number of programs hasdetermined . 3DES solution become insure

in 2030. AES preferred solution for newcomponents. These technology are sort ofmanagement.

2). Smart Grid Transient Environment.

With numerous large and small windturbines and photovoltaic solar arraysnetworked to the Smart Grid, significantamounts of conductors will be utilized toconnect the network as in fig(2). It isplanned that these conductors will havelower impedances, which will allowelectricity to flow with lower losses.However, lower loss conductors will resultin higher frequencies, such as thoseassociated with electrical transients, topropagate further through the electricaldistribution network. Lightning surgescoupled onto electrical systems at thefacility level shows surge currents canhave varying amplitudes and durations.Calculations based on lightning rise times,break over voltages (dielectric withstand)of distribution equipment, and impedanceof the earthing system show that lightningcurrents could exceed 120 kA fig(3)[4].Most transients coupled on the electricaldistribution system will have less energycontent, but the magnitudes of lightning-induced transients will be sufficient todamage most electronic equipment. Whileswitching transients are more commonthan lightning-induced transients, the dataon switching transient is limited. Thecomplexities of switching transients makepredicting amplitude, frequency andduration difficult. Switching transients aredetermined by the characteristicparameters of the electrical system:inductance, capacitance, resistance, andconductance. In general, switchingtransients have a voltage that is less thantwo times the nominal system voltage, e.g.480 V on a 240 V system. With moreequipment located in the environment andconnected to the SMART Grid, there is agreater chance that lightning-inducedtransients will be coupled onto the



SMART Grid. Equipment connected inclose proximity of the induced transienthas a higher probability of being subjectedto higher amplitudes. In the SMART Grid,equipment that is normally connected deepinside a building or residence can nowhave direct exposure to the full energycapability of a lightning or switchingtransient.

Fig.2 Smart Grid

Fig.3.Lightning induced amplitudes.[3].

a). Applying SPDs( surge protectivedevice) to the Smart Grid.

Facilities that contribute to the supply ofelectricity to the Smart Grid have twoservice entrance locations: primary powersource and alternate power source (Figure4). Primary power is supplied from theutility. Alternate power is supplied fromon-site resources (e.g. wind turbine,photovoltaic solar array, etc). Facilitieswith dual source power capabilities requirean automatic transfer switch (ATS) deviceto ensure coordination of both sources. As

shown in Figure 4, the system loads of thefacility can be powered by the primarysource (utility) or the alternate source (e.g.wind turbine, photovoltaic solar array, etc).The ATS also allows for the alternatepower source to be networked into theSMART Grid. Providing surge protectivedevices (SPDs) has become a “bestengineering practice” for protectingequipment and processes from transients.Intercepting transients at the incomingpoint of a facility has shown to be the bestmethod of protection [4]. This requires theplacement of SPDs at the electrical serviceentrance of a facility. Once the transienthas been significantly reduced at theservice entrance, additional protection isrequired at distribution panels and at thepoint of use [4]. Protecting equipmentwithin a facility from transient conditionsrequires SPDs to be applied at the serviceentrance of the primary power andalternate power. To provide the bestprotection, SPDs should be placed as closeto the service entrance as possible. In mostcases, this requires the SPDs to be placedon the input of the service disconnect.SPDs intended to be installed upstream ofthe service disconnect are required to beclassified as a Type 1 SPD [5]. To mitigateany remnant over voltage within thefacility, a Type 2 SPD should be locatedon each subsequent distribution panel.

Fig.4. Illustration of Dual Source Facilitywith SPDs Installed

III. PROBLEM DETECTION ANDMITIGATION.



Many utility customers do not realize thelimited information currently available togrid operators, especially at thedistribution level. When a blackout occurs,for example, customer calls are mapped todefine the geographic area affected. This,in turn, allows utility engineers todetermine which lines, transformers andswitches are likely involved, and what theymust do to restore service. It is not rare,infact , for a utility customer carerepresentative to ask a caller to stepoutside to visually survey the extent of thepower loss in their neighbourhood. It is atestament to the high levels of reliabilityenjoyed by electric utility customers thatmost have never experienced this;however, it is also evidence of anantiquated system. While SCADA andother energy management systems havelong been used to monitor transmissionsystems, visibility into the distributionsystem has been limited. As the grid isincreasingly asked to deliver the abovefour capabilities, however, dispatchers willrequire a real-time model of thedistribution network capable of deliveringthree things: 1) real-time monitoring (ofvoltage, currents, critical infrastructure)and reaction (refining response tomonitored events); 2) anticipation (or whatsome industry specialists call “fast look-ahead simulation”); and 3) isolation wherefailures do occur (to prevent cascades).

IV. CONCLUSION

The electrical grid is poised to transitionfrom a system that was conceived morethan 100 years ago, into a modernengineering marvel. The Smart Grid willrevolutionize generation, distribution andutilization of electrical energy similar tohow the Internet has revolutionizedcommunications. Proponents of the SmartGrid tout that this complex network offossil fuel and renewable energy sourceswill be capable of generating anddistributing electricity at lower costs andlower levels of pollutants. In addition, the

Smart Grid will reduce our dependency onforeign sources of fuel and be morereliable. To protect the Smart Grid, and allthe advanced electronic devices that willbe connected to it, high quality surgeprotective devices (SPDs) with provenperformance, demonstrated safety, andunprecedented reliability are required. TheSPD must be capable of connecting toservice equipment to limit lightninginduced transients from entering thefacility or being coupled onto the SmartGrid. Additional SPDs will be required tolimit any transient overvoltage thatpropagates into the facility. To achieve thevision put forth in this paper, there aremany steps which need to be taken.Primary among them is the need for acohesive set of requirements and standardsfor smart grid security. We urge theindustry and other participants to continuethe work that has begun under thedirection of NIST to accomplish thesefoundational steps quickly.

REFERENCES

[1]. Anthony R. Metke and Randy L. Ekl“Security Technology for Smart GridNetworks” IEEE TRANSACTIONS ONSMART GRID, VOL. 1, NO. 1, JUNE2010.

[2]. Draft smart grid cyber security strategyand requirements, NIST IR7628, Sep.2009 [Online]. Available:http://csrc.nist.gov/publications/drafts/nistir-7628/draft-nistir-7628.pdf.

[3]. E.O. Lawrence Berkley NationalLaboratory (2001). Scoping Study onTrends in the Economic Value ofElectricity Reliability to the U.S.Economy. Available [on-line] athttp://certs.lbl.gov/pdf/47911.pdf.Retrieved 2009, July 31

[4]. Institute of Electrical and ElectronicsEngineers (2005). IEEE RecommendedPractice for Powering and Grounding



Electronic Equipment. IEEE Standard1100, Emerald Book. Piscataway, NJ,USA.

[5]. National Fire Protection Association(2008). National Electric Code. NFPA 70.Quincy, MA, USA.

[6]. Litos Strategic Communication(2009). The Smart Grid: An Introduction.Prepared for the U.S. Department ofEnergy under contract DE-AC26-04NT41817, Subtask 560.01.04.

[7]. The Electricity Forum. Windmills forElectricity Explained. Available [on-line]athttp://www.electricityforum.com/windmills-for-electricity.html. Retrieved 2009,August 21.



Application of Fuzzy Logic for Reduction of CurrentHarmonics in Single-Phase Grid–Connected Inverter

1Preeti Ranjan Sahu ,2Satyaranjan Jena,Member IEEE1Department of Electrical &Electronics Engineering, Roland Institute of Technology, Berhampur (INDIA)

2School of Electrical, KIIT University, Bhubaneswar (INDIA)

Abstract— In the present era distributed generation (DG) systemuses current regulated PWM voltage-source inverters (VSI) forsynchronizing the utility grid with DG source in order to meetthe following objectives: 1) To ensure grid stability 2) active andreactive power control through voltage and frequency control 3)power quality improvement (i.e. harmonic elimination) etc. Inthis paper, fuzzy with hysteresis controller is applied to enhancethe power quality by diminishing current error at higher bandwidth. The studied system is modeled and simulated in theMATLAB/Simulink environment.

Keywords-Distributed generation(DG),Fuzzy logic controllerHysteresis current controller, ,Point of common coupling (PCC),uitility grid, Power quality

I. NOMENCLATURE

Vg grid voltageIo grid currentIref reference currentE current errorVdc dc-link voltagefs switching frequencyLf line inductanceVm maximum voltage amplitudeIm maximum current amplitudeθv voltage phase angleθi current phase angle

II. INTRODUCTION

Distributed generation systems and their interconnectionshould meet certain requirements and specifications wheninterconnecting with existing electric power systems (EPS)[1]. For an inverter-based distributed generator, the powerquality largely depends on the inverter controller’sperformance. Pulse width modulation (PWM) is the mostpopular control technique for grid-connected inverters. Ascompared with the open loop voltage PWM converters, thecurrent-controlled PWM has several advantages such as fastdynamic response, inherent over-current protection, good dc-link utilization, peak current protection etc.

However, the converter performance is largely depends onthe applied current control strategy. Very extensive researchwork has been done besides current control techniques and isavailable in the literature. [2].The common strategies of currentcontrollers can be classified as ramp comparator, hysteresiscontroller, and predictive controller amongst which thehysteresis controllers are widely used because of their inherentsimplicity and fast dynamic response [3]. The main objectives

of the control of grid connected PWM-VSI is to 1) ensure gridstability 2) active and reactive power control through voltageand frequency control 3) power quality improvement (i.e.harmonic elimination) etc. The advantage of fuzzy control isthat it is based on a linguistic description and does not requireany mathematical model of the system. The Fuzzy set theorycan be used to model the natural uncertainties of plant andcontrol variables. The studied system is modeled and simulatedin the MATLAB/Simulink environment.

The paper is organized as follows –Single-phase gridconnected VSI is described in Section III. Analysis ofhysteresis and hysteresis with fuzzy current controller isexplained in the section IV. Section V. dedicated to results anddiscussion, followed by conclusion in Section V.

III. SINGLE PHASE GRID-CONNECTED VSI

Figure 1. Single phase inverter connected to utility grid

The single-phase grid connected inverter shown inFig.1.Which is composed of a dc voltage source (VDC), fourswitches (S1-S4), a filter inductor (Lf) and utility grid (Vg). Ininverter-based DG, the produced voltage from inverter must behigher than the Vg in order to assure power flow to grid. SinceVg is uncontrollable, the only way of controlling the operationof the system is by controlling the current that is following intothe grid.

IV. ANALYSIS OF FUZZY WITH HYSTERESIS CURRENT

CONTROLLER

A. Hysteresis band current controller.

In spite of several advantages, some drawbacks ofconventional type of hysteresis controller are limit cycleoscillations, overshoot in current error, sub-harmonicgeneration in the current and uneven switching [4].In case ofhysteresis controller as shown in fig.2 the error is directly fedto the hysteresis band.



Figure 2. Hysteresis –Band Current Controller

As given by equation (1) the reference line current of thegrid connected inverter is referred to as iref and differencebetween io and iref is referred to as error (e). The hysteresisband current controller assigns the switching pattern of gridconnected inverter.

o refe i i (1)

The switching logic is formulated as follows:

If e >HB then switch S1 and S4 is on

If e <-HB switch S2 and S3 is on

The average load power is computed as:

1

1( ) ( )

n

L s Lj

P v j i jn

(2)

Using Torrey and Al-Zalmel[6] methodology, the referencesource current is computed as:

ref gi kv (3)

Where k is the scaling factor and computed as

2

2 L

m

Pk

V (4)

The switching frequency of the system can be calculated as

odc f g

diV L V

dt (5)

From equation (1)

o refi i e (6)

By rearranging equation (5 and 6) we can calculate

2 fON

dc g

L HBT

V V

(7)

And2 f

OFFdc g

L HBT

V V

(8)

1s ON OFF

s

T T Tf (9)

2 2

4dc g

sdc f

V Vf

V L HB

(10)

Hence, the switching frequency varies with the dc voltage,grid voltage, load inductance and the hysteresis band [7].

B. The fuzzy with hysteresis current controller

The main drawback of hysteresis current controller isuneven switching frequency which causes acoustic noise anddifficulty in designing input filters during load changes. Theswitching frequency can be reduced by reducing the bandwidth of the hysteresis band but at the same time the currenterror will increase which produce more distortion in the outputcurrent. To eliminate drawback upto certain extent fuzzy isused along with hysteresis current controller as shown in fig.3[5].

Figure 3. Block diagram for fuzzy with hysteresis current control for single-phase grid-connected VSI

Figure 4. Fuzzy logic controller

The structure of fuzzy logic controller is given below in fig4.Here the membership function is chosen as triangular asshown in fig 5. The input is taken as error (e) and the changein error (Δe) .Total 49 rules are taken into account as given intable -1.

Figure 5. Membership function

For example

If e is negative small (NS) and Δe is positive big (PB)

Then output is positive medium (PM).

Table-1-Rule base for fuzzy controller



V. RESULTS & DISCUSSION

The section reveals the simulation results for fuzzy withhysteresis current control algorithm applied to single-phasemains connected inverter system. The studied model has beendeveloped and simulated in the MATLAB/simulinkenvironment. For simulation, the Dc-link voltage is taken400V, and the grid voltage is 240V, the inductance of the lineis 5mH and the utility grid frequency is 50Hz.

A. Simulation result steady state

0 0.02 0.04 0.06 0.08 0.1 0.12 0.14 0.16 0.18 0.2-300

-200

-100

0

100

200

300

Time (Sec)

Vol

tage

(V)

0 0.02 0.04 0.06 0.08 0.1 0.12 0.14 0.16 0.18 0.2-15

-10

-5

0

5

10

15

Time (Sec)

Curre

nt (A

)

Actual CurrentReference CurrentError

0 0.02 0.04 0.06 0.08 0.1 0.12 0.14 0.16 0.18 0.2-15

-10

-5

0

5

10

15

Time (Sec)

Currn

t (A)

Grid currentLoad CurrentFigure 6. Simulation result of the fuzzy with hysteresis current controller for

steady state (a) grid voltage (Vg) (b) reference current, actual current andcurrent error.0 0.02 0.04 0.06 0.08 0.1 0.12 0.14 0.16 0.18 0.2

-15

-10

-5

0

5

10

15

Time (Sec)

Cu

rren

t (A

)

Actual CurrentReference CurrentError

0 0.02 0.04 0.06 0.08 0.1 0.12 0.14 0.16 0.18 0.2-15

-10

-5

0

5

10

15

Time (Sec)

Cur

rnt (

A)

Grid currentLoad Current

0 0.02 0.04 0.06 0.08 0.1 0.12 0.14 0.16 0.18 0.2-500

0

500

1000

1500

2000

Time(Sec)

Act

ive

Pow

er(W

att)

Rea

ctiv

e Po

wer

(var

)

Active PowerReactive Power

0 0.02 0.04 0.06 0.08 0.1 0.12 0.14 0.16 0.18 0.20

0.5

1

1.5

Time (Sec)

Pow

er F

acto

r

Figure 7. Simulation result of (a) Grid current and load curren (b) Activepower & reactive power

Fig 7 shows that that the proposed controller is able to controlthe active and reactive power independently.

B. Simulation result for transient state

1) Step change in load

To analyze the performance of the hysteresis and fuzzywith hysteresis controller the load is changed between theperiod 0.06 and 0.14 sec.

0 0.02 0.04 0.06 0.08 0.1 0.12 0.14 0.16 0.18 0.2-15

-10

-5

0

5

10

15

Time (Sec)

Cur

rent

(A

)

Actual currentReference CurrentError

Figure 8. Simulation result of reference current, actual current and error forchange in load for fuzzy with hysteresis current controller

0 0.02 0.04 0.06 0.08 0.1 0.12 0.14 0.16 0.18 0.2-15

-10

-5

0

5

10

15

Time (Sec)

Cu

rren

t (A

)Grid CurrentLoad Current

0 0.02 0.04 0.06 0.08 0.1 0.12 0.14 0.16 0.18 0.2

0

500

1000

1500

2000

Time (Sec)

Act

ive

Po

wer

(w

att)

Rea

ctiv

e p

ow

er(v

ar)

Active Power

Reactive power Power

0 0.02 0.04 0.06 0.08 0.1 0.12 0.14 0.16 0.18 0.20

0.5

1

1.5

Time (Sec)

Pow

er F

acto

r

Figure 9. Simulation result of (a) load current and grid current (b) activepower and reactive power.

From fig 8 we can observe that the distortion in the invertercurrent is less and hence the error is less.

Fig.9 shows the change in active power, load, current andgrid current during the load transient which shows that thedynamic response faster of the Fuzzy logic controller.

2) Change in hysteresis band width

In this case the hysteresis band width is changed at timet= 0.1 sec from HB=1 to HB=3



0 0.02 0.04 0.06 0.08 0.1 0.12 0.14 0.16 0.18 0.2

-10

0

10

Cu

rren

t (A

)

0 0.02 0.04 0.06 0.08 0.1 0.12 0.14 0.16 0.18 0.20

2

4

6

8 x 104

Time (Sec)

Frq

uen

cy

(Hz)

HB=3

HB=3

HB=1

HB=1

Figure 10. Simulation result of grid current ,error and switching frequencyfuzzy with hysteresis controller.

As the band width of the hysteresis controller is increasingthe switching frequency decreases but the current errorincrease so also the distortion in the grid current increases. Inthe proposed controller even if the band width increases thedistortion in grid current and change in error is very less asshown in fig 10, It implies that the switching frequency can bedecreased without hampering the power quality.

In studied current control scheme the inverter is also ableto inject the current in phase with the grid voltage.fig 11. (a)Shows that the current is in phase with the voltage even if thethere is a phase change in the grid voltage at 0.1 sec. Fig 11.(b) Indicates the power factor of the grid current which isunity.

