Copel GT experiences

Authors: G. F. Krefta, COPEL GT, and C. E. F. Pimentel, GE Grid Solutions, Brazil

The measurement of electrical phasors, analogs signal and digital signals that inform the behavior of the power system can be acquired at the same base of time applying the Phasor Measurement Unit (PMU) based on the protocol IEEE C37-118-2011. It means that one disturbance in the electrical power system interconnected can be analyzed from a thousand kilometers far, at same base of time.
Synchrophasors that are components of PMU can be defined as phasors calculated in separate facilities, geographically distant, using the same time reference. There is a need for a single time reference, i.e. Synchronization. Figure 1 shows the concept of Synchrophasors.

In addition, current SCADA systems are not able to show to system operators dynamic behavior of the system, so utilities in general operate their system using large security margins to avoid system blackouts and use data from relays or Digital Disturbance Recorders (DDR) to analyze a system disturbance. However, this approach is not always cost effective, quick or easy to perform, as there is the need to use different systems and equipment, which are not always time synchronized or have same accuracy. Thus, Synchronized Phasor Measurement Systems (SPMS), also known as Wide Area Measurement System (WAMS), were developed and have been increasing their usage in power utilities.
WAMS systems had been recognized as the main alternative to improve real-time monitoring of power systems. These systems, composed by Phasor Measurement Units (PMU), communication system and Phasor Data Concentrators (PDC), have the advantage to have their measurements time-synchronized by a common time base, which makes it possible to compare directly measurement from all system installations to visualized real-time power system conditions in terms of angles between busbars.
Brazilian utility Copel G&T has been facing these challenges in the last years and, in 2015, commissioned its first WAMS where the benefits of this new system and applications are being applied to the company.

Copel's GT Wide Area Measurement System

The system was initially composed by 37 PMU, function that is embedded in their DFR system, distributed in all Paraná State territory, located at Brazilian south region. By now the system is composed by 59 PMU and cover the north, northeast and southeast regions of Brazil while new power plants and substations were built by Copel GT. Two central PDC servers, installed in Copel's control center, receive data from the PMU installed at the substations, and PMU send two streams of data to each PDC. As there is no substation PDC from PMU and central PDC, there are distances over 3,300 km from the substations where the PMU are installed and the control center, where the PDC receives PMU data. Figures 2 and 3 show system architecture.

After the first year of operation, the company stated to increase their PMU database, including also generator measurements in low side and electromechanical signals, such as generator current and voltage, governor position, voltage regulator values, power system stabilizer signal (PSS) and even digital signal to check if operation is performed by Brazilian National Operator (ONS) or locally by Copel. Initial experiences lead the company to adopt these signals, as it was noted that system observability would increase significantly with these signals.

Thus, hydroelectric power plants (totalizing 4,436 MW) were included in the system. Finally, as Copel's wind generation is increasing in the last years, it also included more than 800 MW wind farm, which is in Brazilian northeast region, more than 3,000 km far from Copel's control center. Figure 5 shows an example of the main screen of the WAMS system in operation currently in Copel.
Experience acquired by Copel's personnel in WAMS usage shown that, even though there is a big advantage of using angle differences to operate the system, which nowadays only WAMS systems can perform, there are several additional benefits that these systems offer to utilities, even though if they don't have autonomy to operate their system, as it is Copel GT case in BIPS.
Having a single database, time-synchronized and with the resolution that a PMU can perform, increases significantly the capacity of operators to diagnose and increase operational margins, using their assets better, even though it is not used mainly angle differences. Synchrophasors measurement treated on the PDC operating in the Copel GT is presented as a multifunctional tool included in the concept of smart grid, way beyond the proposed angle between busbars and which came to complement the SCADA system and do not replace it.

The picture in Figure 4 shows resolution difference between SCADA and WAMS. The first does a scan of 4 to 3 seconds while the second system makes a reading every cycle. The result is that when you overlap those two systems measurements, the SCADA system showns a single data point of active power while the WAMS describe an entirely electromechanically oscillation of the system.

Figure 6 shows a SCADA system screen where PMU frequency data were imported from PDC using DNP3 protocol, showing an example of how to integrate both systems.

WAMS results perceived on BIPS on the Argentinian blackout of June 16th

With tools in operation on Copel's system, it was possible to understand what happened in Brazilian Interconnected Power System (BIPS) by the occasion of the blackout of Argentina Power System on June 16, 2019 at 07:06 a.m.
The connection between Brazilian and Argentinian power systems is feasible by the 60/50 Hz Converter Station (HVDC) of Garabi, as shown below. These power systems are not directly connected, as it is on the Uruguayan and Argentinian interconnection, because Brazil operates at the frequency of 60 Hz, while Argentinian system operates in 50 Hz. (Figure 7).
In such a case, where two power systems are interconnected by HVDC, it was expected that controls of the HVDC should provide a compensation in the load flow when it rises or decreases even abruptly as in the rejection of loads as well as in the load taking in order to avoid a power swing.

