Wide-Area Measurement System Development at the ... - IEEE Xplore

This article has been accepted for publication in a future issue of this journal, but has not been fully edited. Content may change prior to final publication. Citation information: DOI 10.1109/TPWRD.2015.2478380, IEEE Transactions on Power Delivery

1

Wide-Area Measurement System Development at the Distribution Level: an FNET/GridEye Example Yong Liu, Lingwei Zhan, Student Member, IEEE, Ye Zhang, Penn N. Markham, Member, IEEE, Dao Zhou, Student Member, IEEE, Jiahui Guo, Student Member, IEEE, Yin Lei, Student Member, IEEE, Gefei Kou, Student Member, Wenxuan Yao, Jidong Chai, and Yilu Liu, Fellow, IEEE Abstract— Electric power grid wide-area monitoring system (WAMS) have been extended from the transmission to distribution level. As the first WAMS deployed at the distribution level, the frequency monitoring network FNET/GridEye uses GPS-timesynchronized monitors called frequency disturbance recorders (FDRs) to capture dynamic grid behaviors. In this paper, the latest developments of monitor design and the state-of-the-art data analytics applications of FNET/GridEye are introduced. Its innovations and uniqueness are also discussed. Thanks to its low cost, easy installation and multi-functionalities, FNET/GridEye works as a cost-effective situational awareness tool for power grid operators and pioneers the development of WAMS in electric power grids. Index Terms—Distribution level, frequency disturbance recorder (FDR), phasor estimation, situational awareness, synchrophasor, wide-area monitoring system (WAMS).

I. INTRODUCTION

P

HASOR measurement unit (PMU) is a relatively new power system situational awareness tool capable of providing real-time, GPS-time-synchronized measurements of grid status at high data rates (two orders of magnitude beyond traditional grid telemetry) [1]-[3]. It reveals unprecedented insights into power system dynamics and is able to improve the power system operators’ situational awareness capability significantly especially with the presence of renewable generations. Nevertheless, PMUs are usually installed in the substations at the transmission level and the associated high manufacturing and installation costs impede their fast deployment. Originally developed in 2003, the frequency monitoring network FNET/GridEye is the first wide-area phasor measurement system ever designed to be deployed at the distribution level [4]-[9]. It uses low-cost high-accuracy frequency disturbance recorders (FDRs) to collect power grid data (frequency, voltage magnitude, voltage phase angle, as This paper was submitted on March 16, 2015, revised on Jun 26, 2015 and accepted on September 7, 2015. This work made use of the Engineering Research Center Shared Facilities supported by the Engineering Research Center Program of the National Science Foundation and DOE under NSF Award Number EEC-1041877 and the CURENT Industry Partnership Program. Penn N. Markham is with Electric Power Research Institute, Knoxville, TN, 37932, USA. (E-mail: [email protected]) Ye Zhang is with Alstom Grid, Redmond, WA, 98052, USA. (E-mail: [email protected]) All the other authors are with Department of Electrical Engineering and Computer Science, the University of Tennessee, Knoxville, TN, 37996, USA. (e-mails: [email protected], [email protected], [email protected], [email protected], [email protected], [email protected], [email protected], [email protected], [email protected])

well as power quality information). Since then, extending the wide-area monitoring system to the distribution level has attracted more and more attention. For example, Ref [10] reported a Micro-PMU concept that aims to monitor the distribution system in real-time to improve its observability and control. As another example, Ref. [11] proposed a distribution level field programmable gate array (FPGA)-based PMU design for post-event analysis. Compared to its recent counterparts, FNET/GridEye is unique in several aspects: Firstly, it takes the voltage at standard 120-V electrical outlets as the signal input. This customer-side single-phase design enables its plug-and-play feature and minimized the manufacture cost and installation effort. Secondly, FNET/GridEye achieves higher measurement accuracy than its transmission and distribution level counterparts. For example, the Micro-PMU’s angle accuracy is expected to be within ±0.05 ̊ [10] while FNET/GridEye has already achieved ±0.005 ̊. Last and most importantly, FNET/GridEye has successfully developed and implemented a suite of data visualization and analytics functions to process a large volume of measurements. There are functions that perform analysis in real-time as data streams arrive from FDRs, and there are functions that work off-line for analysis of archived data. These functions help grid operators to interpret the power grid operation status and take proactive measures to prevent blackouts. It should be noted that though FNET/GridEye can be potentially applied for both transmission and distribution levels, only transmission level applications will be reported in this paper. This is because FNET/GridEye currently does not have enough sensors deployed within a distribution grid. Comparatively, though the applications of Micro-PMU are expected to include both diagnostic and control applications for the distribution grid such as topology detection, state estimation, fault/high impedance locations, protective relaying and Var optimization, they are actually still under development and no real results have been reported yet [10].

Fig.1. Map of FDR locations in North America

FNET/GridEye is widely welcomed by the academia, industry as well as governments and serves more than twenty

0885-8977 (c) 2015 IEEE. Translations and content mining are permitted for academic research only. Personal use is also permitted, but republication/redistribution requires IEEE permission. See http://www.ieee.org/publications_standards/publications/rights/index.html for more information.


