This document discusses challenges in simulating distributed energy resources and microgrids in real-time including bidirectional power flow, integration of new technologies, controls, islanding operations, and communication networks. It also describes a real-time hardware-in-the-loop simulation platform that models a microgrid test system containing generators, loads, energy storage and PV to evaluate commercial microgrid controllers under different operating conditions and grid connection scenarios.

2. • DERs and microgrids pose several challenges to engineers including:
• Shift from radial to bidirectional distribution systems (overvoltage concerns)
• Adoption and integration of new/unknown technologies and systems
• Supervisory power flow, agent-based and subsystem level control
• Islanding
• Compliance with grid interconnection requirements and tests (IEEE 1547)
• “Smart grid” communication considerations (communication networks, cybersecurity)
• [Soon to come] Compliance with IEEE P2030.7 and P2030.8 for microgrid controllers
• Main challenges in performing real-time simulation:
• Decoupling larger systems between processors/FPGAs without introducing delays
• Short Lines
• # of switches (Breakers, relays, converters)
• High-frequency PWM power electronics
• Smart grid functionality, communication protocol support
• Virtual components
Delay Delay

3. Source: MIT-LL, TR-1203: Development of a Real-Time Hardware-in-the-Loop Power Systems
Simulation Platform to Evaluate Commercial Microgrid Controllers

5. • Based on a radial industrial feeder, modelled
using Simulink, SimPowerSystems and OPAL-RT
real-time libraries (ARTEMiS-SSN)
• Specifications:
• 13 x transformers
• 13.8, 4.16, 2.4, 460, 208 kV
• 19 x protection relays
• 10 x dynamic loads
• Min: 4.2 MW, Max: 12 MW
• 2 critical, 4 priority, 4 interruptible
• 2 x 250 hp induction motors
• 2 x Caterpillar diesel generators
• 1 MVA, 4 KVA
• 4 MVA Battery/ESS
• 3.5 MW PV
• Varying irradiance profile

7. • Real-world application with
• 57 switches (breakers, IGBTs, etc.)
• detailed custom component libraries
• Modbus communication streams
• Model runs at 70us on 4 cores
• Core 1-2: Microgrid model
• Core 3-4: Detailed protection relays
• Natural delay used
• Can run faster on 2 cores using latest Xeon E5
processors at 50us!
• ARTEMiS-SSN technique used with groups
selected based
• switch placement (3-12 per group)
• optimal nodal interfaces (e.g. 3 groups at node)

8. • State-Space Nodal solver: ARTEMiS-SSN
• Split circuit into groups and iterate:
• Solve using State-Space technique
• Check for admittance at Nodal interface
• Repeat
• Advantages:
• Very computationally efficient
• Example: solving size 8 vs. 64 matrices
• Isolate switches into groups
• Microgrids have many (breakers, inverters, etc.)
• A large system can be solved on a single processor
or parallelized for performance boost

9. • Test objectives:
• Unit commitment (grid-tied and islanded)
• Peak-shaving, valley-filling, and load-shedding
(grid-tied and islanded)
• Diesel generation fuel optimization (grid-tied and
islanded)
• Loss minimization (islanded)
• Meet power export requirements (grid-tied)
• Optimized energy-storage control (grid-tied)
• Generator-battery hybridization (grid-tied)
• Power factor support at PCC (grid-tied)
• Two-way communication with commercial
generator controllers (grid-tied and islanded)
• Condensed 15-minute Sequence:
• First 7.5 mins: Grid-tied
• @7.5 min, 3.5 MW interruptible loads shed
• @8.3 min, islanding, gensets supply power
• Differences between controllers
• Power import/export
• Use of ESS
• Fuel Use (Vendor 2 > Vendor 1)