{ "id": "texas_ercot_ai_demand", "version": "1.0", "status": "active", "scenario_type": "Power Transition", "name": "Texas ERCOT AI-Driven Grid Expansion Mandate", "subtitle": "Maintaining reliability against 15 GW AI/data-center demand surge by 2031", "region_id": "us", "tags": [ "power-sector", "mandate", "grid-reliability", "battery-storage", "ai-demand", "ercot", "demand-growth" ], "description": "ERCOT \u2014 the near-isolated Texas grid serving 27 million customers \u2014 faces a structural reliability crisis driven by the fastest AI/data-center buildout in US history. Hyperscalers (Microsoft, Google, Amazon, Meta) and co-location operators are adding ~3 GW of new load per year in the Dallas\u2013Fort Worth and San Antonio corridors, with ERCOT's 2025 Capacity, Demand & Reserves report projecting 77 GW of new interconnection requests by 2030. ERCOT operates with fewer than 1 GW of AC ties to neighboring grids, making it the most electrically isolated major grid in North America. The Texas Legislature and Public Utility Commission of Texas (PUCT) mandate that ERCOT maintain a \u226515% planning reserve margin through 2031, even as AI/data-center load grows by 15 GW and summer peaks are amplified by accelerating heat waves. The mandate forces a rapid, clean capacity buildout: large-scale battery storage, West Texas solar expansion, and dispatchable demand-response contracts with AI data centers \u2014 all deployed without new unabated gas, so emissions hold flat at baseline even as load grows 30%.", "baseline": { "year": 2026, "generation_fleet_gw": 117.0, "coal_gw": 4.0, "gas_ccgt_gw": 38.0, "gas_peakers_gw": 14.0, "ccgt_gw": 52.0, "nuclear_gw": 5.4, "wind_gw": 37.0, "solar_gw": 16.0, "bess_gw": 2.6, "coal_capacity_factor": 0.55, "gas_capacity_factor": 0.43, "grid_carbon_intensity_g_per_kwh": 202, "grid_carbon_intensity_note": "Corrected (2026-05-24): 99 Mt \u00f7 490 TWh = 202 g/kWh (2026 total grid average including renewables). The prior value of 340 g/kWh reflected 2021-2022 fossil-fleet dispatch intensity (pre-solar buildout) and overstated CE API emissions outputs by 68%. The operative emissions baseline is gas+coal dispatch: gas 52 GW \u00d7 0.43 CF \u00d7 8760h \u00d7 0.42 t/MWh \u2248 82 Mt + coal 4 GW \u00d7 0.55 \u00d7 8760 \u00d7 0.82 \u2248 16 Mt = 98 Mt \u2248 99 Mt stated.", "annual_generation_twh": 490, "annual_emissions_mt_co2": 99.0, "peak_demand_gw": 88.0, "notes": "Coal: ~4 GW NRG/Luminant (retirements underway). Gas: 38 GW CCGT + 14 GW simple-cycle peakers = 52 GW total. Nuclear: Comanche Peak 2-unit (2.4 GW) + South Texas Project 2-unit (3.0 GW) = 5.4 GW. Wind: 37 GW installed West Texas / Gulf Coast. Solar: 16 GW utility-scale (growing ~4 GW/yr). BESS: 2.6 GW 4-hr deployed. ERCOT AC ties to Eastern Interconnect: <1 GW. AI/data-center demand already growing ~3 GW/yr." }, "target": { "reduction_pct": 2, "deadline_year": 2031, "horizon_years": 5, "required_reduction_mt_co2": 2.0, "ceiling_mt_co2_by_2031": 97.0, "reliability_target": "\u226515% planning reserve margin every summer through 2031", "penalty": { "description": "PUCT reliability failure threshold triggers FERC review and possible mandatory interconnection to SPP/Eastern Interconnect \u2014 loss of ERCOT market sovereignty plus $28B+ economic damage per Uri-scale blackout event.", "mechanism": "FERC Order 1920 implementation + PUCT SB6 reliability penalty" }, "notes": "Reduction target is 2% absolute CO2 reduction vs baseline (99\u219297 MtCO2), framed as a 'net-zero growth' mandate: absorb 15 GW new AI/data-center load WITHOUT dispatching additional gas. BAU path (gas-served AI load) reaches 133 MtCO2 by 2031 \u2014 a 34% increase. The mandate holds emissions flat while serving 30% more load, implying a 25% relative improvement vs BAU." }, "structural_constraints": { "rto_interconnection_queue_yr": 0.5, "rto_queue_threshold_mw": 50, "transmission_thermal_capacity_pct": 81, "peak_demand_gw": 88.0, "demand_growth_cagr_pct": 3.8, "ai_datacenter_demand_gw_per_yr": 3.0, "interconnection_capacity_gw": 0.9, "weather_volatility": 0.65, "permitting": { "bess_brownfield_timeline_months": 18, "solar_greenfield_timeline_months_min": 18, "solar_greenfield_timeline_months_max": 24, "weighted_avg_yr": 1.6, "greenfield_barriers": "West Texas land permitting (mineral rights split), CREZ transmission congestion, ERCOT interconnection study queue pressure" } }, "tech_vectors": [ { "id": "ercot_bess_deployment", "name": "Large-Scale BESS Deployment (4\u20138hr)", "description": "Rapid deployment of 4-hour to 8-hour utility-scale battery storage to replace peaking gas dispatch and provide firm UCAP capacity credit for planning reserves. Texas has a streamlined ERCOT interconnection process for BESS (<50 MW can bypass full study), enabling faster deployment than solar or wind.", "target_capacity_gw": 12.0, "storage_duration_hr": 6.0, "ce_model_mapping": "none", "ce_model_gap": "BESS not in CE TECHS_ABATE; modelled as gas displacement proxy", "estimated_mt_co2": 8.0, "constraints": { "total_lead_time_yr": 1.5, "critical_path": "ERCOT fast-track interconnection + battery supply chain (LFP cells from Korea/China)", "cost_usd_b": 12.0, "cost_per_gw_usd_m": 1000 } }, { "id": "west_texas_solar_plus_storage", "name": "West Texas Solar Expansion + Co-located BESS", "description": "9 GW of new utility-scale solar in the Permian Basin / West Texas region (>6 kWh/m\u00b2/day resource), co-located with 3 GW of 4-hour BESS to smooth the afternoon ramp and extend dispatch into evening peak hours. Uses existing CREZ transmission corridors where available; some new 345kV spurs required.", "target_capacity_gw": 9.0, "bess_colocation_gw": 3.0, "ce_model_mapping": "perovskite_solar proxy", "estimated_mt_co2": 6.0, "constraints": { "total_lead_time_yr": 2.5, "critical_path": "CREZ transmission capacity + land permitting in Reeves/Ward/Pecos counties", "cost_usd_b": 11.7, "cost_per_gw_usd_m": 900 } }, { "id": "ai_datacenter_demand_response", "name": "AI Data Center Dispatchable Demand Response", "description": "Contractual demand-response capacity with AI/hyperscale data center operators in DFW and San Antonio. Data centers can curtail non-critical batch compute loads (training runs) within 10 minutes of ERCOT dispatch signal, providing 3 GW of virtual firm capacity. Pioneered by ERCOT's 'large flex load' program (2024). Avoids capital cost of equivalent generation.", "target_capacity_gw": 3.0, "ce_model_mapping": "none", "ce_model_gap": "Demand-side resource; no generation equivalent in CE TECHS_ABATE", "estimated_mt_co2": 1.0, "constraints": { "total_lead_time_yr": 0.5, "critical_path": "Bilateral contracts + ERCOT DR registration; fastest deployable resource", "cost_usd_b": 0.3, "cost_per_gw_usd_m": 100 } } ], "model_gaps": [ { "gap": "Battery storage not in CE TECHS_ABATE", "impact": "HIGH \u2014 BESS is the critical-path technology; 12 GW of BESS provides 12 GW UCAP credit. Modelled as displaced gas generation proxy.", "mitigation": "GridStabilityService BESS UCAP credit (full credit at \u22654hr duration)" }, { "gap": "ERCOT grid isolation not modelled", "impact": "HIGH \u2014 CE assumes grid-connected dispatch with neighbour support. ERCOT's <1 GW AC interconnection means near-zero emergency import capability.", "mitigation": "GridStabilityService interconnection_gw=0.9 parameter captures isolation penalty" }, { "gap": "AI/data-center demand growth endogeneity", "impact": "HIGH \u2014 CE has no endogenous demand-growth model for AI compute load. Manual CAGR applied (3.8%/yr) based on ERCOT 2025 CDR P50 scenario.", "mitigation": "demand_growth_cagr_pct in structural_constraints; Phase 2 AI demand module planned" }, { "gap": "Demand response as firm capacity", "impact": "MEDIUM \u2014 CE models supply-side only; 3 GW of contractual DR not reflected in abatement math.", "mitigation": "Included in tech_contributions as explicit line item; CE API endpoint accepts demand-side adjustments" }, { "gap": "ERCOT scarcity pricing dynamics", "impact": "MEDIUM \u2014 ERCOT's energy-only market can reach $9,000/MWh during scarcity. Investment signals differ from capacity-market grids (PJM, MISO). CE uses capacity-market logic.", "mitigation": "Non-compliance section captures escalating scarcity cost as proxy for market failure signal" }, { "gap": "Firm unforced capacity (UCAP) parameters not modeled", "impact": "HIGH \u2014 CE does not compute UCAP credits for intermittent resources and storage. Replacing 52 GW of firm gas dispatch with 12 GW BESS + 9 GW solar + 3 GW DR requires UCAP accounting to verify the 15% planning reserve margin is actually met under P90 peak conditions, not just on nameplate.", "mitigation": "GridStabilityService UCAP credit applied at full nameplate for BESS \u22654hr; solar and wind de-rated by CE regional CF factors. UCAP margin calculation in mandate_2031.notes assumes full BESS UCAP credit \u2014 validate against ERCOT CDR P90 UCAP table before investment-grade use." }, { "gap": "Loss of Load Expectation (LOLE) not computed", "impact": "HIGH \u2014 NERC standard is 0.1 days/year LOLE. CE has no probabilistic reliability model; the 23.6% nameplate reserve and 16% UCAP margin are point estimates, not probabilistic outcomes. ERCOT's 1-in-10 peak demand P90 is 12 GW above P50, meaning a single extreme heat wave can eliminate the entire reserve margin.", "mitigation": "Use ERCOT's published LOLE studies (CDR Appendix C) to validate the reserve margin; note CE cannot independently verify the 0.1 day/yr standard for this fleet configuration." }, { "gap": "Sub-hourly dispatch mechanics not modeled", "impact": "MEDIUM \u2014 CE models annual and seasonal average dispatch, not 5-minute SCED intervals. ERCOT's SCED operates at 5-minute clearing; BESS cycling economics (cycle degradation, round-trip efficiency losses, state-of-charge management) are not captured. The 12 GW BESS providing 6-hr continuous dispatch assumed in the scenario requires sub-hourly modeling to verify energy-duration sufficiency during multi-day heat events.", "mitigation": "BESS modeled as gas displacement proxy for emissions; duration economics treated as engineering input rather than CE output. Recommend PLEXOS or similar for sub-hourly validation before commissioning." } ], "analysis": { "critical_path": "west_texas_solar_plus_storage", "abatement_needed_mt_co2": 2.0, "confidence": "medium", "confidence_rationale": "BESS deployment pace well-established in Texas (4 GW deployed 2022\u20132025); primary uncertainty is AI demand growth trajectory (ERCOT P50 vs P90 diverge by 12 GW by 2031)", "tech_contributions": [ { "label": "Utility-Scale BESS (4\u20138hr, 12 GW)", "mt_co2": 8.0 }, { "label": "West Texas Solar + Co-located BESS", "mt_co2": 6.0 }, { "label": "AI Datacenter Demand Response", "mt_co2": 1.0 } ], "estimated_total_mt_co2": 15.0, "estimated_margin_mt_co2": 13.0, "notes": "The 15 Mt 'abatement' figure represents CO2 displacement vs BAU (where 15 GW of new gas would emit +34 MtCO2 by 2031). Actual absolute emissions decline only 2 Mt (99\u219297). The mandate math is expressed as 'net-zero growth': serve +30% load with no emissions increase, which is a more demanding technical challenge than the 2 Mt absolute figure suggests." }, "projections": { "years": [ 2026, 2027, 2028, 2029, 2030, 2031 ], "bau_mt_co2": [ 99.0, 104.0, 110.0, 116.0, 124.0, 133.0 ], "mandate_mt_co2": [ 99.0, 98.5, 98.0, 97.5, 97.2, 96.8 ], "ceiling_mt_co2": 97.0 }, "fleet_evolution": { "baseline_2026": { "coal_gw": 4.0, "gas_ccgt_gw": 38.0, "gas_peakers_gw": 14.0, "nuclear_gw": 5.4, "wind_gw": 37.0, "solar_gw": 16.0, "bess_gw": 2.6, "total_gw": 117.0 }, "bau_2031": { "coal_gw": 2.0, "ccgt_gw": 67.0, "renewables_gw": 62.0, "ders_gw": 4.0, "total_gw": 135.0, "notes": "BAU: 15 GW new gas CCGT to serve AI demand; coal retirements continue; modest solar/wind additions; minimal