Author(s):
Tomas Johannesson -
E.ON Elnät Sverige AB
Magnus Akke -
Dlaboratory Sweden AB
Andreas Vikström -
Dlaboratory Sweden AB
Jakob Hägg -
Dlaboratory Sweden AB
Smart Grid Requirements
We had many requirements for our smart grid analyzer concept (see Figure 1), including flexible hardware and software to implement high-quality disturbance recorders, a communication infrastructure to separate measurement from control, and post processing software with a web interface.
Measurement Equipment
We created a disturbance recorder using off-the-shelf NI hardware and software development tools. Figure 2 shows the measurement hardware, which includes an industrial-grade router and an interface box.
Each unit must have the following:
• 32 channels
• Two neutral-point voltages
• Six phase voltages (two groups of three phases)
• Six phase currents (two groups of three phases)
• 18 zero-sequence currents (maximum of 24 if the phase currents are not used)
• 50 kHz of simultaneous sampling
• 32 GB of USB memory for local backup
For sensitive fault triggering without connection to the earth, the system calculates a negative sequence current on a built-in field-programmable gate array (FPGA) circuit. The system has internal use of 24 bits per channel, but to comply with Comtrade, the output disturbance file uses only 16 bits. There is a hardware reset function by binary output from an industrial 3G router running a virtual private network (VPN).
The system measures from the secondary side of the current transformer. We connected a closed-loop Hall effect current transducer at the location where the secondary current enters the relay protection (see Figure 3).
The installation is safe and easy for field technicians in substations. Technicians can further simplify current sensor installation with split-core current transducers at lower accuracy.
The disturbance recorder saves data in Comtrade format (standard for protective devices). The Comtrade configuration file contains the most important parameters for automatic fault analysis. For the concept to work, we must follow particular settings, such as pre-trigger length.
The disturbance recorders have two types of trigger conditions: direct and derived. They also have a selectable pre-trigger buffer. After a trigger, the system saves the file on the measuring device.
Communication
To benefit from the disturbance data, the technician needs to access it within a few minutes. In Sweden, the 3G network is well developed and reliable. This, together with an industrial-grade router for 3G communication, gives us a stable and relatively high-performance Internet connection with a roughly 2 Mbit/s to 4 Mbit/s download and 1 Mbit/s to 2 Mbit/s upload capacity. Some substations might offer a fiber-based Internet connection.
For security, we use a router with VPN. After a disturbance, recorded data automatically transfers to a server with FTP under the protection of the VPN tunnel.
The device uploads the file via a secure VPN channel to the communication server and transfers it to an application server. This server automatically analyzes the disturbance, creates a report, and sends the report to a supervising network utility.
The application server also saves all disturbances to a database for further data refinement. We can use the database to build applications run by ordinary web browsers for added power utility value (for example, asset management and maintenance planning). With web technology, we do not need extra software at the power utilities.
Automatic Fault Analysis
Technicians can further refine the database of analyzed disturbances with applications integrated in the power utility internal work processes. Specifically, we can base network investment on actual disturbance recording statistics in a larger region during a longer time period. Figures 4 and 5 show two typical faults in distribution networks.
Fault analysis can quickly point out the faulted feeder. Typical data includes type of fault, faulted feeder, estimated fault resistance, extreme current values, and voltages. Figure 6 shows an example of a disturbance and the resulting report.
Conclusion
Using common web development, communication, and measurement equipment standards, we based our system on off-the-shelf NI products for a better value. In 2011 and 2012, we implemented and deployed the proof-of-concept project shown in Figure 1 at E.ON Elnät in Sweden. The project included the installation of several dedicated fault recorders with communication; automatic fault analysis of each disturbance; an application server with a database of analyzed disturbances; and the creation of applications to support grid supervision by rapid automatic analysis, disturbance investigation, maintenance planning, and asset management.
The disturbance recorders have been able to help tracking down the root cause of several interruptions.
Examples:
1) Svalöv August 2010, four cross country faults during one month (double earth fault on different phases and feeders). Root cause was weed in cabinet.
2) Vellinge 10 August 2011, intermittent earth fault. Directional earth fault protection failed, and did not provide any useful information. Root cause was cable fault.
3) Vellinge 6 September 2011, evolving fault sequence with earth faults and short circuits that involved several feeders. The system gave early warnings on 3 and 5 September with short fault transients.
Contact Information:
Magnus Akke
Dlaboratory Sweden AB
August Strindbergs väg 4
Lund
Sweden
Magnus.Akke@dlaboratory.com
Author Information:
Tomas Johannesson
E.ON Elnät Sverige AB
Print
PDF