The THD of the proposed controller is considerably less1.66% as shown in fig.13

3) Phase change in the grid voltage

0 0.02 0.04 0.06 0.08 0.1 0.12 0.14 0.16 0.18 0.2

-200

-100

0

100

200

Time (Sec)

Vol

tage

(V)

Cur

rent

(A

)

Phase=0 Phase=60

(a)

0 0.02 0.04 0.06 0.08 0.1 0.12 0.14 0.16 0.18 0.20

0.5

1

1.5

Time (Sec)

Pow

er F

acto

r

(b)

Figure 11. Simulation result of (a) grid voltage and inverter current frequency(b) power factor

0 0.02 0.04 0.06 0.08 0.1 0.12 0.14 0.16 0.18 0.2-10

-5

0

5

10Selected signal: 10 cycles. FFT window (in red): 1 cycles

Time (s)

0 500 1000 1500 2000 25000

50

100

Frequency (Hz)

Fundamental (50Hz) = 10.76 , THD= 1.66%

Mag

(%

of

Fu

nd

amen

tal)

Figure 12. THD of grid current of fuzzy with hysteresis current comptroller

VI. CONCLUSIONS

The paper presents the control grid connected PWM VSIusing fuzzy with hysteresis controller in the control loop.From the study we observed that, fuzzy with hysteresis currentcontroller can able to enhance the power quality of the gridsystem as it is enable to reduce switching frequency even ifthe band width increased without any significant increase inthe current error. As a result, the THD level of grid current isconsiderably reduced. More over, switching frequency of theinverter system has been reduced, hence switching losses arealso reduced to certain extent.

REFERENCES

[1] Yaosuo Xue; Liuchen Chang; Sren Baekhj Kjaer; Bordonau, J.;Shimizu, T.; , "Topologies of single-phase inverters for small distributedpower generators: an overview," IEEE Transactions on PowerElectronics, vol.19, no.5, pp. 1305- 1314, Sept. 2004.

[2] M.M.P.Kazmierkowaski, L.Malesani: “PWM Current ControlTechniques of voltage source converters-A Survey” IEEE Trns. OnIndustrial Electronics,Vol.45, no.5, oct.1998.

[3] Blaabjerg, F.; Teodorescu, R.; Liserre, M.; Timbus, A.V., “Overview ofControl and Grid Synchronization for Distributed Power GenerationSystems” IEEE Transactions on Industrial Electronics, Vol.:53, Issue:5,2006 , Page(s): 1398 – 1409.

[4] Ho, C.N.-M.,Cheung, V.S.P.,Chung, H.S.-H.” Constant-FrequencyHysteresis Current Control of Grid-Connected VSI without BandwidthControl”,IEEE Trans. on Power Electronics, TPEL,2009 Volume: 24,no. 11 , 2009, Pp:2484 – 2495

[5] Hilloowala, R.M.; Sharaf, A.M.; "A rule-based fuzzy logic controller fora PWM inverter in a standalone wind energy conversion scheme," IEEETrans. on Industrial Applications, vol.32, no.1, pp.57-65, Jan/Feb 1996

[6] D. Torrey and A. Al-Zamel, "Single-phase active power filters formultiple nonlinear loads," IEEE Transactions on Power Electronics, vol.10, no. 3, pp. 263-272, May 1995.

[7] Krismadinata,Rahim N.A.,Selvaraj,J.,” Implementation of HysteresisCurrent Control for Single-Phase Grid Connected Inverter”International. Conference on Power Electronics and Drive Systems,2007. PEDS '07., pp: 1097 – 1101.



Synthesis and Applications of Conjugated polymers & it’s Supramolecular

self assembly for Organic Photovoltaics

Duryodhan Sahu,a Dhananjay Kekuda,b Chih-Wei Chub

a Dept of Chemistry, National Institute of Science and Technology, Berhampur, Odisha, India

b Research Centre for Applied Sciences, Academia Sinica, Taiwan

Abstract: Two novel symmetrical acceptor-donor-acceptororganic dyes (S1-S2) containing 2,7-functionalized Fluorene(electron donor) cores connected with two anchoringcyanoacrylic acid (electron acceptor) termini via differentnumbers (2 or 3) of conjugated thienyl linkers were designedand synthesized. These H-Donor (proton donor) dyes (S1-S2)were complexed with a H-acceptor side-chain homopolymercarrying pyridyl pendants (with 1/2 molar ratio of H-donor/H-acceptor) to produce two novel hydrogen-bonded (H-bonded)polymer networks (PFNA/S1-S2). All the dyes, polymer arecharacterized by both 1H- and 13C- NMR spectroscopy. Inaddition the H-Bonded polymer networks were confirmedthrough FT-IR spectroscopy. There was an overlap of theemission spectra of proton accepting polymer PFNA and theabsorption spectra of the proton donating dyes (S1-S2) thus theenergy transfer from proton accepting polymer PFNA to protondonating dye has been expected. Furthermore the effects of thesupramolecular architecture on optical, electrochemical, andorganic photovoltaic (OPV) properties were investigated.Photophysical properties revealed that the absorption spectra ofH-bonded polymer networks showed broader wavelengthranges and higher λonset values to appeal better light harvesting,though they showed blue shifts in their respective absorptionmaxima (λmax). The H-bonded polymer network PFNA/S2containing S2 dye showed the better photovoltaic performancewith a PCE value of 0.32%, than PFNA.

Keywords— H-bonded polymer network; rod-coilhomopolymer; solar cell; self-assembly; supramoleculararchitecture.

Introduction

Conjugated oligomers and polymers are found to beadvanced materials for the present study of solar cells.Oligomers have their unique advantage of their well definedchemical structure, specific electronic and optical properties.However, the material properties of oligomers are generallyauxalary to those of their polymeric analogues, sinceentanglements of macromolecular chains, which areresponsible for the polymer properties, are lacking.1-2 Meijerand coworkers made use of a supramolecular H-bondedpolymer containing oligo(phenylene vinyene) and ureido-pyrimidinone unit for the applications in OPV devices3 .Afterwards, Lin et.al reported on supramolecular assembliesof H-bonded side-chain polymers which were prepared bycomplexation of solar cell dyes bearing carboxylic acids asprotondonors (H-donors) with side-chain polymers bearing

pyridyl pendants as proton-acceptors for the applications inOPVs 4

In general, different attempts of utilizing conjugatedoligomers/polymers or both have been made to design electrondonor-acceptor architectures to enhance the intramolecularcharge transfer interactions from the electron donors to theelectron acceptors and thus to increase the PCE values. In fact,different solar cell architectures have been developed, usingdifferent types of polymer backbones. However, the interestson oligomers with easy purification processes and lack ofproblems in molecular weight distributions etc. are generallyauxiliary to their polymer analogues due to the better solventprocessabilities and film morphologies in photovoltaic cells.5

Therefore, in order to get the advantages of botholigomeric and polymeric properties, an attractive approach ofwell-defined supramolecular architectures of -conjugatedoligomers with the processabilities of polymers could be anattractive approach in the field of organic photovoltaics. Thus,in this report, In order to get the advantage from physicalproperties of both small molecules and polymers in OPVdevices, symmetrical Fluorene-based conjugated dyescontaining Fluorene cores functionalized at 2,7-substitutedpositions linked to cyano-acrylic acid termini (as double H-donors) via -conjugated oligo-thiophenes were synthesized,which were complexed with a side-chain homopolymerPoly[10-(6-(9,9-diethyl-7-(pyridin-4-yl)-9H-fluoren-2-yl)naphthalen-2-yloxy)decyl methacrylate] (PFNA), bearingpyridyl pendants (as H-acceptors). Then, double H-donors ofdyes and H-acceptors of side-chain homopolymer PFNA canbe self-assembled into H-bonded cross-linking polymers.

1. Synthesis

1.1Synthesis of polymer and Dyes(S1,S2)The H-Accepting polymer (PFNA) and dyes S1 and S2 weresynthesized by simple step by step procedures as shown inscheme 1 and 2 and the detailed procedures are given asfollows:

2, 7-Dibromofluorene. (1) Fluorene (10 g, 60.21 mmol)was dissolved in 100 ml of dry dichloromethane and cooledunder an ice bath, then bromine (7.71 ml, 150.52 mmol) wasadded dropwise over 30min. The reaction mixture was stirredat room temperature for overnight, and a saturated aqueoussolution of sodium bisulfite was added to quench excess ofbromine. The solution was extracted from dichloromethane



and washed with water and little brine. The organic layer wasdried over magnesium sulfate, solvent wasScheme 1: Synthetic procedure for PFNA

removed by rotavapor.The crude product was purified byrecrystallization from hexane to yield white crystals (15.41g,79%). 1H NMR CDCl3, δ(ppm): 7.64(s, 2H), 7.58(d, J = 8.0Hz, 2H), 7.46(d, J = 8.2 Hz, 2H), 3.85(s, 2H)

2, 7-Dibromo-9, 9-diethylfluorene (2). A mixture of 2,7-dibromofluorene (10 g, 30.86 mmol) and potassium-tert-butoxide (10.37 g, 92.58 mmol) were stirred in 120 ml of dryTHF under nitrogen for 1hr, then ethyl iodide (7.46 ml, 92.58mmol) was added drop wise and stirred for additional 3hr.Excess of THF was removed by rota vapor, compound wasextracted with dichloromethane and washed with water and alittle brine. The organic layer was dried over magnesium

sulfate. The solvent was removed by Rota vapor, and thecrude product was purified by column chromatography usingSilica, Hexane as an eluent to yield a white solid (10.7 g,91.2%). 1H NMR (300 MHz, CDCl3), δ(ppm): 7.52 (d, J = 3.0Hz 2H), 7.44 (d, J = 12.0 Hz, 2H), 7.42 (s, 2H), 1.95 (q, J =7.2 Hz, 4H), 0.29 (t, J = 7.5 Hz, 6H),

2-Bromo-9, 9-diethylfluoreneboronic acid (3). 5 g (13.15mmol) of 2, 7-dibromo-9, 9-diethylfluorene was dissolved in100 ml of dry THF and the solution was cooled to -78ºC,under nitrogen atmosphere then 6.8 ml (17.10 mmol) of n-BuLi (2.5M solution in hexane) was added through a syringeand the reaction mixture was allowed to reach roomtemperature, then again cooled to -78ºC and 12.13 ml(52.58mmol) of tri-isopropyl borate was added. The reactionmixture was stirred for additional 12 hr at room temperaturethen 50 ml of 2.0M HCL was added and stirred for another 2hr



. Excess THF was removed and the compound was extractedfrom dichloromethane, washed with water followed by a littlebrine. The organic phase was dried with anhydrousmagnesium sulphate then the solvent was removed usingrotavapor. The crude product was purified by columnchromatography on silica using ethyl hexane: acetate (3:1) aseluent to yield a white solid (2.79 g, 61.5%) 1H NMR (300MHz, DMSO-d6), δ(ppm): 8.07(s, 2H,B-OH) ,7.81(s, 1H),7.77 (s, 2H), 7.75(s, 1H), 7.65(d, J = 1.8 Hz, 1H), 7.51 (dd, J= 10.2 Hz, 1H), 1.98 (q, J = 7.8 Hz, 4H), 0.17 (t, J = 7.2 Hz,6H).

4-iodopyridine (4). 4-Aminopyridine (7 g, 74.37 mmol)was dissolved in 48% aqueous HBF4 (70 ml) and cooled to -10ºC. To the resulting slurry powdered NaNO2 (6.67g,96.6mmol) was added portion wise after each 1min interval atsuch rate that no nitric oxide evolution could be detected.After 30 min the diazonium salt was filtered off. This salt wasadded portion wise to a solution of KI (30.86 g, 185.90 mmol)in 500 ml of an acetone: water (2:3) mixture. The reactionmixture was then decolorized by adding aqueous Na2S2O3,

then neutralized with saturated aqueous Na2CO3 and finallyextracted with dichloromethane. The organic layer was driedover MgSO4 and evaporated, and then the crude product waspurified by column chromatography on Al2O3 usingdichloromethane as eluent. To yield white crystal (10.3 g68%) 1H NMR (300 MHz, CDCl3), δ(ppm): 8.27(d, J = 6.3Hz, 2H) 7.67(d, J = 6.0 Hz, 2H).

2-bromo-9, 9-di-n-ethylfluorene-7-pyridine (5): Mixture ofcompound 3 (2 g, 5.79 mmol), 4 (1.3 g, 6.37 mmol), K2CO3

(1.1 g, 7.55 mmol) were dissolved in 100 ml of toluene andethanol (3:1) and degassed for 10 min. then Pd (PPh3) 4

(100mg, 8.69 × 10-5 mmol) was added and then the resultingmixture was stirred under reflux for 4hr.The solvent wasremoved under vacuum, extracted with dichloromethane andwashed with water followed by a little brine . The organicphase was dried with anhydrous MgSO4 and the solvent wasremoved under vacuum then the compound was purified bycolumn chromatography on Al2O3 using 10:1 mixture ofhexane and dichloromethane as an eluent to yield low meltingwhite solid (1.7g 76%) 1H NMR (300 MHz, CDCl3), δ(ppm)8.66(d, J = 6.0 Hz, 2H), 7.78(d, J = 9.0 Hz, 1H), 7.58 – 7.64(m, 5H), 7.48 (dd, 2H), 2.04(q, 4H),0.35(t, J = 6 Hz, 6H)

10-(6-bromonaphthalen-2-yloxy) decan-1-ol. (6b)6-Bromonepthalene-2-ol (7g, 31.3 mmol), 10-bromo decanol

(9.67g, 40.79mmol), K2CO3 (13g, 94.14), and a catalytic

Scheme2: Synthetic procedure for Dyes (S1, S2)

amount of KI was dissolved in 200 ml of dry acetone andstirred under reflux for 24hr. After cooling to roomtemperature, the potassium salt was filtered off. The solventwas removed by rota vapor then extracted withdichloromethane followed by brine wash. The organic solventwas dried over anhydrous MgSO4 and was removed using rotavapor. The crude product was purified by columnchromatography on silica using mixture of hexane anddichloromethane (3:2) as an eluent, to get a light brown solid.(11.2g, 94%). 1H NMR (300 MHz, CDCl3), δ: 7.90 (s, 1H),7.63 (d, J = 9Hz, 1H), 7.57 (d, J = 9Hz, 1H), 7.47 (d, J = 3Hz,1H ) 7.15 (d, J = 12Hz, 1H) 7.08 (s, 1H) 4.05 ( t, J = 6Hz, 2H)3.64 ( t, J = 6Hz , 2H) 1.79 -1.88 (m, 2H), 1.52 - 1.59 (m, 2H),1.32 – 1.49 (m, 12H) .

(10-(6-bromonaphthalen-2-yloxy) decyloxy)(tert-butyl)dimethylsilane. (7b). Mixture of compound 6b (5g,13.2mmol), imidazole (1.16g, 17.15mmol) were dissolved in100 ml of dry dichloromethane, then degassed for 10min.Tert-Butyl-chloro-dimethyl-silane (2.60, 17.15mmol) was added at0º C, then stirred overnight at room temperature undernitrogen atmosphere, then extracted with dichloromethane andwashed with water .The organic solvent was dried overanhydrous MgSO4 then the solvent was removed usingrotavapor. The crude product was purified by columnchromatography on silica using mixture of Hexane anddichloromethane (4:1) as eluent to get the product (5.53 g,85%) 1H NMR (300 MHz, CDCl3), δ: 7.89 (s, 1H), 7.64 (d, J= 9Hz, 1H) 7.58 (d, J = 6Hz, 1H), 7.48 (d, J = 15Hz, 1H), 7.16(d, J = 15Hz, 1H), 7.07 (s, 1H), 4.64 (t, J = 6Hz, 2H), 3.62 (t, J= 7.5Hz, 2H), 1.79 – 1.84 (m, 2H), 1.45-1.53 (m, 6H), 1.29-1.32 (m 8H), 0.90 (s, 9H), 0.01 (s, 6H).Tert-butyldimethyl(10-(6-(4,4,5,5,-tetramethyl-1,3,2-dioxaborolan-2-yl)naphthalene-2yloxy)decyloxy)silane.8(b).5 g of compound-(7b) (10.13 mmol) was dissolved in 100 mlof dry THF and the solution was cooled down to -78ºC undernitrogen atmosphere then 8.1 ml of n-BuLi (2.5M in Hexane,20.26mmol) was added. The solution was warmed up to roomtemperature than cooled to -78ºC. To this solution 5.51 ml(20.26 mmol) of 2-Isopropoxy-4, 4, 5, 5-tetramethyl-[1, 3, 2]dioxaborolane was added and the resulting solution was stirredfor overnight at room temperature. 50 ml of water was addedand further stirred for 1hr. THF was removed by rotavaporthen extracted with dichloromethane and washed with water.The organic solvent was dried over anhydrous MgSO4 and itwas removed on rotavapor crude product was purified bycolumn chromatography on silica using mixture of hexane anddichloromethane (4:1) as eluent to yield white solid

(3.8 g, 70 %) 1H NMR (300 MHz, CDCl3), δ: 8.27 (s, 1H),7.79(dd, J = 9 Hz, 2H), 7.69(d, J = 6Hz 1H), 7.14(d, J = 3 Hz,1H), 7.10(s, 1H), 4.07 (t, J = 6.5Hz, 2H), 3.60 (t, J = 7.5Hz,2H), 1.79-1.86 (m, 2H), 1.44-1.58 (m, 4H), 1.30-1.38 (m,10H), 0.90 (s, 9H), 0.05 (s, 6H).

4-(7-(6-(10-(tert-butyldimethylsilyloxy) decyloxy)naphthalene-2-yl)-9, 9-diethyl-9H-fluoren-2-yl) pyridine 9(b).



