RTDS Simulation of PV Inverters in a Small Power System
|Sponsor||Dr. Brian Johnson
|Team Name||Simulation Sages|
|Duration||2020 - 2021|
- 1 Executive Summary
- 2 Background
- 3 Problem Definition
- 4 Deliverables
- 5 Specifications
- 6 Design Considerations
- 7 Project Learning
- 8 Final Design
- 9 Validation
- 10 Future Work
- 11 Team Members
- 12 Additional Documentation
The purpose of this project was to build a simulation model of a small distribution power system on the RTDS (Real Time Digital Simulator) with photovoltaic (PV) inverters integrated into different busses on the system. Along with the power system model, the project includes hardware in the loop to control the PV inverters’ power output. This system provides a platform for power engineers and engineering students to study the effects of inverter-based energy sources on small power systems. In the end, our team was able to create a working model of a 13-bus distribution system, with three separate PV inverters connected to different busses on the system. The PV inverters can be switched into system using automation logic from an SEL 487B Bus Protection Relay.
Appropriate power system models are a necessity for effective engineering in the power field. With the rise of inverter based renewable resources such as wind and solar, the need for accurate working models is greater than ever, both for working professionals in the field as well as the future power engineers in the classroom. Inverter-based resources are built upon principles of power electronics and therefore, present problems to the existing power grid that haven’t been present in the past when the grid was dominated by traditional Inertia based generation (hydro, steam turbine, gas turbine, etc.). Since the growth of these sources is unprecedented, it is vital for power engineers to get ahead of the looming problems that will need to be faced in the near future. Accurate models are vital to gain the needed understanding of how these energy sources will affect the existing grid and give engineers an idea as to how to mitigate the adverse effects, and how to add to or alter the existing infrastructure to accommodate these sources.
The RTDS (Real Time Digital Simulator) is a robust and flexible simulation system that can closely model power systems. With a variety of different I/O and communication options available, it can also facilitate the use of other hardware in the simulations. This is where the term Hardware in the Loop comes from. This project addresses the need for a modeling platform for a small distribution system with integrated traditional and inverter-based energy sources, and outside protection and automation hardware to control certain aspects of the inverter-based energy sources.
The goal of this project was to develop a working model of a small distribution system that had the following parameters:
- Uses the RTDS to run the model.
- Integrates three PV inverters on separate busses in the model.
- Uses some means of communication to integrate external hardware into the loop.
- Utilizes automation and/or manual controls from the external hardware to control power output from the PV inverters into the model.
- Utilizes a Human Machine Interface (HMI) to run, control, and analyze the model.
- Accompanying documentation in the form of a lab exercise for future engineering students to use for PV integrated power system analysis.
The scope of the project did not include building an all-inclusive model that could be utilized to analyze any scenario with PV inverters integrated into a small distribution system, but to create a platform from which the specific needs of an engineer could be easily added to the model. It is intended to be a baseline model, as it would be impossible to create a model that could accommodate infinite scenarios. However, it can also stand alone as a simple way to analyze a small system with integrated PV inverters.
Throughout the year we have been required to develop documentation regarding different aspects of the design process and overall project development. Below is a list of files that represent some of the deliverables we have turned in for the project.
File:SimSages Team Contract.pdf
File:Simulation Sages - Team Value Proposition.pdf
File:Simulation Sages Product Requirements.pdf
File:Simulation Sages Design Validation Plan2.pdf
File:SimSages Team Meetings and Notes.pdf
Our team presentations can be found in the Additional Documentation section.
The model we created is based on the IEEE 13 bus system using a Real Time Digital Simulator (RTDS) and implemented with RSCAD software. It incorporates three separate PV inverters with controls for each. We also created a control scheme using an SEL 487B bus protection relay to control real and reactive power output from the PV inverters based on specific grid parameters. A human-machine interface (HMI) was developed for real-time manipulation of the systems controls.
Answering the question "What will the final design look like":
- No real safety standards are necessary due to the project being mostly software based.
- A one line diagram is provided which illustrates the three PV arrays integrated into the IEEE 13 bus system. The diagram of the 13 bus system can be found in the Additional Documentation section under Reference Materials.
- The HMI that was used for this project is a customized RSCAD Runtime HMI. It includes switches to control the breakers that bring the PV inverters onto the grid. It also includes bus voltage plots to monitor the health of the system as these PV inverters are switched in and out of the system
One of the major considerations for our project design was which model to base it on. We had a few different choices. We could have used a model built from scratch that had any number of busses, loads, sources, etc. There were also two pre-made model options, one was the IEEE 13 bus system and the other was the IEEE 34 bus system. Ultimately we decided that since all of the team members were new to RSCAD and working with the RTDS we probably shouldn't try to build a model from scratch as it would be too time consuming. We also decided that a 34 bus system would be too complex considering the goals of the project. The IEEE 13 bus system was the model we chose to base our project on.
We also had to decide how many inverters we wanted to integrate into the system. Dr. Johnson, our client, recommended two or three. We decided to include three PV inverters. This would make the results of the simulation results variable enough to give a good idea as to how the grid behaves based on different power outputs from the inverters. It was also recommended to connect the inverters at the weakest busses. This meant finding which busses had the lowest power factor, or ratio of real power to total apparent power. We could find this by running the IEEE 13 bus system on the RTDS and finding the power factor of each bus.
For weeks the team was roadblocked while working with the PV array model from RTDS. There were some protection elements that weren't allowing the model to function as it should. We decided to go back to a base model instead of pulling the array model from a file that had it integrated into a different system. This made it easier to deconstruct the model and remove the parts that were unnecessary and causing issues.