2

main power grids in the world as of 2015. Fig.1 shows the current FDR installation locations in the North America. All the phasor measurement data collected by FDRs are transmitted to the FNET/GridEye server hosted at the University of Tennessee, Knoxville (UTK), and Oak Ridge National Laboratory (ORNL) for preprocessing, conditioning, storage, and applications. In this paper, the latest development of the FNET/GridEye sensor and server design will be reviewed, and then the stateof-the-art FNET/GridEye applications will be presented categorically. The paper is structured as follows: Section II introduces the latest FDR sensor features; Section III describes the overall framework of the FNET/GridEye server; Real-time applications of FNET/GridEye are presented in Section IV; Section V depicts the various non-real-time data analytics functions; and Section VI concludes the paper. II. SENSOR DESIGN Compared to the transmission level, deploying at the distribution level brings several challenges for the sensor design. Firstly, there are much more harmonics and distortions in the distribution grid. Secondly, despite the noisier environment, the measurement accuracy needs to be higher in order to capture the relatively small variations. Finally and most importantly, the cost of distribution level sensors should be significantly reduced for the sake of wide deployment. Fully addressing these challenges, the FDR achieves the accuracy of ±0.00006 Hz for frequency and ±0.005 ̊ for voltage angle under steady-state conditions with a manufacture cost only tenths of traditional PMUs. More details of the sensor design are presented as follows. Three generations of FDRs have been developed so far. Despite the hardware component difference, all three generations share the same principle. Fig. 2 shows a photo of the currently deployed Generation-II FDR while Fig. 3 demonstrates its hardware layout. As shown in Fig.3, the analog-digital conversion (ADC) component periodically samples the conditioned voltage signal after the voltage transducer and anti-alising filter with the help of digital signal processor (DSP) oscillator pulses. To achieve high sampling precision, these oscillator pulses are regulated by the one pulse per second (PPS) signal provided by the GPS receiver. Phasor values are calculated in the DSP by an improved DFT algorithm[12] [13], time-stamped in the microprocessor, packaged by the Ethernet transceiver, and then transmitted over Ethernet.

Power grid signal

Voltage transducer

Oscillator pulse Data

Antialiasing filter

ADC

Timing

Microprocessor

Data Acquisition

Phasor Calculation Measurement

GPS antenna

Measurement

GPS Receiver

Ethernet Transceiver

LCD

Timing

Communication

Display

Internet

Fig.3. Generation-II FDR hardware block diagram

After launching the Generation-II FDRs successfully, a series of hardware and firmware updates have been added in order to further enhance the devices’ performance and extend their applications. Some of the new features of Generation-III FDRs are briefly introduced below. A. Higher-accuracy phasor calculation Both hardware and firmware have been upgraded to further improve the phasor measurement accuracy for Generation-III FDR [14]-[16]. Specifically, a 16-bit ADC with ultra-high precision bandgap voltage reference is used to replace the 14bit ADC with internal voltage reference in Gen-II FDRs. The signal noise ratio (SNR) of the sampling circuit can be improved from 78 dB to 90 dB as a result. Additionally, an adaptive synchronous sampling algorithm is implemented in Gen-III FDRs, which enables the maximum timing error between two consecutive samples to be less than 10 ns [14]. Due to these upgrades, the steady-state phase angle and frequency measurement error of the Gen-III FDRs is less than 0.005 ̊ and 0.00006 Hz, respectively. The phasor measurement accuracy under dynamic conditions has also been improved by utilizing an improved DFT-based dynamic phasor measurement algorithm in Gen-III FDRs, which relies on multiple-step digital filters and a dynamic error compensation module [15]. The testing results show that the accuracy of Gen-III FDRs far exceeds the PMU Standard C37.118.1-2011 and C37.118.1a2014. B. GPS Timing Backup Because of phasor calculation’s extreme sensitivity to sampling accuracy, GPS signal has been used to help waveform sampling for both PMUs and FDRs. Nevertheless, GPS satellites can be difficult to lock due to weather or other uncontrollable factors, and severe data missing problems are caused as a result. Hence it’s necessary to equip PMUs and FDRs with another accurate timing source in case the GPS signal is not available. Frequency Error (mHz)

0.5

Fig.2. Photo of Generation-II FDR

Digital Signal Processor

PPS

GPS-FDR CSAC-FDR

0

-0.5 0

20

40

60

Time(s)