Mixture of compound 5 (3 g, 7.93 mmol), 8 (4.7 g, 8.7mmol)and K2CO3 (1.424g, 10.3mmol) were dissolved in 120 ml oftoluene: ethanol (3:1) and degassed for 10min. to that 135mgof Pd(PPh3)4 was added and the reaction mixture was stirredunder reflux for 4hrs. After cooled to room temperature thesolvent was removed and extracted with dichloromethane andwashed with water followed by a little brine. The organicphase was dried over anhydrous MgSO4 and the solvent wasremoved using rotavapor. The crude product was purified bycolumn chromatography on Al2O using mixture of hexane anddichloromethane (1:1) as eluent to yield low melting whitesolid (5.1 g, 91%) 1H NMR (300 MHz, CDCl3), δ: 8.68 (d, J =6 Hz, 2H), 8.02(s, 1H), 7.61 -7.86 (m, 11H), 7.15 (d, J = 8.7Hz, 2H), 4.08(t, J = 6.6 Hz, 2H), 3.58 (t, J = 6.6 Hz, 2H), 2.06-2.17 (m, 4H), 1.80-1.87 (m, 2H), 1.40-1.49 (m, 2H), 1.22-1.36(m, 12H), 0.86 (s, 9H), 0.41 (t, J = 3.6, 6H), 0.03 (s, 6H).

10-(6-(9, 9-diethyl-7-(pyridine-4-yl)-9-fluoren-2yl)naphthalene-2yloxy) decan-1-ol(10b) 5 g (7.02 mmol) ofcompound 9(b) was dissolved in 80 ml of dry THF, to this 14ml (14.04 mmol) of 1M TBAF was added under an ice-coldwater bath. The reaction mixture was stirred for 10min at ice-cold water then at room temperature for 24hr. THF wasremoved by rotavapor,then the product was extracted withdichloromethane followed by washing with brine .The organicsolvent was dried over anhydrous magnesium sulphate thenthe solvent was removed using rotavapor. The crude productwas purified by column chromatography on Al2O3 usingmixture of hexane and dichloromethane (1:2) as an eluent toyield white solid (3.5 g, 83%) 1H NMR (300 MHz, CDCl3), δ:8.68 (d, J = 5.7 Hz, 2H), 8.04 (s, 1H ), 7.84 (d, J = 8.1 Hz,1H), 7.79-7.81 (m, 4H), 7.73 (d, J = 1.5 Hz, 1H), 7.70-7.61(m,5H), 7.21 (d, J=2.4 Hz, 1H), 7.18 (s, 1H), 4.11 (t, J = 6.3 Hz,2H), 3.65 (t, J = 6.3 Hz, 2H), 2.11-2.19 (m, 4H), 1.87 (qn, J =6.3 Hz, 2H), 1.27-1.58 (m, 17H), 0.43 (t, J = 7.2 Hz, 3H).

10-(6-(9,9-diethyl-7-(pyridin-4-yl)-9H-fluoren-2-yl)naphthalen-2-yloxy)decyl methacrylate.(11) Compound10(1.3 g 2.37 mmol),1,3dichloro1,1,3,3,tetrabutyldistannoxanedichloro1, 1, 3, 3, tetrabutyldistannoxane (131 mg, 0.23mmol), 2,6-di-tert-butyl-4-methylphenol (52 mg, 0.23 mmol)were taken in a vacuum tight flask and dried by giving undervacuum- nitrogen alternatively. Then vinyl methacrylate (1.47ml, 11.93 mmol) and 1.5 ml of dry THF were added. Theresulting solution was stirred at 50º C for 48hrs .The productwas extracted with dichloromethane followed by washing withbrine. The solvent was dried over anhydrous MgSO4 andremoved by rotavapor. The crude product was purified bycolumn chromatography on Al2O3 using dichloromethane aseluent to yield white solid (1.3g, 89%).1H NMR (300 MHz,CDCl3), δ:8.65 (d, J = 3.3 Hz, 2H), 7.77 (t, J = 7.8 Hz, 2H),7.50 -7.74 (m, 8H), 6.99 (d, J = 8.7 Hz, 2H), 6.08 (s, 1H), 5.52(s, 1H), 4.12 (t, J = 6.9 Hz, 2H), 3.99 (t, J = 6.6 Hz, 2H), 2.13(q, J = 7.2 Hz, 4H), 1.928(s, 3H) ,1.82 (qn, J = 6.6 Hz,2H),1.66 (m, 2H), 1.25-1.46 (m, 12H), 0.37 (t, J = 7.2 Hz,6H).

Synthesis of Poly[10-(6-(9,9-diethyl-7-(pyridin-4-yl)-9H-fluoren-2-yl)naphthalen-2-yloxy)decyl methacrylate] (PFNA).The polymerization was carried out by the free radical

polymerization described as follows. To a Schlenk tube, 1.5 gof monomer 11 was dissolved in 20 wt% monomerconcentration of dry THF (7.5 mL) with AIBN (5 mg, 2 mol% the monomer concentration) as an initiator. The solutionwas degassed by three freeze-pump-thaw cycles and thensealed off. The reaction mixture was stirred and heated at 60°C for 48 h. After polymerization, the polymer wasprecipitated into methanol. The precipitated polymer wascollected, washed with diethyl ether, and dried under highvacuum. Yield: 1.2 g (80%) of PFNA as white solid.

1.2. Synthetic Procedures For S1 and S2 dyes

(2E,2'E)-3,3'-(5',5''-(9,9-Dihexyl-9H-fluorene-2,7-diyl)bis(2,2'-bithiophene-5',5-diyl))bis(2-cyanoacrylic acid)(S1): The dyes S1 and S2 were synthesized by the knownliterature procedure6 to yield dark brown solid (390 mg,82.4%) 1H NMR (300 MHz, DMSO-d6), δ (ppm): 8.49 (s,2H), 7.98 (d, J = 4.5 Hz, 2H), 7.877.81 (m, 4H), 7.70 (dd, J= 3.3 Hz, J = 9.9 Hz, 6 H), 7.62 (d, J = 3.9 Hz, 2H), 2.06 (m,4H), 1.030.96 (m, 12H), 0.66 (t, J = 6.3 Hz, 6H), 0.51 (m,4H). 13C NMR (75 MHz, DMSO-d6); δ(ppm): 164.3, 152.3,147.0, 146.6, 146.3, 142.3, 141.0, 134.6, 132.6, 129.1, 126.2,125.6, 125.4, 121.5, 120.4, 117.3, 98.7, 55.8, 31.5, 29.5, 24.1,22.6, 14.5. MS (FAB): m/z [M+] 853; calcd m/z [M+] 852.22.Anal. Calcd for C49H44N2O4S4: C, 68.98; H, 5.20; N, 3.28.Found: C, 68.62; H, 5.49; N, 3.86.

5',5''-(9,9-Dihexyl-9H-fluorene-2,7-diyl)-di-[2,2':5',2''-terthiophene-5''-(2-cyanoacrylic acid)] (S2)

1H NMR (300 MHz, DMSO-d6), δ (ppm): 8.47 (s, 2H),7.97 (d, J = 4.2 Hz, 2H), 7.837.76 (m, 4H), 7.637.58 (m,8H), 7.49 (d, J = 3.6 Hz, 2H), 7.42 (d, J = 3.9 Hz, 2H), 2.06(m, 4H), 1.000.95 (m, 12H), 0.65 (t, J = 6.3 Hz, 6H), 0.53(m, 4H). 13C NMR (75 MHz, DMSO-d6), δ(ppm): 164.3,152.2, 146.9, 145.8, 144.6, 142.1, 140.7, 139.0, 135.1, 134.7,134.2, 132.8, 128.8, 127.1, 126.1, 125.1, 125.0, 121.3, 120.1,117.2, 98.1, 55.1, 31.0, 29.1, 24.1, 22.1, 14.4, MS (FAB): m/z[M+] 1017; calcd m/z [M+] 1016.19. Anal. Calcd forC57H48N2O4S6: C, 67.29; H, 4.76; N, 2.75. Found: C, 66.61;H, 5.02; N, 2.85.

1.3. Preparation of H-Bonded Polymer Networks(PFNA/S1-S2)

Proton acceptor polymer PFNA and proton donor dye(S1, S2) 1:2 molar ratio were dissolved in minimum amount ofmethylene chloride then sonicated for 5 min to make a clearsolution. Then the solvent was evaporated under ambienttemperature, and was followed by drying in a vacuum oven at60 °C for several hours.(see figure 1) The complexation ofdonor and acceptor through hydrogen bonding occurred duringthe solvent evaporation. Similar procedure was followed forthe other H-bonded polymers PFNA/S2



Figure 1. Synthesis of H-Bonded PolymerNetworks

2.Experimental2.1Materials

Chemicals and solvents were reagent grades and purchasedfrom Aldrich, ACROS, TCI, Strem, Fluka, and LancasterChemical Co. THF and dichloromethane were distilled oversodium/benzophenone and calcium hydride respectively andfreshly distilled before use. Tetra-n-butyl ammoniumhexafluorophosphate (TBAPF6) was recrystallized twice fromabsolute ethanol and further dried for two days under vacuum.N-bromosuccinimide was recrystallized from distilled waterand dried under vacuum. The other chemicals were usedwithout further purification. Chromatography was performedwith Merck silica gel (mesh 70230) and basic aluminumoxide, deactivated with water. The chemical structures for allproducts were confirmed by 1H NMR spectroscopy, massspectra (FAB) and elemental analyses.

2.2. Measurements

1H NMR spectra were recorded on a Varian unity 300MHz spectrometer using d-DMSO as solvents. Elementalanalyses were performed on a HERAEUS CHN-OS RAPIDelemental analyzer. Fourier transform infrared (FT-IR) spectrawere performed a Nicolet 360 FT-IR spectrometer. Thermogravimetric analyses (TGA) were conducted on a Du PontThermal Analyst 2100 system with a TGA 2950 thermogravimetric analyzer at a heating rate of 20 oC/min undernitrogen. Gel permeation chromatography (GPC) analyseswere conducted with a Water 1515 separations module usingpolystyrene as a standard and THF as an eluant. UV-visibleabsorption spectra were recorded in dilute THF solutions (10-5

M) on a HP G1103A spectrophotometer. Thin films of UV-viswere spin-coated on quartz substrates from THF solutionswith a concentration of 1 wt%. cyclic voltammetry (CV)measurements were performed using a BAS 100electrochemical analyzer with a standard three-electrodeelectrochemical cell in a 0.1 M tetrabutylammoniumhexafluorophosphate (TBAPF6) solution at room temperaturewith a scanning rate of 100 mV/s. During the CVmeasurements, the solutions were purged with nitrogen for30s. In H-bonded cross-linking polymers, a carbon working

electrode coated with a thin layer of H-bonded polymer, andfor dyes in solution (DCM) a platinum wire as the counterelectrode, and a silver wire as the quasi-reference electrodewere used, and Ag/AgCl (3 M KCl) electrode was served as areference electrode for all potentials quoted herein. The redoxcouple of ferrocene/ferrocenium ion (Fc/Fc+) was used as anexternal standard. The corresponding highest occupiedmolecular orbital (HOMO) and lowest unoccupied molecularorbital (LUMO) levels were calculated using Eox/onset andEred/onset. The onset potentials were determined from theintersections of two tangents drawn at the rising currents andbackground currents of the cyclic voltammetry (CV)measurements.

Figure 2. Structure of H-Donor dyes (S1, S2) and H-Accepting Polymer

2.3 Fabrication and Characterization of OPV Devices

Solar cells were fabricated on indium tin oxide (ITO)-coated glass substrates through a following procedure: TheITO coated glass substrate was first cleaned with a detergentand sonicated in an ultrasonic bath, and they were driedovernight in an oven. After this cleaning procedure, thesubstrates were subjected to UV ozone cleaning for 15minutes. PEDOT: PSS (Baytron PH) was spin coated onto thesubstrates at 4000 rpm. The films were dried on a hotplate at120 ºC for 30 minutes. The samples were then transferred tothe nitrogen filled glove box for the active layer deposition. Asolution containing hydrogen bonded polymer networks(PFNA/Sx, where x vary from 1,2) and PCBM were prepared(2 wt%, 1:1 ) and were kept overnight digitally controlledhotplate at 60 ºC for the uniform mixing. The solutions wereused for the active layer deposition and were spin coated onthe substrates at 1500 rpm for 60 sec. The films were allowedto dry in a covered Petri dish. No thermal annealing treatmentwas given to the active layers prior to the cathode deposition.Finally, a 20 nm Ca and 50 nm Al cathodes were depositedusing a thermal evaporator at a base pressure of 110-6 Torr.The device area was 0.1 cm2. The devices were thentransferred to the nitrogen filled glove box. The current-voltage characteristics of the devices were measured using HP



4156 semiconductor parameter analyzer. Air mass 1.5 Global(AM 1.5 G) solar simulator was used for the photoillumination of the devices.

2.4. Structural Characterization

H-acceptor side-chain homopolymer PFNA, dyes S1, S2were satisfactorily characterized by 1H NMR, FAB, andelemental analyses. The weight average molecular weight(Mw) of H-acceptor polymer PFNA was determined by gelpermeation chromatography (GPC) with THF as the elutingsolvent and polystyrene as a standard. Figure 2 shows thestructure of dyes S1, S2 and polymer PFNA. The numberaverage molecular weight (Mn) and polydispersity index (PDI)of polymer PFNA were 12200 g /mol and 1.72, respectively.The decomposition temperatures (Td) of H-bonded polymernetworks PFNA /S1-S2 and H-acceptor polymer were shownin Figure 4. H-bonded cross-linking polymers (Figure 3)showed good thermal stabilities with decompositiontemperatures of H-bonded polymer networks PFNA/S1 andPFNA/S2 are 382 and 374 °C, respectively which areadequate for their applications in polymer solar cells and otheroptoelectronic devices.

Figure 4. TGA thermograms of H-Bonded polymer Network(PFNA/S1, PFNA/S2)

3. Result and Discussion

3.1. Optical Properties

UV-vis absorption spectra of H-acceptor polymer PFNAand H-donor dyes S1, S2 in DCM solutions (10-5 M) areshown in Figure 5 and their data are listed in Table 1. H-acceptor polymer PFNA showed the maximum absorptionwavelength of 360 nm with a bandgap of 2.95 eV, H-donordyes S1, S2 showed the maximum absorption wavelengths452 and 479 nm respectively and was attributed to theintramolecular charge transfer (ICT) from the donor Fluorenecore to the cyanoacrylic acid termini. The increase in theabsorption spectra from S1 to S2 and from PFNA/S1 toPFNA/S2 is attributed to the increase in conjugation withincrease in a symmetric thiophene unit.7

Figure 3. Graphical Representation of H-Bonded PolymerNetwork

As shown in Figure 5, the maximum absorption peak of H-bonded polymer networks PFNA/S1S2 on solid quartz filmswere blue shifted in contrast to those of H-donor dyes (S1-S2)and the blue shifts were ascribed to the dilution effect of H-acceptor polymer as solid solvent for dyes as solutes in the H-bonded cross-linking polymers.8 However, the broaderabsorption spectra of H-bonded cross-linking polymers(compared with PFNA) extending to longer wavelengthregions (Figure 5) favored better light harvesting andincreased the photocurrent response region. In addition, thebroadening of absorption spectra in the H-bonded cross-linking polymers increased their λonset, thus the decrease ofoptical band gaps in solid films was observed

.

Figure 5. UV-visible absorption spectra of H-acceptorpolymer (PFNA), H-donor dyes (S1-S2), and H-bondedpolymer networks (PFNA/S1-S2) in solid films coated ontofused quartz plates

100 200 300 400 500 600

60

70

80

90

100

weig

ht (%

)

Temperature (oC)

PFNA/S1 PFNA/S2

400 500 600 7000.0

0.2

0.4

0.6

0.8

1.0

1.2

Abs

orba

nce

(a.u

.)

Wavelength (nm)

PFNA S1 S2 PFNA/S2 PFNA/S1



Table 1. Optical Properties of PFNA, S1, S2 , PFNA/S1 andPFNA/S2

Polymer λabs, film (nm)b λonset, film

(nm)Eg

opt, film (eV)

c

PFNAS1S2

PFNA/S2

360452479465

420542580686

2.952.282.131.80

PFNA/S1 431 661 1.87a In DCM dilute solutions.b Spin coated from DCM solutions.c The optical bandgaps were obtained from the equation Eg

opt,

film = 1240/onset, film.

3.2. Electrochemical Properties

Inorder to go deep into the energy lavels of H-bondedpolymer network and charge injection processes of the H-bonded polymer networks in the BHJ solar cell devices, cyclicvoltametry (CV) measurements are employed. The cyclicvoltograms of polymers are illustrated in Figure 6 and therelated data are summarized in Table 2. Cyclic voltammetry(CV) measurements were performed at a scanning rate of 100mV/s in a solution of 0.1 M tetrabutylammoniumhexafluorophosphate (Bu4NPF6) dissolved in acetonitrile atroom temperature using a standard three-electrodeelectrochemical cell. A platinum disk working electrode, a Ptwire counter electrode, and an Ag/AgCl reference electrodewere used. As shown in Table 2 the HOMO and LUMO forPFNA/S1 and PFNA/S2 were 5.25, 5.36 and 3.25, 3.45respectively which is in the desirable range to be used inphotovoltaic application. Furthermore the electrochemicalband gap for PFNA/S1 and PFNA/S2 were 1.81 and 1.91 eVrespectively which is in the acceptable range of error withrespect to the optical band gap.

-2.0 -1.5 -1.0 -0.5 0.0 0.5 1.0 1.5 2.0

Potential (V)

PFNA/S1 PFNA/S2

Figure 6. Cyclic voltammograms of H-Bonded polymernetwork (PFNA/S1-S2) in solid films at a scan rate of 100mV/s.

Table 2. Electrochemical Properties of H-bonded Polymernetworks

PolymerEox,onset

(V)aEred,onset

(V)aHOMO(eV)b

LUMO(eV)b

Egcv,

(eV)PFNA/S1 0.90 -0.91 5.25 3.25 1.81PFNA/S2 1.01 -0.90 5.36 3.45 1.91

a Onset oxidation and reduction potentials measured by cyclicvoltammetry in solid filmsb HOMO/LUMO = [-(Eonset - 0.45) - 4.8] eV, where 0.45 V isthe value for ferrocene vs. Ag/Ag+ and 4.8 eV is the energylevel of ferrocene below the vacuum.