BESS beyond current buildout" }, "mandate_2031": { "coal_gw": 2.0, "gas_ccgt_gw": 52.0, "nuclear_gw": 5.4, "renewables_gw": 62.0, "ders_gw": 14.6, "total_gw": 136.0, "notes": "Mandate: No new gas CCGT. Firm thermal = existing gas 52 GW + nuclear 5.4 GW (no CCUS deployed). Renewables = wind 37 + solar 25 (new 9 GW). DERs = BESS 2.6 + new 12 GW. Peak demand 2031 = ~110 GW (88 \u00d7 1.038^5). Nameplate reserve = (136/110-1) = 23.6%. UCAP margin with BESS credit \u2248 +16% (passes 15% threshold)." } }, "non_compliance": { "trigger_year": 2031, "mandate_cost_label": "~$23\u201324B", "mandate_cost_description": "BESS + solar capex (5yr build, amortized)", "mechanism": "PUCT reliability failure triggers FERC-directed grid interconnection study (loss of ERCOT market sovereignty); escalating economic damage from scarcity events. Data-center operators face SLA penalties and potential relocation pressure. AI workload displacement to non-Texas regions undermines state economic strategy.", "affected_exports_usd_b": 145.0, "embedded_emissions_mt_co2": 25, "max_annual_cost_usd_b": 13.5, "five_year_cumulative_usd_b": 37.2, "tax_schedule": [ { "year": 2031, "rate_usd_per_t": 120, "annual_cost_usd_b": 3.0, "cumulative_usd_b": 3.0 }, { "year": 2032, "rate_usd_per_t": 185, "annual_cost_usd_b": 4.6, "cumulative_usd_b": 7.6 }, { "year": 2033, "rate_usd_per_t": 265, "annual_cost_usd_b": 6.6, "cumulative_usd_b": 14.2 }, { "year": 2034, "rate_usd_per_t": 380, "annual_cost_usd_b": 9.5, "cumulative_usd_b": 23.7 }, { "year": 2035, "rate_usd_per_t": 540, "annual_cost_usd_b": 13.5, "cumulative_usd_b": 37.2 } ], "affected_sectors": [ { "name": "AI / Hyperscale Data Centers", "icon": "fa-server", "export_value_usd_b": 85.0, "exports_usd_b": 85.0, "embedded_mt_co2": 12.0, "jobs": 185000, "notes": "~200 GW of planned Texas data center capacity (per CBRE 2025); $85B annual economic value at risk from reliability failure; 12 MtCO2 embedded from emergency diesel backup + coal dispatch during scarcity events" }, { "name": "Petrochemical Refining", "icon": "fa-industry", "export_value_usd_b": 42.0, "exports_usd_b": 42.0, "embedded_mt_co2": 8.5, "jobs": 64000, "notes": "Gulf Coast refineries and petrochemical plants (ExxonMobil Baytown, BASF, LyondellBasell); highly sensitive to power interruptions; continuous-process risk from blackout events" }, { "name": "Semiconductor Fabrication", "icon": "fa-microchip", "export_value_usd_b": 18.0, "exports_usd_b": 18.0, "embedded_mt_co2": 4.5, "jobs": 29000, "notes": "Samsung Austin (Taylor), Texas Instruments, NXP; wafer fabrication requires uninterruptible power; single Uri-scale event destroys in-process inventory" } ] }, "action_items": [ { "id": "ai_01", "audience": "utility_grid_operator", "action": "ERCOT: require large load interconnection applicants (>100 MW) to post financial assurance (performance bonds or letters of credit) sized to cover grid upgrade costs before queue study commencement \u2014 reducing speculative queue entries that inflate upgrade cost estimates for all participants.", "rationale": "ERCOT's interconnection queue currently has 300+ GW of requested capacity, the majority of which will not be built. Speculative entries inflate network upgrade cost estimates for real projects and delay legitimate interconnection. Financial assurance requirements (implemented by MISO and PJM) have reduced speculative entries by 60\u201370%.", "defensible_basis": "FERC Order 2023 (interconnection reform); MISO and PJM financial assurance precedent; ERCOT Nodal Protocol revision docket (2025). ERCOT has independent rulemaking authority \u2014 PUCT docket filing is the required pathway.", "urgency": "immediate", "no_regret": true }, { "id": "ai_02", "audience": "corporate_industrial_buyer", "action": "Hyperscale data centre developers committing to Texas: evaluate behind-the-meter generation (co-located CCGT + solar) structures rather than pure grid-pull configurations \u2014 reducing load concentration risk on ERCOT transmission and providing price hedge against ERCOT scarcity pricing events ($9,000/MWh cap).", "rationale": "ERCOT's nodal pricing can reach the $9,000/MWh administrative cap during scarcity events. A 500 MW AI campus drawing from the grid faces energy cost volatility of $0\u20139,000/MWh. Behind-the-meter generation provides price stability and avoids $500M/yr in potential scarcity cost exposure during the scenario's high-demand years (2028\u20132031).", "defensible_basis": "ERCOT market price data (2023 summer: multiple $5,000+/MWh hours); PUCT retail electric rules for behind-the-meter generation; EIA combined-cycle LCOE at Texas natural gas prices. Behind-the-meter generation is commercially available and electrically straightforward.", "urgency": "near_term", "no_regret": true }, { "id": "ai_03", "audience": "sovereign_policymaker", "action": "Texas PUC: update ERCOT's demand forecast methodology to include AI/GPU cluster load profiles \u2014 current forecasting models AI demand as a flat industrial load, missing the highly variable ramping characteristics of GPU training clusters that can swing 0\u2013100% in minutes.", "rationale": "GPU training workloads have fundamentally different demand profiles than traditional industrial load: near-instantaneous full-load engagement, short duration, repeat cycles. ERCOT's reserves are sized on historical demand ramp rates. AI cluster ramps that exceed historical parameters could trigger under-frequency events if reserves are undersized.", "defensible_basis": "ERCOT 2025 Capacity Demand and Reserves (CDR) report; Lawrence Berkeley National Laboratory data centre load characterisation (2024); NERC reliability standard for frequency response (BAL-003). Methodology update within ERCOT planning authority \u2014 no new regulation.", "urgency": "immediate", "no_regret": true }, { "id": "ai_04", "audience": "renewable_energy_developer", "action": "Texas wind and solar developers: begin EIS processes for 345kV transmission corridors from the Permian Basin and Panhandle wind/solar zones to the Austin-San Antonio data centre corridor NOW \u2014 this transmission is the critical path for serving AI load with clean energy and EIS takes 4\u20136 years.", "rationale": "The scenario requires 34 GW of renewable equivalent for AI load by 2030. Existing transmission from West Texas wind zones to Austin is already congested. New 345kV corridor EIS filed in 2026 targets energisation by 2031\u20132032. Filing later pushes clean energy delivery past the scenario's mandate window.", "defensible_basis": "ERCOT 2025 Long-Term System Assessment (transmission constraint identification); DOE National Transmission Needs Study 2024 (Texas corridor needs); NEPA EIS average timeline data. EIS filing is a private-sector application \u2014 within developer authority.", "urgency": "immediate", "no_regret": true } ], "sources": [ "ERCOT Capacity, Demand & Reserves Report 2025 (CDR) \u2014 P50 and P90 scenarios", "PUCT SB6 Weatherization and Reliability Standards (2021, amended 2024)", "ERCOT Large Flexible Load program \u2014 2024 demand response tariff update", "EIA Texas State Energy Profile 2024 \u2014 generation mix and emissions", "CBRE Texas Data Center Market Report Q1 2025 \u2014 hyperscaler buildout pipeline", "NERC 2024 Long-Term Reliability Assessment \u2014 ERCOT region", "ACC/ERCOT Settlement Interconnection Queue \u2014 April 2025" ], "created": "2026-05-05", "last_updated": "2026-05-19", "author": "CE Scenario Engine v3.7", "fiscal_transition": { "entity_name": "ERCOT / PUCT (Texas Grid Operator / Public Utility Commission)", "price_label": "ERCOT Average Residential Retail Rate (\u00a2/kWh)", "price_unit": "\u00a2/kWh", "framing": "Phase 1 (2026\u20132028): Absorb AI demand surge without new gas dispatch. ERCOT is the largest competitive electricity market in the US (117 GW, 88 GW peak). Three major hyperscalers (Microsoft, Google, Meta) have announced 15 GW of new AI data centre load in Texas by 2031 \u2014 a 17% load increase on a market that already operates at the tightest planning reserve margins in North America. Without BESS and utility solar to serve this load, ERCOT's only dispatchable option is gas \u2014 pushing BAU emissions to 133 Mt CO2 by 2031. The mandate holds emissions flat at 97 Mt while serving 30% more electricity, requiring solar+BESS to capture all incremental AI load. Phase 2 (2028\u20132031): AI demand response + weatherization lock-in. PUCT SB6 weatherization mandate ensures winter reliability after the 2021 Uri event ($28B economic damage). Demand response contracted from AI data centres (1 GW) provides a novel market-clearing mechanism: data centres defer batch AI workloads during peak events, acting as virtual dispatchable load reduction equivalent to a gas peaker.", "phase_1": { "label": "BESS + Solar Sprint to Absorb AI Load", "years": "2026\u20132028", "annual_capex_usd_b": 2.78, "capex_sources": { "ira_itc_solar_30pct": "IRA \u00a745 ITC 30% on West Texas solar ($5.4B CAPEX): $1.62B tax equity; monetised by Vistra Energy, NRG, AES tax equity banks", "bess_project_finance": "Private BESS project finance: $3.5B at 6.2% (investment-grade offtake contracts with hyperscalers; Goldman Sachs, BlackRock infrastructure)", "doe_lpo_bess": "DOE LPO \u00a717 grid-scale battery $1.5B at 4.2%; ERCOT reliability programme; 5 GW BESS portfolio financing", "texas_rep_ppas": "Texas REP (NRG/Vistra/TXU) long-term PPA: $4.0B equity in solar+BESS; 15-year PPA at 5.8\u00a2/kWh for AI data centre offtake", "ercot_weatherization_bond": "Texas DOT + PUCT SB6 weatherization bond $1.2B; winter hardening of gas fleet; electric heat tracing mandate", "transmission_ercot_cdp": "ERCOT competitive transmission (CREZ 2.0): $2.8B; West Texas Panhandle to Austin/Dallas corridors; regulated transmission recovery" }, "peak_domestic_financing_gap_usd_b": 0.85, "peak_financing_gap_year": 2028, "entity_deficit_trajectory": [ { "year": 2026, "deficit_usd_b": 0.28, "note": "BESS procurement launch; weatherization completion; ERCOT SCED market redesign for demand response; PUCT SB6 compliance" }, { "year": 2027, "deficit_usd_b": 0.62, "note": "BESS construction peak (5 GW under build); 9 GW AI load added; summer 2027 reliability stress \u2014 reserve margin 14.2% (below target)" }, { "year": 2028, "deficit_usd_b": 0.85, "note": "Peak stress: 12 GW AI load active; West Texas solar+BESS construction; grid right at reliability threshold; PUCT emergency reliability procurement" }, { "year": 2029, "deficit_usd_b": 0.52, "note": "8 GW BESS operational; 6 GW West Texas solar delivering; reserve margin recovers to 16.8%; AI demand response certified by ERCOT" }, { "year": 2031, "deficit_usd_b": 0.15, "note": "Mandate achieved: 97 Mt with 15 GW AI load absorbed; summer reserve margin 18.2%; winter weatherized; ERCOT sovereignty preserved" } ], "price_trajectory": [ { "year": 2026, "price": 12.5, "note": "ERCOT average residential retail (\u00a2/kWh); gas at $3.50/MMBtu; post-Uri normalization; competitive REP market" }, { "year": 2027, "price": 12.8, "note": "3 GW/yr AI load growth; summer 2027 scarcity events (5 \u00d7 $9,000/MWh caps); BESS begins dampening spikes" }, { "year": 2028, "price": 13.2, "note": "Peak rate; 9 GW AI load; BESS+solar construction CAPEX; mild supply crunch in summer peak hours" }, { "year": 2029, "price": 12.8, "note": "Solar+BESS online; West Texas 4 GW solar at $18/MWh LCOE \u2014 below gas marginal cost; peak-hour prices stabilize" }, { "year": 2031, "price": 12.2, "note": "Mandate achieved; solar LCOE $18/MWh < retained gas $35/MWh; BESS dampening all scarcity events; AI demand response reducing peak pricing" } ], "fx_reserve_risk": "Not applicable \u2014 USD domestic. ERCOT is isolated from Eastern and Western Interconnects (900 MW DC tie only). This isolation \u2014 ERCOT sovereignty \u2014 is Texas's primary energy