IEEE TRANSACTIONS ON SMART GRID, VOL. 3, NO.

4, DECEMBER 2012

2019

Microgrid Generation Capacity Design With Renewables and Energy Storage Addressing Power Quality and Surety
Qiang 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

AbstractMicrogrids are receiving attention due to the increasing need to integrate distributed generations and to insure power quality and to provide energy surety to critical loads. Since renewables need to be in the mix for energy surety, a high renewable-energy penetrated microgrid is analyzed in this paper. The standard IEEE 34 bus distribution feeder is adapted and managed as a microgrid by adding distributed generation and load proles. The 25 kV system parameters are scaled down to 12 kV and renewable sources including solar PV and wind turbines, an energy storage system, and a diesel generator for islanded mode have been added to the 34-bus system. The distribution generations (DG) and renewables are modeled in detail using PSCAD software and practical constraints of the components are considered. The monitoring of the microgrid for measuring power quality and control requirements for these DGs and storage are modeled to maintain the power quality of the system when loads are varied. Renewable sources are modeled with seasonal variation at different locations. The microgrid is monitored at number of buses and the power quality issues are measured and indexes are calculated. This paper proposes a generalized approach to design (determine the capacity requirements) and demonstrates the management of microgrids with metrics to meet the power quality indexes. Index TermsDistributed generation, high penetration renewable, microgrid, power quality, renewable energy, smart grid.

I. INTRODUCTION

HE GLOBAL electrical energy demand is growing gradually. 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 solar PV power generations are rapidly utilized as alternate energy sources in many parts of the world [2]. According to the American Wind Energy Association (AWEA), wind energy is now the largest new source for electricity production. Installed wind energy capacity in the U.S. was at 40 GW by the end of 2010 [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; accepted September 04, 2012. Date of current version December 28, 2012. This work was 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 WisconsinMilwaukee, Milwaukee, WI 53201 USA (e-mail: nasiri@uwm. edu). V. Bhavaraju is with Eaton Corporation Innovation Center, Milwaukee, WI 53216, USA (e-mail: VijayBhavaraju@eaton.com). T. Abdallah is with the U.S. Army, Champaign, IL 61822 USA (e-mail: tarek. abdallah@erdc.usace.army.mil). Color versions of one or more of the gures in this paper are available online at http://ieeexplore.ieee.org. Digital Object Identier 10.1109/TSG.2012.2223245

2010 [7]. The installed PV capacity in 2010 was eight times of the capacity in 2006. Since deregulation of electrical energy system has been lowering the investment in large power plants, the need for new electrical power sources could be very high in the near future. Renewable energy systems have many benets 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 systems such as wind and solar PV has introduced many technical issues, including power quality, reliability, safety and protection, load management, grid interconnections and controls, new regulations, 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 Katrina type event. A power system has a high level of surety if it delivers energy to the essential and critical loads for extended periods while providing security, reliability, safety, sustainability, and cost effectiveness. Renewable energy sources and other DG can be utilized in a microgrid by proper design and management. A primary goal of a microgrid is to operate a cluster of DG that are placed in an area power system to provide power and energy with higher reliability, surety, and quality to the local loads [3]. Most microgrids are designed to be connected to the utility grid. In case of grid power outage, they isolate themselves from the grid and manage the local loads, voltage, and frequency [14]. In a microgrid, the sources and customers are within the microgrid and some of the sources could be renewable sources such as wind and solar. In the presence of renewable sources, there is an opportunity to reduce the fossil fuel consumption within a microgrid. However, renewable power sources are intermittent and can cause power reliability or quality concerns [4]. In this paper, the performance of a microgrid in both islanded and grid connected mode is analyzed. Managing microgrids with high renewable penetration to achieve power quality and surety is addressed. Controls have been implemented for the components of the system under study to properly utilize the microgrid in both islanded and grid connected modes. High penetration of renewable energy sources adds complexity to the system. In addition, the effect of energy storage element location is analyzed in this paper. Utilities have been using the power reliability indexes, namely, System Average Interruption Duration Index (SAIDI), System Average Interruption Frequency Index (SAIFI), and Customer Average Interruption Duration Index (CAIDI) to evaluate the reliability of power provided to their customers. An analysis of SAIDI, SAIFI, and CAIDI is provided 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 conguration of the microgrid studied in this paper in islanded mode.