80

100

Fig. 4. Comparison between CSAC and GPS timing



3

The atomic clock has not been considered as an option until recently because of its formidable price tag. In 2011, the chipscale atomic clock (CSAC) was made available and affordable for civil use. Since then, CSACs such as Quantum™ SA.45s have been utilized in industrial areas that require extremely high-precision timing. This product was also tested on FDRs to explore its potential as a PMU and FDR timing source. The test results showed that, achieving a 100 nano second-level accuracy, CSAC proves to be an accurate and reliable timing source for power system wide-area monitoring. Fig.4 shows a frequency measurement comparison between a FDR with a GPS module (including a GPS receiver and antenna) and a FDR equipped with a CSAC. It’s obvious that the measurement accuracies of two timing sources are similar. Our experiments also demonstrated that, with the help of CSAC, FDRs’ holdover capability can be up to a day compared to a few seconds of certain PMUs. Please note that CSAC still requires an external timing source, such as a GPS module, for disciplining and is only intended to be used when the GPS timing source is temporarily lost. That means a CSAC can only be considered as a backup, rather than a replacement of a GPS module. C. Non-contact Sensor Design In open areas far from any substation, it is difficult to install and maintain a traditional PMU device due to the lack of needed facilities. However, there is a need to monitor those areas’ overhead transmission lines for some applications. It’s known that over-head transmission lines generate electric fields in the right-of-way of their corridors. This electric field is of the same frequency as the transmission line AC voltage waveform and can be strong enough to be captured by an IEEE 644-1994 Standard-defined free-body meter without being totally submerged by environmental noise [17]. Alternative electric field 2.2MΩ

2.2MΩ

+12V

Copper panel

OP177

Data Center

OP177

+ -12V 10kΩ

10µF

The measurements collected by FDRs are transmitted to data centers at UT and ORNL where all the FDR measurement data are concentrated and processed. Physically, a FNET/GridEye data center operates on several dedicated server machines, e.g. data server, application server, web server, backup server, etc. Functionally, a data center can be treated as a multi-layer data management system as shown in Fig. 6. The first and most important component of a data center is the data concentrator, where real-time measurements are extracted from the network TCP/IP data package, interpreted, error checked, time-aligned, and then streamed into different layers for application or storage. The second layer of a data center hierarchy is composed of the real-time application agent and the data storage agent. The real-time application agent consists of various real-time application modules (e.g., disturbance detection, oscillation detection).

+12V

+ 10kΩ

III. SERVER FRAMEWORK

-

10kΩ

10µF

10kΩ

D. Power Quality Monitoring Power quality is an essential issue for electric utilities. Deployed at the distribution level, FDRs have the unique advantage of promoting a better understanding of distributionlevel power quality issues. Specifically, a modified Periodogram method is adopted to calculate power spectral density (PSD) of the power grid signal in frequency domain [14]. Furthermore, the method recommended by IEC 61000-430 Standard is employed to detect voltage sag and swell. Therefore, with minor changes of the FDR design, nextgeneration FDRs can also function as a powerful power quality analyzer by estimating harmonics composition and detecting voltage sag and swell. Additionally, the harmonics analysis results can help to better understand how the phasor calculation is tampered by the harmonics.

-12V

FDR Signal Input

Firewall &Router

Free-body meter Fig.5. Illustration of non-contact FDR design

Amplifier

Taking advantage of this phenomenon, a non-contact FDR design was developed [18]. To form a free-body meter, copper is printed on both sides of a printed circuit board (PCB) (as shown in Fig.5). According to Faraday’s law, a small alternative potential will be induced between the two copper panels by the varying electric field in the atmosphere. This voltage signal will be magnified by an amplifier circuit and fed into an FDR as the input signal. The total power consumption of such an FDR device is less than 5 Watts. With a battery and a wireless communication module, a portable non-contact FDR will be available. This FDR design can take equally accurate measurements of power grid frequency as traditional PMUs while greatly increasing the flexibility and portability of phasor measurements.

Realtime Data streaming

Real-time Applications

FDR 1 FDR 2 FDR n

Firewall &Router

Non Real-time Applications

Data Concentrator Realtime Data streaming

Data Storage Layer 1

Layer 2

Layer 3

Fig.6. FNET/GridEye data center structure

The third layer is the non-real-time application agent. Applications implemented on this layer (e.g., event replay applications, frequency statistical analysis, and other userrequested web services) are operated on various saved data formats (e.g., txt, mysql data) instead of real-time streaming data. This multi-layer data management system successfully deals with the various time requirements of different functionalities and accomplishes the efficient collection, storage, and utilization of real-time data. In the following two



4

sections, various real-time and non-real-time applications run on the data center will be introduced respectively. Please note that since FNET/GridEye relies on the public Internet for data communication right now, the latency and other communication issues are understandably more severe than traditional WAMS. Only real-time monitoring and diagnostic applications have been implemented online so far because they have relatively lower requirements for communication. If the measurement data is intended to be used for real-time control purposes, internal or dedicated communication networks will be needed. IV. REAL-TIME APPLICATIONS As mentioned above, FNET/GridEye applications can be divided into real-time and non-real-time applications roughly by their response time frame. Real-time applications require response within seconds or even sub-seconds after receiving the measurement data, while non-real-time applications have more flexible timing requirements or are upon request [19]. In this section, some of the important real-time applications are presented. A. Real-time Visualization of Measurement Data Real-time visualization of wide-area measurement data is one of the FNET/GridEye system’s most important applications. Correlating streaming wide-area frequency and voltage angle measurements with corresponding FDR geographical location information, the FNET/GridEye system creates an intuitive real-time visualization tool that helps operators better interpret what happens in the power grid in realtime. More importantly, an electric utility usually has access to information about its own system but very limited access outside its control area. FNET/GridEye provides full coverage and thus presents a whole picture of the entire North American power system. Fig. 7 and Fig. 8 shows a snapshot of the realtime frequency and angle contour map of the North American power grid respectively. These maps can be accessed online through the FNET/GridEye web services.

