Grid Topology and DER Effects¶

Research Question¶

How does the choice of feeder (IEEE 13-bus, 34-bus, 123-bus, etc.), and the locations and capacities of loads and distributed energy resources (DERs), affect datacenter impact and controllability?

Overview¶

Distribution feeder topology determines how datacenter power consumption propagates to bus voltages. Shorter, stiffer feeders (like IEEE 13) have lower voltage sensitivity, while longer feeders with downstream loads (like IEEE 34 and 123) exhibit larger voltage drops. The presence of DERs (solar PV, storage) and time-varying loads further complicates the voltage landscape.

OpenG2G includes three standard IEEE test feeders with pre-configured datacenter, PV, and load scenarios:

System	Buses	Voltage	DC Sites	PV Sites	Regulators	Characteristics
IEEE 13	16	4.16 kV	1	1	1 bank (3-phase)	Short, stiff feeder; internal DC load changes dominate
IEEE 34	37	24.9 kV	2	2	2 banks (per-phase)	Long rural feeder; PV and time-varying loads
IEEE 123	132	4.16 kV	4	3	4 regulators	Large suburban feeder; 4 zones with mixed disturbances

Scripts¶

Script	Purpose
`run_baseline.py`	Run baseline simulation for a specific system
`run_ofo.py`	Run OFO simulation for a specific system
`plot_topology.py`	Visualize system topology with DC, PV, and regulator locations

Usage¶

Visualize System Topology¶

# IEEE 13-bus
python examples/offline/plot_topology.py --system ieee13

# IEEE 34-bus
python examples/offline/plot_topology.py --system ieee34

# IEEE 123-bus
python examples/offline/plot_topology.py --system ieee123

The topology plot shows buses (colored by zone), DC locations (stars), PV systems (yellow triangles), and voltage regulators (red diamonds). It automatically parses connections and regulator definitions from the OpenDSS files.

Compare Systems¶

Run baseline and OFO on each system and compare voltage statistics:

# IEEE 13
python examples/offline/run_ofo.py --system ieee13

# IEEE 34
python examples/offline/run_ofo.py --system ieee34

# IEEE 123
python examples/offline/run_ofo.py --system ieee123

Key Observations¶

Electrical distance matters: Buses farther from the substation have higher voltage sensitivity to load changes, making them harder to regulate but also more responsive to batch-size control.
Regulator placement: Systems with regulators between the substation and DC buses (like IEEE 34) offer an additional control degree of freedom via tap changes.
DER interaction: PV injection can cause overvoltage at nearby buses while DC loads cause undervoltage — the net effect depends on relative locations and magnitudes.
Zone structure: In multi-zone systems (IEEE 123), different zones may have independent voltage dynamics, requiring per-zone controller tuning.

Configuration¶

The topology and DER configuration is defined inline in each experiment function (see run_ofo.py):

Feeder constants (in systems.py): dss_case_dir, dss_master_file, source_pu, initial_taps, exclude_buses
Experiment parameters (inline in each script): dc_sites (datacenter locations and power levels), pv_systems (solar PV locations and peak capacities), time_varying_loads (additional load locations and magnitudes)