06397646

IEEE TRANSACTIONS ON SMART GRID, VOL. 3, NO. 4, DECEMBER 2012 2019

Microgrid Generation Capacity Design WithRenewables and Energy Storage Addressing

Power Quality and SuretyQiang Fu, Luis F. Montoya, Ashish Solanki, Student Member, IEEE, Adel Nasiri, Senior Member, IEEE,

Vijay Bhavaraju, Member, IEEE, T. Abdallah, and David C. Yu, Senior Member, IEEE

Abstract—Microgrids are receiving attention due to the in-creasing need to integrate distributed generations and to insurepower quality and to provide energy surety to critical loads.Since renewables need to be in the mix for energy surety, a highrenewable-energy penetrated microgrid is analyzed in this paper.The standard IEEE 34 bus distribution feeder is adapted andmanaged as a microgrid by adding distributed generation andload profiles. The 25 kV system parameters are scaled down to 12kV and renewable sources including solar PV and wind turbines,an energy storage system, and a diesel generator for islandedmode have been added to the 34-bus system. The distributiongenerations (DG) and renewables are modeled in detail usingPSCAD software and practical constraints of the components areconsidered. The monitoring of the microgrid for measuring powerquality and control requirements for these DGs and storage aremodeled to maintain the power quality of the system when loadsare varied. Renewable sources are modeled with seasonal varia-tion at different locations. The microgrid is monitored at numberof buses and the power quality issues are measured and indexesare calculated. This paper proposes a generalized approach todesign (determine the capacity requirements) and demonstratesthe management of microgrids with metrics to meet the powerquality indexes.

Index Terms—Distributed generation, high penetration renew-able, microgrid, power quality, renewable energy, smart grid.

I. INTRODUCTION

T HE GLOBAL electrical energy demand is growing grad-ually. It is expected that the demand will be doubled in

20 years [1]. Moreover, due to price volatility, limited supply,and environmental concerns of fossil fuels, wind, and solarPV power generations are rapidly utilized as alternate energysources in many parts of the world [2]. According to theAmerican Wind Energy Association (AWEA), wind energy isnow the largest new source for electricity production. Installedwind energy capacity in the U.S. was at 40 GW by the end of2010 [11]. PV industry is also experiencing a large growth.Production capacity of solar PV reached 16 GW at the end of

Manuscript received July 28, 2011; revised November 30, 2011; acceptedSeptember 04, 2012. Date of current version December 28, 2012. This workwas supported in part by the U.S. Army Corps of Engineers (ERDC/CERL)under Contract W9132T-11-C-0022. Paper no. TSG-00273-2011.Q. Fu, A. Solanki, L. F. Montoya, A. Nasiri, and D. C. Yu are with University

of Wisconsin—Milwaukee, Milwaukee, WI 53201 USA (e-mail: [email protected]).V. Bhavaraju is with Eaton Corporation Innovation Center, Milwaukee, WI

53216, USA (e-mail: [email protected]).T. Abdallah is with the U.S. Army, Champaign, IL 61822 USA (e-mail: tarek.

[email protected]).Color versions of one or more of the figures in this paper are available online

at http://ieeexplore.ieee.org.Digital Object Identifier 10.1109/TSG.2012.2223245

2010 [7]. The installed PV capacity in 2010 was eight timesof the capacity in 2006. Since deregulation of electrical energysystem has been lowering the investment in large power plants,the need for new electrical power sources could be very highin the near future.Renewable energy systems have many benefits for energy

surety as transport of diesel, natural gas, or coal is not involved.However, their utilization does not come without challenges.The higher penetration of intermittent renewable energy sys-tems such as wind and solar PV has introduced many technicalissues, including power quality, reliability, safety and protec-tion, load management, grid interconnections and controls, newregulations, and grid operation economics [2].Energy surety is a term derived for defense applications and

needs during a natural calamity such as a 2005 Hurricane Ka-trina type event. A power system has a high level of suretyif it delivers energy to the essential and critical loads for ex-tended periods while providing security, reliability, safety, sus-tainability, and cost effectiveness.Renewable energy sources and other DG can be utilized in a

microgrid by proper design and management. A primary goalof a microgrid is to operate a cluster of DG that are placed inan area power system to provide power and energy with higherreliability, surety, and quality to the local loads [3]. Most mi-crogrids are designed to be connected to the utility grid. In caseof grid power outage, they isolate themselves from the grid andmanage the local loads, voltage, and frequency [14].In a microgrid, the sources and customers are within the mi-