Fig. 7. FNET/GridEye real-time frequency contour map

Fig. 8. FNET/GridEye real-time angle contour map

B. Disturbance Recognition and Location Disturbances such as losing generators or transmission lines occur in the power grid frequently—often on an hourly basis. To prevent a single disturbance from escalating into large-area blackouts, the first step is to detect and locate the disturbance in the fastest manner. Continuously screening the streaming frequency measurement data, the FNET/GridEye disturbance recognition module uses the rate of average frequency change df/dt as an indicator of power system disturbance. If df/dt exceeds a predefined threshold, the disturbance detection module will be triggered. Please note that the thresholds vary for different power grids and different seasons. Once a frequency disturbance is detected, it will be categorized into generation trip, load shedding, line trip etc. based on its characteristics using an artificial neural network (ANN) algorithm [20] and then a geometrical tri-angulation algorithm making use of the time difference of arrival (TDOA) will be utilized to locate this disturbance [21] [22]. If it is a generation trip, the net active power change ΔP can be estimated by multiplying the average frequency deviation 𝛥f and the empirical coefficient beta value 𝛽. ABC points of the frequency curve will also be calculated. All these information as well as a frequency plot will be presented in a FNET/GridEye event report (as shown in Fig.9) and sent out automatically to service subscribers, such as utility operators, in seconds. C. Inter-area Oscillation Detection and Modal Analysis Small-signal stability is a key concern of power system operators. Utilizing a frequency or angle-based oscillation detection algorithm, FNET/GridEye is effective in monitoring the low-frequency inter-area oscillations that occur in a power grid [23]. Employing a multi-channel matrix pencil algorithm, a modal analysis of each oscillation event can also be performed. As shown by Fig. 10, once an oscillation is detected, the frequency and relative angle of involved FDRs will be plotted and the FDRs that record the largest oscillation amplitude will be listed. The frequency and damping ratio information of dominant modes will also be computed and given. Please note that this is also an automatic service that can be sent to service subscribers in real time.



5

Fig. 9. FNET/GridEye disturbance recognition and location report

Fig. 10. FNET/GridEye oscillation report

D. Ambient Data-Based Oscillation Mode frequency and Damping Ratio Estimation Power system ambient data is the natural response of the system due to small-magnitude disturbances, random load switching, etc. Despite the embedded higher noise, rich

oscillation modal information can still be abstracted. This FNET/GridEye application module employs an empirical mode decomposition (EMD) filter [24] to de-trend the ambient signal first and then utilizes an auto-regressing moving-average (ARMA) model method [25] to abstract the oscillation mode frequency and damping ratio information. Most importantly, by



6

use of a multi-channel parallel processing design, this function module has the capability to process hundreds of streaming FDR measurements at the same time, which gives a whole picture of the entire North American power grids oscillation information in real-time for the first time ever. Fig. 11 shows the oscillation mode frequency and damping ratio calculation results of one FDR.

based power system models always have limitations of the amount of detail that can be completely and accurately included. Taking advantage of the real-time wide-area measurements, one of phasor measurements’ novel applications is to develop data-driven models that can be updated online to estimate or predict system responses [28]-[30]. -3

x 10

Voltage Angle (rads)

1 0 -1 -2 -3 Actual Estimated

-4 10

20

30 40 Time (s)

50

60

Fig.13. Estimated angle response from a measurement data-driven model

Fig. 11. Online oscillation mode frequency and damping ratio display

E. Islanding and Off-grid Detection Islanding is an extremely dangerous phenomenon that occurs when one or more generators are no longer working synchronously with the rest of the power system. Islanding can escalate to blackouts if no control action is taken in time. The FNET/GridEye islanding detection module calculates the integration of frequency difference between each FDR and the system average and sends out islanding warnings to grid operators if one or more FDR detect abnormally large frequency differences, which indicates certain sub-systems (or generators) have become islanded from the others [26][27]. Fig. 12 shows an islanding event detected by this module.

Fig.13. is the estimated voltage angle response of an event from such a data-driven model. It shows a good match with the actual measurement. This process can be summarized as: firstly, identify the transfer function model between the “output FDR” (green one in Fig. 14) and the “input FDR” (yellow ones in Fig.14) using measurement data of certain contingencies, and then use the trained transfer function model and the measurement data from the “input FDR” to estimate the response of the “output FDR” of other contingencies. This estimation technique can be used for missing measurement data interpolation and wrong data identification.