policy identity. The mandate preserves ERCOT sovereignty by avoiding the FERC Order 1920 mandatory interconnection trigger (which would fire if ERCOT fails reliability standards). The fiscal risk is bilateral: gas lock-in fails reliability (FERC trigger) OR clean energy transition fails supply adequacy (also FERC trigger).", "sovereign_debt_trajectory": { "baseline_debt_gdp_pct": null, "transition_peak_debt_gdp_pct": null, "peak_year": null, "stabilized_debt_gdp_pct": null, "stabilization_year": null, "imf_dsa_threshold_pct": null, "notes": "US domestic scenario \u2014 sovereign debt not applicable. Texas does not have a state income tax; general obligation debt is limited. PUCT SB6 weatherization bond ($1.2B) is state-level debt within Texas bond cap. ERCOT is a non-profit 501(c)(4); has no debt-issuing authority \u2014 all capital raises are via market participants (REPs, generators, transmission)." }, "imf_compatibility": "Not applicable \u2014 US federal/state mandate. PUCT and FERC are the regulatory authorities. ERCOT reliability standards (15% planning reserve margin) are codified in PUCT rules. SB6 weatherization requirements are Texas state law. FERC Order 1920 applies at interconnect level \u2014 ERCOT's isolation means FERC has limited jurisdiction unless interconnection is triggered.", "key_risks": [ "Winter Storm Uri repeat: a Cat.2 polar vortex event dislodges gas wellhead production (as in February 2021), removing 30\u201335 GW of gas generation in 2\u20133 days; BESS can provide 8\u201310 hours of support but not multi-day duration; without weatherized gas wellheads (PUCT SB6), a Uri repeat would again cost $28B+ and trigger FERC intervention", "AI data center demand acceleration: 3 GW/yr is the current ERCOT-filed interconnection pipeline; if Microsoft, Google, or OpenAI accelerate to 5 GW/yr, the supply gap widens faster than BESS+solar can be deployed \u2014 reliability failures in summer 2028 or 2029 are possible", "CREZ 2.0 transmission cost: West Texas solar requires $2.8B of new 765kV transmission to reach load centres; competitive transmission (eTrans mechanism) requires PUCT approval; landowner opposition on 450-mile corridors has blocked two previous projects; if delayed 2 years, West Texas solar cannot deliver before 2030", "BESS thermal management: lithium-ion BESS in Texas summer (ambient 42\u00b0C+) requires active cooling systems \u2014 a significant O&M cost item; multiple BESS fires in Texas 2024 have increased insurance costs and added fire safety regulations; regulatory delays possible for large-format BESS in ERCOT" ] }, "phase_2": { "label": "Demand Response + Weatherization Lock-in", "years": "2028\u20132031", "savings_label": "Gas Avoided Cost + Uri-Scale Event Avoided (annual)", "savings_context": "vs BAU 15 GW AI load served by gas dispatch (133 Mt trajectory) + Uri-scale event probability", "primary_savings_usd_b_annual": 1.8, "import_label": "Gas Fuel Cost Avoided (2031 vs BAU gas-served AI trajectory)", "import_context": "15 GW AI load clean-served; 37 Mt CO2 avoided vs BAU; gas dispatch reduced from BAU 133 Mt to 97 Mt", "import_exposure_end_usd_b": 1.52, "entity_fiscal_trajectory": "ERCOT market economics increasingly favour solar+BESS as the marginal cost setter in peak hours. By 2029, West Texas solar at $18/MWh LCOE is below gas marginal cost \u2014 solar+BESS sets the price floor and cap simultaneously. REP PPAs at 5.8\u00a2/kWh for AI datacentre load are locked-in at below-market rates. ERCOT's energy-only market design (no capacity payments) incentivises BESS for ancillary services (ORDC, reg-up) \u2014 BESS earns $35\u201365/MW/h in ancillary services, partially self-funding O&M.", "export_competitiveness": "Texas AI/data centre sector ($42B/yr server farm investment pipeline): hyperscaler decisions anchor on grid reliability + clean power certification. Without mandate, hyperscalers face a choice between Texas gas-heavy power (CBAM-ineligible for EU data processing) and clean-power states (Virginia, California, Oregon). With mandate, Texas retains $15B/yr of announced hyperscaler CAPEX. Texas semiconductor manufacturing (TSMC Taylor fab, Samsung Austin) benefits from grid reliability and clean power certification.", "resilience_dividend": "PUCT SB6 weatherization eliminates the gas wellhead failure mode that caused Uri ($28B). BESS provides 8\u201310 hours of black-start capability \u2014 a grid restart capability ERCOT did not have in 2021. AI demand response (1 GW contracted) provides a new class of dispatchable demand reduction: data centres with interruptible load clauses can reduce 1 GW on 5-minute notice during peak events, equivalent to a $400M gas peaker built at zero capital cost.", "bond_market_outlook": "ERCOT market participants (Vistra: BBB, NRG: BBB+, AES Texas: BBB-) benefit from long-term AI data centre PPA offtake (investment-grade counterparties: Microsoft AAA, Google AA+, Meta A). Texas transmission utilities (Oncor: A, CenterPoint: A2) maintain stable credit as CREZ 2.0 expansion is cost-recoverable through PUCT tariff. Texas state credit (Aaa/AAA) unaffected \u2014 energy policy is market-driven." }, "counterfactual_inaction": { "label": "Gas-Served AI Load + FERC Interconnection Trigger", "framing": "Without mandate: 15 GW AI load served by gas dispatch (BAU trajectory: 133 Mt CO2 by 2031). ERCOT fails 15% planning reserve target in summer 2028 or 2029 \u2014 triggers FERC Order 1920 review and possible mandatory SPP/Eastern Interconnect coupling. Loss of ERCOT market sovereignty is the single largest political risk in Texas energy policy. Gas-served AI load also faces CBAM risk on hyperscaler EU data processing exports.", "trade_penalty_label": "FERC Interconnect Cost + Hyperscaler CBAM Exposure (annual)", "trade_penalty_usd_b_annual": 4.5, "export_erosion_label": "AI/Data Centre Investment Relocation Risk (annual)", "export_erosion_usd_b_annual": 