PMU-WAMS research and application in Brazil

1 Research background

Studies on PMU applications in Brazil was initiated in the early 1990s by the Study Committee of the Interconnected Operation Coordination Group(GCOI)prior to the deregulation of the Brazil energy sector.Difficulties faced by the Brazilian economy during this period and the restructuring process of the electric energy sector delayed the project until 1999.However,some issues still need to be addressed to minimize the risks associated with such large scale project and to avoid potential future problems.

To address the risks and challenges of the BIPS,as well as to improve power system security and stability,the system operator and utilities have called for strengthening of power system dynamic security monitoring and real-time dynamic monitoring and management of the power system.

This project aims to find a suitable WAMS for the Brazil power grid and implement an actual pilot project where PMU data can be measured from some substations in the Brazil power grid.It also aims to facilitate further research on WAMS technology and methods to implement the technology in the control center of an actual power grid using the WAMS pilot project.The project is expected to deliver practical technical methods for power grid widearea dynamic monitoring,on-line power system accident detection and analysis and to help the operation personnel of power networks carry out fast response and prompt handling of abnormal situations in the power grid [2,3].

2 Overall project design

2.1 Main research contents

(1)The main research topics are:Research on key techniques and method for online real-time monitoring,including:

·wide-area dynamic monitoring method;

·on-line low frequency monitoring method;

·on-line disturbance detection method;

·event-driven case management method.

(2)Building WAMS master station for a pilot project in the above research tasks:

·To realize online monitoring and analysis of grid dynamic process;

·To provide high-quality snapshot for any following application.

2.2 Project objectives

The goals of this project are:

(1)To conduct research on advanced technology based on PMU data and to form application achievements.

(2)To conduct research and develop a stable and practical WAMS design and implementation scheme.

(3)To implement online real-time dynamic monitoring and analysis based on technical research in BIPS;

(4)To develop a solid database and technique for online integrated monitoring,analysis,protection and control in the future.

2.3 System structure

The system consists of the following:

(1)PMU in substations

Six sets of PMUs in substations,which are mainly responsible for acquisition and transfer of power grid dynamic data to a central control and management center.

(2)WAMS main station

The WAMS main station is mainly responsible for PMU data acquisition,monitoring,and analysis [4].

3 PMUs site selection(General introduction)

The location of the PMUs to be installed in the system was proposed taking the following into consideration:

(1)Observability of the highest number of electromechanical oscillation modes present in the National Interconnected System(SIN),namely(i)North-northeast vs.southeast;(ii)North vs.northeast;(iii)Mode of plants in the 440 kV system;(iv)South vs.southeast;(v)Mode of plants in the Acre and Rondonia system; and(vi)Mode of plants in the Mato Grosso system.

(2)Coverage in detecting disturbances in SIN(contingencies in the transmission system,loss of synchronism between areas of SIN,islanded plants or groups of plants,sub or over frequencies in the islanded systems,etc.),focusing on the main regions of the Brazilian system and the corresponding interconnections.

(3)Ease of implementation,operation,and maintenance of PMUs,focusing on bays of lines or transformers and giving priority to substations of the concessionaires whenever possible.

The project foresees the installation of six PMUs.Aimed at bringing more flexibility to the continuity of the project and considering that the substations have quite different conditions regarding communication infrastructure and aspects of O & M,some additional alternatives were established in terms of number of substations that is nearly the same in each region or SIN area,as well as the indication of a seventh region or SIN area as an alternative.Fig.1 shows the locations of PMUs and WAMS.

Fig.1 Locations of PMUs and WAMS

4 PMU

4.1 General

PMUs are indispensable in key substations and power stations for the implementation of WAMS technology and to enhance the dynamic monitoring and analytical capability of the power system to the dispatching center.PMU measures the phasor values of current and voltage at different locations in a synchronized power network.These phasor values receive a high precision time stamp and together with power flow data,frequency,and so forth,which are also time stamped,are transmitted to a central analysis station.The central analysis station performs dynamic monitoring and analysis of the power system.WAMS will be the main data source for real time dynamic data platform of the center,and will be combined with the EMS and the automatic security control system step by step to enhance the dynamic and robust monitoring of the power system.

