ERCOT’s Large Load Working Group (LLWG) met on 8/14 with a clear through-line: voltage ride-through (VRT) for large loads is now a grid-reliability issue with real operational, planning, and market implications. The chair opened by noting broad industry engagement, from traditional data centers to crypto miners—and a growing set of technical ideas to solve VRT at both the grid and the load.
The Four ERCOT Studies (and why they matter)
1) System Operating Limits (SOL/IROL) for Large-Load VRT
Objective: Find specific fault locations where large-load tripping during voltage dips could violate frequency or voltage stability, and define operational limits plus mitigation plans.
- Method: Start with a DWG case, update load models via the LL RFI responses, and simulate planning events.
- Outputs (by year-end):
- A list of fault locations + total coincident load loss that would breach stability criteria.
- Identification of SOLs/IROLs and associated mitigation plans (first preference is non-load-shed actions like switching or committing generation).
- Notes: These are not expected to be GTCs (i.e., not SCED-controlled constraints). Details will be ECEII; only high-level findings will be shared.
Why this matters: If your site trips on voltage dips, it can end up in the “monitored set” for operational limitations.
2) Frequency Limits vs. Load Tripping (beyond the “2,600 MW worst-case”)
Objective: Quantify how the “safe” megawatts of instantaneous load loss change with system inertia, downward PRC, and other conditions; assess whether a downward PRC ancillary service would materially help.
- Method: TSAT snapshots across different historical system conditions; same acceptance criterion (steady-state frequency must remain ≤ 60.4 Hz).
- Early signal: Downward PRC helps only modestly (think hundreds of MW), and very fast droop alone likely doesn’t fix the core problem. Deeper physics-based explanation coming when final results are ready.
Why this matters: Banking on a new down-PRC product to “solve it” for large loads is unlikely to be a silver bullet.
3) Transmission Solutions: Sync-Cons, Grid-Forming Statcoms & Friends
Objective: Estimate the relative benefit of devices (sync-cons in West Texas 2026–27, potential grid-forming statcoms) and broader grid upgrades in reducing VRT-driven instability risk.
- Method: 2030/31-ish DWG case focused on Far West/West/ Panhandle; test devices/upgrades to see how much they raise stability limits and shrink impacted areas.
- Scope: Not a project recommendation list, a criteria-building study to tell planners where devices/upgrades pay off most.
Why this matters: Helps you and your TSP evaluate which device strategies meaningfully reduce “voltage sag → mass load trip” risk near your site.
4) Evaluating a Load-Side VRT Standard (proposed NOGRR)
Objective: Test the draft large-load VRT spec (how deep, how long, and critically how fast loads must return) and refine parameters so frequency and voltage both remain acceptable.
- Trade-off: Frequency wants fast load restoration; voltage recovery prefers not too fast. ERCOT will tune parameters with sensitivity studies.
- Process: Draft standard soon for stakeholder discussion; final will reflect study results.
Why this Matters: This is likely to become the baseline compliance target for large electronic loads across ERCOT.
Two technology spotlights to know about
- Siemens Energy (eSTATCOM with supercapacitors): Sub-10 ms response at the POI to cancel sub-second oscillations and smooth short ramps, protecting nearby synchronous machines and the grid. It won’t deliver long energy (minutes), but does reshape fast dynamics (including short post-fault effects) and provides reactive/voltage support.
- Advanced Crypto Services (device-level approach): Classic PSU “big capacitors” won’t practically hold an entire miner up for 160 ms. A more viable path is keeping the control board alive with targeted caps + firmware changes to restore full load within ~1 s (instead of full reboots). A structured lab program with grid simulators is planned to quantify what’s achievable at the miner level.
Takeaway: Expect hybrid mitigation—device-level logic + site-level power electronics (e.g., eSTATCOM) + operational playbooks.
Data center & crypto load realities (ERCOT analytics)
- Data centers (non-crypto): Three signatures stood out—storage (very flat), cloud compute (modest diurnal movement), and compute clusters. Across 54 ≥25-MW sites, daily variability is small relative to nameplate. First-half 2025 growth ~5%, mostly during outage season; once energized, sites tend to stay flat.
- Crypto mining: ~4.5 GW currently observed aggregate, with installed capability ~5.6 GW and some migration to AI underway. Crypto is strongly price-responsive and near-4CP responsive; large coincident ramps are real. ERCOT is monitoring for regulation/AS impacts; no immediate AS methodology change adopted for 2026, but growth will be reevaluated.
Why this matters: Your load model and operational behavior (price-/4CP-responsive) directly inform study assumptions, SOL/IROL exposure, and any future operating limits near you.
What large-load owners/developers should do now
- Respond with quality to ERCOT’s modeling RFI
- Provide dynamic load models that reflect real UPS logic, reclosing behavior, and any internal controls (e.g., “flooring” GPU load).
- If final equipment is TBD, submit best-estimate now and update upon PO.
- Self-assess VRT
- Can your site ride through sub-second voltage dips without wholesale transfer to UPS?
- If UPS transfer is required, do you have a return-to-grid timeline that balances fast frequency support and voltage recovery?
- Plan mitigation options
- Device-level: control/firmware to keep brains alive; fast restart logic.
- Site-level: evaluate eSTATCOM/supercap solutions for sub-second smoothing and voltage support; quantify required MW/MWs.
- Operational: write playbooks (reclose coordination with TSP, staged restoration curves, price/4CP policies that avoid synchronized ramps).
- Engage your TSP early
- Share models, expected behaviors, and any planned device mitigation so TSP/ ERCOT can reflect it in studies.
- Track the standard & FAQ
- Watch for the draft large-load VRT NOGRR, LLWG study updates, and the living FAQ. Align internal designs and EPC scopes now.
How ZEG experts Accelerate Large-Load Planning
Large-Load VRT is moving from “interesting” to “operational.” If you own or develop large electronic loads in ERCOT, data centers, HPC/AI, or crypto, now’s the moment to turn this LLWG guidance into a defensible plan.
How ZEG helps:
- ERCOT Rule & Guide Alignment
Translate emerging large-load VRT expectations into a practical compliance roadmap (what to submit, when, and how to defend it). - Modeling & Validation Support
Assemble and refresh planning-grade dynamic models (PSS®E/TSAT/PSCAD as applicable) that reflect your true behavior at the POI and support TSP/ERCOT reviews. - Electrical & Stability Studies
Run the studies that underpin credible ride-through capability (fault response, restoration timing sensitivities, voltage/VAR control interactions). - Submission & Stakeholder Support
Package filings cleanly for ERCOT/TSP processes and facilitate technical dialogue so your assumptions are understood and accepted. - OEM & Integrator Coordination
Herd the cats—UPS, inverter, and control vendors—so device-level logic and site-level mitigation hang together. - Peer Review & QA
Pre-flight checks to catch the avoidable deficiencies that slow endorsements and energization.
Ready to make your site VRT ready?
Let our experts shoulder the heavy lift. Share your site(s) and stage (RFI → design → energization), and we’ll outline a concrete plan you can act on immediately. Contact us today.
Learn more about large load infrastructure, explore our Comprehensive Guide on Large Load Infrastructure here, or view the LLWG meeting materials here.