3.3. Photovoltaic Properties

In order to prove the potential application of the H-bondedpolymer networks (PFNA/S1-S2) as electron donors inphotovoltaic devices, we fabricated devices by spin-coatingfrom 2 wt% dichlorobenzene solutions of polymer blendscontaining H-bonded polymer networks PFNA/S1-S2 (aselectron donors) and PCBM (as an electron acceptor) in 1:1weight ratio with a configuration of ITO/PEDOT:PSS/H-bonded polymer networks:PCBM (1:1 w/w)/Ca/Al (under AM1.5 irradiation, 100 mW/cm2). Figure 7 shows the currentdensity versus voltage (J-V) curves and the results areillustrated in Table 3. As shown in Table 3, the short circuitcurrent density (Jsc), open circuit voltage (Voc), fill factor (FF),and PCE values of the OPV devices containing H-bondedpolymer networks PFNA/S1-S2 were 1.76, 1.88 mA/cm2,0.55, 0.58 V, 27, 29 %, 0.26, 0.32%, respectively. Thedecrease in HOMO values of PFNA/S2 facilitated higher Voc

value than that of PFNA/S1. The broader absorption spectrain PFNA/S2, harvested higher photocurrent and appealed ahigher Jsc value of 1.88 mA/cm2 than that of PFNA/S1.

Figure 7. Current density-voltage curves of illuminated solarcells incorporating H-bonded polymer networks /( PFNA/S1-S2with PCBM in 1:1 (w/w) ratio under AM 1.5G, 100mW/cm2.

0.0 0.1 0.2 0.3 0.4 0.5 0.6

-2.4

-2.0

-1.6

-1.2

-0.8

-0.4

0.0

J(m

A/c

m2 )

Voltage (V)

PFNA/S1PFNA/S2



As a contributing parameter for the net PCE value, the fillfactors of the OPV devices were low, which could beassociated with the large series resistance of the devices. Onthe whole, the OPV device containing blended H-bondedpolymer network PFNA/S1:PCBM resulted the better PCEvalue of 0.32% with Jsc = 1.88 mA/cm2, Voc =0.58 V, and FF =29%, which was mainly because of its broader absorption andbetter utilization of the solar spectrum. In general, theapproach in self-assembly of H-donor dyes with H-acceptorpolymers via hydrogen bonding was proven to enhance OPVproperties by the unique non-covalent-bonded practicalapplications.Table 3.Photovoltaic Performance of PSC Devicesa Basedon H-Bonded Polymer Networks PFNA/S1-S2 MeasuredUnder AM 1.5 Irradiation, 100 mW/cm2

Polymernetwork

Voc (V) Jsc

(mA/cm2)FF (%) PCE (%)

PFNA/S1 0.55 1.76 0.27 0.26PFNA/S2 0.58 1.88 0.29 0.32

a PSC devices with the configuration of ITO/PEDOT:PSS/(PFNA/S1-S2):PCBM/Ca/Al where H-bonded polymernetwork:PCBM = 1:1 by weight.

4. Conclusion

To facilitate getting the physical properties of both smallmolecules and polymers in OPV devices, the concept ofsupramolecular architectures by complexation of H-donordyes (S1-S2) with a side-chain H-acceptor homopolymer(PFNA) was applied to produce H-bonded cross-linkingpolymers. Photophysical properties revealed that theabsorption spectra of H-bonded polymer networks showedbroader wavelength ranges and higher λonset values to appealbetter light harvesting, though they showed blue shifts in theirrespective absorption maxima (λmax). The H-bonded polymernetwork PFNA/S2 containing S2 dye showed the betterphotovoltaic performance with a PCE value of 0.32%, thanPFNA. Thus, the results by complexation of small moleculesand polymers via H-bonding are very encouraging for thefuture research on donor-acceptor supramoleculararchitectures of oligomeric dyes H-bonded with processable -conjugated polymers in organic solar cell applications. Thusthe supramolecular architectures by the complexation ofconjugated polymers with surface modified nanoparticles arein progress.

5. References

1. (a) K. Walzer, B. Maennig, M. Pfeiffer and K. Leo,Chem. ReV., 2007, 107, 1233; (b) S. Gu¨nes, H.Neugebauer and N. S. Sariciftci, Chem. ReV., 2007,107, 1324; (c) Y.-J. Cheng, S.-H. Yang and C.-S.Hsu, Chem. ReV., 2009, 109, 5868.

2. (a) H. X. Zhou, L. Q. Yang, A. C. Stuart, S. C. Price,S. B. Liu and W. You, Angew. Chem. Int. Ed., 2011,50, 2995; (b) T.-Y. Chu, J. P. Lu, S. Beaupre, Y. G.Zhang, J.-R. Pouliot, S. Wakim, J. Y. Zhou, M.Leclerc, Z. Li, J. F. Ding and Y. Tao, J. Am. Chem.Soc., 2011, 133, 4250; (c) S. C. Price, A. C. Stuart, L.Q. Yang, H. X. Zhou and W. J. You, Am. Chem. Soc.,2011, 133, 4625; (d) J.-H. Park, D. S. Chung, D. H.Lee, H. Kong, I. H. Jung, M.-J. Park, N. S. Cho, C. E.Park, and H.-K. Shim, Chem. Commun., 2010, 46,1863.

3. Jonkheijm, P.; Duren, J. K. J. V.; Kemerink, M.;Janssen, R. A. J.; Schenning, A. P. H. J.; Meijer, E.W. Macromolecules 2006, 39, 784.

4. Liang, T. C.; Chiang, I. H.; Yang, P. J.; Kekuda, D.;Chu, C. W.; Lin, H. C. J. Polym. Sci. Part A: Polym.Chem. 2009, 47, 5998.

5. (a) G. D. Wei, S. Y. Wang, K. Sun, M. E. Thompsonand S. R. Forrest, Adv. Energy Mater., 2011, 1, 184;(b) R. Fitzner, E. Reinold, A. Mishra, E. Mena-Osteritz, H. Ziehlke, C. Körner, K. Leo, M. Riede, M.Weil, O. Tsaryova, A. Weiß, C. Uhrich, M Pfeifferand P. Bäuerle, Adv. Func. Mater., 2011, 21, 897; (c)Y. Matsuo, Y. Sato, T. Niinomi, I. Soga, H. Tanaka,and E. Nakamura, J. Am. Chem. Soc., 2009, 131,16048;(d) L.-Y. Lin, Y.-H. Chen, Z.-Y. Huang, H.-W. Lin, S.-H. Chou, F. Lin, C.-W. Chen, Y.-H. Liu,and K.-T. Wong, J. Am. Chem. Soc., 2011, 133,15822.

6. Sahu, D.; Padhy, H.; Patra, D.; Yin, J,-F.; Hsu, Y.-C.; Lin, J.-T. Tetrahedronn.; 2010, 67, 303-311

7. Sahu, D.; Tsai, C.-H.; Wei, H.-Y.; Chu, C.-W. J.Mater. Chem. 2012, 22, 7945.

8. Liang, T. C.; Chiang, I. H.; Yang, P. J.; Kekuda, D.;Chu, C. W.; Lin, H. C. J. Polym. Sci. Part A: Polym.Chem. 2009, 47, 5998.



Demand Response in Smart MicroGridsG. Sivaranjani, Smitanjali Bhukta

School of Electrical Sciences, National Institute of Science & Technology, Odisha, India

Abstract—The mismatch between supply and demand inmicrogrids can be overcome by effectively utilizingdistributed energy resources and/or encouraging demandside management. Demand response is one of the populartechniques to demand side management. In this paper anagent based architecture to simulate virtual markets enablingcustomers of the market toparticipate in demand response and trade power usingintelligent trading strategy which makes the market morerealistic and microgrids smarter is proposed. The proposedconcept is verified on a system with two microgrids. Theproposed MAS is developed using Java ApplicationDevelopment (JADE) framework.

Index Terms— Multi Agent Systems, Energy marketssimulation, Smart grid, Demand response, Micro grids.

I. INTRODUCTION

Smart distribution grids can be realized with customerdriven microgrids. A Microgrid is an aggregation ofdistributed energy resources (DERs) and loads. DERsinclude distributed generation (DG) and distributed storage(DS). Effective management of a microgrid is challengingwith non-dispatchable sources like wind and photovoltaic[1-3]. In general microgrids are designed to operate in eithergrid connected or isolated modes. Also, these are providedwith intelligence to island from main grid intentionally or incase of a fault. Apart from benefits obtained by integratingDERs into the distribution system, there are economic,commercial and technical challenges. Some of the technicalchallenges with respect to power management wereaddressed in the literature with viable options [4-5].

In the presence of non-dispatchable generation in microgrid,there is always an issue of supply-demand mismatch. One ofthe known solutions to such problem is to equip microgridswith backup power (diesel generators). The operating costand environmental effects by such backup system aresignificantly high compared to green DERs and hence it isessential to embed certain intelligence to effectively utilizeintegrated distributed resources.

The gap between supply and demand can also be reduced byincorporating demand side management at the distributionend of the power system. Participation ofconsumers/customers in reducing the peak demand and

shifting of the peak according to the generation will reducethe gap between demand and supply. One such mechanismis “demand response” in which, a customer can participatein load management voluntarily. Incorporating demandresponse (DR) in existing microgrid is a challenging issueas it involves smart communication, negotiation, andcontrol. This paper presents an application of demandresponse in microgrids by implementing a virtual marketenabling

customers to participate in demand response. The customerswith appliances/loads with low priority may participate indemand response by allowing the microgrid intelligentsystems to operate their appliances when the cost ofelectricity is low. The main participants in residential sectorinclude kettle, dishwasher, washing machine etc. Thecustomer participating in demand response shall provide theutility/microgrid with information like starting, end andoperating time. The microgrid intelligent system identifiessituations where the cost of the power is low and startsoperating the appliance continuously for the duration ofoperational time specified by customer.

The key intention of the proposed mechanism is to enablemicrogrids to participate in the simulated market andthereby effectively utilize the DERs. The developedintelligence routes the power from surplus locations todeficient. However the cost at which power will be traded isdecided by pool members following an auction protocol.

In this paper multi agent system (MAS) is used to modelmarket scenario with energy buyers and energy sellers.JADE (Java Application DEevelopment framework), is usedto develop the proposed MAS architecture. The proposedarchitecture consists of several intelligent agents to makethe developed market realistic and to incorporate demandresponse. Each customer in the market can participate indemand response by providing control over his/herequipment to developed intelligent system. Power tradingby pool members using double auction protocol is discussedin section II. The developed intelligent system starts aninstance of bilateral negotiation with energy sellers for eachreceived demand response following the procedure alongwith architecture is given in section III. The implementationof the proposed architecture using JADE [7] on a test casewith two microgrids forming pool is presented in SectionIV.



II. AGENT BASED STRATEGYA. Multi-Agent Systems

A multi-agent system is a system with two or more agentsor intelligent agents. An agent is “a software or hardwareentity that is placed in some environment and is able toautonomously react to the changes in that environment”.This nature of agent is called autonomy. An intelligent agentis an agent who exhibits pro-activity i.e. goal directedbehavior and is able to interact with other intelligent agentsand reactivity. The key features of agent technology are dataand method encapsulations.MAS has been applied to widerange of engineering applications. The application of MASis a construct robust, flexible and extensible systems or as amodeling approach.

B. Continuous Double Auction (CDA) Markets:

A CDA is a market place with agents selling goods calledsellers and agents buying goods called buyers. The sellersand buyers in any CDA market trade single type of goodslike energy or power. An “ask” is the price placed by aseller to sell one unit of the goods. A “bid” is the priceplaced by a buyer to purchase a unit of goods. At any timein the market, the current lowest “ask” is called anoutstanding ask and is generally represented by oa.Similarly, the current highest bid in the market is called anoutstanding bid and is represented by ob. A valid ask islower than the present oa and ask not lower than oa isinvalid ask. A valid bid is a bid higher than present ob and abid not higher than ob is called invalid bid. CDA progressesin rounds and in each round invalid asks and bids areneglected by the market. In each round at most one unit ofgoods will be cleared and hence a CDA run will havemultiple rounds and it terminates when all possible matcheshave made. A match will be made in the market when ob ishigher than or equal to oa and match will be made betweenseller who submitted oa and the buyer submitted ob. Thematch price is equal to the average of oa and ob.For each trader in the CDA market there is an acceptableprice range. PL is the lowest acceptable price in the marketand PU is the highest acceptable price in the market. Incompetitive grid connected energy markets this range willbe termed as Grid buying price and Grid selling price.The role of agents in agent based CDA markets is torepresent their owners, who may be buyers or sellers. Theagent can be an auctioneer, who conducts the auction. Inpractical, an agent in CDA market is a software intelligencethat can be viewed as a delegate of his/her owner to achievegood profit.

C. Bidding Strategy:

Each trading agent in CDA follows a bidding strategy whichallows him/her to squeeze maximum profit from the market.The bidding strategy followed by any trading agent merely

requires two types of data, the global and local data. Theglobal data includes acceptable market price range, presentoa and ob, winner of the last round, matching price historyand supply to demand ratio. The local data includesexpected profit margin, how many units of goods to trade,the forecast of future market and risk attitude. Afterperceiving this information, trading agents start placing theirbids/asks. Depending on whether the previously placedbid/ask got accepted or not, each trading agent computesnext bid/ask to be placed in the market.In the agent based CDA market proposed here, tradingagents follow the intelligent bidding strategy. This biddingstrategy is derived from standard bidding strategy called“Zero-Intelligence-plus(ZIP)” strategy.In a CDA market, profit motivated trading agents willinitially quote at a prices which brings them maximumprofit. If ‘p’ is the bid/ask price placed by a trading agent,then the price should always be better than or equal to aprice called limit price (LM) of the trader. Placing askslower than limit prices gives loss to the selling agents andplacing bids more than limit price gives loss to buyingagents. The relative difference between p and LM is calledprofit margin (PM). For selling agent increasing profitmargin increases the ask value, whereas decreasing theprofit margin reduces ask value. On the other hand, for abuying agent increasing profit margin decreases lower bidand decreasing profit margin results higher bid. The strategyfollowed by intelligent trading agents is similar to those ofZIP trading agents. However, the amount of increase/decrease in profit margin is different from ZIP strategy.Qualitatively the intelligent strategy followed by tradingagent is as follows:

a. An intelligent selling agent raises its profitmargin whenever the last ask was accepted.On the other hand, a selling agent lowers profitmargin if it is still active and the last ask wasrejected.

b. An intelligent buying agent raises its profitmargin whenever the last bid was accepted andis less than LM of the buying agent. On theother hand, a buying agent lowers profitmargin if it is still active and the last bid wasrejected.

The quantitative explanation of the intelligent tradingstrategy is similar to that of ZIP bidding strategy except thetarget price used in the bid/ask updating. The proposedintelligent strategy computes the target price Ti(t) asTi(t)= Ai(t)*S + Bi(t) * e(t) ……………(1)The Ai(t) , Bi(t) are the random values generated asmentioned above. S is standard target which is current oafor sellers and current ob for buyers. The e(t) in (1) is calledeagerness and is calculated as the present ratio of supply todemand for seller agents and demand to supply for buyeragents.



III. MICRO-GRID ARCHITECTUREThe proposed architecture to embed demand response intomicrogrid is shown in Fig.1. the architecture has agentscalled Global Intelligent Agent(GIA) , Microgrid IntelligentAgent (MIA), Demand response agent (DRA), LoadAgent(Lxy) and Generation Agent (Gxy). The agent GIA isfurther having agents named Recording Agent (RA) andGlobal Auctioneer Agent (GAA). The agent DRA is havingtwo internal agents named Floating Buyer ManagementAgent (FBMA) and Bi-lateral Contract Management Agent(BCMA) to serve the loads participating in demandresponse. The agent MIA is also having two internal agentsi.e. Priority Management Agent (PMA) and LocalAuctioneer Agent (LAA). The detailed explanation of theagents in the architecture is given in the following sections.

A. Load and Generation Agents

The agents in the bottom level of the architecture are Lxyand Gxy which represents Load and Generation entities. Theterm ‘xy’ in ‘Lxy’ indicates load agent’s location andassociation. For example L12 indicates load number one ofthe second microgrid. Similarly Gxy indicates generationnumber ‘x’ of the microgrid ‘y’. The agents ‘Lxy’ and‘Gxy’ have intelligence to bid in the auction conductedeither by GAA or LAA. Upon requested by MIA, each loadand generation agent collects owner choice and informsback to the MIA. Apart from collecting owner choice ‘Gxy’has intelligence to sign bilateral contracts and to remind theowner regarding the signed contracts. Upon receiving theend of auction signal ‘Lxy’ and ‘Gxy’ will inform the ownerregarding the set points for the next interval. Theintelligence developed for Lxy and Gxy will allow theseagents to participate in global market and local marketsimultaneously.

B. Microgrid Intelligent Agent

The next hierarchal agent in the architecture is MIA whichis responsible for conducting auction among local agents bymaintaining equal supply and demand in the local market.MIA makes local market cheaper and provides a privilege of‘participation’ to high priority loads. The whole functioningof MIA is attended by Priority Maintenance Agent (PMA)

and Local Auctioneer Agent (LAA).The PMA of each MIAis responsible for maintaining priorities of the loads havingassociation with MIA. The PMA calculates priority index ofeach load agent by considering how frequently a load agentis participating in demand response and the share of the loadagent in totally recorded demand response. PMA issues gatepass to high priority loads to enter local market. On theother hand, PMA verifies the type of generation of eachgeneration agent willing to participate in the auction andallows low generation cost agents to enter local market. TheLAA of each MIA is responsible for conducting auctionamong the traders who got gate pass to local market. Eachtrader in the local market follows the bidding strategydiscussed before. Upon receiving bids from traders (loadand generation agents), the LAA identifies the outstandingbid (ob) and outstanding ask (oa) in the market. LAA clearsa portion of the market following the protocol mentioned insection II and subtracts an amount called ‘clearing size’from current supply and demand in the local market. Theclearing size is not a static value in the local market and itvaries from round to round. The clearing size (CS) is theminimum among entitlements of the traders who submittedoa and ob. Upon finding the clearing size, LAA calculatesthe matching price by taking average of oa and ob.