crogrid and some of the sources could be renewable sourcessuch as wind and solar. In the presence of renewable sources,there is an opportunity to reduce the fossil fuel consumptionwithin a microgrid. However, renewable power sources are in-termittent and can cause power reliability or quality concerns[4]. In this paper, the performance of a microgrid in both is-landed and grid connected mode is analyzed. Managing micro-grids with high renewable penetration to achieve power qualityand surety is addressed. Controls have been implemented forthe components of the system under study to properly utilizethe microgrid in both islanded and grid connected modes. Highpenetration of renewable energy sources adds complexity to thesystem. In addition, the effect of energy storage element locationis analyzed in this paper. Utilities have been using the power re-liability indexes, namely, SystemAverage Interruption DurationIndex (SAIDI), System Average Interruption Frequency Index(SAIFI), and Customer Average Interruption Duration Index(CAIDI) to evaluate the reliability of power provided to theircustomers. An analysis of SAIDI, SAIFI, and CAIDI is pro-vided in this paper for the microgrid system by monitoring the

1949-3053/$31.00 © 2012 IEEE

2020 IEEE TRANSACTIONS ON SMART GRID, VOL. 3, NO. 4, DECEMBER 2012

Fig. 1. The configuration of the microgrid studied in this paper in islanded mode.

voltage at key load locations. The proposed microgrid config-uration and load profiles in this paper can be used to evaluateand compare the performance of various control methods andsource management.

II. MICROGRID CONFIGURATION

In order to accurately study the behavior of the renewableenergy systems and diesel generator and their effects on thevoltage and frequency in a microgrid, a standard 25 kV IEEE34 bus system is adopted in this paper [5], [6]. Fig. 1 showsthe configuration of the microgrid. The original system is a 60Hz, 24.9 kV, 12 MVA with different fixed loads connected tothe utility main at bus 800 and no DG on the system. The loadtypes include constant active/reactive power loads and constantdistributed impedance loads (three-phase and single-phase). Inorder to match the properties of the system with a microgridunder construction at Fort Sill, OK, the nominal voltage of thesystem is changed to 12 kV and other components of the systemincluding loads and line impedances have been scaled accord-ingly. The base parameters of the system are changed to 12 kV, 6MVA. The transformer on bus 832 is scaled down to 12 kV/4.16kV and the two voltage regulators at bus 832 and 814 are alsoscaled to 6.9 kV, phase voltage.The power ratings of the fixed PQ loads are reduced to half

of their original values. The same also applies to the single-phase PQ loads. To scale the constant impedance loads, theirimpedances are reduced to half. Since the voltage is also half ofthe original value, their power rating is reduced to half. Thereare two types of the distribution lines in this system, namely,lumped line impedance and distributed line impedance. For thelines with lumped impedance, to keep the same voltage drop,the line impedances have been halved.The case for distributed line impedance is different. Three

methods have been considered to modify the line impedances,when scaling from 24.9 kV to 12 kV system: i) halving the R/Lmatrix, ii) halving the length of lines, and iii) halving the lengthof line and quadrupling the capacitance matrix. Methods i) andii) yield similar results but the voltage drop is larger than the

original case. Method iii) cuts the line power flow in half andat the same time keeps the nodal voltages in per unit the same.Therefore, we have used method iii) to scale the distributed lineimpedances [16].After scaling the microgrid, three types of power sources are

added a 250 kW solar PV plant, two 750 kW wind turbines, a1.5 MVA diesel generator, and two 250 kW, 500 kWh zinc-bro-mide energy storage elements. During the islanding mode, thegrid connection at bus 800 is replaced with a diesel generator(Fig. 1). Solar PV and wind turbines are modeled in PSCADin current mode. They need a reference voltage from the dieselgenerator to provide power. Energy storage model is also devel-oped in PSCAD according to experimental test results [9] withthe storage inverter operating in current mode.It should be noted this 34-bus distribution system has signifi-

cant power losses due to long distribution lines. For instance,the line between buses 806 and 814 is 49730 feet long withimpedance of . In addition, the line be-tween buses 852 and 854 is 18415 feet long with impedance of

. These losses require additional generationcapacity to supply the demand.

