Multi‐Layered Simulation using the SGRS Simulator; Interaction of TE and Flywheel Controlled Dynamic System Kevin Bachovchin, Martin Wagner, Marija Ilic kbachovc@andrew.cmu.edu, mwagner1@andrew.cmu.edu, milic@andrew.cmu.edu 10th CMU Electricity Conference Pre-conference Workshop March 30, 2015 Outline Link multi‐time scale simulations Adaptive load management (ALM): slow time‐scale Fast dynamics with flywheel control: fast time‐scale Market needs to specify reactive power set point Reactive power set point is critical to whether there is an equilibrium Depending on the reactive power set points, sometimes dynamics cannot be stabilized Demonstrate stable and unstable responses on Smart Grid in a Room Simulator 2 Interactive Communication Communication for multi time‐scale simulation with ALM and fast dynamics for generators Source: M. R. Wagner, K. D. Bachovchin, M. D. Ilić, "Computer Architecture and Multi Time-Scale Implementations for Smart Grid in a Room Simulator," EESG Working Paper No. R-WP-1-2014, March 2015. 3 Importance of Reactive Power Typically the market only specifies the active power set point However the reactive power is critically important to the equilibria and stability of the system Power Factor PF = 0.99 (Without Shunt Capacitor) Power Factor PF=0.2 (Without Shunt Capacitor) 3 3 2 2 [0.9615, -0.18 rad] 1 2 (rad) 1 2 (rad) Manifold of Active Power Balancing Eqn at Bus 2 Manifold of Active Power Balancing Eqn at Bus 2 Manifold of Reactive Power Balancing Eqn at Bus 2 0 0 -1 -1 [0.182, -1.27 rad] -2 -2 -3 -3 0 0.5 1 1.5 2 2.5 3 Voltage V2 (pu) 3.5 4 4.5 5 0 0.5 1 1.5 2 2.5 3 3.5 4 4.5 Voltage V2 (pu) Source: X. Miao, K. D. Bachovchin, M. D. Ilić, "Effect of Load Type and Unmodeled Dynamics in Load on the Equilibria and Stability of Electric Power System," EESG Working Paper No. R-WP-1-2014, March 2015. 4 5 Flores Island – Market Control Based on prices, market computes active power set points P* from each component 5 Flores Island – Dynamics Control Since currently the market does not specify reactive power set points Q*, data for Q* is randomly created Place a voltage source inverter and the variable speed drive on the hydro and diesel generator buses Control the sum of the power out of the hydro and diesel generators to match the active and reactive power set points 6 Flores Island – Variable Speed Drive Control Determine the variable speed drive set points based on the set points from the market Controllable inputs for the variable speed drive are u k g k xk , yck 1 , yk ref T u k u1d u1q u2 d u2 q T x k i1d i1q qC1 iS 2 d iS 2 q iR 2 T y ck 1 vd vq y k ref 2 ref i1d ref i1q ref T Set points for the variable speed drive are y k ref h k yck 2 , rk ref T y ck 2 iGd iGq T r ref PG * QG * PG * Vd i1d ref iGd Vq i1q ref iGq QG * Vq i1d ref iGd Vd i1q ref iGq Solve for i1d ref and i1q ref 7 Simulation Results – Hydro/Diesel Generator Bus Total Power out of Hydro Generator Bus Total Power out of Diesel Generator Bus 0.2 P P* power(pu) power(pu) 0.2 0.1 0 0 20 40 60 80 0.1 P P* 0 100 0 20 40 0.02 0.01 Q Q* power(pu) power(pu) 80 0 20 40 60 t(sec) 80 100 Q Q* 0 -0.02 0 100 t(sec) t(sec) -0.01 60 0 20 40 60 80 100 t(sec) 8 Simulation Results – Wind Generator Bus Stable Case: Unstable Case: Reactive Power Load Consumption 10 5 0 0 20 40 60 80 100 power(pu) power(pu) Reactive Power Load Consumption 10 5 0 0 20 40 80 100 t(min) t(min) Wind Generator Bus Voltages Wind Generator Bus Voltages 1.2 1.2 1 1 0.8 vB2d vB2q 0.6 0.4 voltage(V) voltage(pu) 60 0.8 vB2d 0.6 vB2q 0.4 0.2 0 0.2 -0.2 0 -0.2 -0.4 0 20 40 60 t(min) 80 100 -0.6 0 20 40 60 t(min) 80 9 100 Conclusions Demonstrated multi time‐scale simulation with feedforward market controller and feedback flywheel controller Showed that with a high reactive power load, the system may not reach a stable equilibrium Future Work Market level controller specify the reactive power set point Design fast dynamic control, so that we don’t need voltage source inverter at each bus Combine with other time‐scale simulations (AGC) 10