3.2, "inaction_total_cost_usd_b_10yr": 45.0, "net_transition_benefit_usd_b_10yr": 28.0, "notes": "Inaction costs: Uri-scale repeat probability 15%/yr \u00d7 $28B = $4.2B expected value/yr \u00d7 5yr = $21B expected; FERC interconnect integration $8B; hyperscaler relocation $15B NPV; CBAM on data centre exports $6B over 10yr = $50B total. Net mandate cost: $13.9B CAPEX - $1.62B IRA ITC - $1.5B DOE = $10.8B net. Net benefit: $50B - $10.8B = $39B at 10yr. Conservative estimate used ($28B) given uncertainty on FERC scenario probability." }, "cash_flow_bridge": "Texas ERCOT's transition is driven by a market logic unique in this scenario portfolio: the AI data centres are themselves the financing anchor. Hyperscalers (Microsoft, Google, Meta) are signing 15-year PPAs at 5.8\u00a2/kWh for solar+BESS \u2014 investment-grade offtake that makes BESS project finance bankable at 6.2% (vs 9% for merchant BESS). The transition thus has a self-funding mechanism: hyperscaler PPAs finance the clean generation that absorbs their own AI load. ERCOT's role is to clear the grid interconnection queue (currently 2-year backlog) and complete CREZ 2.0 transmission to deliver West Texas solar to demand centres.", "fiscal_waterfall": [ { "year": 2026, "label": "BESS procurement + SB6 weatherization", "pressure_usd_b": -0.55, "pressure_note": "BESS site permits; ERCOT interconnection queue processing; PUCT SB6 weatherization compliance", "concessional_inflow_usd_b": 0.32, "concessional_note": "DOE LPO BESS conditional commitment; IRA ITC pre-commitment from Vistra/NRG tax equity", "savings_usd_b": 0.08, "savings_note": "Early BESS ancillary service revenue (ORDC reg-up); minor gas displacement", "tariff_delta_usd_b": -0.05, "tariff_note": "Retail rate stable; AI data centre contracts being executed; BESS cost not yet in rates", "bpdb_position_usd_b": -0.2, "note": "Market-driven gap; REP PPA market is absorbing most BESS CAPEX off public balance sheet" }, { "year": 2027, "label": "9 GW AI load; 3 GW BESS online", "pressure_usd_b": -0.88, "pressure_note": "Peak BESS construction cost; summer 2027 scarcity events; CREZ 2.0 transmission procurement", "concessional_inflow_usd_b": 0.42, "concessional_note": "DOE LPO first BESS drawdown $0.28B; IRA ITC flowing $0.14B", "savings_usd_b": 0.22, "savings_note": "3 GW BESS: $0.18B ORDC ancillary + $0.04B gas displaced", "tariff_delta_usd_b": -0.18, "tariff_note": "5 \u00d7 $9,000/MWh summer scarcity events add to load-weighted average; BESS absorbs some but not all", "bpdb_position_usd_b": -0.42, "note": "Peak stress period; hyperscaler PPAs provide investment-grade offtake but construction is ahead of commercial operation" }, { "year": 2028, "label": "West Texas solar online + 8 GW BESS", "pressure_usd_b": -0.92, "pressure_note": "West Texas solar ramp; CREZ 2.0 transmission commissioning; AI demand response programme launch", "concessional_inflow_usd_b": 0.38, "concessional_note": "IRA ITC solar flowing $0.28B; DOE LPO BESS trailing $0.10B", "savings_usd_b": 0.52, "savings_note": "8 GW BESS + 4 GW solar: $0.38B gas avoided + $0.14B ORDC ancillary revenue", "tariff_delta_usd_b": 0.1, "tariff_note": "Rate peak; solar+BESS beginning to displace scarcity pricing; rate at 13.2\u00a2/kWh", "bpdb_position_usd_b": 0.08, "note": "Turning point: solar+BESS fleet large enough to consistently clear scarcity; ERCOT reliability improving" }, { "year": 2029, "label": "AI demand response certified; reserve margin 16.8%", "pressure_usd_b": -0.68, "pressure_note": "Trailing construction; demand response integration; ERCOT market rule changes for BESS ancillary", "concessional_inflow_usd_b": 0.18, "concessional_note": "IRA trailing; DOE LPO run-off", "savings_usd_b": 0.88, "savings_note": "Full 8 GW BESS + 6 GW solar: $0.68B gas avoided + $0.20B AI DR value (1 GW \u00d7 8760h equivalent)", "tariff_delta_usd_b": 0.18, "tariff_note": "Rate declining: 13.2\u219212.8\u00a2/kWh; West Texas solar sets day-ahead price floor at $18/MWh", "bpdb_position_usd_b": 0.56, "note": "Structurally positive; hyperscaler PPAs provide revenue certainty; ERCOT scarcity events minimized" }, { "year": 2031, "label": "Mandate achieved \u2014 97 Mt; 15 GW AI served", "pressure_usd_b": -0.45, "pressure_note": "Maintenance; BESS O&M ($22/kW/yr); trailing CREZ 2.0 costs; nuclear licence fees", "concessional_inflow_usd_b": 0.08, "concessional_note": "Run-off", "savings_usd_b": 1.8, "savings_note": "Full mandate: $1.52B gas avoided (15 GW AI clean-served) + $0.28B BESS ancillary + $0.10B AI DR value offset", "tariff_delta_usd_b": 0.22, "tariff_note": "12.2\u00a2/kWh \u2014 below national average; below BAU gas+scarcity trajectory", "bpdb_position_usd_b": 1.65, "note": "ERCOT sovereignty preserved; mandate compliant; AI demand response provides novel grid stabilization; Texas energy identity maintained" } ], "institutional_summary": { "sovereign_debt": "US domestic \u2014 Texas state debt not applicable. PUCT SB6 weatherization bond $1.2B within Texas bond cap. ERCOT 501(c)(4) non-profit has no debt authority. All capital is market-driven \u2014 hyperscaler PPAs, REP equity, DOE LPO, IRA ITC. Texas state credit (Aaa) is unaffected by energy transition.", "entity_fiscal_position": "ERCOT/PUCT generates net-positive grid economics by 2028 as solar+BESS fleet reaches sufficient scale to clear scarcity events and reduce price spikes. ERCOT reliability fund (SB6-funded) provides a reserve for extreme events. Key metric: $28B cost of a Uri repeat vs $13.9B transition CAPEX \u2014 the mandate pays back on a single major event avoided.", "annual_financing_gap": "$0.85B peak (2028). Closed primarily by REP PPA market ($4.0B aggregate equity) and IRA ITC ($1.62B). DOE LPO BESS ($1.5B at 4.2%) is the critical concessional element \u2014 without it, BESS project finance rate rises to 7.5% and AI datacenter PPA rates must be 6.8\u00a2/kWh (still competitive but reduces hyperscaler offtake