C. Demand Response AgentNext hierarchical agent in the architecture is DemandResponse Agent (DRA) and its role is to receive and servethe demand response options by load agents. Like MIA,DRA also contains two internal agents viz. Floating BuyerAgent (FBMA) and Bi-lateral Contract Management Agent(BCMA) which will address the key roles of the DRA.Upon receiving demand response options from load agents,MIA communicates all demand response participations toBCMA of DRA. For each demand response option received,BCMA initiates bilateral negation with all the sellersirrespective of their current participation status. For doingbilateral negotiation BCMA creates a form called ‘bi-lateralcontract form’. The bilateral contract form contains timeintervals, price and participating load information. Afterreceiving bi-lateral contract form, each seller agent leavesits signature against the favorable intervals in thewillingness record of the bilateral contract and sends back toBCMA. BCMA collects all the signed bilateral contracts byseller agents and consolidates willingness records. Whileconsolidating the willingness record, BCMA takes thefollowing the steps into account.

a. Arrange all the contract forms in the order of arrivaltime.b. Create an empty willingness record.c. Allot the interval signed intervalsd. If more than one seller agent is willing to make thecontract in same interval, then consider the number of otherintervals they have signed and allot to the agent who hassigned for less number of intervals.



In Fig. 2(a), (b) and (c) a demand response with startinginterval ‘5’ and dead line of 8 has been considered withoperational time (OT) of 3. In order to serve such demandresponse option, BCMA creates a bi-lateral contract havingwillingness record with DS as 5 and DL as 8 and sends toall sellers (say G12, G22, and G32). The received bilateralcontract forms signed by seller agents G12, G22 and G32 areshown at the left of Fig. 2(a), 2(b) and 2(c) respectively. Ithas been assumed that the signed contract sent by G12 hasreached first.In Fig. 2 (a) the interval ‘5’ is allotted to the seller agentG12. Fig.2 (b) shows updating consolidated willingnessrecord with the willingness record sent by G22. G22 hassigned in two intervals 7 and 8. Since 7 and 8 have not beenallotted so far, G22 is allotted with the intervals 7 and 8. Fig.2 (c) shows updating consolidated willingness record withthe willingness record sent by G32. G32 has signed in interval7. Since interval 7 has already been allotted to G22, BCMAverifies back and allots 7 to G32 because G32 has signed forless number of intervals and hence the consolidatedwillingness record becomes {5=G12,6=null,7=G32,8=G22} asshown at the right of Fig. 2. (c).

Once consolidating the received contract forms is over,BCMA maps the consolidated willingness record to an arrayof ‘1’s and ‘0’s. While mapping consolidated willingnessrecord, BCMA places ‘1’ if the interval is allotted otherwiseplaces zero. The mapped on solidated willingness record isshown in Fig. 2 (d). Once the mapped willingness record isprepared, BCMA identifies the best set of intervals duringwhich the load agent participating in demand response is tobe served. In order to find best set of intervals, BCAMcalculates sum of the mapped values of all possible

combinations of intervals with each combination havingintervals of OT. OT is the operation time specified by theload agent participating in demand response. Uponcalculating all possible combinations, BCMA traces thecombination for giving maximum value. This combinationis the optimal duration to serve the load participating indemand response. If more than one such combinationsresults the same value then the combination which will beserved first is to be considered. BCMA informs the result ofnegotiation to load agents participating in demand response,generation agents who have been given the chance ofserving load and FBMA. FBMA examines the resultcommunicated by BCMA and identifies the number ofvirtual buyers to be created. The virtual buyers are alsocalled floating buyer agents (FBAs). FBMA creates FBAfor each zero in the best set of intervals so as tocontinuously feed the load participating in demandresponse. The key role of a FBA is to purchase power fromglobal market on behalf of load agent participating indemand response and after finishing the job assigned byFBMA these agents will terminate on their own.

D. Global Intelligent Agent :

The GIA in top level of the proposed architecture isresponsible for initiating all local markets and conductingauction with global scope. The GIA also records thesuccessful contracts and thereby informs correspondingtraders. Like DRA of the top level architecture, GIA alsocontains two agents viz. Recording Agent (RA) and GlobalAuctioneer Agent (GAA). The key functions of the GIAmentioned areactually addressed by GAA and RA. The GAA of GIAconducts auction among the buyer agents willing topurchase power from global market and the seller agentswilling to sell power in global market. The buyer agentsmay include FBAs and /or the load agents from themicrogrids with deficit power generation. The GAAconducts auction following CDA protocol similar to LAA.Unlike LAA, the GAA clears the market with fixed clearingunit size. The clearing unit size is fixed for one CDA runand is equal to greatest common divisor of all theentitlements of the traders articipating in auction. Theclearing size will be fixed at the beginning of the first roundand remains fixed for all the rounds. At the end of theauction GAA matches all unmatched traders to grid. TheRA of GIA is instrumental in recoding all successfultransactions among global members. RA is also responsiblefor sending auction close signal to all the tradersparticipated in the global auction market.

IV. AN EXAMPLE

This section illustrates the working of the proposedarchitecture with the help of a case study. During eachdemand interval (15min duration), the first five minutes is



considered as auction period during which the auction(CDA) will be conducted among the agents willing to sell orpurchase power in the next demand interval. Fig. 3 shows asystem with two interconnected microgrids each having twodistributed generators and two loads. Generation and loadagents in each microgrid can participate simultaneously inboth local and global markets. If excess power is available,then the intelligent agent of the corresponding microgridinitiates global market participation by allowing agents withhigh cost of generation to global market. MIA allows lowpriority load agents to enter global market if power deficithappens. For the case study shown in Fig. 3, a market pricerange of (GBP=9 cents/kWh, GSP=13.5 cents/kWh) isconsidered. Hence in each market (local or global) thebids/asks should be in this range. Table 1 and 2 show theglobal auction market prices and participation share of eachtrading agent without demand response on the previous day.The interval numbers in table 1 indicate the demandintervals starting from 6 am. The price history of the globalmarket is an essential index for customers to startparticipating in demand response. It is also considered thatL21 is having low priority loads of total capacity 30kW.

In the beginning of auction period of the interval ‘0’, theGAA of GIA issues “auction start” command to each MIAand FBMA of DRA. Upon receiving auction start signalfrom GIA, MIAs will request their load and generationagents to submit their bids/asks. While submitting the loadrequirement data L11 and L21 check the global history. Sincethe market is in initial stage and no demand response hasrecorded so far, the priorities of all load agents are same andhence any one of the load agents can be forwarded to globalmarket.Due to non availability of the power in local market frominterval 1 to 3 on the previous day L21 was forwarded toglobal market and L21 purchased 30 kW of power fromglobal market at a price 12.3 cents/kWh. Therefore, L21

chooses to participate in demand response and submitsdemand response option with DS=1, DL=7, OT=3 as (1, 7,3). L11 (50kW), L21

(30kW, (1, 7, 3)), G11 (40kW) and G12 (30kW) of MG1 willsubmit the load requirement, demand response andgeneration availability information to intelligent agent(MIA1). The PMA of MIA1 updates priority record toaccount new demand response participation and forwards



demand response option to BCMA of DRA. However, theinformation providedby the agents in MG1 may not be same as previous day. Insuch situation, MIA 1 will issues gate pass as per theupdated history of the agents. Similarly L12, L22, G12, andG22 submit requirement and availability of power to MIA2.Due to surplus power available MIA2 forwards G22 toglobal market.MIA1and MIA2 issue gate pass to highestpriority loads and low cost generation agents. While givinggate passes MIA maintains supply to demand ratio as unity.LAA of the MIAs and GAA of GIA start conducting auctionamong the agents entered local and global auction marketsrespectively. On the other hand BCMA creates bilateralcontract for the received demand response option andinitiates negotiation with all generation agents. The bilateralcontract form is having information size=30k, price= 11.25cents/kWh, and willingness record with DS=1, DL=7. Theprice 11.25 is calculated by taking the average of GBP andGSP. As a reply to the bilateral contract form, eachgenerator compares their price history for the previous daywith the current price offered in the contract and sendswillingness to BCMA. In the present case study G11 sends{4}, G12 sends {6}, G21 send {6, 7} as their willingness. TheBCMA consolidates the signed contract forms sent bygeneration agents by following the procedure given insection III as {1=null, 2=null, 3=null, 4=G11, 5=null, 6=G12,7=G21}. The BCMA maps the consolidated willingnessrecord and identifies the best set of intervals as {4, 5, 6}.The BCMA communicates this result to concerned agentsand FBMA. FBMA creates a virtual buyer FBA to purchasepower for interval 5 from global market in order to maintaincontinuity in serving the load. During the interval 4 thenewly created FBA will participate in global auction andpurchase power on behalf of L21. Upon successfullyconducting and recording all successful matches, GIA andMIA terminates the auction by sending auction close signalto all participated agents. The FBAs participated in theglobal auction market will terminate from agent platformafter receiving auction close from RA ofthe GIA. Fig. 4 shows the flow chart of the aforementionedmechanism.

The above MAS will be simulated using Java ApplicationDEvelopment (JADE) framework.

V. FUTURE WORK.

In this work an agent based architecture for trading andpower management in microgrids are presented. Theproposed system uses continuous double auction algorithmfor trading. Priority for customers participating in demandresponse at trading level is novel. The concept mentionedabove has to be simulated with two microgrids system usingJADE framework and can be extended for multiplemicrogrids.

VI. REFERENCES

[1] H.S.V.S. Kumar Nunna and Suryanarayana Doolla, “Demand Response in Smart Microgrids”, IEEE PESInnovative Smart Grid Technologies- India,2011.

[2] H.S.V.S. Kumar Nunna and Suryanarayana Doolla,“Demand Response in Smart Distribution System WithMultiple Micro-grids”, IEEE Transactions on Smart Grid,Vol.3, No. 4, pp 1641-1649, Dec 2012.

[3] J. M. Guerrero, F. Blaabjerg, T. Zhelev, K.Hemmes, E.Monmasson, S. Jemeï, M. P. Comech, R. Granadino,and J.I. Frau, “Distributed generation: Toward a newenergy paradigm ,”IEEE Ind. Electron.Mag., pp. 52-64, Mar. 2010.

Authors are solely responsible for Plagiarism Copy right reserved DGSG-2013@NIST[Type text]


Implementation of TOU for Smart PowerConsumption

CH. VENKATESWRA RAO 1, ARUN KUMAR RATH 2, S.S. TULASIRAM 3

1 PHD Scholar, JNTU, Kakinada, A.P.,2 Asst Professor in EEE, GIET Gunupur Orissa,

3 Professor, JNTU, Hyderabad

Abstract - Smart grid encourages consumers to prefer suitable energy in co operation with power gridat the most suitable time. The main aim of this, is the load sharing by motivating consumers tooperate only the most essential appliances at peak load periods and to transfer the operation of lessneeded appliances at off peak hours when tariff may be lower. A case study has been developed toprovide information to enable the user Mr. Krishna Morthy who is a LT consumer, The results andanalysis which have been presented in this has been generated by a real time data collected forMogulthur, Andhra pradesh.

I. INTRODUCTION

The smart energy information managementsystem developed with ability to record, store andprocess power consumption data of every majorappliance in the domestic and industries. Thepower consumption data is accessible through thesmart way like consumer energy portal, emailsand on handheld devices.Consumers can track their power usage bydevice, room or appliance, which helps betterregulate power consumption by analyzing minutewise, half-hourly, hourly, daily and monthly andsmart appliances automatically response toincreased demand in smart grid.

The case studies is developed for LT consumers,build complete power analytics, developeddifferent types reports and save the powerconsumption amount by using power during offpeak or mid peak. The TOU (Time of Use) Tariffis used reduce peak loads. The power analyticsreports provide the deep insight into theirconsumption patterns like on peak, off peak, midpeak and cost per appliances.

Appliances wise consumption for LTconsumer – The following case study has beendeveloped to provide information to enable theuser to act Mr. Krishna mort is a LT consumerand category is LT1-domestic . The belowanalysis provides deep insight for Mr. KrishnaMoorty to take effective decisionsThe consumers does not know about their Dailyenergy consumption, daily/hourly/half hourly/minute wise consumption, which appliances arecausing more energy consumptionHow they could manage their consumption betterThe below reports clearly demonstrated the netsavings by the use of TOU tariffs and thedesigned reports like cost by appliances.

Appliances wise consumption for LTconsumer - In the LT consumers in Mogulturucircle we have consider Krishns Moorty withservice no.000017 and physically the data iscollected by appliances wise and analyzed thepower consumption by appliance for KrishnaMoorty and provided the deep insight on overallconsumption.Table Appliances wise consumption for LT consumers inmonth of September 2012

Application ID Billed units kWhclother washer 9dryer+iron 11fans+cooler 40Lights 32

Others 18Pump 12Refrigeration 15TV+ Computer 44water heater 9

190



Figure-1: Appliances wise consumption for LT consumersin month of September 2012

Table Appliances wise consumption for LTconsumers in month of October 2012

Application IDBilled units

kWhclother washer 12dryer+iron 14fans+cooler 48lights 32others 20pump 17refrigeration 15TV+ Computer 52water heater 10

220

Table Appliances wise consumption for LT consumers inmonth of October 2012Table Appliances wise consumption for LT consumers inmonth of novmber2012

Application ID

BilledunitskWh

clother washer 7dryer+iron 13fans+cooler 43Lights 35Others 23Pump 9Refrigeration 19TV+ Computer 47water heater 12

208

3%

6%

21%

17%11%

4%

9%

23%

6% LT Consumer - krishna moorty

clother washer dryer+iron fans+coolerlights others pump

Table Appliances wise consumption for LT consumers inmonth of November 2012

Cost Wise Consumption For LT consumer - Inthe LT consumer in Mogulturu circle we haveconsider KRISHNA MOORTY with service no.sc1020/21 and physically the data is collectedappliances wise and analysed the power

consumption cost by Appliances wise andanalyzed the power consumption cost byappliance for KRISHNA MOORTY andprovided deep insight on overall consumption

Table cost wise consumption for LT consumer in the monthof September 2012

Appliance IDCost

clother washer 32.4dryer+iron 39.6fans+cooler 144

lights 115.2others 64.8pump 43.2refrigeration 54TV+ Computer 158.4water heater 32.4

684



5%6%

21%

17%9%

6%8%

23%

5% LT Consumer - krishna moorty

clother washer dryer+iron

Figure cost wise consumption for LT consumer in themonth of September 2012Table cost wise consumption for LT consumer in the monthof October 2012

Appliance ID Costclother washer 69dryer+iron 80.5fans+cooler 276lights 184others 115pump 97.75refrigeration 86.25TV+ Computer 299water heater 57.5

1265

Figure cost wise consumption for LT consumer in themonth of October 2012Table cost wise consumption for LT consumer in the monthof November 2012

Appliance ID Costclother washer 40.25dryer+iron 74.75fans+cooler 247.25lights 201.25others 132.25pump 51.75refrigeration 109.25TV+ Computer 270.25water heater 61Overall result 1188

Figure cost wise consumption for LT consumer in themonth of November 2012

Appliance wise consumption for six months inkWh units for LT consumer - The KrishnaMoorty consumption pattern for six months byappliance wise and analyzed the month wisecomparison

ApplicationID

clothswasherbilledunits

Dryer +ironbilledunits

Fans +coolerbilledunits

lightsbilledunits

othersbilledunits

pumpbilledunits

Refrigeratorbilledunits

Tv +computerbilledunits

waterheaterbilledunits

overallresult

Sep-12 9 11 40 32 18 12 15 449

Oct-12 12 14 48 32 20 17 15 5210

Nov-12 7 13 43 35 23 9 19 4712

54 67 260 204 124 69 105 26567



Monthly peak wise consumption for LTconsumer

The consumer can visualize every monthpower consumption sliced into the MID

PEAK,OFF PEAK and ON PEAK and canperform what if analysis on moving from the onpeak load to off peak load. it will be direct saving

to consumers in TOU Tariff and reduce thedemand – supply mismatch.Krishna moorty Can login to consumer energyportal and he can do the different types ofanalysis.Table monthly peak wise consumption LTconsumer for six months

peakcategory

MidpeakbilledunitsKWH

offpeakbilledunitsKWH

onpeakbilledunitsKWH

Overallresult

Sep-12 102 47 41 190Oct-12 154 53 13 220Nov-12 132 55 21 208Overallresult

725 327 165 1217

0500

41153 41183 41214

102 154 13247 53 5541 13 21

Mid peak billed units

off peak billed units

on peak billed units

Figure monthly peak wise consumption LTconsumer for six monthsTable monthly peak wise consumption in themonth of September 2012Peak category billed units

kwhMid peak 102Off peak 47On peak 41Overall result 190



figure monthly peak wise consumption in themonth of September 2012Table monthly peak wise consumption in themonth of October 2012Peak category billed units


Table monthly peak wise consumption in themonth of October 2012Table monthly peak wise consumption in themonth of november 2012Peak category billed units


Figure monthly peak wise consumption in themonth of november 2012Daily- Month peak wise consumption for LTconsumer

Daily data provides enough details forconsumers to relate day- to- day activities toelectricity usage.examples include vacationdays,holyday events,working days versus nonworking days,extreme weather days versus milddays and so on. Only the days that are unusual

stand out and those are the days the consumercare about.Krishna moorty Can login to consumer’s energyportal and he can do the different types ofanalysis.Daily- month peak wise consumption in themonth of September 2012

Day Mid peakbilled units

off peakbilled units

on peakbilled units

Overallresult

1 3 2 1 62 7 1 1 93 3 2 1 64 8 1 1 10



5 3 3 1 76 2 2 1 57 6 4 0 108 4 1 1 69 4 2 2 810 2 1 1 411 2 1 1 412 4 2 0 613 7 1 1 914 3 2 2 715 4 3 1 816 3 1 1 517 9 2 1 1218 2 1 2 519 3 2 1 620 6 1 1 821 3 3 3 922 4 2 1 723 3 1 1 524 2 2 1 525 8 2 0 1026 3 1 1 527 8 2 1 11

28

4 1 2 7

29 8 2 1 1130 6 2 1 9

134 53 33 220

Table Daily- month peak wise consumption in the month of October 2012Day Mid peak

billed unitsoff peakbilled units

on peakbilled units

Overallresult

1 3 2 1 62 5 1 1 73 3 2 2 74 8 1 1 105 3 2 1 66 2 3 1 67 6 4 0 108 4 1 1 6

02468

1012

1 3 5 7 9 11 13 15 17 19 21 23 25 27 29on peak billed units



9 3 2 1 610 2 1 1 411 2 1 1 412 4 2 1 713 7 2 0 914 3 1 2 615 4 3 1 816 3 1 1 517 5 2 1 818 2 2 3 719 3 3 1 720 6 2 1 921 3 2 1 622 4 2 1 723 3 2 0 524 2 2 1 525 8 1 0 926 3 1 1 527 8 2 1 1128 4 2 2 829 3 1 1 530 6 2 1 9Overall Result 122 55 31 208

Hour wise consumption for LT customerHour wise consumption instantly knowing

through consumer energy portal and also HANautomatically control the power consumptionwith help of smart appliances. The consumer can

visualize the load pattern of the powerconsumption and take necessary action.