voltage at key load locations. The proposed microgrid conguration and load proles in this paper can be used to evaluate and compare the performance of various control methods and source management. II. MICROGRID CONFIGURATION In order to accurately study the behavior of the renewable energy systems and diesel generator and their effects on the voltage and frequency in a microgrid, a standard 25 kV IEEE 34 bus system is adopted in this paper [5], [6]. Fig. 1 shows the conguration of the microgrid. The original system is a 60 Hz, 24.9 kV, 12 MVA with different xed loads connected to the utility main at bus 800 and no DG on the system. The load types include constant active/reactive power loads and constant distributed impedance loads (three-phase and single-phase). In order to match the properties of the system with a microgrid under construction at Fort Sill, OK, the nominal voltage of the system is changed to 12 kV and other components of the system including loads and line impedances have been scaled accordingly. The base parameters of the system are changed to 12 kV, 6 MVA. The transformer on bus 832 is scaled down to 12 kV/4.16 kV and the two voltage regulators at bus 832 and 814 are also scaled to 6.9 kV, phase voltage. The power ratings of the xed PQ loads are reduced to half of their original values. The same also applies to the singlephase PQ loads. To scale the constant impedance loads, their impedances are reduced to half. Since the voltage is also half of the original value, their power rating is reduced to half. There are two types of the distribution lines in this system, namely, lumped line impedance and distributed line impedance. For the lines 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/L matrix, ii) halving the length of lines, and iii) halving the length of line and quadrupling the capacitance matrix. Methods i) and ii) yield similar results but the voltage drop is larger than the

original case. Method iii) cuts the line power ow in half and at the same time keeps the nodal voltages in per unit the same. Therefore, we have used method iii) to scale the distributed line impedances [16]. After scaling the microgrid, three types of power sources are added a 250 kW solar PV plant, two 750 kW wind turbines, a 1.5 MVA diesel generator, and two 250 kW, 500 kWh zinc-bromide energy storage elements. During the islanding mode, the grid connection at bus 800 is replaced with a diesel generator (Fig. 1). Solar PV and wind turbines are modeled in PSCAD in current mode. They need a reference voltage from the diesel generator to provide power. Energy storage model is also developed in PSCAD according to experimental test results [9] with the storage inverter operating in current mode. It should be noted this 34-bus distribution system has signicant power losses due to long distribution lines. For instance, the line between buses 806 and 814 is 49730 feet long with impedance of . In addition, the line between buses 852 and 854 is 18415 feet long with impedance of . These losses require additional generation capacity to supply the demand. III. SYSTEM POWER PROFILE The data used for loads, wind, and solar PV plants are actual measured data of existing systems that are scaled for the microgrid in this paper. The system includes a total of 53 loads, consisting of xed and variable PQ loads and xed impedance loads. The load prole for a single load at bus 848 and total microgrid 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 is 1120 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 the microgrid or grid. The PV power model contains a 24 hour insulation prole for the summer of 2002. Maximum Power Point Tracking (MPPT) for the panels was developed and simulated using PSCAD software. Fig. 3 shows the output power prole

FU et al.: MICROGRID GENERATION CAPACITY DESIGN WITH RENEWABLES

2021

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

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

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

for the 250 kW system. A control method is developed to curtail the PV power when the total renewable generation is more than the total load demand. In this case, the diesel power is at minimum and it only establishes the voltage reference for the microgrid. The wind turbine power prole is also modeled using measured wind speed data near the city of Milwaukee, WI. The turbine is modeled using PSCAD software considering the turbine efciency factor and the mechanical and electrical efciencies. The inverter modeled as a current source similar to the PV inverter. Fig. 4 shows the power prole for a 0.75 MW wind turbine for a 24 hour period. Controls have been implemented for both wind and solar PV inverters to adjust the output reactive power, within the power rating of the inverter, in order to regulate the terminal voltage. If the voltage is still beyond the

