Recent Sandia Secure, Scalable Microgrid Advanced Controls Research Accomplishments

Image of Mesa-del-Sol-microgrids_web-1024x852.jpg

A recently completed Grand Challenge Laboratory-Directed Research and Development project resulted in a secure, scalable microgrid (SSM) research facility that is actively engaged in developing technologies to solve many of the nation’s most complex challenges in satisfying its electric energy needs. Initial focus has been on enabling resilient and reliable performance when incorporating high penetration levels of stochastic renewable-energy (RE) sources in power systems. Theories and concepts from multiple fields were integrated to advance solutions to this challenge, including

  • Hamiltonian Surface Shaping and Power Flow Control™-based nonlinear distributed control theory,
  • informatics/agent-based algorithms,
  • communication theory, and
  • power-electronic system theory.

With the advancement of microgrid power systems that include high levels of RE penetration, comes the need to determine energy-storage requirements. Energy storage is an important design component in microgrids that include high RE resource penetration levels—to maintain system stability, because of the stochastic nature of the sources. Storage devices can be distributed close to the source and/or at the microgrid bus (centralized).

Some solutions to these challenges have recently been published in an online journal, Electrical Power and Energy Systems, “Energy storage requirements of DC microgrids with high penetration renewables under droop control,” W. Weaver, R. Robinett III, G. Parker, D. Wilson. This paper highlights a novel design and analysis approach of energy-storage devices in a DC microgrid system with a high penetration level of RE sources that have high variability. The focus of the study is the effects and trade-offs in placing energy storage close to the sources, or at a centralized bus storage device.

A common default design is to include one large energy-storage device on the system’s main bus to help regulate voltage and to provide power when RE resources are unavailable. By distributing the energy storage amongst the sources and the bus, the design engineer can optimize peak power requirements and bus voltage variations.

In addition, a load-sharing control scheme is required when two or more energy resources contribute power onto a common bus or grid. This control scheme can be centralized if an interconnected communication system is present. However, this can create a single point of failure. Droop control employs a completely decentralized control structure that enables high flexibility and resilience. Droop control is a common technique for distributed control of electrical sources in a microgrid where the control implements a virtual impedance such that the load current is distributed between the sources proportional to the droop settings.

Due to the novel nature of the energy storage unified power-flow control (UPFC) architectures, a patent was also recently awarded to Wilson and Robinett, “Computing an operating parameter of a unified power flow controller,” US Patent #8930034, Jan. 6, 2015.

Image of SSM-GC-LDRD-model-1024x241.jpg
(a) A one-machine infinite bus (OMIB) model with a unified power-flow control architecture and wind-turbine generation. (b) OMIB power requirements. (c) OMIB energy-storage requirements.

A simple wind integration with the electric power grid demonstrates this capability and performance (Wilson and Robinett, IEEE MSC 2011). In the figure, (a) shows a one machine infinite bus (OMIB) model with a UPFC and wind turbine generator. The power and energy storage requirements are given in figures (b) and (c), where the wind variability is compensated with the UPFC device to provide a constant power output. These technical accomplishments are part of a base set of technologies presently being used to solve national and international power system problems with highly stochastic sources and loads.temporary placeholder

  In a power-distribution system (grid), electricity demand (load) often varies. Historically, a system operator handled this uncertainty with dispatachable power from a central generating station (e.g., the operator throttles a fossil-fueled generator up/down to meet a load increase/decrease). Because the electricity production of some RE resources, such as wind or solar, varies based on the availability of the resource, which does not necessarily coincide with consumer demand, new methods of meeting load are required to reliably incorporate large amounts of these ‘stochastic’ (nondeterministic, random) RE resources into a power grid.