Krishna moorty can logon to consumer’senergy portal and he can do the different types ofanalysis.



Figure hourly daily consumption in the month of sept.2012

Hourly-daily consumption in the month of nov.– 2012TOU tariff calculation for LT consumer

Power companies to charge consumersdifferent prices for electricity on the basis of thetime of use the electricity used so that it willencourages consumers to shift some of theirelectricity uses to lower cost,non peak hoursmany consumers able to reduce their electric bills

and also reduce stress on the electricitygeneration and transmission infrastructure.

In the LT case study by seeing thesereports he can able to shift the small amount ofpeak load into mid peak and half peak anddirectly reduce the electricity bill in the TOU

tariff compare to regular Tariff and in thefollowing table will provides the net savings toKrishna moorty.

Table TOU Tariff calculation monthly wise forLT consumer

billedunits

regularcost

TOUcost

netsavings

Sep-12 190 684 674.38 9.62Oct-12 220 1265 1275.39 11.15Nov-12 208 1196 1175.31 20.69Overallresult

1217 5740 5618.84 142.7



Figure TOU Tariff calculation monthlywise for LT consumer

Table TOU month wise consumption

monthregularcost

TOUcost

net savings

Sep-12 684 674.38 9.62Oct-12 1265 1253.85 11.15Nov-12 1196 1175.31 20.69Overallresult

5740 5597.3 142.7

Figure month wise TOU Tariff calculation costwise for LT consumers

ConclusionIt will reduce the demand supply mismatch, smartgrids will be able to self heal, provide highreliability and power quality, be resistant to cyberattacks, operate with multi-directional powerflow, increase equipment utilization, operate withlower cost and offer customers a variety ofservice choices. Smart appliances should be usedin conjunction with smart grid for reducing thepeak demand. Real time information feedbackregarding peak load conditions sent to smartappliances at customer site. Reduced variabilityin consumption leads to lower breakdowns andlower operating costs. The different types TOUtariff can be designed by analyzing the differentload patterns. Master data standardization can bedeveloped for strong analysis.

REFERENCES

1. A R Metke and R L Ekl, “SecurityTechnology for Smar Grid Networks,”Smart Grid, IEEE Transactions on vol 1,pp. 99-107, 2010

2. T Sauter and M Lobasov, “End-to-EndCommunication Architecture for SmartGrids,” Industrial Electronics, IEEETransactions on, vol 58, pp. 1218-1228,2011

3. Z Ming-Yue, “TransmissionCharacteristics of Low-VaoltageDistribution Networks in China under theSmart Grids Environment,” PowerDelivery, IEEE Transactions on, vol 26,pp. 173-180, 2011

4. Z Pei L Fangxing and N Bhatt, “NextGeneration Monitoing, Analysis andControl for The Future Smart ControlCentre,” Smart Grid, IEEE Transactionson, vol 1, pp. 186-192, 2010

5. M Liserre, T Sauter and J Y Hung,“Future Energy Systems : IntegratingRenewable Enery Sources into the SmartPower Grid through IndustrialElectronics,” Industrial ElectronicsMagazine, IEEE, vol 4, pp. 18-37, 2010

6. D Y Raghavendra Nagesh, A Sowjanya,Dr S S Tulasiram, “Real Time DecisionSupport for Energy Management” –Proceedings of the World Congress onEngineering 2008 Vol I Wce 2008, July2-4, 2008, London, UK. Isbn : 978-988-98671-9-5

7. R W Uluski, “Interactions between AMIand Distribution Management System forEfficiency/Relaiability Improvement at aTypical Utility”

8. Jian Wu, Yong Cheng, and Noel N. ,“Overview of Real Time DatabaseManagement Systems Design for PowerSystem SCADA System”

9. Jim See, Wayne Carr, Pe, and Steven E.Collier, “Real Time Distribution Analysisfor Electric Utilities”



Voltage mode control for improving MPPT performance in PV systemO.P. Suresh1, B.Subudhi1 and A.K.Panda2

1Dept. of Electrical Engg., National Institute of Technology Rourkela-769008, India2Dept. of Electronics & Communication Engg., NIST Berhampur- 761008 India

Abstract

In photo voltaic power systems, both photovoltaic (PV) modules, switching mode converters and power conditioning systemspresent non linear and time variant characteristics, which results in difficult control problem. So the main theme of this paper is toreview the different control techniques for maximum power point tracking (MPPT). These control techniques not only boost thelow voltage but also convert the solar dc power into high quality ac power for driving autonomous loads without any filter. Pulsewidth modulation (PWM) used in these control strategies can reduce the switching loss, voltage and current stress of the switchingdevices. In this study dynamics of control techniques are also presented.

I. Introduction

The growing demand for energy, together with the increasedprice of oil products and the attention paid to environmentalpollution, have progressively increased the interest inrenewable energy sources. Many applications employing thistechnology have been developed, such as solar powergeneration, solar vehicle construction, battery charging,water pumping, satellite power systems, and so on.Nowadays, PV systems, which make the conversion of solarenergy to electrical energy possible, have become widelydiffused, both in grid connection and stand aloneconfigurations. These are especially beneficial in remoteareas that are isolated from the power distribution network.

Fig.1 Renewable Energy as share of total primaryenergy consumption

Among all the energy sources, renewable energy sharesonly 8% of the total energy consumption as shown in Fig.1,but solar energy is only 1% in 8% of renewable energywhich shows the utilization is far low

Two principal types of photo voltaic power system exist,which are classified by their functions and configurations,namely: 1) Standalone PV system and 2) Grid-connected PVsystem [1]. The operation of Standalone PV system isindependent of electric utility grid.

Recently, there have been an increasing number of grid-connected systems, which are in parallel with the electricutility grid [2] and supply solar power to the utility via thegrid. Most of the solar power systems installed is gridconnected structures, so the control techniques discussed inthis paper are related to grid-connected systems [3-5]. Themain problem with linear techniques is exact control isobtained as the PV crrent equation is linearised with someapproximations. So non linear techniques has been reviewedin this paper along with the linear techniques for obtainingthe accurate and better control in PV system withoutapproximating the PV equation. The problem is defined insection II. Eight control techniques for MPPT are reviewedin this paper, namely PWM technique using current modecontrol in booth boost converter and inverter in section III,voltage mode control in boost converter[6] in section IV,and last section briefs results, conclusion and references.

II. Problem Formulation

Since the PV array has to be operated at maximum poweroutput at all radiations and temperature, it is necessary totrack the voltage corresponding to the maximum powerpoint. In almost all the linear controllers, this method isgenerally used and MPPT algorithms are generally used fortracking purpose. But still the MPPT efficiency is not uptothe required mark. Typically, the PV array is connected to apower converter that can vary the current coming from thePV array. So, based on the variation of PV current, manynon-linear controllers have been proposed in literature, a few



non-linear controllers are reviewed in this paper along withlinear controllers.

III. Control overview in PV system

In this section control in both dc-dc converter and in inverterhas been described. Fig.2 shows the grid connected PVsystem. The PV system includes a PV array, boost converter,and inverter connected to grid. As the insolation andtemperature is varying continuously during 24 hours time,the main objective in PV system is to track the maximumpower for all insolation and temperature. The PV voltage pvv

and the PV current pvi are taken in the MPPT block for

picking the maximum power point voltage pvv for

calculating the maximum power from the instantaneousvalues of pvv and pvi at different insolation and temperature.

The algorithm used in MPPT for trapping the reference

voltage pvv is P&O or I&C or hill-climbing. Since the PV

reference output voltage pvv is very small, it is required to

step-up the voltage, so the dc-dc converter used is boostconverter with PWM technique in current control modecontrol.

The reference voltage pvv obtained is given to PI controller

to obtain the reference current command Li . Fig.2 shows the

switching control both in dc-dc converter and in inverter.

The inductor current Li and the reference current command

Li is used to calculate the duty ratio D in pulse width

modulator as per the equations proposed in [7]. The driversignal obtained from PWM is used to drive the powerMOSFET bS in boost converter.

Coming to the inverter case, the output in boost converter

dcv is compared with our required reference dc voltage dcv

and fed to the PI controller, the output of PI controller is

used to adjust the grid voltage gv , to obtain the reference

grid current gi . The grid current gi and the reference grid

current gi is used to calculate the control signal proposed in

[8] for driving the four MOSFET switches s1,s2,s3,s4 in fullbridge inverter. The PI controllers are considered by usingthe first order transfer function in which only one zero andone pole is considered.

Fig.2 Basic block diagram of single phase grid connected PV system



In this control, the main observation that has to be

considered is current Li is following the reference current Li ,

PV voltage pvv is following the reference voltage pvv and

grid current gi is following the reference grid current gi to

shape the fundamental sinusoidal wave to avoid theharmonics generated on the grid side. The linear PIcontrollers are used for obtaining all the above referenceswhich controls the power switches (MOSFET) in the currentor voltage mode control. But the response obtained with thelinear PI controllers are not satisfactory which leads to thedevelopment of non-linear controllers. Three MPPTalgorithms which are commonly used for tracking thereference voltage of PV array at maximum power arediscussed to a brief extent along with flowcharts.

Perturb and observe (P&O) algorithm is dominently used inpractical PV systems for the MPPT control due to its simpleimplementation,high reliability and tracking efficiency.Fig.3 shows the flow chart of (P&O) algorithm. The presentpower ( )p k is calculated with the present values of PV

voltage ( )V k and current ( )I k , and is compared with the

previous power ( 1)p k . If the power increases, keep the next

voltage change in the same direction as the previous change.Otherwise, change the voltage in the opposite direction asthe previous one.

In hill climbing algorithm shown in Fig.4, instead ofperturbing the voltage change, duty cycle perturbation istaken, and the remaining part of algorithm is taken same.Fig.4 Flow chart of Hill-climbing algorithm

The duty ratio in this case is defined for boost converter as

pv pv

pv

V VD

V

, (1)

where pvV is the MPPT output voltage and pvV is the solar

panel voltage.

During rapidly changing irradiation and temperature, there isnecessity to track the maximum power point, so Incrementalconductance(I&C) algorithm is best suited. Fig.5 shows theflow chart of the incremental conductance algorithm whichis based on the fact that the sum of instantaneous

conductance I

V

and the incremental conductance I

V

is

zero at the MPP, positive at the left of MPP, and negative on

the right. The condition I I

V V

is usually replaced by

I I

V V

, where is a small positive number.

Fig.3 Flow chart of P&O algorithm



Fig.4 Flow chart of Hill Climbing algorithm

Fig.5 Flow chart of I&C algorithm

With this algorithm, at steady state the operating point maybe located at a certain point in the interval BC or oscillating

between the interval AB and CD as shown in Fig.6. If thestep size pvV C is increased for rapid dynamic response to

extract more energy under rapidly changing atmospheric(e.g. solar radiation and temperature) conditions, thetracking accuracy is decreased due to higher oscillationaround the MPP and vice-versa. The perturbation step-size isautomatically tuned according to the inherent PV arraycharacteristics. If the operating point is far from MPP, thestep size is increased to achieve fast dynamic response,whereas the operating point is near to the MPP, the step sizebecomes very small to reduce the steady state oscillation andimprove energy converter efficiency of the PV powersystem. From Fig.6 is obtained based on the followingequations

0,

0,

0,

dpleft of M PP

dvdp

at M PPdv

dpright of M PP

dv

(2)

Fig.6 I&C MPPT operating point trajectory

From Fig.5, it can be seen that the perturbation step sizeshould be positive (negative) to locate the MPP whenoperating on the left(right) of MPP, considering equation[

1], it is evident that the dp

dvis essentially suited for step size

IV .Voltage mode control

Fig.7 shows the PV system with PWM boost converter involtage mode control. Varying the duty cycle D can controlthe DC output voltage of a converter. Voltage mode PWM



control mainly consists of three parts, namely output voltageerror amplifier, a saw tooth generator and a PWMcomparator. In voltage control mode also for tracking theMPPT , the above three algorithms can be used. In thismethod, error amplifier compares the output voltage of dc-dcboost converter with the MPPT reference voltage whichgenerates the error voltage ev given by

2 2

1 1

1+e pv AZ Z

v v vZ Z

(3)

The feedback dynamics is determined by the error amplifierconsisting of 1Z and 2Z .The output error voltage ev is

compared with the saw tooth signal carrierv in the PWM

comparator. The output of the comparator is the signal thatdrives the power switch S in boost converter. Based on thevariation of duty ratio generated by PWM comparator, thepower switch S is operated by calculating the duty ratiowhich is given by

e

carrier

vD

v (4)

The duration of the on-time is determined by the timebetween the reset of the saw tooth signal and the intersectionof the error voltage with the positive going ramp signal.

When the output is lower than the nominal dc output value, ahigh error voltage is produced which increases the duty ratiocausing the subsequent increase in output voltage in voltagemode control and if error voltage is reduced which causesthe decrease in duty ratio such that the output voltage isdecreased to balance the total output voltage to maintainconstant.

The three algorithms Perturb & observe, hill climbing andincremental conductance algorithms used in the current andvoltage control modes are tabulated according to theperformance parameters in Table.1.

Table.1 Comparison of parameter performance of PV algorithm

Algorithm/Parameter Perturb & Observe Hill-Climbing Incremental Conductance

Dependence Voltage variation Duty ratio variation Conductance variation

Complexity Low Low Medium

Analog/Digital Both Both Digital

Periodic tuning No No Yes

Sensed parameters Voltage, current Voltage, current Voltage, current

Convergence speed Relatively low Relatively low Relatively high



Fig.7 Voltage mode control in Boost converter for switching power MOSFET

Results:

The boost converter output voltage is compared withthe reference voltage in the error amplifier and theerror voltage Ve is noted, which again is comparedwith the carrier voltage of fixed frequency (20KHz)and fixed carrier voltage (15V) in the PWMcomparator. The pulses obtained from the PWMcomparator are used for switching power devices. Thewaveform results for different operating voltages havebeen shown in Fig.8(a-i). In Fig.8(a,d,g), the outputvoltage Vs time for Vin=10V,20V,30V have beenplotted, which resulted in output voltages of 25V, 45Vand 72V respectively. Fig.8(b,e,h) shows the Errorvoltage Ve),carrier voltage(Vcarrier) Vs time which giveserror voltages of 8.5V,7.6V,6.4V for Vin=10V,20V and30V respectively. When error voltage is intersectingwith the fixed carrier frequency, switching pulses aregenerated which gives switching pulses as shown inFig.8(c,f,i). So the switching is very important to getthe required output according to the load variations.Error amplifier and PWM comparators in which thefirst order pole and zero is used to obtain the errorvoltage and PWM switching pulse.

Fig.8.a output voltage Vs time for Vin=10V

Fig.8.b Error voltage Ve),carrier voltage(Vcarrier) Vs time forVin=10V



Fig.8.c PWM switching pulses Vs time for Vin=10V

Fig.8.d output voltage Vs time for Vin=20V

Fig.8.e Error voltage Ve),carrier voltage(Vcarrier) Vs time forVin=20V

Fig.8.f PWM switching pulses Vs time for Vin=20V

Fig 8.g output voltage Vs time for Vin=30V

Fig.8.h Error voltage Ve),carrier voltage(Vcarrier) Vs time forVin=30V

Fig.8.i PWM switching pulses Vs time for Vin=30V

Fig.8[a-i] waveforms of boost converter switching

Conclusion

The above mentioned technique improve the maximumpower point tracking. voltage control mode discussed abovetracks the maximum power by using the converters(boostconverter and inverter). Different algorithms are proposedand voltage mode control along with error amplifier is also



explained for improving MPPT. Voltage control mode givenbetter results compared to MPPT algorithms.

References:[1] B. M. T. Ho and H. S.-H. Chung, “An integrated

inverter with maximum power tracking for grid-connected PV systems,” IEEE Trans. PowerElectron., vol. 20, no. 4, pp. 953–962, Jul. 2005.

[2] M. Calais, J. Myrzik, T. Spooner, and V. Agelidis,“Inverters for single-phase grid connectedphotovoltaic systems—an overview,” in Proc.IEEE Power Electronic Specialists Conf., Jun.2002, pp.1995–2000.

[3] S.B. Kjaer, J.K. Pedersen, and F. Blaabjerg, “Areview of single-phase grid-connected inverters forphotovoltaic Modules,” IEEE Trans. Ind.Appl.,vol. 41, no. 5, pp. 1292–1306, Sep./Oct. 2005.