maximum allowed (1.05 p.u.), the active power is curtailed to lower the voltage at the source terminal to ensure continued energy availability. The diesel generator plays a very important role in the Microgrid. It is the main source to control the voltage and frequency of microgrid in the islanded mode. Other sources use it as a reference for frequency. Whenever a load is applied to or removed from the microgrid, the voltage and frequency experience a transient before settling at the steady state values. The magnitude and duration of this transient depends on the generator exciter and engine governor controls. During sudden changes in the load, the diesel generator must be able to maintain the voltage and frequency within the limits. The same is also true when there is 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 that the excitation system of the machine is modeled with sufcient details. The desired model must be suitable for representing the actual excitation equipment performance for large, severe disturbances as well as for small perturbations. IEEE Standard 421.5 recommends three distinctive types of excitation systems including dc type excitation systems, ac type excitation systems, and static type excitation systems. Due to the fairly small size of the machines in this paper, we have chosen AC8B: Alternator Supplied Rectier 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 starts at no load. At 10 s, a 100% block load with a power factor of 0.8 is applied. The speed (frequency) and terminal voltage of the machine swing as indicated. At 20 s, the load is removed and the voltage and frequency spike. It should be noted that these results are only for the diesel generator and a load and the exciter and governor controls are adjusted to represent typical performance. Results for the IEEE 34 bus conguration is discussed in Sections IVVIII. IV. GENERATION CAPACITY REQUIREMENT A rst order approach is used in this paper to size the generation capacities. The maximum load demand of the system, as shown in Fig. 2, is 1.42 MW. To size the diesel generator and energy storage, it is suggested that they should meet the total load demand considering line losses without renewable energy sources. Two energy storage systems with total rating of 0.5 MW for two continuous hours are considered for the

2022

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

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

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

system. This storage system is a zinc-bromide battery that has been tested and modeled by the research team [10]. The diesel generator rating is selected at 1.5 MVA considering 0.4 MW line loss to meet the total demand. The ratings of the renewable sources are selected so that their total capacity does not exceed the total system demand. The total demand considering line loss is estimated at 1.75 MW. Wind speed and solar irradiation pattern near city of Milwaukee have been used to calculate the capacity factors for both wind and solar PV systems. The capacity factors for location and device specic data have been calculated at 0.29 for solar PV and 0.34 for wind turbines [10]. Considering an average energy cost of $210/MWh for solar PV and $90/MWh for wind energy [8], required capacities for PV and wind energy are determined to meet the peak load 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 power of all the sources to regulate the frequency and voltage of the system. The secondary control is to maximize the power capture from the renewable energy sources and minimize the energy delivered by the diesel generator. A droop control is designed for battery and diesel generator to coordinate the sources in order to regulate the system frequency. This method is utilized by several references [12], [13]. The diesel generator power is forced to a minimum when the frequency reaches 61 Hz as shown in Fig. 6. The battery starts absorbing power when frequency exceeds 60.5 Hz and provides power when the frequency falls under 59.5 Hz. However, the battery has a discharge rating of twice as its charging capability. A proportional integrator (PI) controller is designed to curtail the wind power when the frequency exceeds 60.8 Hz. For the solar PV, the curtailment starts at 60.9 Hz. The sources are divided into frequency droop-controlled and PI-controlled in order to prevent oscillations in the

Fig. 7. The reactive power droop mechanism for diesel generator, solar PV, and battery in the microgrid.

system frequency and to maximize the energy delivered by the renewable sources [15]. When a source delivers active power, the voltage at its terminal and adjacent buses rises due to line impedances. However, if the source consumes reactive power, the terminal voltage may decrease due to line reactance. This concept has been utilized to regulate the bus voltages. A droop control mechanism has been dened for diesel, solar PV, and battery. The diesel ceases to produce reactive power when its terminal voltage reaches 1.05 p.u., as shown in Fig. 7. It will generate 1 p.u. of reactive power when the terminal voltage drops to 0.95 p.u. The battery inverter and solar PV inverter will also provide reactive power when their terminal voltages drop under 1 p.u. The solar PV inverter absorbs reactive power to lower the voltage when it exceeds 1 p.u. The reactive power of the wind generator is regulated using a PI controller in order to prevent ghting between sources to adjust the voltage. It controls the reactive power to adjust the terminal voltage at 1 p.u. Three conditions are applied in addition to the mechanisms mentioned above for active and reactive power control. The rst condition relates to the apparent power ratings of the sources. In any condition, the total delivered apparent power must not exceed the rating of the source. The second condition is for the reactive power delivery of the inverter based sources and battery. To reduce the stress on the diodes of the inverter, the reactive