Supported by GPS and DSP technology,PMU can measure synchronous phasor of voltage and current from all the nodes of the entire power system,and can provide data sources and approaches for the implementation of power system monitoring,substation automatic monitoring and control,stability control,and self-adaptive relay protection of the power system.

Data gathered by PMU sub-station acquisition cabinets is transferred to the sub-station controller of the main cabinet,and this controller then transmits the data to the main station of WAMS.If there is no unified timing system or the timing system is not sufficiently accurate in the field,special synchronous clock equipment can be employed to meet the requirement of the synchronous phasor measurement unit for high-precision time setting.Fig.2 shows the general structure diagram of a PMU sub-station system using a synchronous phasor measurement unit.

Fig.2 General structure diagram of PMU substation system

4.2 Functional description

The PMU adopts a decoupling algorithm based on clock synchronization and frequency tracking,which ensures simultaneous and highly accurate data sampling.The device can carry out synchro phasor measurement based on a standardized transmission protocol,IEEEC37.118.1-2011 [5,6,7].

4.3 Hardware

Fig.3 shows the hardware arrangement.

Fig.3 Overall hardware structure diagram

5 Power grid wide-area dynamic monitoring

5.1 Introduction

Power grid wide-area dynamic monitoring is the key monitoring technology of WAMS; it provides on-line and real-time monitoring capability for power grid basic operating status based on PMU data and information.

Power grid wide-area dynamic monitoring mainly provides various intuitive,accurate,and convenient visualized means for schedulers to carry out monitoring of the grid’s dynamic behaviors.Meanwhile,the technology can process the grid’s running data in a simple way,provide limit violation alarms,and monitor grid running stability.[8]

Power grid wide-area dynamic monitoring application monitors the grid’s dynamic behaviors with respect to the generator,line,bus voltage,and frequency based on PMU’s real-time data.Power grid wide-area dynamic monitoring includes the following relevant technologies:

(1)Bus voltage replacement;

(2)Bus line voltage calculation;

(3)Grid-wide frequency selection;

(4)Grid-wide reference point of phase angle selection;

(5)Shift transformer data phase;

(6)Determining the reasonability of PMU’s real-time data;

(7)Displaying PMU’s real-time dynamic data frame by frame;

(8)PMU measurement/time curve;

(9)Relative phase angle monitoring for limit violation.

5.2 General description

Fig.4 shows the position of the power grid wide-area dynamic monitoring application in WAMS.

Fig.4 Position of the power grid wide-area dynamic monitoring application in WAMS

The data acquisition subsystem receives real-time dynamic report data from the PMU,and makes subsequent use of the data after processing the messages.Then,the real-time data is transmitted to the real-time data cache pool and the real-time data service provides real-time data to the PMU through the real-time data buffer pool for the computation and analysis of power grid operation state monitoring application.The power grid dynamic monitoring application is based on complete real-time data service analysis and calculation of the running state of the power grid.

The power grid wide-area dynamic monitoring application relies on PMU’s dynamic data to monitor gridwide dynamic behaviors; therefore,the data acquisition application should ensure normal operation and its dynamic data should be properly stored so that the wide-area dynamic monitoring application can obtain dynamic data.In addition,the output of the wide-area dynamic monitoring application serves as data sources for other applications.For example,power data(that cannot be directly acquired from PMU)required by the low frequency oscillation monitoring application should be provided by the wide-area dynamic monitoring application after calculation.

5.3 System function architecture

After the FES of the master station receives dynamic data from the PMU substation and sends it to the real-time data buffer,the power grid wide-area dynamic monitoring application can directly obtain measurements from the buffer.The measurements will be re-stored to the buffer or real-time database after calculation and processing by the wide-area dynamic monitoring application.The inputs of all modules of the power grid wide-area dynamic monitoring application are dynamic data of the buffer.Fig.5 shows the application modules and data flowchart of power grid widearea dynamic monitoring.

Fig.5 Application modules and data flowchart of power grid wide-area dynamic monitoring

6 Low-frequency oscillation online monitoring

6.1 General description

Low-frequency oscillation online monitoring application utilizes real-time data collected by the PMU to identify lowfrequency oscillations that occur in the grid.When lowfrequency oscillation occurs,alarm signal is sent out in real time,and the online analysis module is initiated to analyze the characteristics of the low-frequency oscillation,identify the dominant oscillation mode,and track mode changes such that dispatchers have sufficient time to take measures to prevent further deterioration of the oscillation,avoid serious accidents,and improve stability of the power grid.