III. SYSTEM POWER PROFILE

The data used for loads, wind, and solar PV plants are actualmeasured data of existing systems that are scaled for the mi-crogrid in this paper. The system includes a total of 53 loads,consisting of fixed and variable PQ loads and fixed impedanceloads. The load profile for a single load at bus 848 and total mi-crogrid load are shown in Fig. 2. The peak load occurs at 7 P.M.and it is 1420 kW. The minimum load occurs at 2 A.M. and it is1120 kW.The solar PV system is modeled using solar irradiation data

from Solar Advisor Module (SAM) for the city of Milwaukee,WI. The inverter is modeled as a current source connected to themicrogrid or grid. The PV power model contains a 24 hour in-sulation profile for the summer of 2002. Maximum Power PointTracking (MPPT) for the panels was developed and simulatedusing PSCAD software. Fig. 3 shows the output power profile

FU et al.: MICROGRID GENERATION CAPACITY DESIGN WITH RENEWABLES 2021

Fig. 2. Load power profile for a single day: (a) typical load profile on bus 848and (b) total load for the microgrid.

Fig. 3. The power profile for a 250 kW solar PV plant.

for the 250 kW system. A control method is developed to cur-tail the PV power when the total renewable generation is morethan the total load demand. In this case, the diesel power is atminimum and it only establishes the voltage reference for themicrogrid.The wind turbine power profile is also modeled using mea-

sured wind speed data near the city of Milwaukee, WI. The tur-bine is modeled using PSCAD software considering the turbineefficiency factor and the mechanical and electrical effi-ciencies. The inverter modeled as a current source similar to thePV inverter. Fig. 4 shows the power profile for a 0.75 MWwindturbine for a 24 hour period. Controls have been implementedfor both wind and solar PV inverters to adjust the output reac-tive power, within the power rating of the inverter, in order toregulate the terminal voltage. If the voltage is still beyond the

Fig. 4. The power profile for a 0.75 MW wind turbine.

maximum allowed (1.05 p.u.), the active power is curtailed tolower the voltage at the source terminal to ensure continued en-ergy availability.The diesel generator plays a very important role in the Micro-

grid. It is the main source to control the voltage and frequencyof microgrid in the islanded mode. Other sources use it as a ref-erence for frequency. Whenever a load is applied to or removedfrom themicrogrid, the voltage and frequency experience a tran-sient before settling at the steady state values. The magnitudeand duration of this transient depends on the generator exciterand engine governor controls. During sudden changes in theload, the diesel generator must be able to maintain the voltageand frequencywithin the limits. The same is also truewhen thereis sudden change in the renewable energy generations.There is one diesel generator in the microgrid system with the

rating of 1.5 MVA connected to bus 800 in an islanded mode.In order to accurately study the behavior of the synchronous

machine for the power system stability studies, it is essential thatthe excitation system of the machine is modeled with sufficientdetails. The desired model must be suitable for representingthe actual excitation equipment performance for large, severedisturbances as well as for small perturbations. IEEE Standard421.5 recommends three distinctive types of excitation systemsincluding dc type excitation systems, ac type excitation systems,and static type excitation systems. Due to the fairly small size ofthe machines in this paper, we have chosen AC8B: AlternatorSupplied Rectifier Exciter with Digital Control #2 exciter type.Fig. 5 shows the simulations results for the diesel generator

when load steps occur at 10 and 20 seconds. The machine startsat no load. At 10 s, a 100% block load with a power factor of0.8 is applied. The speed (frequency) and terminal voltage ofthe machine swing as indicated. At 20 s, the load is removed andthe voltage and frequency spike. It should be noted that these re-sults are only for the diesel generator and a load and the exciterand governor controls are adjusted to represent typical perfor-mance. Results for the IEEE 34 bus configuration is discussedin Sections IV–VIII.

IV. GENERATION CAPACITY REQUIREMENT

A first order approach is used in this paper to size the gen-eration capacities. The maximum load demand of the system,as shown in Fig. 2, is 1.42 MW. To size the diesel generatorand energy storage, it is suggested that they should meet thetotal load demand considering line losses without renewableenergy sources. Two energy storage systems with total ratingof 0.5 MW for two continuous hours are considered for the


Fig. 5. Simulation results for the diesel generator when load steps are appliedat 10 s and 20 s; from top, electrical frequency (Hz) and terminal voltage (pu).

system. This storage system is a zinc-bromide battery that hasbeen tested and modeled by the research team [10].The diesel generator rating is selected at 1.5 MVA consid-

ering 0.4 MW line loss to meet the total demand. The ratingsof the renewable sources are selected so that their total capacitydoes not exceed the total system demand. The total demand con-sidering line loss is estimated at 1.75MW.Wind speed and solarirradiation pattern near city of Milwaukee have been used tocalculate the capacity factors for both wind and solar PV sys-tems. The capacity factors for location and device specific datahave been calculated at 0.29 for solar PV and 0.34 for wind tur-bines [10]. Considering an average energy cost of $210/MWhfor solar PV and $90/MWh for wind energy [8], required capac-ities for PV and wind energy are determined to meet the peakload demand. Capacities of 0.25 MW for solar PV and 1.5 MW(two 0.75 MW turbines) for wind energy are calculated.