FU et al.: MICROGRID GENERATION CAPACITY DESIGN WITH RENEWABLES

2023

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

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

Fig. 9. Active and reactive power delivered by one of the 0.75 MW wind turbines.

power generation is limited to 0.5 p.u. when the active power of the sources is under 0.29 p.u. This active power level is the border line for power factor (PF) of 0.5. The third condition curtails the real power of the inverterbased sources when the voltage exceeds a certain value (e.g., 1.04 p.u.) and source is at its maximum apparent power capability. To lower the voltage, the active power is reduced using a PI controller and reactive power is absorbed. The third condition reduces the power delivered by the renewable sources but it is necessary to regulate the voltage in the system. Figs. 811 show the active and reactive power delivered by the solar PV, wind, storage, and diesel generator. The solar PV power is almost at its maximum capability as shown in Fig. 8. There are some instances (e.g., during noon) that PV has to provide reactive power in order to improve the voltage and consequently, it has to reduce the active power. The voltage problem is mainly caused by a fast varying wind power. All the sources including storage and diesel contribute reactive power to improve 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 the system voltage drops, the capability to export active and reactive power 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 power but the PV inverter cannot contribute 1 p.u. of reactive power since its terminal voltage is under 1 p.u. The energy storage inverter charges and discharges the battery to support the system frequency in different cases. For instance in Fig. 10, after 8 P.M., the storage supports the system with full power since there is a large drop in wind power. When the wind energy is at maximum (e.g., before 5 A.M.), it absorbs power to adjust frequency. The diesel generator supports the

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

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

system when renewable energy systems do not provide enough power to support the demand. It also provides reactive power to support the voltage. Fig. 12 shows the voltage at bus 822, which experiences the lowest voltage prole in the system due to long lines. The voltage stays in the acceptable range during all 24 hours. However, voltages of solar PV and wind buses drop under 0.92 p.u. in several instances, as shown in Fig. 13. To improve the voltage prole for the renewable sources, a voltage regulator has been added between buses 832 and 852. There are several impedance loads in the system as mentioned earlier. By improving the voltage, the demand of those impedance loads increases. Right after 8 P.M., when wind power decreases to zero, diesel and storage cannot meet the system demand. Therefore, the frequency of the system collapses as shown in Fig. 14. Due to long transmission lines between diesel generator and major loads, the loss in the system increases

2024

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

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

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

Fig. 16. Voltage prole of bus 822 when a regulator is added to the system and energy storage is moved to bus 832.

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

when renewable sources are not providing power. That adds more 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 device closer to these sources. For instance, if the storage is moved from bus 828 to bus 832, it can provide voltage support to renewable sources and supply the adjacent loads when power from these sources decrease. In addition, this will reduce the loss in the long distribution line between buses 852 and 854. Fig. 15 shows the voltage proles of the sources in the system when the storage is move to bus 832. All voltages remain in the 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 drop in wind power. The voltage prole of bus 822 is provided in Fig. 16. In addition, this method limits the energy delivery of wind turbine connected to bus 840. An alternative solution to the problem is to move the storage to bus 832 but do not add the regulator between buses 832 and 852. This solution results in better voltage prole for the system and allows more energy delivery from renewable sources. Fig. 17 shows the voltage prole of the sources when the storage is moved to bus 832. All the

Fig. 17. Improved voltage prole of renewable sources when storage is moved to bus 832.

voltages are again in acceptable range. The voltages of all loads are also in the acceptable range. Fig. 18 shows the total load, generation, and loss in the system. The total loss of the system is less than the case when storage is placed on bus 828. Another solution to the problem is to apply load shedding at buses near renewable energy surces to improve voltage prole. This method is not desirable since it lowers the energy surety of the system and reduces the energy delivered by renewable sources. VI. SYSTEM MODELING IN GRID TIE MODE The system has also been modeled during grid connected mode. In this mode, the diesel generator is disonnected. The wind energy and solar irridation proles are the same as in islanded mode. Fig. 19 shows the active and reactive power prole of the PV. The active power of the PV is not curtailed due to

FU et al.: MICROGRID GENERATION CAPACITY DESIGN WITH RENEWABLES

2025

Fig. 18. Total generation, load, and loss in the system when storage is moved to bus 832. Fig. 21. Active and reactive power of the wind turbine on bus 840 in grid connected mode.