Fig.14. Illustration of the measurement data-driven model development

V. NON-REAL-TIME APPLICATIONS

Fig. 12. Islanding event detected by FNET/GridEye

F. Measurement-based Model Construction Phasor measurement data provide first-hand knowledge of power system dynamic behaviors. On the other hand, circuit-

Besides the various real-time applications, FNET/GridEye also developed a series of non-real-time data analytics applications, some of which are briefly introduced in this section.



7

A. Event Replay and Post-event Analysis Blackouts cause disastrous losses in large areas. For instance, the 2008 Florida blackout led to the loss of 22 transmission lines, 4,300 MW of generation, and 3,650 MW of customer service or load in two-thirds of the Florida area. FNET/GridEye and PMU measurements of this blackout were

used to replay this event, one screen shot of which is shown in Fig. 15. It can be clearly noticed that this event originated from the Florida area, and then propagated to the entire Eastern Interconnection (EI). This event replay movie facilitates the blackout’s post-event analysis and helps avoid similar blackouts in the future.

(a)

(b)

(c)

(d)

(e)

(f)

Fig.15. Florida blackout replay enabled by FNET/GridEye frequency measurement

B. Measurement-aided Model Validation As one of the most important non-real-time applications, FNET/GridEye frequency measurements have been utilized to validate the dynamic model of U.S. power grids, such as the Eastern system. By comparing the real frequency response recorded by FNET/GridEye to the dynamic simulation results from the EI multi-regional modeling working group (MMWG) model, it is revealed that the current EI dynamic model is far from accurate enough to give a creditable frequency response. Our validation experience tells us that the parameters of

machine inertias, governor settings, loads etc. need to be carefully tuned to match the real system response recorded by FNET/GridEye [31]. In Fig. 16, incorporating the previously un-modeled governor dead-bands into the EI model, its frequency response simulation accuracy was significantly improved [32]. This example reveals the great potential of FNET/GridEye measurement in large-scale power system model validation.



8

over a year (as shown by Fig.18) [34] and the impact of social events like the FIFA World Cup or the NFL SuperBowl on the power grids can be analyzed [35].

Fig.16. FNET/GridEye frequency measurement model validation

Fig.18. EI system disturbance distribution over a year in 2012

E. Other Non-Real-time Applications FNET/GridEye also offers solutions to some other unique problems. For example, the unpredictable generation-load profile makes the power grid frequency at any time essentially stochastic or “unique.” When a digital device is used to record sound or video, it captures not only the wanted data but also the power grid frequency signal from the electrical grid, which is within a narrow range around 60 Hz in the U.S. Therefore, the continuous time-stamped frequency measurements from FNET/GridEye can be used as the benchmark to authenticate a suspected digital recording. 60.06

Frequency(Hz)

C. Electromechanical Speed Map Development Electromechanical waves propagate at different speeds in the power grids. This phenomenon has long been observed in the U.S. power grids and can be explained by different generator and load densities in different regions [33]. Accurate calculation of the electromechanical wave propagation speed is important to understand the power grid dynamic characteristics. FNET/GridEye provides a measurement-based solution to this problem: the propagation speed of an electromechanical wave traveling from one FDR to another can be calculated easily by dividing the distance between those two FDRs by the time difference of arrival. With as many as hundreds of almost evenly deployed FDRs and electromechanical waves detected in the EI system, average propagation speed between any two FDRs can be obtained, based on which EI system propagation speed contour map is available (as shown by Fig. 17) [20]. In this way, a trustworthy estimation of the propagation speeds over the EI territory can be obtained making use of FNET/GridEye frequency measurements.

60.04

Replacement of 10s data Replacement of 30s data Replacement of 50s data No tampering FDR measurement

60.02

60

59.98 0

20

40

60 Time(s)

80

100

120

Fig.19. Audio authentication using FNET/GridEye measurement

Fig.17. Electromechanical speed map calculated by FNET/GridEye

D. Historic Data Statistical Analysis Since the FNET/GridEye system went online in 2004, a large amount of data have been collected from FDRs located within the United States and around the world. Employing some data mining techniques, FNET/GridEye historic data can be extremely informative. For instance, statistical analysis is able to demonstrate how the disturbances in the EI system distribute

As shown by Fig. 19, the electrical grid frequency signal can be abstracted from audio recordings by short-time Fourier transform-based digital filtering method [36], which will be matched to the FNET/GridEye frequency measurement of the same time period. If any tampering has ever been applied to the recording, such as the replacement of 10, 30 and 50 seconds data in Fig.19, the mismatch will accurately pinpoint where it occurred. VI. CONCLUSION Wide implementation of the traditional PMU-based WAMS is limited by its high cost and deployment effort, not to mention the lack of mature applications and implementation experience etc. Deployed at the distribution level, FNET/GridEye was built as a pilot WAMS specifically applied to power system