V. SYSTEM MODELING IN ISLANDED MODE

To model the described microgrid system in islanded mode,the primary control is to adjust the active and reactive powerof all the sources to regulate the frequency and voltage of thesystem. The secondary control is to maximize the power capturefrom the renewable energy sources and minimize the energy de-livered by the diesel generator. A droop control is designed forbattery and diesel generator to coordinate the sources in order toregulate the system frequency. This method is utilized by sev-eral references [12], [13]. The diesel generator power is forcedto a minimum when the frequency reaches 61 Hz as shown inFig. 6. The battery starts absorbing power when frequency ex-ceeds 60.5 Hz and provides power when the frequency fallsunder 59.5 Hz. However, the battery has a discharge rating oftwice as its charging capability. A proportional integrator (PI)controller is designed to curtail the wind power when the fre-quency exceeds 60.8 Hz. For the solar PV, the curtailment startsat 60.9 Hz. The sources are divided into frequency droop-con-trolled and PI-controlled in order to prevent oscillations in the

Fig. 6. The active power droop mechanism for diesel generator and battery inthe microgrid.

Fig. 7. The reactive power droopmechanism for diesel generator, solar PV, andbattery in the microgrid.

system frequency and to maximize the energy delivered by therenewable sources [15].When a source delivers active power, the voltage at its ter-

minal and adjacent buses rises due to line impedances. However,if the source consumes reactive power, the terminal voltage maydecrease due to line reactance. This concept has been utilizedto regulate the bus voltages. A droop control mechanism hasbeen defined for diesel, solar PV, and battery. The diesel ceasesto produce reactive power when its terminal voltage reaches1.05 p.u., as shown in Fig. 7. It will generate 1 p.u. of reactivepower when the terminal voltage drops to 0.95 p.u. The batteryinverter and solar PV inverter will also provide reactive powerwhen their terminal voltages drop under 1 p.u. The solar PVinverter absorbs reactive power to lower the voltage when it ex-ceeds 1 p.u.The reactive power of the wind generator is regulated using

a PI controller in order to prevent fighting between sources toadjust the voltage. It controls the reactive power to adjust theterminal voltage at 1 p.u.Three conditions are applied in addition to the mechanisms

mentioned above for active and reactive power control. The firstcondition relates to the apparent power ratings of the sources. Inany condition, the total delivered apparent power must not ex-ceed the rating of the source. The second condition is for the re-active power delivery of the inverter based sources and battery.To reduce the stress on the diodes of the inverter, the reactive


Fig. 8. Active and reactive power delivered by 0.25 MW solar PV.

Fig. 9. Active and reactive power delivered by one of the 0.75 MW wind tur-bines.

power generation is limited to 0.5 p.u. when the active powerof the sources is under 0.29 p.u. This active power level is theborder line for power factor (PF) of 0.5.The third condition curtails the real power of the inverter-

based sources when the voltage exceeds a certain value (e.g.,1.04 p.u.) and source is at its maximum apparent power capa-bility. To lower the voltage, the active power is reduced usinga PI controller and reactive power is absorbed. The third condi-tion reduces the power delivered by the renewable sources butit is necessary to regulate the voltage in the system.Figs. 8–11 show the active and reactive power delivered by

the solar PV, wind, storage, and diesel generator. The solar PVpower is almost at its maximum capability as shown in Fig. 8.There are some instances (e.g., during noon) that PV has to pro-vide reactive power in order to improve the voltage and conse-quently, it has to reduce the active power. The voltage problemis mainly caused by a fast varying wind power. All the sourcesincluding storage and diesel contribute reactive power to im-prove the voltage at those instances.It should also be noted that the inverter-based sources have a

maximum current capability to protect the switches. When thesystem voltage drops, the capability to export active and reactivepower drops proportionally. This can be seen in Fig. 8 at 8 P.M.when the PV terminal voltage is low due to drop in wind powerbut the PV inverter cannot contribute 1 p.u. of reactive powersince its terminal voltage is under 1 p.u.The energy storage inverter charges and discharges the bat-

tery to support the system frequency in different cases. For in-stance in Fig. 10, after 8 P.M., the storage supports the systemwith full power since there is a large drop in wind power. Whenthe wind energy is at maximum (e.g., before 5 A.M.), it absorbspower to adjust frequency. The diesel generator supports the

Fig. 10. Total active and reactive power delivered by two 0.25 MW batterystorage systems.