A major consideration that we had to decide was the communications scheme through which the RTDS could communicate with the SEL-487B and vice versa. Initially it was suggested to use Generic Object Oriented Substation Event (GOOSE) messaging to communicate both ways between the RTDS and the relay. It turned out that GOOSE was quite complex to set up from the RTDS side. After much trial and error, it was found that the RTDS didn't have the necessary firmware installed to allow the communications scheme to work properly. Through some consultation we also tried using DNP communication, but this was also somewhat complex and time didn't permit us to learn a whole new communications protocol and finish the project on time. Ultimately we decided to use analog outputs from the GTAO card on the RTDS to send low level bus voltage signals to the relay. For communication to the RTDS from the relay we used the simple output contacts of the relay connected to the digital inputs on the RTDS. This setup allowed for a simple, easy to duplicate and control communications scheme that fit well into the final product.
Our main need for learning at the beginning of this project was to learn how to use RSCAD to build models to run on the RTDS. There is a fairly steep learning curve for this software. After finding a PV inverter model that would be suitable for integration in the IEEE 13 bus system, we found it difficult to find how to make the system register the effects of the inverter.
With help from our mentor from SEL, Asad Mohammad, we found that there were issues with the protection logic of the PV array model. We haven't found the root cause for this issue, but we have found that by bypassing the protection value "blockx8" we can get acceptable results from the PV inverter.
Once the PV inverters were working properly we had to integrate them into the IEEE 13 Bus model. We decided to use a phased approach to this goal. First we made sure that we connect a single PV array to a single bus. Then we added two arrays to a bus. Since we will be using three arrays on the 13 bus system we will then add a third array to the bus. Once all three of these arrays were working properly we integrated the three arrays onto three separate buses on the IEEE 13 bus system.
The next major step in our project will be to establish a connection to the SEL RTAC and create an HMI for manipulation of the inverter controls. We initially had issues connecting to the available RTAC but with some help from Asad we were able to establish a connection after changing some IP addresses. We then were able to connect the RTAC to the RTDS. Although we had a connection between the RTAC and RTDS we decided to bypass the RTAC altogether and use the SEL-487B for control of the inverters. This made the system less complex and easier for a student to learn and manipulate. It also made it so we could bypass the issues we were having with GOOSE messaging since we could use low level analog inputs to the relay and digital output contacts to the RTDS.
The final design can be separated into four main parts:
- The main distribution system model with the integrated PV inverters
- The hardware in the loop with associated communication scheme
- The human-machine interface where the simulation can be run and controlled
- The functionality of the system
Distribution System Model
The main distribution system model was built around the IEEE 13-bus system. This is a generic system, well known in industry. It was a good fit for our project scope, as it was large enough to keep the integrity of the simulation results (it isn't a tiny "toy" system), but it was also small enough to only utilize a single RTDS core to run the simulation. The PV inverters added to the three busses were taken from a microgrid model. These inverter models were quite complex and had many extra features that were beyond the scope of this project, including protection logic. Some of these features had to be stripped away from the model in order for the inverters to be successfully integrated into the 13-bus system.
Hardware in the Loop
The hardware used for this project was the SEL-487B bus protection relay. The original plan was to use an SEL RTAC, but complications in the communications between devices drove us to change our plans and use a protection relay instead. This allowed us to use the analog output feature of the RTDS to send low-level signals to the relay. These signals are then processed by the relay and through the use of undervoltage elements and pushbuttons, automatic and manual closing of the inverter breakers is accomplished. The close breaker output signal is sent through the 487B's output contacts and sent to the RTDS digital input board. From there it is processed by the GTFPI board and the close command is sent to the appropriate breaker in the model.
Human Machine Interface
Originally we planned to use a feature of the SEL RTAC to build a controller HMI for this system. However, as explained above, the breakdown of communications with the RTAC required us to reconsider our options. Built in to RSCAD is a robust and easily customizable HMI called Runtime. Once models in RSCAD are compiled, Runtime is the portion of the program that actually runs the simulation. In Runtime you can add plots, meters, switches, buttons, annunciator lights, and so on. We built an HMI that monitors bus voltages for each bus on which a PV inverter is located. It also has switches that control the breakers for each inverter. We also included a highly capacitive load that could be switched on to the system to demonstration how the automation features of the 487B work.
When the model is compiled and run on the Runtime HMI there are a few options for where to start. The PV inverters should start disconnected (open breaker) from the system. The inverters can be manually switched into the system one at a time through the HMI switches labeled accordingly or through the use of three pushbuttons on the front panel of the 487B. The bus voltages begin to sag as the PV inverters are switched into the system. This demonstrates some of the instability that these inverters can cause to the system.
To demonstrate the automation logic that was programmed into the 487B all PV breakers should be open. When the "Big C" load is switched on, the 487B's undervoltage elements will assert based on the large bus voltage sag. This will close the three output contacts assigned to the three PV breakers. Through the digital inputs and GTFPI processor card on the RTDS, a "Close Breaker" command is sent to the corresponding PV breakers.
The Design Validation Plan can be found in the image to the right (also in the Deliverables section). It shows the major project requirements and how they were tested to make sure they met specifications. Some of the requirements changed, as was the case with the RTAC HMI and control, though the same testing could be used to validate the different hardware we ended up using. The lab procedure documentation part of the validation is left for future work.
Since the software learning curve was larger than originally anticipated, we were unable to complete a writeup of a lab procedure. Future work in this project will include that writeup, as the project was originally intended to be used as a teaching tool for power engineering students. Additionally, future work can focus on the expansion and tuning of the model we developed. We designed this project not to be a large, all-encompassing model, but rather a base to be built on and dropped into other simulations as the need arises. To this end, future work on the simulation could focus on fine tuning of the inverter models and branching into different communication protocols as well as making this system into a learning experience for students.
Major: Electrical Engineering
Major: Electrical Engineering
Major: Electrical Engineering