As it was seen on Copel's WAMS, HVDC of Garabi substation didn´t operated as expected at that day during the blackout of Argentina, as it was perceived an oscillation between the entirely BIPS in order of 0.537Hz, as shown in Figure 8. For this occasion, as shown at the report, Brazil was sending 1,000MW to Argentina. The load rejection produced a power swing at Brazilian system as we can see at the graphic in Figure 8.
Behavior of the generators owned by Copel during the oscillation of the Brazilian system can be evaluated and compared directly by Copel's GT WAMS.

Salto Caxias hydroelectric power plant is in Brazilian south region, at Paraná state, where Copel operates and maintains the electrical system that is interconnected with the rest of BIPS by the 500 kV transmission system. This power plant has four hydraulic generators of 345 MVA operating in 16 kV. As this is the closer hydroelectric power plant owned by Copel to the Garabi station, it was used the measurements of this generators to evaluate the response of the BIPS in this blackout. (Figure 9).

WAMS Results Perceived on BIPS in a Wind Farm with an SVC nearby Substation
A typical example of systemic control used nowadays is the Static Compensator of Reactive Power (Static VAR Compensator - SVC).
An SVC system inserts capacitors in the system to raise the voltage and inserts inductors to decrease the voltage, thus controlling the voltage around a set point with steps pre-defined. The control is done automatically using thyristors or manually.

This alternative is attractive to avoid building more overhead transmission lines, which are becoming increasingly difficult to install due to public appeal and restrictions on environmental releases.
Despite the great benefit of these technologies, they have the drawback of introducing the risk of sub-synchronous oscillation in its various forms. These oscillations must be monitored and mitigated for a healthy operation of the electrical system. While studies, planning, design, filtering, and system protection reduce the risk of an occurrence with damage by the resonance effect, monitoring can also be an element of strategic mitigation of those problem.
Copel GT has four wind farms at the northeast of Brazil with the capacity to produce 800 MW. Those wind farms are part of a total of 16 GW of BIPS Wind Farms installed in the northeast.

Another example of PMU application is to monitor those wind farms of Copel GT.
Figure 10 shows the oscillation of voltage of Cutia – 230kV Wind Farm substation whose controller of wind generators seems to have interacted with the controller of a SVC installed at the interconnecting substation of Ceará Mirim II – 230kV. Because of this oscillation some aerogenerators are disconnected incorrectly by overvoltage protection.

WAMS results as a Tool for Checking and Reviewing Wind Farm Simulation Models

Simulation based on dynamic behavior models is a fundamental tool for planning and proper operation of an energy system. It is used for stable and failover contingency analysis to determine if the system is operating within safety margins and supply quality standards.
The inadequacies of the model can thus have real and significant consequences for the power system. Excessively conservative limits may lead to costly and inefficient solutions, while the erroneous results of the stability assessment can lead to separation or blackout. Validation and improvement of system models are vital and likely to become more challenging and resource-intensive while power systems become more complex as we move towards a low carbon future.

WAMS data based on Synchrophasors Measurement are ideally suited for model validation, being a continuous time-aligned record of disturbed and steady state power systems.
The example shown in Figure 11 can be used as a study case of information provided by WAMS to check power systems models. It is related to the contribution to a fault current provided by Cutia wind farm near to its connection to Ceará Mirim II substation. WAMS shown the injection of active and reactive power by the wind farm during the occurrence of a fault in the system.

Technical requirements for wind power plants described in the BIPS National System Operator (ONS) procedures were developed to preserve the active power, and its reduction is only acceptable in specific voltage and frequency values, when special protection systems are started. However, WAMS records shown that the Cutia Wind Farm absorbed active power while injected reactive power at BIPS during this system occurrence.
From the records shown in Figure 12, where it can be concluded that a capacitive load was perceived during the fault, one can conclude that expected polarization of protection relay in the direction of the fault would be compromised and relay could stay inoperative depending on how it was commissioned.

WAMS in a Process Bus (9-2LE) Environment
As a final example of WAMS application, it is demonstrated in Figure 13 integration between conventional and Process Bus PMU, where measurement of voltages and currents are acquired from optical CT and Merging Units.
This is a pilot project, where PMU using IEC 61850-2-9LE can be compared with the conventional PMU measuring current and potential inductive transformers at same time.  



Gilmar Francisco Krefta born in Curitiba, PR- Brazil. Graduated in Electrical Engineering by UTFPR in 1985. Specialization in Digital Technology in 1998 by UTFPR. He holds a master's degree in Energy Systems from UFPR in 2008. He was a System Protection Engineer at Copel G & T from 1988 to 2016. He currently works at the Operations Center as an Electrical Studies Engineer at Copel G & T.

Carlos Eduardo Ferreira Pimentel born in Guarapuava, PR-Brazil. Graduated in Electrical Engineering by UTFPR in 2011, has more than 7 years of experience in the electrical sector. Major areas of experience are Power Systems Protection and Control, WAMS, TWFL and Digital Substations. Currently serves as Lead Services Application Engineer at GE Grid Solutions.