[4] S. Saha and V. P. Sundarsingh, “Novel grid-connected photovoltaic inverter,” Proc. Inst. Elect.Eng., vol. 143, no. 2, pp. 143–56, 1996.

[5] B.K.Bose, “Energy, environment, and advances inpower electronics,” IEEE Trans. Power Electron.,vol. 15, no. 4, pp. 688-701, Jul. 2000.

[6] Power Electronics by Muhammad H.Rashid, thirdedition, pearson ducation, prentice hall.

[7] Mattavelli, P.; “Digital control of dc-dc converterswith inductor current estimation” Applied powerelectronics conference and exposition,2004.nineteenth annual IEEE volume:1,year2004,pages:74-80.

[8] kasa,N.; Iida,T.chen,L:, ” Flyback InverterControlled by Sensorless Current MPPT forPhotovoltaic Power System”, Industrial electronics,IEEE Transactions on Aug. 2005 ,Volume : 52 ,Issue:4 page(s): 1145-1152.



Last Mile Connectivity to a Web of ThingsAn Energy Management Perspective

Saswat Mohanty

Department of Electronics and TeleCommunicationsNational Institute of Science and TechnologyBerhampur, Odisha-761008 [email protected]

Abstract-Today last mile data connectivity is veryessential considering the growing data centricoperations that we require to automate and optimizeour daily activities.To minimize the non renewableenergy consumption as compared to renewableenergy consumption at the last mile we need to beconnected to the remote locations beyond the reachof prevailing data communication channels like fiberoptic cables and coaxial cable as the add up to amassive cost overhead. We will study the possibilitiesof Power Line Communications and how it can beused to bridge the gap between remote network ofintelligent devices which regulate and monitor theenergy consumption .As the intelligent devices willbe connected to a global database over the internetwe will configure these Web Of Things over the

internet.

I. INTRODUCTION

Power Line Communication (PLC) is acommunication technology that enables sending dataover existing power cables. This means that, with justpower cables running to an electronic device (forexample) one can both power it up and at the same timecontrol/retrieve data from it in a half-duplex manner.

For the purpose of understanding, PLC can be broadlyviewed as:

1. Narrowband PLC

2. Broadband PLC

Narrowband PLC works at lower frequencies (3-500kHz), lower data rates (up to 100s of kbps), and haslonger range (up to several kilometers), which can beextended using repeaters. Broadband PLC works athigher frequencies (1.8-250 MHz), high data rates (upto 100s of Mbps) and is used in shorter-rangeapplications.

Recently, narrowband Power Line Communication hasbeen receiving widespread attention due to itsapplications in the Smart Grid. Another application thatnarrowband PLC has been used in is smart energygeneration, particularly in micro-inverters for solarpanels.

Broadband PLC, in contrast, has mainly foundacceptance as a last-mile solution for Internetdistribution and home networking. With its high datarates and no additional wiring, broadband PLC is seenas an exciting and effective technology for multimediadistribution within homes. This optimism in the marketis reflected by the recent acquisitions of Intellon byAtheros, Coppergate by Sigma, DS2 by Marvell, andGigle by Broadcom, all in the Home Area Networking(HAN) segment.

There is another way to classify Power LineCommunication and that is:

1. PLC over AC lines

2. PLC over DC lines

While most companies are currently geared towardsproviding AC-PLC solutions, PLC in DC lines also hasapplications. Two such applications are PLC over theDC-bus in distributed energy generation, and PLC intransportation (electronic controls in airplanes,automobiles and trains). This use reduces wiringcomplexity, weight, and ultimately cost ofcommunications inside vehicles. However, in thisarticle, we will be dealing mostly with narrowband PLCover AC lines.

The term Internet of Things (IoT), which is rapidlygaining ground in the scenario of modern wirelesstelecommunications, has recently become popular toemphasize the vision of a global infrastructure ofnetworked physical objects. The basic idea of thisconcept is the pervasive presence around us of a varietyof things or objects – such as Radio-Frequencyidentication (RFID) tags, sensors, actuators, mobilephones, etc. – which, through unique addressingschemes, are able to interact with each other andcooperate with their neighbors to reach commongoals[1-2].

The IoT-idea is not new. However, it only recentlybecame relevant to the practical world, mainly becauseof the progress made in hardware development in thelast decade. The decline of size, cost and energyconsumption, hardware dimensions that are closely



linked to each other, now allows the manufacturing ofextremely small and inexpensive low-end computers.

Unquestionably, the main strength of the IoT idea isthe high impact it will have on several aspects ofeveryday-life and behavior of potential users. From thepoint of view of a private user, the most obvious effectsof the IoT introduction will be visible in both workingand domestic fields. In this context, domestics, assistedliving, e-health, enhanced learning are only a fewexamples of possible application scenarios in which thenew paradigm will play a leading role in the near future.

Similarly, from the perspective of business users, themost apparent consequences will be equally visible infields such as, automation and industrial manufacturing,logistics,business/process management, intelligenttransportation of people and goods.

Internet of Things technology for smart grid isforming a tremendous smart network between peopleand equipment, by all kinds of information sensingequipment or distributed the recognition (RFID device,Infrared sensors , the global positioning system, laserscanning, etc.),which combined with

the existing network technology, databasetechnology ,middleware technology, etc.[3], thefunctions are as follows:

• Running status of the electrical equipment in powersystem. (Temperature, humidity, air pressure, etc.)

• Electrical parameters monitoring of all networknodes in power system;

• Main equipment health state in power system;

• Real-time tracking of operating personnel and repairpersonnel;

• Technical personnel’s management information;

• Environmental protection index and servicecondition of environmental protection equipment.

So, the real-time monitoring of key equipmentoperation status; the network interaction and the instantconnection between users ,between user and powersystem, the meta-analysis of data information ,and theimplementation of data real-time high-speedbidirectional transmission can improve reliability andavailability of power system, which realize theoptimization of operation and management. Thus,observability, controllability, self-recoverability forpower system can be realized, for association betweenequipment, personnel management, and two-wayexchange between personnel and equipment [4].

This paper analysis the application requirement of IoTat smart grid, combined with its fundamental function.On the base of that, the author initially puts forward thestructure and function of IoT under the smart gridframework, and analysis the open issues.

II. POWER LINE COMMUNICATION

PLC is like any other communication technologywhereby a sender modulates the data to be sent, injectsit onto medium, and the receiver de-modulates the datato read it. The major difference is that PLC does notneed extra cabling, it re-uses existing wiring.Considering the pervasiveness of power lines, thismeans with PLC, virtually all line- powered devices canbe controlled or monitored.

When discussing communication technology, it isoften useful to refer to the 7-layer OSI model. SomePLC chips can implement only the Physical Layer ofthe OSI model, while others integrate all seven layers.One could use a Digital Signal Processor (DSP) with apure software realization of the MAC and an externalPHY circuit, or an optimized System-on-Chip (SoC)solution, which includes the complete PLC – MAC andPHY.

Modulation Schemes used in PLC

A variety of modulation schemes can be used in PLC.Some of these are Orthogonal Frequency DivisionMultiplexing (OFDM), Binary Phase Shift Keying(BPSK), Frequency Shift Keying (FSK), Spread-FSK(S-FSK) and proprietary schemes too (for exampleDifferential Code Shift Keying (DCSK) from Yitran).In the table below, BPSK, FSK, SFSK and OFDM arecompared on the basis of two important criteria –bandwidth efficiency and complexity (cost).

OFDM in particular offers high data rates, but requirescomputational horsepower to churn out Fast FourierTransforms (FFT) and Inverse-FFT (IFFT), as requiredby the scheme. On the other hand, BPSK, FSK arerobust and simple but offer lower data rates. Thecurrent trend is to move towards OFDM with PSKmodulation (G3 and probably P1901.2). Such heavycomputation will require DSP capability, whereas FSK,PSK and SFSK can be accomplished by amicrocontroller.

Various standards have been developed in order toensure reliable communications and inter-operability,especially for the smart grid and home networking.Examples of such standards are:

These, along with the organizations that govern themlike CENELEC, FCC, ARIB, Homeplug PowerAlliance specify ranges for operation of PLC. If aworldwide standard for PLC were to be established, thiswould have a positive impact on adoption of PLC. Sofar, the G3-PLC standard is touted as the most robustscheme available, and the IEEE 1901.2 working groupis committed to developing a universally acceptablestandard.

Where:



Figure 1. Last Mile Connectivity to IoT using PLC

CENELEC - European Committee forElectrotechnical Standardization.

ARIB – Association of Radio Industries & Businesses

EPRI – Electric Power Research Institute

FCC – Federal Communications CommissionFrequencies

Different regions of the world have differentfrequency bands allocated to narrowband PLC. Thetable below summarizes the different frequenciesavailable for narrowband PLC communication in therespective region.

Applications of PLC

Earlier, we saw that PLC is widely used in the SmartGrid and in micro-inverters. As the market gets familiarwith this technology, PLC should see wider adoption inother applications like lighting (e.g. traffic light control,LED dimming), industrial (e.g. UPS communicating toa network device, irrigation control), machine-to-machine (e.g. vending machines, a hotel’s reception-to-room communication), telemetry (e.g. offshore oil rigs),transport (e.g. Electronics in cars, trains and airplanes)and indeed, applications of PLC are only limited byone’s creativity. In this article, we will find out a littlemore about PLC in data distribution.

Last Mile Connectivity to Smart Grid using PLC

The ‘Smart Grid’ is essentially modernization of thetransmission and distribution aspects of the electricalgrid. This intelligent power distribution infrastructureenables two-way communication between theconsumers and the utility. The consumers use homenetworks to communicate with their smart meter, whichfurther communicates with the utility (AdvancedMetering Infrastructure-AMI). The Smart Griddefinition does not stop at energy utilization; supply ofenergy to the grid from Distributed Generation (DG)sources such as solar and wind fall into the samecategory. The DG system also includes Vehicle-to-Grid(V2G) - bi-directional sharing of electricity betweenElectric Vehicles (EVs) and Plug-in Electric HybridVehicles (PHEVs) and the electric power grid. In this

article, we will talk about AMI, Smart Appliances andV2G.

Advanced Metering Infrastructure:

The whole measurement and collection system thatincludes meters at the customer site, communicationnetworks between the customer and a service provider,such as an electric, gas, or water utility, and datareception and management systems, that make theinformation available to the service provider, arereferred to as AMI. The Smart Meters transmit thecollected data through commonly available fixednetworks such as Power Line Communications (PLC),Fixed Radio Frequency (RF) networks, and publicnetworks (e.g. landline, cellular, paging) which isaggregated by a concentrator, sent to the utility and thento a Meter Data Management System for data storage,analysis and billing . Studies show that NarrowbandPLC is best suited for AMI with over a 100 million NB-PLC devices installed to date.

Figure 2. TI Smart Meter with digital data interface

Utilities are investing billions in AMI systems. PLCsolution for data transmission needs no newinfrastructure, unlike wireless, as it uses the existingpower cables. Power line carrier systems have longbeen a favorite at many utilities because it allows themto reliably move data over an infrastructure that theycontrol. Utilities may also use public cellular as thebackhaul for the AMI data due to its footprint, zeroimplementation cost and low monthly fee. But on manyoccasions they may not be able to provide 100 percentcoverage of a utility’s entire customer base.

Figure 3. TI PLC Modulator interfaced with TI



Smart Mater to collect metering data over PLC

Alternatively, using wireless networks, RF solutionsor PLC for data transmission will solve this issue. Ruralutilities or the utilities located at challenging locations(for e.g. mountainous terrains) which are ill-served bywireless will have a difficulty communicating with theconsumers. Additionally, wireless and RF solutionshave reduced data rates in presence of interference likeBluetooth devices, cordless phones, concrete objects,hills and even trees. PLC can communicate to anylocation connected via the power line and has no line-of-sight requirement for data transmission. One of themost important considerations, due to the volume ofnetwork traffic inherent to the smart grid network, iscongestion mitigation. As compared to wirelesssolutions based on ZigBee or Wi-Fi, PLC-based AMIhave a proven track record of being better suited toavoid network congestion in emergency situations.Another oft cited requirement is that of redundancy inthe communication channel – with the ubiquity ofpower lines, deploying a redundant channel becomesmore economical.

III. WEB OF THINGS – THE INTERNET OFTHINGS CONCEPT

The users of today’s networked world are swampedwith information coming from a myriad of applicationsand services present on their devices, communicationinfrastructures, and on the Internet. In the near future,this information overload will be magnified many timeswhen the Internet of Things (IoT) becomes a reality, i.e.objects, smart devices, services, sensors, and RFIDs caninteract with the user and among themselves to provideservices or information. Despite the simplicity of theoperational principles of IoT technology, the design of acomplete IoT system encompasses complex interactionsnot only between different layers of the open systemsinterconnection model, but it also involves severalmarket, privacy, security, and business issues.

This heterogeneous landscape calls for a middlewareplatform which is able to consider all these complexvariables in a flexible and modular way, which is ableto provide a starting point for future upgrades andinnovations, and which considerably reduces theimplementation costs of IoT solutions.

Internet Of Things used towards last mile connectivityin Smart Grid Architecture

Electric power systems contain three majorsubsystems, power generation, power delivery, andpower utilization. Recently, IoT have been widelyrecognized as a promisingtechnology that can enhanceall these three subsystems, making IoT a vitalcomponent of the next-generation electric powersystem, the smart grid. In this section, an overview of a

few potential opportunities of applying IoT in smartgrid is discussed.

A. Smart Grid

The term “smart grid” is somewhat qualitative sincethere are various proposed implementations that havevarying levels of sophistication. However, standardamong all implementations is the use of advancedsensor and communications technologies to enablebetter use of assets provide improved reliability andenable consumer access to a wider range of services.There are some defining features that exist in mostsmart grids

• A smart grid will provide an interface betweenconsumer appliances and the traditional assets in a

power system (generation, transmission anddistribution),

• A smart grid will be at least semi-autonomous,

• A smart grid will optimize the assets of the powersystem,

• A smart grid will support better integration ofdistributed generation into the conventional centralizedpower system.

B. The application of IoT as a last mile connectivitysolution in Smart Grid

As an Indispensable basic link of smart gird peripheralinformation perception, sensor network has broadapplication space in the power system, which will playan important role in grid construction, power systemsafety production management,operation andmaintenance, information collection, safety monitoring,measurement and user interaction, etc., and improveInformation perception depth and breadth and density inthe smart grid.

Application requirements of IoT in smart grid is clear,the main scenarios are:

• In the power generation area, IoT can be used forunit monitoring, distributed power plant monitoring,plant area monitoring, pollutants and gas emissionsmonitoring, energy consumption monitoring, coalmaterial monitoring, pumped storage power plantmonitoring, wind power plant monitoring, powerprediction, photovoltaic power plantmonitoring,biomass electricity generation, energystorage monitoring, power connection,etc.

• IoT is also widely used for transmission linemonitoring, tower protection, intelligentsubstation,distribution automation, conditionmonitoring, operation and equipment management, etc.

• IoT is mainly used for smart meter and seniormeasure, smart power consumption, multi-networkconvergence, electric vehicle and charging, energy



efficiency monitoring and management, power demandside management, etc.

This shows that, every link in generation,transmission,transformation, distribution, consumptionin smart grid needs technical support of IoT. Mostbusinesses of smart grid are associated with IoT.

C. Research in the Achitecture of IoT at user sideof smart grid using PLC as interface

The final goal to realize the integrated smart gridoperation management system is constructing smartgrid IoT systematic management system, while IoTsystem at user side is the key part of smart grid IoToperation management system, which can be dividedinto the following sections. The structure of IoT appliedin smart grid

• Perception layer construction of IoT at user side ofsmart grid: The key to the IoT construction ofdistribution network and user network is the dataacquisition and the safety monitoring during dataacquisition.

Network layer construction of IoT at user side ofsmart grid: When realizing real-time data acquisition tosmart grid, it is noted that the way of data acquisition isdiverse, such as carrier, fiber,radio,GPRS,etc.

• Management layer/ decision auxiliary layerconstruction of IoT at user side of smart grid: Themanagement of IoT at user side of smart grid needs datacurve, history curve, peak alarm, as well as automatictopology and information bidirectional treatmentfunction from user side to distribution side.

Figure 4. IoT Node Based on ARM Microcontroller

IV. COMPATIBILITY ISSUES BETWEEN THEPLC AND THE IoT.

Although the enabling technologies described abovemake the IoT concept feasible, a large research effort isstill required.

we summarize the open research issues in IoTresearch, the causes for which they are specicallycrucial for IoT scenarios and the sections when suchissues will be discussed [9]. Besides, there are sometechnical issues for smart grid and the compatibilitybetween IoT and smart grid.

• There is information islands exist in electric powercorporations, which makes the information integration

difficult. For example, many utilities have implementedseparate systems such as a Trouble Call System (TCS),Supervisory Control And Data Acquisition (SCADA)System, Outage Management System (OMS), AssetManagement System (AMS), Geographic InformationSystem (GIS), just to name a few. In order to performtheir designated tasks, often these systems requireinformation that is resident in some other system andcannot be easily shared with the system that needs it.

• In order to access IoT to smart grid, needs to realizethe communication protocol conversion of IoTheterogeneous system and the safety access of sensornodes.

• The intellectualization of power system has not beencomplete, a lot of the in use device need to be upgraded,which need tremendous volume of works.

Figure 5. Testing IoT Node Performance in the Lab

• Because of the access of IoT, massive dataprocessing and storage are needed. The computer dataprocessing ability has to meet higher and stricterrequirements, and the corresponding hardware costs ofpower system are also enormous.

Various applications would include remotemonitoring, outage management (which includes faultdetection of MV equipment), Demand Response (i.e.managing customer consumption of electricity inresponse to grid supply conditions), island detection(i.e. ensuring that local grids are not being powered bythe DG system when there is no power present from theelectric grid) and fraud/theft detection.