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

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

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

the presence of the utility grid. The PV inverter provides reactive power to support its bus voltage when it drops as shown in Fig. 20. The voltages of renewable sources are also more stable in this mode due to support from the grid. Fig. 21 shows the active and reactive power of the 0.75 MW wind turbine connected to bus 840. When the voltage of the bus 840 exceeds 1.04, the turbine inverter absorbs reactive power to lower the voltage. If the voltage is not limited or if the rated apparent power is reached, the active power is curtailed to limit the bus voltage. The active power delivery is much less than the maximum capbility as shown in Fig. 4, due to voltage problems. This shows that the turbine must be placed on a line with less impedance but the distribution system and microgrid have typically high impedance lines. The active and reactive power received from the grid at bus 800 is shown in Fig. 22. The active power prole of the grid is very similar to that of diesel generator in islanded

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

mode, as shown in Fig. 11. However, the reactive power delivery is much more since the apparent power rating limitation does not apply. Fig. 23 shows the voltage prole of bus 822 in grid connected mode, which is in acceptable range. VII. POWER QUALITY ASSESSMENT The IEEE 34 bus system adapted for microgrid in this study has 53 loads and ve generators. As decribed above, controls have been implemented for renewable source inverters and energy storage inverters to control the voltage in the system and increase the energy delivery of these sources. In the islanded

2026

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

mode, the diesel generator regulates the system frequency and helps stabilize the voltage proles. The energy storage device provides active and reactive power support to the system when the power from renewable energy sources uctuates. In addition, the storage device plays an important role when the renewable energy do not provide signicant power, e.g., between 8 P.M. and 10 P.M., in Fig. 10. It provides maximum power to support the nearby loads. This will reduce the burden on diesel generator and also reduces the total distribution power loss in the system. In the grid connected mode, the energy storage device provides reactive power support to stabilize the system voltages. The results of our study have shown that the best case for this system is to remove the regulator between buses 832 and 852 and move the storage device to bus 832. This case provides more stable voltage prole and yields more energy from reneable 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 usually measured over the course of a year, and the median value for North American utilities is approximately 1.50 hours. It is described as follows. (1) SAIFI is the average number of interruptions that a customer would experience. SAIFI is measured in units of interruptions per customer. It is usually measured over the course of a year, and the median value for North American utilities is approximately 1.10 interruptions per customer. (2) CAIDI is the Customer Average Interruption Duration Index and is described as