Fig. 11. Active and reactive power delivered by 1.4 MVA diesel generator.

Fig. 12. Voltage profile at bus 822 of the system.

system when renewable energy systems do not provide enoughpower to support the demand. It also provides reactive powerto support the voltage. Fig. 12 shows the voltage at bus 822,which experiences the lowest voltage profile in the system dueto long lines. The voltage stays in the acceptable range duringall 24 hours. However, voltages of solar PV and wind buses dropunder 0.92 p.u. in several instances, as shown in Fig. 13.To improve the voltage profile for the renewable sources,

a voltage regulator has been added between buses 832 and852. There are several impedance loads in the system as men-tioned earlier. By improving the voltage, the demand of thoseimpedance loads increases. Right after 8 P.M., when wind powerdecreases to zero, diesel and storage cannot meet the systemdemand. Therefore, the frequency of the system collapses asshown in Fig. 14. Due to long transmission lines between dieselgenerator and major loads, the loss in the system increases


Fig. 13. Voltage profile of all sources in the system.

Fig. 14. Active and reactive power, voltage, and frequency of the diesel gen-erator when a collapse in the system occurs.

when renewable sources are not providing power. That addsmore stress on the diesel generator to supply the demand.A solution to this problem and to improve the voltage near

renewable energy sources is to move the energy storage devicecloser to these sources. For instance, if the storage is movedfrom bus 828 to bus 832, it can provide voltage support to re-newable sources and supply the adjacent loads when powerfrom these sources decrease. In addition, this will reduce theloss in the long distribution line between buses 852 and 854.Fig. 15 shows the voltage profiles of the sources in the systemwhen the storage is move to bus 832. All voltages remain inthe acceptable range. However, the voltage of some loads, e.g.,load at bus 822, drops under 0.88 p.u. when there is a large dropin wind power. The voltage profile of bus 822 is provided inFig. 16. In addition, this method limits the energy delivery ofwind turbine connected to bus 840. An alternative solution tothe problem is to move the storage to bus 832 but do not addthe regulator between buses 832 and 852. This solution resultsin better voltage profile for the system and allows more energydelivery from renewable sources. Fig. 17 shows the voltage pro-file of the sources when the storage is moved to bus 832. All the

Fig. 15. Voltage profile of sources when a regulator is added and storage ismoved to bus 832.

Fig. 16. Voltage profile of bus 822 when a regulator is added to the system andenergy storage is moved to bus 832.

Fig. 17. Improved voltage profile of renewable sources when storage is movedto bus 832.

voltages are again in acceptable range. The voltages of all loadsare also in the acceptable range. Fig. 18 shows the total load,generation, and loss in the system. The total loss of the systemis less than the case when storage is placed on bus 828. An-other solution to the problem is to apply load shedding at busesnear renewable energy surces to improve voltage profile. Thismethod is not desirable since it lowers the energy surety of thesystem and reduces the energy delivered by renewable sources.

VI. SYSTEM MODELING IN GRID TIE MODE

The system has also been modeled during grid connectedmode. In this mode, the diesel generator is disonnected. Thewind energy and solar irridation profiles are the same as in is-landed mode. Fig. 19 shows the active and reactive power pro-file of the PV. The active power of the PV is not curtailed due to


Fig. 18. Total generation, load, and loss in the system when storage is movedto bus 832.

Fig. 19. Active and reactive power of the solar PV in grid connected mode.

Fig. 20. Voltages at renewable source buses in grid connected mode.

the presence of the utility grid. The PV inverter provides reac-tive power to support its bus voltage when it drops as shown inFig. 20. The voltages of renewable sources are also more stablein this mode due to support from the grid. Fig. 21 shows the ac-tive and reactive power of the 0.75 MWwind turbine connectedto bus 840. When the voltage of the bus 840 exceeds 1.04, theturbine inverter absorbs reactive power to lower the voltage.If the voltage is not limited or if the rated apparent power isreached, the active power is curtailed to limit the bus voltage.The active power delivery is much less than the maximum cap-bility as shown in Fig. 4, due to voltage problems. This showsthat the turbine must be placed on a line with less impedancebut the distribution system and microgrid have typically highimpedance lines. The active and reactive power received fromthe grid at bus 800 is shown in Fig. 22. The active power profileof the grid is very similar to that of diesel generator in islanded

Fig. 21. Active and reactive power of the wind turbine on bus 840 in grid con-nected mode.

Fig. 22. Active and reactive power delivered by the grid to the system.