9

dynamics monitoring. It is considered as a perfect complement of the traditional PMU-based WAMS and proves to be a costeffective situational awareness tool for electric utilities, independent system operators, and regulatory agencies. So far it serves more than tens of electric utilities in the U.S. thanks to its low cost, easy deployment and useful data analytics functionalities. The latest developments of FNET/GridEye were reviewed in this paper and its innovations and uniqueness can be summarized as follows: 1. FNET/GridEye sensors are single-phase plug-play synchrophasor devices installed at standard 120-V outlets. These low cost and easy installation features enable them to be deployed potentially in a large volume. Compared to traditional PMUs, it allows the coverage of an electric grid at a minimum cost. 2. Thanks to the application of chip-scale atomic clock, non-contact design, adaptive sampling technique and dynamic phasor estimation algorithm etc, the accuracy, reliability and flexibility of FNET/GridEyes sensors can be much higher than its existing counterparts. 3. While most of the existing phasor applications are still at the stage of conceptual design or lab test, all the FNET/GridEye data analytics applications presented in this paper are based on real measurement data and some are even running online to serve the industry. 4. As a complemental WAMS, the visualization, detection and analytics functions FNET/GridEye provides can be easily integrated into existing electric utility control centers to enhance situational awareness capabilities for power system operators. In the future, FNET/GridEye will continue to explore its applications in power system dynamic monitoring and pioneer the development of WAMS in power systems. Particularly, with more sensors deployed, the applications of FNET/GridEye in the distribution grid and microgrid operations will be investigated.

[6] [7]

[8]

[9]

[10]

[11]

[12]

[13]

[14]

[15]

[16]

[17]

[18]

[19]

[20]

VII. ACKNOWLEDGEMENT The authors would like to express gratitude to all the past FNET/GridEye group members for their pioneering work that contribute to the successful implementation of FNET/GridEye. Also, the authors would like to thank all the past and current FNET/GridEye sponsors and hosts.

[21]

[22]

[23]

VIII. REFERRENCES [1]

[2]

[3] [4]

[5]

P. Bonanomi, “Phase angle measurements with synchronized clocks principle and applications,” IEEE Trans. Power App. Syst., vol. PAS-100, no. 12, pp. 5036–5043, Dec. 1981. A.G. Phadke, J.S. Thorp, and M.G. Adamiak, "A new measurement technique for tracking voltage phasors, local system frequency, and rate of change of frequency," IEEE Trans. Power App. Syst., vol. PAS-102, no.5, pp.1025-1038, May 1983. A.G. Phadke, "Synchronized phasor measurements in power systems," IEEE Comput. Appl. Power, vol.6, no.2, pp.10-15, April 1993. Q. Bin, L. Chen, V. Centeno, X. Z. Dong, and Y. Liu, “Internet based frequency monitoring network (FNET),” in Proc. 2001 IEEE Power Eng. Soc. Winter Meet., pp. 1166–1171. Z. Zhong, C. Xu, B. J. Billian, L. Zhang, S.-J.S. Tsai, R. W. Conners, V. A. Centeno, A. G. Phadke, and Y. Liu, “Power system frequency

[24]

[25]

[26]

[27]