TABLE I POWER QUALITY PARAMETERS FOR THREE CASES ISLANDING MODE

OF THE

SYSTEM

IN

annual case. The generation and load proles belong to a specic day in month of April. In other months, the solar PV, wind, and load prole would be different. However, battery storage and diesel generator are available to compensate for any power reduction in power generation. In fact, the annual variations of power proles are much smaller in scale than the variations of wind power prole during a 24-hour period. VIII. CONCLUSIONS Generation capacity sizing and power quality evaluations for a microgrid in islanded and grid connected modes have been presented in this paper. Standard IEEE 34 bus system is adapted as a microgrid and diesel, solar PV, wind generations, and energy storage have been added to the system. Average models for the sources have been implemented to run the whole system using PSCAD software for 24 hours. The voltages at different nodes have been monitored to calculate power quality indexes for the system. Various locations for energy storage elements have been examined and their effect on the system performance is presented. The models provided in this paper can be used to properly size the renewable generations to provide energy surety and 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 Contract W9132T-11-C-0022. Any opinions, ndings, and conclusions or recommendations expressed in this material are those of the author(s) and do not necessarily reect the views of the U.S. Army Corps of Engineers. REFERENCES
[1] F. Blaabjerg, Z. Chen, and S. B. Kjaer, Power electronics as efcient interface in dispersed power generation systems, IEEE Trans. Power Electron., vol. 19, no. 5, pp. 11841194, 2004. [2] T. Degner, J. Schmid, and P. Strauss, Distributed Generation With High Penetration of Renewable Energy Sources Final Public Report of Dispower Project, 2006, ISBN 3-00-016584-3. [3] J. Huang, C. Jiang, and R. Xu, A review on distributed energy resources and microgrid, Renewable Sustainable Energy Rev., vol. 12, no. 9, pp. 24722483, 2008. [4] N. Saito, T. Niimura, K. Koyanagi, and R. Yokoyama, Trade-off analysis of autonomous microgrid sizing with PV, diesel, and battery storage, in Proc. IEEE Power Energy Soc. Gen. Meet., Tokyo, Japan, 2009. [5] R. C. Dugan and W. H. Kersting, Induction machine test case for the 34-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 machine test case for the 34-bus test feederSteady state and dynamic solutions, in Proc. IEEE Power Eng. Soc. Gen. Meet., Montreal, QC, Canada, 2006. [7] L. Sherwood, U.S. solar market trends, 2010, Integrated Renewable Energy Council, Jun. 2011.

(3) In the modeled system, we extend the results of 24-hour study to a year to calculate the power quality parameters. Although the applied methods may not be accurate, they provide a good estimation of the system performance. Power quality is evaluated for three cases, i) case-1 without regulator between buses 832 and 852 and with storage element at bus 828, ii) case-2 with regulator and energy storage at bus 832, and iii) case-3 without regulator and storage at bus 832. Voltage monitoring unites have been implemented on each bus to analyze the voltage proles. Table I shows the power quality parameters for these three cases. A load is considered to be interrupted when its terminal voltage falls under 0.88 p.u. For case 3, all the voltages in the system are above this limit and therefore there are no interruptions. It should be noted that these values only reect the power quality when all the components of the system function normally. No scheduled or unscheduled component maintenance is considered. In addition, the results are only for a single wind and solar PV power prole. Other proles may yield different results. Although the load and generation proles are considered for a 24-hour period, the model and results can be extended to an

FU et al.: MICROGRID GENERATION CAPACITY DESIGN WITH RENEWABLES

2027

[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 ZincBromide energy storage for vehicular applications, IEEE Trans. Ind. Electron., vol. 57, no. 2, pp. 624632, 2010. [10] A. Esmaili and A. Nasiri, Energy storage for short-term and long-term wind 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. Vandevelde, A control strategy for islanded microgrids with DC-link voltage control, IEEE Trans. Power Del., vol. 26, no. 2, pp. 703713, Apr. 2011. [13] L. Xu and D. Chen, Control and operation of a DC microgrid with variable generation and energy storage, IEEE Trans. Power Del., vol. 26, no. 4, pp. 25132522, 2011. [14] R. H. Lasseter, Smart distribution: Coupled microgrids, Proc. IEEE, vol. 99, no. 6, pp. 10741082, Jun. 2011. [15] I.-Y. Chung, W. Liu, D. Cartes, E. Collins, and S.-I. Moon, Control methods of inverter-interfaced distributed generators in a microgrid system, IEEE Trans. Ind. Appl., vol. 46, pp. 10781088, 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. IEEE ISGT 2012 Conf., Washington, D.C..

