Abstract-As we transition towards a power grid that is increasingly based on energy from renewable resources like solar and wind, the intelligent control of distributed energy resources (DER) including photovoltaic (PV) arrays, controllable loads, energy storage and plug-in electric vehicles (EVs) will be critical to realizing a power grid that can handle both the variability and unpredictability of renewable energy sources as well as increasing system complexity. In addition to providing added system reliability, DERs acting in coordination can be leveraged to address supply and demand imbalances through demand response (DR) and/or price signals on the electric power grid by enabling continuous bidirectional load balancing. Intelligent control and integration has the capability to reduce or shift demand peaks and improve grid efficiency by displacing the amount of backup generation needed and offsetting the need for spinning reserves and peaking power plants.Realizing such a decentralized and dynamic infrastructure will require the ability to solve large scale problems in real-time with hundreds of thousands of DERs simultaneously online. Because of the intractable scale of the optimization problem with variables and constraints for every DER, load and generator online at each time period, we use an iterative decentralized method to operate each DER independently and autonomously within this environment. This method was developed in [1] using a distributed algorithm referred to as the Alternating Direction Method of Multipliers (ADMM). Specifically, we consider a commercial site equipped with with on-site PV generation, partially curtailable load, EV charge stations and a stationary battery electric storage (BES) unit for backup. The site operates as a small microgrid that can participate in the wholesale market on the power grid or operate off-grid in an islanded state. The ADMM algorithm is deployed within a Model Predictive Control (MPC) framework to allow the microgrid to distribute the optimization among the individual DERs and dynamically adapt to changes in the operating environment while responding to external real-time wholesale prices and potential contingency situations. At each time step, embedded controllers model their own DERs as optimization problems with local objectives subject to individual constraints and forecasts. They then use the ADMM algorithm to solve the problem and obtain a control schedule across the MPC horizon. The local objectives are augmented with a regularization term that includes a simple exchanged message between neighbors in the microgrid. This is the only communication required between DERs. Through the exchange of these messages, the decentralized method rapidly converges to an optimal solution for the entire microgrid when each DER is able to locally solve its own problem efficiently in parallel. Once solved, the controllers execute the first step of the schedule and await the next time step at which point they re-solve the problem using any new information that arrives...

No abstract

No abstract

Ground Fault Overvoltage can occur in situations in which a four-wire distribution circuit is energized by an ungrounded voltage source during a single phase to ground fault. The phenomenon is well-documented with ungrounded synchronous machines, but there is considerable discussion about whether inverters cause this phenomenon, and consequently whether inverters require effective grounding. This paper examines the overvoltages that can be supported by inverters during single phase to ground faults via theory, simulation and experiment. It identifies the relevant physical mechanisms, quantifies expected levels of overvoltage, and makes recommendations for optimal mitigation. It concludes that under many circumstances, effective grounding of inverters is not necessary to prevent ground fault overvoltage.