Figure 6. Comparision between PLC and Wireless

Monitoring Data from IoT using

Last Mile Connectivity

Smart Appliances:

A Home Area Network (HAN) is a communication-enabled home where all electrical appliances areconnected in a mesh through Wireless, RF or PLC.Electrical appliances, today, are connected in a networkwith two-way communication enabled, with each otheras well as the substation. These Smart Appliances allowautomation and control from single or multiple accesspoints.

Home appliances like washing machine, dryer,dishwasher, oven and stove, refrigerator, freezer, airconditioner and water heater, talk to the Smart Meter,(Refer to previous section for details) which gathersinformation of peak pricing hours from the utilitythrough PLC. The appliances can then switch OFF/ONaccording to the price variations. This is a win-winsituation for the consumer who saves on the electricitybill, and the utility, which can better manage peakdemands. PLC also enables appliance monitoring andHVAC control – leading to further energy awarenessand savings.

The sensors on the appliance side which are part of anIoT are connected to the Monitor via the Power Line.Any changes on the appliance end will be reflected onthe LCD Display, which can be viewed and changed asrequired. Consumers will literally have control of theirentire house at their fingertips.

PLC is a considerably more effective in homenetworks. HAN being realized by Wireless/Zigbee, willneed new infrastructure to be installed. Moreover,penetrating physical barriers like walls within one floor,or reaching out to different floors is a challenge forWireless. Wireless networks often face performanceissues, like mentioned in the previous section, due toRF interference caused by devices like microwave oven,cordless phones or even Bluetooth devices at home.PLC on the other hand can reach out to every nodeconnected via the power line. It converts virtually everysocket in ones home into an access point, in many waysincorporating the best of wired and wirelesscommunication.

Smart Power in Data Centers:

The Problem: With the rise of cloud computing andinternet services, data centers and collocation facilitiescontinue to show consistent double digit growth. Datacenter downtime is completely unacceptable due to lossof revenue and reputation that it causes. Such downtimeis primarily caused by UPS battery failure, UPSoverloading, and circuit breaker failure. Another criticalconcern for data centers is energy efficiency of devicesbecause of rising unit electricity costs and additionalcooling costs. Simultaneously, companies whooutsource their computing requirements want access toall performance metrics of their systems includingpower at various levels. Adding communicationbetween devices is a challenge, because wireless cannotwork reliably in the data center environment whilewired communication would exacerbate the problem ofcable clutter.

Figure 7. PLC

V. CHALLENGES

PLC, of course has its challenges. Firstly, Power lineswere not designed to carry data, and actually behave aslow pass filters. Modeling the PL channel is difficult –it is a very harsh and noisy transmission medium,frequency-selective, time varying, and is impaired bycolored background noise and impulsive noise. Thusmaintaining signal integrity over power lines requiresrobust signaling techniques and hardware. Secondly,the structure of the grid differs across and withincountries and the same applies for indoor wiringpractices. There is no universal standard either for PLCor the grid; steps to ensure interoperability of devicesneed to be taken. Thirdly, questions are being raisedtoday about the digital security of personal informationthat is sent over the power lines because these can betapped into. Thus establishing privacy safeguards andequally important - convincing the public of these isanother large-level issue that is being addressed. Lastly,PLC faces competition from other means ofcommunication - both wired and wireless, andultimately the choice of technology will be decided by amix of cost, complexity, and feasibility. Today, themajor competing technologies to narrowband PLC areZigbee, Wi-Fi, GPRS and RS-232.

CONCLUSION

In the above article, we introduced Power LineCommunication, as a data bridging technology between



the smart grid and the internet of things that collectinformation and help in implementation of the smartgrid. The various types of PLC, modulation schemes,standards, and frequencies in use today were discussed.Finally, we presented the various applications of PLC –in energy generation, the smart grid, data-center powerdistribution networks, and, in LED lighting. Finally, thechallenges of PLC were briefly emphasized.We alsodiscussed how the Internet of things can be interlacedwith the PLC and finally the Smart Grid.The challangesin the above approach were discussed .

REFERENCES

[1] D. Giusto, A. Iera, G. Morabito, L. Atzori (Eds.), TheInternet of Things,Springer, 2010. ISBN: 978-1-4419-1673-0.

[2] Luigi Atzori, Antonio Iera,GiacomoMorabito, TheInternet of Things: A survey, Computer Networks, 2010.

[3] Li Xun 㧘 Gong Qing-wu,Qiao Hui. The application ofIOT in power systems. Power System Protection andControl, 2010,38(22):232-236

[4] Chen Shu-yong 㧘 Song Shu-fang 㧘 Li Lan-xin 㧘 et alSurvey on smart grid technology[J].Power SystemTechnology, 2009, 33(8)㧘 1-5.

[5] Ovidiu Vermesan. Outlook on Future IoTApplications.

[6] Modern grid initiative:http://www.netl.doe.gov/moderngrid/.

[7] Cameron W. Potter. Building a Smarter Smart GridThrough Better Renewable Energy Information.

[8] Pan Mingming, He Weimin, Yu dongfang. Researchof the internet of things constuction at smart grid[J].Telecommunications Science, 2010,

[9] Luigi Atzori , Antonio Iera , Giacomo Morabito. TheInternet of Things:A survey[J]. Computer Networks 54(2010) 2787–2805



Appendix-A

Renewable Energy Certification (REC)Eligibility Criteria

Existing RE generators who have power purchase agreement (PPA) with the distributionlicensees would not have the option of participating till the validity of their PPA. Hence,renewable energy producers who have opted for the preferential tariff agreement with thedistribution licensees are NOT eligible for the REC route. As per the CERC guidelines, agenerating company is eligible to apply for registration and issuance of RECs if it meets thefollowing criteria:

1. It has obtained accreditation from the State Agency;

2. It sells the generated electricity either

To the distribution licensee of the area in which the eligible entity is located, at a price notexceeding the pooled cost of power purchase of such distribution licensee or;To any other licensee or to an open access consumer at a mutually agreed price, orThrough power exchange at a market determined price;

3. All REC based captive power producers shall be eligible for their entire energy generationincluding self consumption.Therefore, in the light of above discussion, a solar power producerreally has two options for selling the produced power: Either through the preferential tariffagreement or through the REC mechanism that utilizes market forces and feeds the RenewablePurchase Obligations (RPOs).



Appendix-B

How to optimise building orientation?Building orientation plays a major role in the building envelop heat gain. A proper building orientation ofthe building will reduce the solar heat gain to your building and will make good use of day light as well.Hence it is important to consider the building orientation as well as material selection for the buildingenvelop to optimise the building energy consumption. Maximum heat gain in the building is from thesouth façade/windows, hence the designers can also consider the extended shading devices in order toreduce the heat gain.

The plate size of the building also affects the building heat gain. If there is a building having the squareplate size, the envelope surface area (peripheral area) will be minimum and if the building is having along rectangular plate size, the envelop area will be larger. The larger surface area will lead to the largerheat gain in the building and the smaller surface area will have the lower envelop heat gain. However theday lighting in a square building plate area will not be good as much of the floor area in a square buildingis far from the perimeter lighting. Hence the building plate size selection is a trade off between theenvelop heat gain and the day-lighting accessibility.



The results indicates that rectangular shaped Building consumes more energy than the Squarebecause of large area exposure to Heat. If we change the orientation of rectangular building and



put the smaller surface area towards south this reduces the energy consumption in the building4062 mWh 3852 mWh.

There are few lessons we can conclude from this exercise is that the building envelop designingis a trade-off between the external heat in the building and the day-lighting requirements. Anoptimum solution for this is a site specific requirement. If the building is a data centre and youdo not have too much of day lighting requirements, then go for a larger plat size and reduce thebuilding envelop surface area and thereby reducing the overall envelop heat gain. However if thebuilding is a commercial building, then the day-lighting requirement is a critical and you need tofind out a trade-off solution for the optimum plate size of the building in order to reduce thebuilding heat gain.



Appendix-C

How to structure a bankable solar PV project?The increased interest in solar projects in recent years have raised the concerns of the investorstowards the structuring a bankable solar project. On one side while there is a lack of clarity ofcash flows in the REC route, the banks and financial institutions are reluctant in financing theseprojects and consider these projects as non-bankable. Some of the investors have gone highlyaggressive in the reverse bidding and find it difficult to get the financed from the Banks. There isa need for the investors to understand that how the project structuring can be done in order tomake it bankable project.

Structuring a project as a bankable project requires detailed consideration on the technical, legaland economic aspects of the project. Every bank and financial has its own set of criteria throughwhich it assess the bankability of a project. However the basic requirement is that the projectshould have a stable and visible cash flow throughout the entire financing period of the project.The assessment by any banks or financial institution is done in typical four phases, which includethe

a) Initial qualitative assessment, b) Assessment of cash flow models c) Quantitative riskassessment d) Assessment of financial structure.

The major aspect which any bank or financial institution assess for the bankability of a solarproject is the local site conditions. The local site conditions are typically evaluated on the basis



of solar radiation available at the site and expected yield (Annual Energy Generation) whichdictates the cash flow of the project. It is not only important to assess the energy yield for asingle year, there is a requirement to have the energy yield assessment on long-term basis forwhich you are required to have the historical solar radiation data at least for 10 years for the sitelocation. So that the trend can be forecasted to give the expected future energy generation for thebanker. The data is not generally viable through the free data source. However some of the paiddata sources such as 3tier or solar GIS provides this data on cost basis, which makes yourdetailed project report (DPR) as a bankable DPR. Following example show the historical data forsome of the typical paid site done by us.

In order to make the DPR bankable the financial institute also asks to conduct the energy yieldassessment and electricity generation assessment at P50, P90 basis. Which gives reasonablecomfort to the bank towards the expected generation from the solar plant. The typical P50 andP90 for the site can be depicted as follows.



Apart from generation the financial institutes also sees for technical assessment so that they canassure there is no nearby shading from nearby hills which may affect the generation. In fact yourDPR should conduct detailed shadow analysis of the site including the inter row shadow whichgives a high ranking in terms of technical assessment to the bank or financial institution.

While you are submitting your application in the bank, there is a need to provide the detailsabout nearby infrastructure such as access roads, power evacuation, water availability etc.

Apart from technical aspects the bank-ability of solar project is heavily dependent on powerpurchase agreement. The PPA should be on long-term basis on a clarity of cash flow at least forthe long tenure. The lender assess PPA and the party’s financial strength/pay ability, risk of thebuyer. In this regard the PPAs with NVVN are considered bankable. The PPAs with state powerutilities are also bankable, however some of the financial institutions have discomfort with



regards to balance sheet (negative) of the state power utilities which increases the off takerpayment risk in the project.

Sometimes the PPA are signed with private parties as part of capital/ third part sale and thebalance sheet of captive power buyer with third party becomes a critical factor in the bankabilityof the solar project. The technology selection and EPC also play an important role with thebankability of a solar project. For example if the modules and inverters are supplied from thereputed supplier. It is considered as a less risky proportion by the banker. If the EPC is notexperienced or the structuring of EPC is not done through a single EPC contractor. It reduces thebankability of the project, hence de bundling of EPC contract is not advisable because banks finddifficult to put a single point responsibility in terms of performance ratio guarantee in case ofmultiple/de bundled contractor.

Finally you should take care of permits, agreements, licences, land papers and other statutoryclearances available so that there is no compliance risk seen by the lender. If you structure yourproject with all above aspects the project can be bankable, will achieve financial closure. Manytimes people approach bank first and then start accumulating this information, however if all thisinformation is compiled in advance with bank/financial institution then it will not take much timein releasing your loan.



Appendix-D

Global Smart Grid Technologies and GrowthMarkets 2013-2020

October 30, 2013 · by firstgreenadmin · in Home, News. ·

Rate This

GTM Research forecasts the cumulative value of the smart grid market to surpass $400 billionby 2020, growing with an average compound annual growth rate of over 8%. This new reportexamines the global smart grid market from the perspective of growth and market share oftechnologies and services, featuring a top-down and bottom-up financial forecast along withhighlights of regional trends, analysis and opportunity across the world’s established andemerging smart grid markets. By examining the major markets and technologies, the reportidentifies the biggest growth opportunities for the next 8 years and those markets which are mostattractive to investors and utility vendors



It examines the global smart grid market from a comparative perspective, looking at the expectedgrowth rate of the individual regions as well as the cumulative global growth rates of discretesmart grid technologies. Then, each of the regions is examined in more detail, withcomprehensive forecasting on each area for the period of 2013 to 2020, based on the key metricsof market segment value, percent share of the overall smart grid market, and compound averagegrowth rate (CAGR).



Appendix-E

Building Integrated Photo Voltaic (BIPV)Building-integrated photovoltaic (BIPV) systems are photovoltaic (PV) solar energy systemsthat are specifically designed to blend in with the architecture of a building, combining theeconomic and sustainability benefits of distributed solar energy generation with the aestheticappeal of a seamless integration into the overall building design.

BIPV system replaces the conventional building façade materials from different parts of thebuilding’s exterior and significantly improves a building’s economic and ecological balance.

Different types of BIPV panels based on crystalline silicon and thin film technologies that can beincorporated in the buildings include:

1. Transparent PV panels2. Flexible PV Panel3. Opaque – Fixed PV panels

BIPV modules can be integrated into a building in different application types, e.g. in Buildingfacades, canopies, pergola’s, roofing, skylights, railings, parking structures, fencing and others.



The feasibility of a BIPV system in a project will depend on various design considerations fore.g., areas available, type of PV module used- Crystalline or Thin film, Building orientation andanticipated energy production.

BIPV modules support structure is the building itself along with other conventional structuralframing. It replace’s the conventional façade material like glazing/granite/marble and alsoprovides onsite electricity which will reduce carbon intensive energy consumption, therebyreducing the net installation cost of the system. BIPV System also reduces thermal radiationsinto buildings and allows natural daylight pass through and further assists to improve sound andheat insulation of the building.

Onsite electricity generation maximizes the energy efficiency by eliminating the transmissionlosses that generally occur when electricity is supplied through the national grid.

The cost of a BIPV system is similar to a solar PV system, but by replacing the conventionalstructural glazing material from the building façade, the net installation cost decreases. Alongwith this benefit, accelerated depreciation and onsite electricity generation the payback periodcan be as low as 2 to 3 years and free onsite electricity is available through the life of the PVpanels.

Many aesthetically elegant and well-integrated BIPV systems have been installed worldwide andhave achieved a good market acceptance. Few project example photographs are as shown below.



BIPV system installed in a building envelope provides a highly visible public expression of thecompany’s environmental commitment and sustainability goals.

The adoption level of BIPV technology is gradually growing in India with increasing number ofsystem awareness and showcase of developments by many solar market players.

With the conventional energy prices rising very fast, BIPV is a very cost effective solution forintroduction of solar PV into building envelope and is one the crucial step in starting a Net ZeroEnergy building community.



Appendix-F

Grid integration of wind power; what can belearnt from Germany?

Germany has about 65 GW of installed capacity of which about 20 GW is contributed by windpower and balance is through Coal, Gas, Solar, Biomass and Hydro. Variability of wind poweras always been a concern to the grid operators and Germany has successfully handled gridintegration of wind through various approaches. The major initiative taken by Germany to handlesuch a large capacity was that they ensured balancing of power in order to absorb the fluctuationsof wind power.

The transmission network in Germany is predominantly at 380KV level and there are four gridoperators to handle the overall transmission system. Apart from transmission operators there areabout 70 regional grid operators which operate at voltage level of 220KV, 110KV, and 20KV.The regional grid operators do not buy or sale powers rather they provide the wheeling facilityfor the electricity. German grid operators have very high capability of forecasting the electricitydemand as well as generation and accordingly the balancing reserves are planned to fill thebuffer. While the system has about 20GW of wind power there is large variation in wind powergeneration on regional basis which can go up to 1GW of variability.



The grid operators are very well equipped with forecasting the generation as well as demand andthere are special electricity trading products available in real time so that the projectedfluctuations can be deviated through price sensitive electricity trading products. This productsinclude intraday trading and wind reserve which are applicable for two hour ahead production.

In some areas at some times the wind power is more than the gridcapability and wind farms are subjected to shut down and stop producing electricity which



effects up to 10% of their revenues of wind power generators. German law providescompensation to wind power generators for not absorbing the electricity during such events.There are additional initiatives to absorb such high wind power, this initiatives include theconstruction of energy storage facilities in the form of MW class battery storage systems, 300MW pump hydro systems as well as compressed air storage systems.

Apart from these initiatives Germany has also started adopting demand side management andload shifting options to handle the variability of the grid. Germany is also trying to implementthe joint operation of various decentralized plants to handle the variability of wind power. Forexample wind power plants are being integrated with CHP, PV and Micro hydro as a jointoperation plant which is controlled in combination from one central point. The country is alsoadopted smart grid solutions which include uses of plug in hybrid electrical vehicles, installingenergy management systems in buildings advanced metering technologies and bettercommunication system.



The systems of the four German interconnected companies with parts of Denmark, Luxembourgand Austria together form the German control bloc. The load frequency controller for theGerman control unit is located in Brauweiler (Amprion GmbH, Transmission System Operation).It controls the exchange of electrical energy vis-à-vis the UCTE system, including theCENTREL system (Poland, Slovakia, Czech Republic, Hungary). Within the German controlbloc, each interconnected company controls the import/export with the neighboring systems forhis own control zone. The surplus wind energy can fed to other zones through this interconnectedGerman bloc.



While India is also facing the grid integration issues for wind power and there is need to learnfrom German experience and adopt the solutions in the form of better scheduling andforecasting, investment in balancing reserves, upgrading the transmission line for congestionmanagement and use of storage facilities the integration of southern grid with NEW grid willalso help in evacuating power from wind farms located in southern there is a need to introducevarious demand side management (DSM measures) and innovative electricity trading to takebenefit of real time demand variations. There is no single formula for grid integration, but it isneed to take technical, policy and market mechanisms to handle large scale wind power in theIndian electricity system.



Documents

Microgrid-“A Smart Grid for Community Users” · PDF fileProceedings of National Seminar on “Dispersed Generation & Smart Grid” 2 ... the power grid, and if a microgrid