Objective: This project will provide support to the power systems industry in developing test methods, tools, and performance characterization and metrics for time synchronization to securely achieve the precision time stamps necessary for applications such as system analysis and diagnosis, system state estimation and distributed multivariate control, and enabling operators to respond to fast dynamic changes (to effectively maintain load and resource balance using distributed energy resources) in the power grid by2017.
What is the new technical idea? Research in the area of improved control and estimation for power systems has recently seen a significant increase in interest. This has been motivated by government and industry requirements for grid, efficiency improvements and increased integration of non-conventional, renewable sources of energy. There is also significant industry interest in improving network stability and load regulation within distribution circuits. All four of these improvements are particulary relevant in the face of decentralized power generation within microgrids.
The current power grid, however, is not well equipped to handle non-traditional electrical network architecture where generation is not centralized and may be bidirectional. Further, distribution circuits may operate as islanded microgrids. While there are several challenges in the decentralized generation and regulation of power, a fundamental concern is the real possibility of uncoordinated interaction between microgrids resulting in large reactive power flows - consequently causing instability in the network. In order to achieve coordination and guarantee reactive support, it is essential to first ensure synchronization of clocks. Precisely synchronized clocks are needed to:
- ensure fidelity of error and event logs used for diagnosis;
- accurately estimate the state of the system based on its time stamped measurements; and
- develop control algorithms that can apply corrective action in a timely manner.
The new technical idea is that scalable implementation of control and estimation algorithms for large power networks depend on their amenability to decentralization and parallelization, thus requiring precise synchronization between the various interacting components.
Timing accuracy is vitally important to fuse data from heterogeneous sources (diagnostics), to make accurate estimates of state (situational awareness) and to ensure safety of decentralized control schemes (situational intelligence). Practical timing considerations are rarely considered during design but pose a significant challenge at the time of implementation.
What is the research plan? The research plan is to develop a set of software tools, along with metrics and guidance on timing requirements and its implications, which will significantly aid industry to integrate future algorithmic improvements in control, estimation and diagnosis.
The plan also includes benchmarking the analysis against highly dynamic aspects of the power network e.g., mitigation of voltage collapse (using VAR compensation) and suppression of switching transients(using high bandwidth demand-response), where decentralized control methods have been currently deemed infeasible.
Specifically the project will focus on applications in the low voltage (<11KV) distribution network or a microgrid, where closed-loop time constants are less than one second, requiring clocks accurate to a few milliseconds.
The plan can be organized into the following steps to address some of the timing challenges listed in the NIST report on "Technology, Measurement, and Standards Challenges for the Smart Grid."
Step-1. Continues the support of the IEEE C37.238 Conformance Testing and standards development effort. Provide a venue for plug-fest(s) in the conformance testing effort with IEEE Conformity Assessment Program (ICAP) and University of New Hampshire's Inter-Operability Laboratory (UNH-IOL)as needed.
Step-2. Augment the testbed with practical use case scenarios based on a survey of priorities with other power systems researchers (IEEE Power Systems Relay Committee (PSRC) Working Group, Smart Grid Interoperability Panel (SGIP), Oak Ridge National Laboratory (ORNL)). As a starting point, use cases will be matched with scenarios listed in the steering committee report, such as, distributed VAR compensation, coordinated peak load mitigation (demand response) and regulation of distributed microsources.
Step-3. Based on the chosen use cases, prototype a semantic modeling representation (e.g. leveraging IEC 61850) of the distribution circuit. The semantic representation would enable dynamic discovery of new components, synthesis of generic model types and inclusion of context specific performance parameters (including timing performance) to aid model-based control, estimation and diagnosis functions.
Step-4. Integrate state estimation and control algorithms into the testbed via embedded simulations and real hardware (where possible). Timing performance data can be obtained from the hardware already installed in the testbed.
Step-5. Develop metrics for assessing the impact of time synchronization on the power systems use case developed in Step-2.
Step-6. Develop a closed-loop implementation (estimation+control+diagnosis) in conjunction with collaborating partners (ORNL) and catalog benchmark measurements for the metrics in Step-5.
Step-7. Explore methods to dynamically adapt models developed in Step-3 during operation. Since the model types are semantically coded and classified, it is possible to update them intelligently e.g., update the model of a dynamic slack generator to accommodate phase errors introduced by timing uncertainty.
Step-8. Implement our control algorithms on at least one industrially certified active VAR compensator such as a self-commutated voltage source converter.
Step-9. Support the Smart Grid Cybersecurity Team in security testing and detection of vulnerabilities introduced by the timing infrastructure.