monitoring network (FNET) implementation,” IEEE Trans. Power Syst., vol. 20, no. 4, pp. 1914–1921, Nov. 2005. Y. Liu, “A US-wide power systems frequency monitoring network,” in Proc. 2006 IEEE Power Eng. Soc. Gen. Meet., p. 1-8. R. M. Gardener and Y. Liu, “FNET: A quickly deployable and economic system to monitor the electric grid,” in Proc. 2007 IEEE Conf. Technol. Homeland Security, pp. 209–214. Z. Lin, Y. Liu, Y. Jia, Y. Zhang, T. Xia, Y. Liu, and F. Wen, "Dynamic performance test of single-phase phasor measurement units," in Proc. 2011 IEEE Power Eng. Soc. Gen. Meet., pp. 1-7. Y. Zhang, Y. Liu, L. Chen, J. Guo, and Y. Liu, "Visualization of distribution level voltage magnitude pattern trend in EI system using FNET data," in Proc. 2014 IEEE T&D Conference and Exposition, pp. 1-5. A. Von Meier, D. Culler, A. McEachern, and R. Arghandeh, "Microsynchrophasors for distribution systems," in Proc. 2014 IEEE PES Innovative Smart Grid Technologies Conference (ISGT), pp.1-5, Feb. 2014. B. Pinte, M. Quinlan, A. Yoon, K. Reinhard, and P.W. Sauer, "A onephase, distribution-level phasor measurement unit for post-event analysis," in Proc. 2014 Power and Energy Conference at Illinois (PECI), pp.1-7, Feb. 2014. L. Wang, J. Burgett, J. Zuo, C. Xu, B. J. Billian, R. W. Conners, and Y. Liu, “Frequency disturbance recorder design and developments,” in Proc. 2007 IEEE Power Eng. Soc. Gen. Meet., pp. 1–7. T. Xia and Y. Liu, "Single-phase phase angle measurements in electric power systems," Power Systems, IEEE Trans. Power Syst., vol.25, no.2, pp.844-852, May 2010. L. Zhan, J. Zhao, S. Gao, J. Cullis, Y. Liu, and Y. Liu, “Universal grid analyzer design and development,” in Proc. 2015 IEEE Power Eng. Soc. Gen. Meet, pp. 1-5. L. Zhan, Y. Liu, J. Culliss, J. Zhao, and Y. Liu, "Dynamic single-phase synchronized phase and frequency estimation at the distribution level," IEEE Trans. Smart Grid, vol. 6, no. 4, pp.2013-2022, July 2015. L. Zhan and Y. Liu, "Improved WLS-TF algorithm for dynamic synchronized angle and frequency estimation," in Proc. 2014 IEEE Power Eng. Soc. Gen. Meet, pp. 1-5. G. Wijeweera, B. Bahreyni, C. Shafai, A. Rajapakse, and D.R. Swatek, "Micromachined electric-field sensor to measure AC and DC fields in power systems,", IEEE Trans. Power Del., vol.24, no.3, pp.988-995, July 2009. Y.C. Zhang, W. Yao, J. Culliss, G. Zhang, Z. Teng, and Y. Liu, "Electrical field based wireless devices for contactless power gird phasor measurement," in Proc. 2014 IEEE PES T&D Meet., pp.1-5. Y. Zhang, P. Markham, T. Xia, L. Chen, Y. Ye, Z. Wu, Z. Yuan, L. Wang, J. Bank, J. Burgett, R.W. Conners, and Y. Liu, "Wide-Area frequency monitoring network (FNET) architecture and applications," IEEE Trans. Smart Grid, vol.1, no.2, pp.159-167, Sept. 2010. P. N. Markham, “Data Mining and Machine Learning Applications of Wide-Area Measurement Data in Electric Power Systems,” Ph.D. dissertation, Dept. of EECS, Univ. of Tenn., Knoxville, TN, 2012. R. M. Gardner, J. N. Bank, J. K. Wang, A. J. Arana, and Y. Liu, “Non-parametric power system event location using wide-area measurements,” in Proc. 2006 IEEE Power Syst. Conf. Expo., pp. 1668–1675. R. M. Gardner, J. K. Wang, and Y. Liu, “Power system event location analysis using wide-area measurements,” in Proc. 2006 Power Eng. Soc. Gen. Meet., p. 7. Z. Yuan, T. Xia, Y. Zhang, L. Chen, P.N. Markham, R.M. Gardner, and Y. Liu, "Inter-area oscillation analysis using wide area voltage angle measurements from FNET," in Proc. 2008 IEEE Power Eng. Soc. Gen.Meet., pp.1-7, 2008. N. E. Huang, Z. Shen, S.R. Long, M.C. Wu, H. H. Shih, Q. Zheng, N. Yen, C.C. Tung, and H.H. Liu, “The empirical mode decomposition and the Hilbert spectrum for nonlinear and non-stationary time series analysis,” Proceedings of Royal Society, vol. 454, no. 1971, pp.903-995, March 1988. D. Trudnowski, J. Pierre, N. Zhou, J. Hauer, and M. Parashar, "Performance of three mode-meter block-processing algorithms for automated dynamic stability assessment," IEEE Trans. Power Syst., vol.23, no.2, pp.680-690, May 2008. Z. Lin, T. Xia, Y. Ye, Y. Zhang, L. Chen, Y. Liu, K. Tomsovic, T. Bilke, and F. Wen, "Application of wide area measurement systems to islanding detection of bulk power systems," IEEE Trans. Power Syst., vol.28, no.2, pp.2006-2015, May 2013. J. Guo, Y. Zhang, M. A. Young, M. J. Till, A. Dimitrovski, Y. Liu, P. Williging, and Y. Liu, "Design and implementation of a real-time off-grid



10

[28]

[29]

[30]

[31]

[32]

[33]

[34] [35]

[36]

operation detection tool from a wide-area measurements perspective," IEEE Trans. Smart Grid, vol. 6, no. 4, pp.2080-2087, July 2015. Y. Liu, K. Sun, and Y. Liu, “A Measurement-based power system model for dynamic response estimation and instability warning”, Electric Power Systems Research, vol. 124, pp. 1-9, July 2015. C. Li, Y. Liu, K. Sun, Y. Liu, and N. Bhatt, "Measurement based power system dynamics prediction with multivariate autoregressive model," in Proc. 2014 IEEE T&D Conference and Exposition, pp.1-5. F. Bai, Y. Liu, Y. Liu, K. Sun, N. Bhatt, A.D. Rosso, E. Farantatos, and X. Wang, "Measurement-based correlation approach for power system dynamic response estimation," IET Generation, Transmission & Distribution, vol. 9, no.12, pp.1474-1484, Aug. 2015. L. Chen, “Wide-area measurement application and power system dynamics,” Ph.D. dissertation, Dept. of EECS, Univ. of Tenn., Knoxville, TN, 2011. G. Kou, S. Hadley, and Y. Liu, "Dynamic Model Validation with Governor Deadband on the Eastern Interconnection," Oak Ridge National Laboratory (ORNL), Oak Ridge, TN, USA, Feb. 2014. J. S. Thorp, C. E. Seyler, and A. G. Phadke, “Electromechanical wave propagation in large electric power systems,” IEEE Trans. Circuits Syst. I: Fundam. Theory Appl., vol. 45, no. 6, pp. 614–622, Jun. 1998. Y. Ye, “Wide-area Situational Awareness Application Developments,” Ph.D. dissertation, Dept. of EECS, Univ. of Tenn., Knoxville, TN, 2011. Y. Lei, Y. Zhang, J. Guo, D. Zhou, and Y. Liu, “The impact of synchronized human activities on power system frequency”, in Proc. 2014 IEEE Power Eng. Soc. Gen. Meet., pp.1-5. Y. Liu, Z. Yuan, P. N. Markham, R. W. Conners, and Y. Liu, “Application of power system frequency for digital audio authentication,” IEEE Trans. Power Del., vol. 27, no. 4, pp. 1820-1828, 2012.