Fig. 23. Voltage profile at bus 822 of the system during grid connected mode.

mode, as shown in Fig. 11. However, the reactive power de-livery is much more since the apparent power rating limitationdoes not apply. Fig. 23 shows the voltage profile of bus 822 ingrid connected mode, which is in acceptable range.

VII. POWER QUALITY ASSESSMENT

The IEEE 34 bus system adapted for microgrid in this studyhas 53 loads and five generators. As decribed above, controlshave been implemented for renewable source inverters and en-ergy storage inverters to control the voltage in the system andincrease the energy delivery of these sources. In the islanded


mode, the diesel generator regulates the system frequency andhelps stabilize the voltage profiles. The energy storage deviceprovides active and reactive power support to the system whenthe power from renewable energy sources fluctuates. In addi-tion, the storage device plays an important role when the re-newable energy do not provide significant power, e.g., between8 P.M. and 10 P.M., in Fig. 10. It provides maximum power tosupport the nearby loads. This will reduce the burden on dieselgenerator and also reduces the total distribution power loss inthe system. In the grid connected mode, the energy storage de-vice provides reactive power support to stabilize the systemvoltages. The results of our study have shown that the best casefor this system is to remove the regulator between buses 832and 852 and move the storage device to bus 832. This case pro-vides more stable voltage profile and yields more energy fromreneable sources.SAIDI, SAIFI, and CAIDI are the parameters used by the

utility companies to evaluate the power quality and reliability.SAIDI is the average outage duration for each customer served.SAIDI is measured in units of time, often minutes. It is usuallymeasured over the course of a year, and the median value forNorth American utilities is approximately 1.50 hours. It is de-scribed as follows.

(1)SAIFI is the average number of interruptions that a customer

would experience. SAIFI is measured in units of interruptionsper customer. It is usually measured over the course of a year,and the median value for North American utilities is approxi-mately 1.10 interruptions per customer.

(2)

CAIDI is the Customer Average Interruption Duration Indexand is described as

(3)

In the modeled system, we extend the results of 24-hour studyto a year to calculate the power quality parameters. Although theapplied methods may not be accurate, they provide a good es-timation of the system performance. Power quality is evaluatedfor three cases, i) case-1 without regulator between buses 832and 852 and with storage element at bus 828, ii) case-2 withregulator and energy storage at bus 832, and iii) case-3 withoutregulator and storage at bus 832. Voltage monitoring unites havebeen implemented on each bus to analyze the voltage profiles.Table I shows the power quality parameters for these three cases.A load is considered to be interrupted when its terminal voltagefalls under 0.88 p.u. For case 3, all the voltages in the systemare above this limit and therefore there are no interruptions.It should be noted that these values only reflect the power

quality when all the components of the system function nor-mally. No scheduled or unscheduled component maintenanceis considered. In addition, the results are only for a single windand solar PV power profile. Other profiles may yield differentresults.Although the load and generation profiles are considered for

a 24-hour period, the model and results can be extended to an

TABLE IPOWER QUALITY PARAMETERS FOR THREE CASES OF THE SYSTEM IN

ISLANDING MODE

annual case. The generation and load profiles belong to a spe-cific day in month of April. In other months, the solar PV, wind,and load profile would be different. However, battery storageand diesel generator are available to compensate for any powerreduction in power generation. In fact, the annual variations ofpower profiles are much smaller in scale than the variations ofwind power profile during a 24-hour period.

VIII. CONCLUSIONS

Generation capacity sizing and power quality evaluations fora microgrid in islanded and grid connected modes have beenpresented in this paper. Standard IEEE 34 bus system is adaptedas a microgrid and diesel, solar PV, wind generations, and en-ergy storage have been added to the system. Average modelsfor the sources have been implemented to run the whole systemusing PSCAD software for 24 hours. The voltages at differentnodes have been monitored to calculate power quality indexesfor the system. Various locations for energy storage elementshave been examined and their effect on the system performanceis presented. The models provided in this paper can be used toproperly size the renewable generations to provide energy suretyand reach at certain power quality indexes.

ACKNOWLEDGMENT

This material is based upon work supported by the U.S.Army Corps of Engineers (ERDC/CERL) under ContractW9132T-11-C-0022. Any opinions, findings, and conclusionsor recommendations expressed in this material are those of theauthor(s) and do not necessarily reflect the views of the U.S.Army Corps of Engineers.

REFERENCES

[1] F. Blaabjerg, Z. Chen, and S. B. Kjaer, “Power electronics as efficientinterface in dispersed power generation systems,” IEEE Trans. PowerElectron., vol. 19, no. 5, pp. 1184–1194, 2004.

