(463c) Stochastic Scheduling for Microgrid Power Systems with Constrained External Power Exchange

Authors: 
Zachar, M., University of Minnesota
Daoutidis, P., University of Minnesota

Microgrid power systems are of interest as a potential avenue for increasing the penetration of distributed renewables, increasing resiliency, decreasing the carbon intensity of energy supply, and facilitating the liberalization of the power market [1].  Existing literature on microgrids has covered topics such as the optimal sizing and technology selection [2-3], the effect of stochasticity on microgrid design [4], real-time control to maintain internal power quality [5], and scheduling to ensure robust power supply and minimal operating expenses [6-8].

An aspect that has received little attention in microgrid scheduling is the nature of the power exchange with the surrounding macrogrid.  Due to the stochastic nature of renewable power and load, power exchange at this interface may be erratic which could lead to macrogrid instability and place undue stress on utility-scale plants [1].  In this work, we propose a novel formulation of the scheduling problem in which we ensure that power exchange with the macrogrid is well behaved.  Specifically, commitments for power exchange with the macrogrid are made in a day-ahead fashion and have a limited ramp rate so that the microgrid operates as a “good citizen” of the grid.  Thus, from an external perspective, the microgrid has limited stochasticity (realized power exchange must lie close to the commitments made) and mild ramping (which is conducive to tracking by large power plants).

A case study is carried out for a prototype microgrid consisting of photovoltaics, microturbines, a battery bank, and a bi-directional connection to the macrogrid.  Measured residential load and rooftop PV production data over a 1-year period is used.  The unit commitment problem is formulated as a chance-constrained mixed integer linear program over a 48 hour receding horizon.  The chance constraints ensure that the first-stage variables (discrete unit states and day-ahead power exchange commitment) are able to adequately satisfy the power balance subject to the inherent stochasticity.  Using forecasts for power demand and renewable availability, the chance-constrained power balance is transformed into a set of linear inequalities.  The schedule resulting from this optimization is compared to two alternative formulations: the deterministic unit commitment (where only the nominal power balance is considered) and the naïve unit commitment (where only the nominal power balance is considered and grid exchange is not scheduled ahead of time).

The performance is analyzed with respect to the operational cost, amount of PV power curtailed, the probability of maintaining the power balance without violating power exchange commitments, and accuracy of predicted second-stage variables (e.g. unit setpoints and storage levels).  Computation time is also considered since this unit commitment is intended to be one layer of an on-line hierarchical control system.  Finally, additional benefits of this method are discussed such as the intelligent determination of reserve capacity via the chance constrained power balance (as opposed to determining reserve margins a priori as in traditional power systems scheduling).

[1] Colson, C. M., and M. H. Nehrir. "A review of challenges to real-time power management of microgrids." Power & Energy Society General Meeting, 2009. PES'09. IEEE. IEEE, 2009.

[2] Liu, P., M. C. Georgiadis, and E. N. Pistikopoulos. "An energy systems engineering approach for the design and operation of microgrids in residential applications." Chemical Engineering Research and Design 91.10 (2013): 2054-2069.

[3] Koutroulis, E., et al. "Methodology for optimal sizing of stand-alone photovoltaic/wind-generator systems using genetic algorithms." Solar energy 80.9 (2006): 1072-1088.

[4] Zhou, Z., et al. "A two-stage stochastic programming model for the optimal design of distributed energy systems." Applied Energy 103 (2013): 135-144.

[5] Piagi, P., and R. H. Lasseter. "Autonomous control of microgrids." IEEE Power Engineering Society General Meeting, 2006.

[6] Khodaei, A. "Resiliency-oriented microgrid optimal scheduling." IEEE Transactions on Smart Grid 5.4 (2014): 1584-1591.

[7] Zhang, Y., N. Gatsis, and G. B. Giannakis. "Robust energy management for microgrids with high-penetration renewables." IEEE Transactions on Sustainable Energy 4.4 (2013): 944-953.

[8] Trifkovic, M., et al. "Dynamic real‐time optimization and control of a hybrid energy system." AIChE Journal 60.7 (2014): 2546-2556.