After low-frequency oscillation has occurred,case-save application is triggered automatically to record the entire real-time dynamic data and analysis results of the PMU during oscillation process to rich study case and accumulate experience for the power grid operation and management.Retrieval and analysis of low-frequency oscillation events can reproduce historical scenes for offline analysis.

The specific functions and features are as follows:

(1)Monitoring the network security status of lowfrequency oscillations in real time.Once low-frequency oscillation is identified,it launches the monitoring screen of low-frequency oscillation automatically,refreshes oscillation curves and tracks changes in oscillation mode throughout the oscillation process real-time.

(2)Recording the entire oscillation process automatically when low-frequency oscillation occurs.

(3)Using data measured by the PMU and saved in the historical library to perform data scanning of a specified period,invert historical low-frequency oscillation process,and replicate historical scenes.Analyzing the characteristics of low-frequency oscillation,identifying the dominant oscillation mode,tracking mode changes offline.

6.2 Technical features

Low-frequency oscillation online monitoring application utilizes real-time data collected by PMU to reflect the actual dynamic process of the grid,achieve real-time early warning and online analysis of low-frequency oscillation event of the grid.The entire system has the following technical characteristics:

(1)Monitoring network security status of low-frequency oscillations in real-time.Once low-frequency oscillation is identified,it launches the monitoring screen of lowfrequency oscillations automatically,refreshes oscillation curves and tracks changes in oscillation mode throughout the oscillation process real-time.

(2)Recording the entire oscillation process automatically when low-frequency oscillation occurs.

(3)Using data measured by the PMU and saved in the historical library to perform data scanning of a specified period,invert historical low-frequency oscillation process,and replicate historical scenes.Analyzing the characteristics of low-frequency oscillation,identifying the dominant oscillation mode,tracking mode changes offline.

(4)The entire system performs the function of crossplatform,can run on Unix platform as well as on Windows platform,and its hardware can be easily configured.

6.3 Analysis algorithm

This subsection highlights the calculation process and information flow.

6.3.1 Calculation process

(1)Preprocessing of measured data of monitoring equipment,forming analysis,and calculation of input data.

(2)Prony analysis of data after data preprocessing to obtain the oscillation mode characteristic quantity group,identify the characteristic quantities of these groups’ oscillation mode amplitude frequency response characteristics of the absolute amount of the groups’ dominant mode.This mode characteristic quantities of the oscillation curve group are the dominant oscillation mode characteristic quantities groups.

(3)Determine whether the dominant oscillation mode of curves obtained are in the frequency range of low-frequency oscillation,and list dominant oscillation modes that meet the frequency limit of the dominant oscillation mode according to the amount of energy,identifying the most serious oscillation mode.

(4)Determine low-frequency oscillation security state monitoring generator according to the pre-alarm damping ratio threshold value and contact lines.

(5)Determine the current low-frequency oscillation security state of the grid according to the security state of monitoring generators,contact lines,and transformers.

6.3.2 Information flow

Fig.6 shows the low-frequency oscillation monitoring information flow.

7 Power grid disturbance identification

7.1 Introduction

With the interregional connection of power systems,the scale of power grid will be further enlarged,and the structure will become increasingly complicated.Although a large power grid has outstanding capabilities,it also has prospective safety problems.For example,local power grid disturbances,particularly short circuits,may spread to a wide area,and if handled improperly,the problem may escalate into a systematic event and even cause system corruption and then a blackout.

WAMS is a new generation power grid monitoring system that is formed based on the PMU.It uses computer technologies and modern high-speed digital communication networks to implement synchronous sampling of gridwide data,real-time recording and long-distance real-time transmission,serving as a platform for online power grid disturbance identification.

By processing and filtering PMU’s real-time dynamic data,disturbance identification applications can obtain information indicative of the disturbance characteristics and sort disturbances such as short circuits and load shedding based on information in order to send an alarm message to the scheduler upon a disturbance.