Knoxville in 2014. He is currently working towards his Ph.D. degree in power systems at the University of Tennessee, Knoxville. Wenxuan Yao received his B.S. degree in electrical engineering from Hunan University, Changsha, China, in 2011. He is pursuing the Ph.D. degree in the Department of Electrical Engineering and Computer Science, University of Tennessee, Knoxville. Yin Lei (S’13) received the B.S. and M.S. degrees in electrical engineering from Xi’an University of Technology in 2009 and University of Southern California in 2011, respectively. Currently, she is pursuing her Ph.D. degree in the Department of Electrical Engineering and Computer Science at the University of Tennessee, Knoxville. Jidong Chai received the B.S. degree from Tianjin University in 2011, and the M.S. degree in electrical engineering from the University of Tennessee, Knoxville, in 2014. He is currently working towards his Ph.D. degree in power systems at the University of Tennessee, Knoxville. Yilu Liu (S’88–M’89–SM’99–F’04) received her M.S. and Ph.D. degrees from the Ohio State University, Columbus, in 1986 and 1989. She received the B.S. degree from Xian Jiaotong University, China. Dr. Liu is currently the Governor’s Chair at the University of Tennessee, Knoxville and Oak Ridge National Laboratory (ORNL). She is also the deputy Director of the DOE/NSF-cofunded engineering research center CURENT. Prior to joining UTK/ORNL, she was a Professor at Virginia Tech. She led the effort to create the North American power grid Frequency Monitoring Network (FNET) at Virginia Tech, which is now operated at UTK and ORNL as GridEye. Her current research interests include power system wide-area monitoring and control, large interconnection-level dynamic simulations, electromagnetic transient analysis, and power transformer modeling and diagnosis.

IX. BIOGRAPHY Yong Liu (S’11) received his Ph. D. degree in electrical engineering (power system direction) from the University of Tennessee, Knoxville, in 2013. He received his M.S. and B. S. degree in electrical engineering from Shandong University, China, in 2007 and 2010, respectively. He is currently a research assistant professor in the DOE/NSF-cofunded engineering research center CURENT and Department of Electrical Engineering and Computer Science at the University of Tennessee, Knoxville. His research interests are wide-area power system measurement, power system dynamic analysis and renewable energy integration. Lingwei Zhan (S’13) received the B.S. and M.S. degrees in electrical engineering from Tongji University in 2008 and 2011, respectively. Currently, he is pursuing his Ph.D. degree in the Department of Electrical Engineering and Computer Science at the University of Tennessee, Knoxville. His research interests include PMU, synchrophasor measurement algorithm, wide-area power system monitoring, renewable energy sources, FACTs, and HVDC. Ye Zhang (S’13) received the B.S. degrees in software engineering from Shanghai Jiaotong University, China, in 2006, M.S degree in computer engineering from Polytechnic University of Milan, Italy, in 2008, and Ph.D. degree from the University of Tennessee, Knoxville, in 2014. She is currently with Alstom Grid. Her research interest lies in the visualization study of power system dynamics. Penn N. Markham (S’03–M’12) received the B.S. degree in electrical engineering from Virginia Polytechnic Institute and State University, Blacksburg, in 2006 and the Ph.D. in electrical engineering and computer from the University of Tennessee, Knoxville, in 2012. He is currently with Electric Power Research Institute. Dao Zhou (S’13) received the B.S. degree in electrical engineering from North China Electric Power University in 2007, and the M.S. degree in electrical engineering from Virginia Polytechnic Institute and State University in 2010. He is currently pursuing the Ph.D. degree in electrical engineering in University of Tennessee, Knoxville. Jiahui Guo (S’13) received his B.S degree in Electrical Engineering from Tsinghua University, Beijing, China, in 2011, and the M.S. degree in electrical engineering from the University of Tennessee, Knoxville in 2014. He is currently working towards his Ph.D. degree in power systems at the University of Tennessee, Knoxville. Gefei Kou (S’14) received the B.S. degree from Tianjin University in 2011, and the M.S. degree in electrical engineering from the University of Tennessee,