Qiang Fu (S10) was born in Anhui Province, China, in 1984. He received the B.S. and M.S. degrees from Chongqing University, Chongqing, China, in 2006 and 2009, respectively. Currently, he is working toward the Ph.D. degree in the Electrical Engineering Department, 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 worked as an intern for Rockwell Automation and Eaton Co., respectively. His research mainly focuses on modeling and assessing of microgrid as well as probabilistic analysis of power system. He is a coauthor of the book Architecture, Programming, and Interfacing for the Freescale DSP 56F8346 published by China Machine Press. He has published more than ten conference and journal papers and awarded one Chinese patent. Luis F. Montoya (S10) was born in Bogota, Colombia, in 2011. He received the B.S. degree from the National University of Colombia, Bogota, and received his M.S degree from the University of Wisconsin, Milwaukee, in 2012. He worked for the laboratory of Hydraulic Designs in 2010 at the National University of Colombia as an intern. He was also a Research Assistant in the Power System lab at the University of Wisconsin-Milwaukee (UWM), where he worked on IAPS project for UWM & Eaton Corp. He is currently working as an electrical engineer for Eaton Corp in South Carolina. His research interests are optimization techniques and methodologies for microgrid optimization and renewable energy penetration analysis for power system. Ashish Solanki was born in Anand, Gujarat, India, in 1985. He received his B.S. degree from S. P. University, Gujarat, in July 2007 and the M.S. degree from Gannon University, Erie, PA, in December 2008, all in electrical engineering. Currently, he is working toward the Ph.D. degree in electrical engineering from the University of Wisconsin, Milwaukee. He worked for Anthony IT, Inc. in 2009. He also worked for Tapco, Milwaukee, WI, as an Intern in Summer 2010. His area of research are microgrids, wind energy, solar PV, energy storage systems, and power quality and power management in microgrids.

Adel Nasiri (SM06) was born in Sari, Iran, in 1974. He received the B.S. and M.S. degrees from Sharif University of Technology, Tehran, Iran, in 1996 and 1998, respectively, and the Ph.D. degree from Illinois Institute of Technology, Chicago, in 2004, all in electrical engineering. He worked for Moshanir Power Engineering Company, Tehran, from 1998 to 2001. He also worked for ForHealth Technologies, Inc., Daytona Beach, FL, from 2004 to 2005 on an automated syringe lling device. He is currently an Associate Professor in the Department of Electrical Engineering and Computer Science at the University of Wisconsin, Milwaukee. His research interests are renewable energy systems including wind and solar energy, energy storage, and microgrids. He has published numerous technical journal and conference papers on related topics. He also holds four patent disclosures. He is a coauthor of the book 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 IEEE TRANSACTIONS ON INDUSTRY APPLICATIONS, Associate Editor of the International Journal of Power Electronics, and Editorial Board Member of the Journal of Power Components and Systems. He was the general Chair of 2012 IEEE Symposium on Sensorless Electric Drives and has been a member of organizing committees for ECCE and IECON conferences and a reviewer of IEEE journals. He is also a member of IEEE Industry Applications, Industrial Electronics, Power Electronics, Power and Energy, and Vehicular Technologies Societies. Vijay Bhavaraju (M00) was born in Vizag, India, in 1954. He received the B.S. degree in electrical engineering from IIT-Madras in 1976, the M.S. degree in 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&M University, 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 commissioning off-shore and land rigs. He developed three important products: the mud-pump synchronizer, the auto-drill, and block controller while working at Tech Power Controls (later acquired by NOV). He was at Ford-Ecostar during 1998 to 2004 working on inverters for microturbines, photovoltaics, and fuel cells. Since 2005 he has been at the Innovation Center, Eaton Corporation, Milwaukee, WI. He is involved in different projects related to inverters for solar, batteries, and microgrids. He led a team that released the 250 kW PV inverter. He is currently leading an Army microgrid project as Principal Investigator. Dr. Bhavaraju was a member of IEEE 1547 standard during 20002004. He is currently a member of the IEC Project Team for Microgrid for Disaster Preparedness and Recovery. T. Abdallah is an Electrical Engineer in the Energy Branch of the U.S. Army Engineer Research and Development Centers Construction Engineering Research Laboratory (ERDC-CERL). He has been part of the Energy Branch at ERDC-CERL since April 2003 and oversees research and development programs focused on design and development of alternative, renewable, and emerging energy sources to military applications including fuel cells, hydrogen storage, and synthetic fuels. He graduated from Dartmouth College with an M.S. in electrical engineering with emphasis in power electronics and magnetics. David C. Yu (M84) received the Ph.D. degree in electrical engineering from the University of Oklahoma, Norman, in 1983. Currently, he is a full Professor in the Department of Electrical Engineering and Computer Science at University of Wisconsin, Milwaukee. His research interests include power distribution system analysis, renewable energy, and microgrid analysis.