[2] T. Degner, J. Schmid, and P. Strauss, Distributed GenerationWithHighPenetration of Renewable Energy Sources Final Public Report of Dis-power Project, 2006, ISBN 3-00-016584-3.

[3] J. Huang, C. Jiang, and R. Xu, “A review on distributed energy re-sources and microgrid,” Renewable Sustainable Energy Rev., vol. 12,no. 9, pp. 2472–2483, 2008.

[4] N. Saito, T. Niimura, K. Koyanagi, and R. Yokoyama, “Trade-offanalysis of autonomous microgrid sizing with PV, diesel, and batterystorage,” in Proc. IEEE Power Energy Soc. Gen. Meet., Tokyo, Japan,2009.

[5] R. C. Dugan and W. H. Kersting, “Induction machine test case for the34-bus test feeder -description,” in Proc. IEEE Power Eng. Soc. Gen.Meet., Montreal, QC, Canada, 2006.

[6] N. Samaan, T. McDermott, B. Zavadil, and J. Li, “Induction machinetest case for the 34-bus test feeder—Steady state and dynamic solu-tions,” in Proc. IEEE Power Eng. Soc. Gen. Meet., Montreal, QC,Canada, 2006.

[7] L. Sherwood, “U.S. solar market trends, 2010,” Integrated RenewableEnergy Council, Jun. 2011.


[8] Energy Information Administration, “Annual energy outlook 2011”[Online]. Available: http://www.eia.doe.gov/oiaf/aeo/index.html

[9] E. Manla, A. Nasiri, C. Rentel, and M. Hughes, “Modeling of Zinc-Bromide energy storage for vehicular applications,” IEEE Trans. Ind.Electron., vol. 57, no. 2, pp. 624–632, 2010.

[10] A. Esmaili and A. Nasiri, “Energy storage for short-term and long-termwind energy support,” in Proc. IEEE IECON 2010, Glendale, AZ.

[11] “AWEA US wind industry annual market report year ending 2010,”2010 [Online]. Available: http://awea.org

[12] T. L. Vandoorn, B. Meersman, L. Degroote, B. Renders, and L. Vande-velde, “A control strategy for islandedmicrogrids with DC-link voltagecontrol,” IEEE Trans. Power Del., vol. 26, no. 2, pp. 703–713, Apr.2011.

[13] L. Xu and D. Chen, “Control and operation of a DC microgrid withvariable generation and energy storage,” IEEE Trans. Power Del., vol.26, no. 4, pp. 2513–2522, 2011.

[14] R. H. Lasseter, “Smart distribution: Coupled microgrids,” Proc. IEEE,vol. 99, no. 6, pp. 1074–1082, Jun. 2011.

[15] I.-Y. Chung, W. Liu, D. Cartes, E. Collins, and S.-I. Moon, “Controlmethods of inverter-interfaced distributed generators in a microgridsystem,” IEEE Trans. Ind. Appl., vol. 46, pp. 1078–1088, May 2010.

[16] Q. Fu, A. Solanki, L. F. Montoya, A. Nasiri, D. Yu, and V. Bhavaraju,“Managing intermittent renewables in a microgrid,” in Proc. IEEEISGT 2012 Conf., Washington, D.C..

Qiang Fu (S’10) was born in Anhui Province, China,in 1984. He received the B.S. and M.S. degrees fromChongqing University, Chongqing, China, in 2006and 2009, respectively. Currently, he is working to-ward the Ph.D. degree in the Electrical EngineeringDepartment, University of Wisconsin, Milwaukee.He worked for QingDao SunSong Co. Ltd. as a

part-time Research Engineer from 2007 to 2009.During the summer of 2011 and 2012, he workedas an intern for Rockwell Automation and EatonCo., respectively. His research mainly focuses on

modeling and assessing of microgrid as well as probabilistic analysis ofpower system. He is a coauthor of the book Architecture, Programming, andInterfacing for the Freescale DSP 56F8346 published by China Machine Press.He has published more than ten conference and journal papers and awardedone Chinese patent.

Luis F. Montoya (S’10) was born in Bogota,Colombia, in 2011. He received the B.S. degreefrom the National University of Colombia, Bogota,and received his M.S degree from the University ofWisconsin, Milwaukee, in 2012.He worked for the laboratory of Hydraulic De-

signs in 2010 at the National University of Colombiaas an intern. He was also a Research Assistant inthe Power System lab at the University of Wis-consin-Milwaukee (UWM), where he worked onIAPS project for UWM & Eaton Corp. He is cur-

rently working as an electrical engineer for Eaton Corp in South Carolina. Hisresearch interests are optimization techniques and methodologies for microgridoptimization and renewable energy penetration analysis for power system.