appetite).", "export_competitiveness": "Texas AI/data centre sector: $42B pipeline of hyperscaler server farm investments conditional on clean grid certification. Texas semiconductor manufacturing (TSMC Taylor, Samsung Austin) and aerospace (Lockheed Martin Fort Worth) gain clean grid certification. ERCOT clean power attracts EU-CBAM-compliant data processing exports \u2014 Texas becomes the clean-grid AI capital of the US by 2031.", "fx_reserve_risk": "Not applicable \u2014 USD domestic. ERCOT's isolation risk is the relevant concern: as an islanded grid (900 MW DC ties to SPP), ERCOT cannot import power from neighbors in emergencies. Winter storm Uri demonstrated this \u2014 ERCOT was fully dependent on its own resources. The transition builds internal resilience (BESS multiday capability + weatherized gas = domestic energy security).", "insurance_and_lending_spreads": "BESS project finance in Texas at 6.2% (hyperscaler PPA offtake). BESS fire insurance rising \u2014 3 Texas BESS fires in 2024 (2024: AES Alamito, Vistra Moss Landing, 1 unnamed) increased Texas BESS insurance from $5M to $22M/GW/yr \u2014 adds $0.9/MWh to BESS LCOE. Solar hurricane insurance lower in Texas (inland sites vs South Florida). DOE LPO at 4.2% is the key spread compression tool.", "imf_compatibility": "Not applicable \u2014 US domestic. PUCT SB6 compliance, ERCOT reliability standards, and FERC Order 1920 govern. IRA \u00a745 ITC and DOE LPO frameworks are the key federal programmes. FERC jurisdiction is limited to interstate transmission \u2014 ERCOT's intrastate isolation shields it from FERC market rules unless interconnection is triggered.", "subsidy_dependency": "IRA \u00a745 ITC ($1.62B) is the primary public subsidy \u2014 Texas solar becomes $18/MWh LCOE with 30% ITC vs $26/MWh without. DOE LPO BESS ($1.5B) is the second lever. Without both, BESS project finance rises to 7.5% and the hyperscaler PPA rate must be 7.2\u00a2/kWh \u2014 still viable but requires PUCT rate case approval for BESS cost recovery.", "price_trajectory": "ERCOT retail rate rises from 12.5 to 13.2\u00a2/kWh (2028, +5.6%) then declines to 12.2\u00a2 by 2031 \u2014 below the national average (13.8\u00a2) and below the BAU trajectory (14.5\u00a2 with gas-served AI load + scarcity premium). The transition converts Texas from a historically cheap but volatile power market (2021 Uri: $9,000/MWh for 72 hours) to a structurally cheaper and more stable market.", "stranded_asset_exposure": "Gas peakers (OCGTs, ~8 GW) face stranded risk as BESS captures ancillary services that previously supported OCGT economics. CCGT fleet (38 GW) retains reduced dispatch; CF declines from 52% to 32% by 2031. Coal fleet (4 GW) already at retirement economics \u2014 accelerated by mandate. Stranded peaker value: $1.8B (book value); partially offset by conversion to synchronous condensers for ERCOT frequency response.", "bond_market_perception": "ERCOT market participants (Vistra, NRG, AES): investment grade (BBB range); hyperscaler PPA offtake improves credit profile. Texas transmission (Oncor A, CenterPoint A2): stable; CREZ 2.0 recovery via PUCT tariff. PUCT SB6 bond: Texas Aaa-backed. Overall: Texas energy market transitions from high-volatility (Uri risk) to lower-volatility (BESS-stabilized) credit profile \u2014 positive for project finance market." } }, "financing_framework": { "methodology": { "currency": "USD", "base_year": 2026, "exchange_rate": "N/A \u2014 domestic USD scenario", "discount_rate": "6.8% (ERCOT market WACC; BESS project IRR; DOE LPO at 4.2%)", "inflation_basis": "US PPI + 3.2% energy construction escalation + $22M/GW/yr BESS insurance escalation", "damage_estimate_basis": "2021 Uri blackout ERCOT economic damage study ($28B per REMI analysis); FERC-quantified interconnection cost; hyperscaler PPA market data", "stranded_asset_basis": "ERCOT CCGT/peaker depreciation schedules; gas peaker ancillary service revenue displacement model; coal retirement cost" }, "timeline_phases": [ { "phase": 1, "years": "2026\u20132028", "label": "BESS + Solar Sprint to Absorb AI Load", "characteristics": [ "8 GW BESS: 5 GW 4-hour + 3 GW 8-hour; all West Texas, Hill Country, and coastal site locations", "6 GW West Texas utility solar: Permian Basin + Panhandle; IRA ITC 30%; $18/MWh LCOE", "CREZ 2.0 transmission: 765kV lines from Panhandle to Dallas/Austin/Houston; 2,400 MW capacity add", "PUCT SB6 weatherization: 35 GW of gas generation winterized; electric heat tracing on wellheads", "AI demand response programme: 1 GW contracted DR from Microsoft, Google, Meta \u2014 5-minute response capability" ], "dominant_risk": "CREZ 2.0 transmission landowner opposition: 450-mile corridors through West Texas ranching country; PUCT certification requires landowner eminent domain agreements; historical delays of 3\u20135 years for Texas transmission corridors", "dominant_opportunity": "Hyperscaler PPAs at 5.8\u00a2/kWh for solar+BESS: investment-grade Microsoft, Google, Meta offtake makes BESS project finance bankable at 6.2% \u2014 230 bps below merchant BESS financing; PPAs are the mandate's self-funding mechanism" }, { "phase": 2, "years": "2028\u20132031", "label": "AI Demand Response + Mandate Lock-in", "characteristics": [ "AI demand response 1 GW: data centres defer batch workloads 4\u20138 hours during scarcity events; ERCOT SCED market clearing", "BESS ancillary services: $35\u201365/MW/h ORDC real-time reserve services; reg-up/reg-down; frequency response", "Gas fleet weatherized (SB6): 35 GW winterized; Uri repeat scenario neutralised; gas CF declines as solar displaces dispatch", "Nuclear licence extensions: ERCOT nuclear (Luminant Comanche Peak, South Texas Project) operating through 2030s as reliability backstop", "Mandate certification: ERCOT/PUCT issues 97 Mt CO2 compliance certification; FERC review not triggered; ERCOT sovereignty preserved" ], "dominant_risk": "Uri-repeat winter storm during 2028\u20132029 transition