Fig.6 Low-frequency oscillation monitoring information flow

7.2 System architecture

7.2.1 Module architecture

In terms of applications,power grid disturbance identification is mainly completed by data extraction modules,data processing modules,and data analysis modules.Fig.7 shows the specific structure.

The data extraction modules perform three functions,namely to extract real-time dynamic data,historical data and dynamic files from the PMU.The data processing modules include the FES system communication module,database communication module,alarm service module,and numerical filtering module.The database communication module stores disturbance information,the alarm service module reports an alarm in the event of a power grid disturbance,and the numerical filtering module filters PMU’s real-time data using average median filtering.The data analysis modules include the short-circuit identification module and load shedding module.The short circuit identification module is responsible for monitoring and analyzing short circuits,whereas the load shedding identification module is responsible for monitoring and analyzing load shedding.

Fig.7 Power grid disturbance identification function diagram

7.2.2 Data flow

Power grid disturbance identification has the following inputs:PMU’s real-time data,historical data,and dynamic files.The outputs include alarms,results,flow curves,and case saving start messages,where results and flow curves are sent to the graphics system,which displays the results and curves using an MMI interface for users to understand the flow of power grid disturbance.Fig.8 shows the power grid disturbance identification flowchart.

Fig.8 Power grid disturbance identification flowchart

8 WAMS event-driven case management

8.1 Function introduction

After detecting and identifying power grid accidents in WAMS using analysis functions(such as identification of power grid disturbance and online monitoring of lowfrequency oscillation),in most cases,it is always necessary to store data related to the accident process so that the operators can deduce reversely in the wake of the accident.Such data are often important and precious case data,which are necessary for analysis and research on relevant problems and characteristics of a grid system.Therefore,such case data will often be permanently stored rather than cyclically covered as in conventional historical data.The event-driven case data management function is designed to satisfy this demand.

The event-driven case data management performs data extraction,data query,background data service and other functions.It is a storage warehouse and centralized management platform for event data in WAMS.

The data extraction function can contribute to automatic analysis and integration of data at different time scale and space scale,forming a multi-state logging data about events,so that the original event logging can be available for postevent analysis and other applications.

Data query is designed in the foreground centralized management interface,so that users can arrange the order or inquire about event data based on time as well as plant and station name in the interface.Data such as event summary,brief data,and PMU dynamic data can be displayed according to the classification.

Background data service is exclusively used to provide services for case data storage and system management,including background access,query,and management.The data service is configured with a client/server model,and user applications can be accessible via the network for ease of configuration and management [8].

8.2 System structure

Upon the completion of power grid event analysis function in WAMS(online monitoring functions such as disturbance identification and low-frequency oscillation),the case data storage function will be triggered via the system message to complete permanent storage of power grid event case data at the designated measuring point and designated historical period,and form a complete case storage in combination with the main application analysis results.

The storage functions of case data are mainly composed of:

(1)Automatic Storage Function of Application-triggered Case Data.The case data storage service receives case data storage message sent by WAMS application function,and can be used for the triggered storage of historical case data and integration of analysis information according to the designated measuring points mentioned in the message.

(2)Storage Function of Artificially-triggered Case Data.The case data storage function is triggered and activated based on the artificially-designated measuring point and historical period(interface interaction)to achieve permanent storage of artificially-designated case data.

(3)Display Function of Storage Case.This function provides a special interface for case query,query and list case data of the designated period,and displays description result information and the case data curve corresponding to the case.

9 On-line monitoring of PFR performance of the unit

9.1 General description

When the grid frequency(cycle)deviates from the rated value in primary frequency regulation(PFR),the regulation control system of the generator in the grid will automatically control the increase(frequency drop)or decrease(frequency increase)of the active power to limit the variation characteristics of the grid frequency.It is an important link for the regulation and control of both the active power and frequency of a power system.It reflects the emergency capacity of a grid in case of load emergency,and plays an important role in the safe and stable operation of the power system.

In recent years,the gradual maturity and development of synchronized phase measuring technique based on global positioning system(GPS)technology work out the problem of synchronous data acquisition in different spatial positions,realize the synchronous acquisition of wide-area measured information of unit,and provide a new opportunity for online monitoring and analysis of unit PFR performance.

