Skip to content

Latest commit

 

History

History
43 lines (37 loc) · 2.28 KB

roadmap.md

File metadata and controls

43 lines (37 loc) · 2.28 KB

Under development and future work

In progress

Feature What it will do
Modular CMLD motor load The modular composite load model (CMLD) as an option beside CLOD for the motor share of a facility.
Faster black-box validation A shorter validation run, by reusing PSS/E processes and keeping every worker busy.
Full-facility testing in a real system Upload a PSS/E case and test a whole facility, several load models together that represent a data center, in that system.

Future work

Known gaps and planned improvements. The first two are described in detail in Model behavior and limitations.

  • Review preset defaults against the PSS/E findings:
    • VRPA for the five presets that ship with VRPA = 1.00 (ERCOT, PJM, MISO, NERC LLWG SAR, NERC PRC-029-1);
    • DROOPA per preset, where a standard asks for constant consumption during frequency events;
    • overvoltage reconnect timing (TRUPA/RAMPUA) to match each standard's recovery window.
  • Editor warnings when VRPA is below a trip stage (the block cycles), VRPA ≥ VRUPA (never reconnects), or IPMAXA < LFM (the block cannot draw its own load).
  • Recovery voltage in the live chart. A control for the voltage (or frequency) the event recovers to, to show a VRPA window or a cycling block directly. The simulator already supports it.
  • Multi-event disturbances in the live chart: sags that change depth, or recover briefly and return, with the model's resetting stage timers.
  • Standard-model refinements: washout and lead-lag transients in the live chart.
  • Keep the PERC2 warnings current. The PSS/E validation found issues in the PERC2 models that their authors may fix or document (see each model's validation page). The editor's warnings will be revisited when new versions of PSS/E or the DLL appear.
  • Parameterize load models from PSCAD benchmark performance. Derive model parameters from how a PSCAD model performs against a benchmark.
  • More load models, and more ride-through standards as utilities and NERC publish them.
  • Out-of-sample PSS/E validation on randomized scenarios.

Suggestions for this list are welcome on the project's GitHub Issues page.