Ashish Solankiwas born in Anand, Gujarat, India, in1985. He received his B.S. degree from S. P. Univer-sity, Gujarat, in July 2007 and the M.S. degree fromGannon University, Erie, PA, in December 2008, allin electrical engineering. Currently, he is working to-ward the Ph.D. degree in electrical engineering fromthe University of Wisconsin, Milwaukee.He worked for Anthony IT, Inc. in 2009. He also

worked for Tapco, Milwaukee, WI, as an Intern inSummer 2010. His area of research are microgrids,wind energy, solar PV, energy storage systems, and

power quality and power management in microgrids.

Adel Nasiri (SM’06) was born in Sari, Iran, in 1974.He received the B.S. and M.S. degrees from SharifUniversity of Technology, Tehran, Iran, in 1996 and1998, respectively, and the Ph.D. degree from IllinoisInstitute of Technology, Chicago, in 2004, all in elec-trical engineering.He worked for Moshanir Power Engineering

Company, Tehran, from 1998 to 2001. He alsoworked for ForHealth Technologies, Inc., DaytonaBeach, FL, from 2004 to 2005 on an automatedsyringe filling device. He is currently an Associate

Professor in the Department of Electrical Engineering and Computer Science atthe University of Wisconsin, Milwaukee. His research interests are renewableenergy systems including wind and solar energy, energy storage, and micro-grids. He has published numerous technical journal and conference papers onrelated topics. He also holds four patent disclosures. He is a coauthor of thebook Uninterruptible Power Supplies and Active Filters, (CRC, Boca Raton,FL).Dr. Nasiri is currently the chair of IEEE IAS/IES Milwaukee Section,

Editor of IEEE TRANSACTIONS ON SMART GRID, Associate Editor of IEEETRANSACTIONS ON INDUSTRY APPLICATIONS, Associate Editor of the Inter-national Journal of Power Electronics, and Editorial Board Member of theJournal of Power Components and Systems. He was the general Chair of 2012IEEE Symposium on Sensorless Electric Drives and has been a member oforganizing committees for ECCE and IECON conferences and a reviewer ofIEEE journals. He is also a member of IEEE Industry Applications, IndustrialElectronics, Power Electronics, Power and Energy, and Vehicular TechnologiesSocieties.

Vijay Bhavaraju (M’00) was born in Vizag, India, in 1954. He received theB.S. degree in electrical engineering from IIT-Madras in 1976, the M.S. degreein power system operation and controls degree from S. V. University Tirupathi,India, in 1988, and the Ph.D. degree in power electronics from Texas A&MUniversity, College Station, in 1994.He worked at L&T, and Kirloskar in India and taught electrical engineering

at REC Warangal and GITAM before immigrating to the United States in 1990.After his Ph.D., he worked in the oil industry designing and commissioningoff-shore and land rigs. He developed three important products: the mud-pumpsynchronizer, the auto-drill, and block controller while working at Tech PowerControls (later acquired by NOV). He was at Ford-Ecostar during 1998 to 2004working on inverters for microturbines, photovoltaics, and fuel cells. Since 2005he has been at the Innovation Center, Eaton Corporation, Milwaukee, WI. Heis involved in different projects related to inverters for solar, batteries, and mi-crogrids. He led a team that released the 250 kW PV inverter. He is currentlyleading an Army microgrid project as Principal Investigator.Dr. Bhavaraju was a member of IEEE 1547 standard during 2000–2004. He

is currently a member of the IEC Project Team for Microgrid for Disaster Pre-paredness and Recovery.

T. Abdallah is an Electrical Engineer in the Energy Branch of the U.S. ArmyEngineer Research and Development Center’s Construction EngineeringResearch Laboratory (ERDC-CERL). He has been part of the Energy Branchat ERDC-CERL since April 2003 and oversees research and developmentprograms focused on design and development of alternative, renewable, andemerging energy sources to military applications including fuel cells, hydrogenstorage, and synthetic fuels. He graduated from Dartmouth College withan M.S. in electrical engineering with emphasis in power electronics andmagnetics.

David C. Yu (M’84) received the Ph.D. degree inelectrical engineering from the University of Okla-homa, Norman, in 1983.Currently, he is a full Professor in the Department

of Electrical Engineering and Computer Science atUniversity of Wisconsin, Milwaukee. His researchinterests include power distribution system analysis,renewable energy, and microgrid analysis.

06397646

Documents