Currently,on-line monitoring and off-line test of PFR characteristics are adopted for study on PFR characteristics based on PMU data.On-line monitoring of PFR is mostly based on the unit output and the characteristic curve of frequency variation to assess the unit PFR performance,which is necessary for engineering applicability.However,the response performance has large dynamic characteristics because the unit PFR is affected by the operating conditions of the unit and other factors.Therefore,the PFR performance assessment index obtained cannot completely reflect the real effect of the unit PFR.The relatively single monitoring function assessment index cannot completely and thoroughly reflect the PFR capability of the unit after a sudden disturbance in the grid or accident and cannot correctly tap the frequency regulation potential of the unit to offer more assistance for the operation and management of the grid.In the current situation,it is urgent to conduct an indepth study on monitoring,analysis and assessment of PFR performance of the unit to provide more comprehensive,reasonable,and scientific guidance for the supervision and management of the unit PFR performance of a power plant.It is also urgent to fully tap the PFR potential to improve the production and operation safety of the Brazil Power Grid.

9.2 Technical route

On the basis of the unified platform of technical support system for smart grid dispatching,the comprehensive analysis technique of PFR parameters in different time periods of frequency fluctuation under different typical operating conditions of the unit is investigated using widearea measured information such as unit frequency,output,PFR enabling/disabling signals and PFR actuating signal to form a PFR lean assessment index in different time periods of frequency fluctuation under different typical operating conditions.

Establish a comprehensive analysis and lean assessment system for PFR based on wide-area measured information,realize comprehensive analysis and lean assessment of PFR rapidity,efficiency and deviation of unit groups and provide more comprehensive,reasonable and scientific guidance to the PFR analysis and assessment [9,10].

Fig.9 shows the functional structure of comprehensive analysis and lean assessment of PFR performance of the unit.

Fig.9 Functional structure diagram of comprehensive analysis and lean assessment of PFR performance of the unit

10 Conclusions

This project selects the lines and substations that can obviously monitor system disturbance information through system analysis.Six sets of PMUs transmit data to WAMS in the dispatch center.The PMUs are used for storage,analysis,and display of real-time data of the power grid,as well as to complete wide-area dynamic monitoring and analysis of the power grid and for recording and inversion of synchronous disturbance data.

This project involved the study of key technologies and methods of dynamic monitoring and analysis of grid operation based on smart grid dispatch technology support system,in which the real-time dynamic behaviors of the network was monitored by wide-area synchronous dynamic information uploaded by PMU.The research results include:wide-area network dynamic monitoring method,low-frequency oscillation online monitoring method,network disturbance identification online detection method,and event-based case management method.It is beneficial for system users to understand the dynamic operating characteristics of Brazil grid so that the users’ ability to control the overall operating state of the grid will be improved effectively.These are necessary to ensure the safe operation of the power grid.

Through the key technology research of this project and the construction of WAMS pilot system,power system monitoring,analysis and control functions are expanded from static to dynamic.The dynamic monitoring of the power grid realizes monitoring of real-time dynamic processes of the power grid.It has diversified,accurate,and convenient visual means to monitor the dynamic behaviors of the power grid,and can monitor key parameters for safe operation of the power grid such as voltage,frequency,power angle,and power flow of the contact line,which enhances real-time monitoring and analysis ability of the dynamic process of the power grid.

Grid disturbance identification ensures the extraction of dynamic characteristics,identification of grid disturbance,and actual alarming.

Low-frequency oscillation can realize monitoring,warning,and analysis of low-frequency oscillation of the system,which improves the ability to deal with the abnormal state of the power grid quickly and reduce the risk of large-scale power outage.Thus,social and economic losses can be avoided.

Case management software can realize automatic analysis and integration of data of different time scales and different forms,and realize data query,browsing,curve display,contrastive analysis and management functions,which can be used for post-analysis and playback to help operators understand the dynamic characteristics of the grid.

Furthermore,this project strengthens the support of auxiliary service of the power market and provides a new monitoring method for the dynamic operation of the power grid for operators,which helps operators improve the accuracy of power network analysis and effectively improve the ability to control the overall operation state of the power grid,thus improving the transmission capacity and the reliability of power grid operation.

It can improve wide-area dynamic monitoring level of a large power grid in Brazil and reduce the probability of large-scale power failure,which will have a profound impact on technical standards,norms and product access areas of the Brazil power grid in the future.