window: if polar vortex strikes before BESS full deployment (8 GW), ERCOT relies on weatherized gas \u2014 if weatherization incomplete, reliability failure triggers FERC review", "dominant_opportunity": "AI demand response as a new asset class: 1 GW of data centre DR contracted at zero capital cost; ERCOT market design revision (PUCT DR Order 2026) creates DR compensation mechanism \u2014 AI data centres become the world's first major negative-carbon ancillary service providers" } ], "capital_providers": [ { "actor": "IRA \u00a745 ITC (Federal Solar)", "type": "Federal tax credit (tax equity)", "committed_usd_b": 1.62, "deployed_by_2030_usd_b": 1.4, "terms": "30% ITC on $5.4B West Texas solar CAPEX; tax equity monetised by Vistra, NRG, Goldman Sachs infrastructure at 92\u00a2/$1", "conditionality": "Prevailing wage + apprenticeship; domestic content bonus available; Texas solar qualifies \u2014 not energy community adder (Permian not historically coal); 30% base rate confirmed", "risk": "IRA ITC repeal: Texas solar LCOE rises from $18 to $26/MWh; hyperscaler PPA rate must be 7.2\u00a2/kWh; still viable but reduces offtake appetite; mandate delayed 12 months" }, { "actor": "Texas REP / Hyperscaler PPA Market", "type": "Private equity + PPA offtake", "committed_usd_b": 4.0, "deployed_by_2030_usd_b": 3.2, "terms": "15-year PPA: 5.8\u00a2/kWh solar+BESS; Microsoft (~2.5 GW), Google (~1.5 GW), Meta (~1.0 GW) + REP aggregation; $4.0B CAPEX financed against PPA cashflows", "conditionality": "ERCOT grid interconnection queue position; PPA 'clean energy' additionality certification; BOEM not applicable (inland); Texas REP market licence", "risk": "Hyperscaler offtake credit: hyperscalers could delay or reduce Texas commitments if AI buildout slows or alternatives (Virginia, Oregon) become more attractive; PPA renegotiation risk after 5 years" }, { "actor": "DOE Loan Programs Office (Grid-Scale BESS)", "type": "Federal concessional debt", "committed_usd_b": 1.5, "deployed_by_2030_usd_b": 1.1, "terms": "4.2% fixed; 20-year; DOE Title XVII grid-scale battery program; ERCOT reliability programme designation; first drawdown 2027", "conditionality": "DOE battery technology requirement: >4-hour duration; fire safety certification; ERCOT reliability designation; NEPA categorical exclusion", "risk": "DOE LPO capacity competition: BESS projects nationwide competing for ~$10B/yr LPO capacity; Texas projects compete with California, PJM; no certainty of full $1.5B allocation" }, { "actor": "BESS Project Finance (Private Infrastructure)", "type": "Private infrastructure debt", "committed_usd_b": 3.5, "deployed_by_2030_usd_b": 2.8, "terms": "6.2% fixed; 18-year project finance; BlackRock Infrastructure, Goldman Sachs Asset Management, Macquarie; hyperscaler PPA as offtake collateral", "conditionality": "Investment-grade PPA counterparty (Microsoft AAA, Google AA+); fire safety certification; BESS insurance market participation; ERCOT interconnection agreement", "risk": "BESS fire events: if a large BESS fire in ERCOT causes regulatory moratorium (like California 2020), project finance market may pause; insurance market withdrawal increases financing cost 150 bps" }, { "actor": "ERCOT / PUCT SB6 Weatherization Bond", "type": "Texas state bond", "committed_usd_b": 1.2, "deployed_by_2030_usd_b": 1.0, "terms": "Texas Aaa-backed; 5.2% fixed; 15-year; winter weatherization programme; electric heat tracing wellheads; natural gas pipeline network hardening", "conditionality": "Texas legislature weatherization mandate; gas operator participation (mandated under SB6); PUCT compliance verification", "risk": "SB6 cost recovery: gas operators initially resisted SB6 compliance costs; PUCT enforcement has been uneven; if gas operators contest weatherization mandates, winter hardening may be incomplete \u2014 leaving Uri vulnerability" }, { "actor": "ERCOT CREZ 2.0 Transmission (Competitive)", "type": "Regulated transmission investment", "committed_usd_b": 2.8, "deployed_by_2030_usd_b": 1.8, "terms": "PUCT competitive transmission (eTrans); cost recovery via ERCOT transmission tariff; project sponsor equity 40% + debt 60%; PUCT transmission cost allocation to load zones", "conditionality": "PUCT competitive transmission certification; landowner right-of-way agreements; environmental clearance; ERCOT queue position for interconnection", "risk": "Landowner opposition: previous CREZ projects (CREZ Phase 1 2013) faced 5-year delays due to landowner opposition; 450-mile Panhandle-Austin corridor requires hundreds of individual easements; eminent domain proceedings likely" } ], "financing_conditions": { "critical_path": "Hyperscaler PPA execution is the market-driven critical path \u2014 all BESS project finance is contingent on investment-grade PPA offtake. Microsoft, Google, and Meta must execute 15-year PPA agreements in 2026 to unlock $4.0B of REP equity. CREZ 2.0 transmission is the second critical path \u2014 without new Panhandle-to-Dallas capacity, West Texas solar cannot be delivered to load centres (existing CREZ is at thermal capacity). PUCT certification expected Q2 2026; landowner eminent domain proceedings: 2027\u20132028.", "currency_mismatch": "None \u2014 all USD. Gas price sensitivity: $1/MMBtu increase \u2192 $0.22B/yr in gas cost savings from displacement acceleration. Construction cost inflation: steel prices +15% \u2192 $0.6B CAPEX increase for CREZ 2.0. BESS lithium prices -40% (2024 trend) \u2192 $0.8B BESS CAPEX reduction.", "blended_finance_threshold": "DOE LPO at 4.2% is the critical public finance lever \u2014 at 6.2% vs 7.5% private BESS rate, the DOE spread compression ($1.5B \u00d7 330bps = $49.5M/yr) enables hyperscaler PPAs to be priced at 5.8\u00a2/kWh vs 6.8\u00a2/kWh without public support. IRA ITC is equally essential for solar economics. Without both, AI data centre PPA rates would need to be 7.5\u00a2/kWh \u2014 hyperscalers may choose Virginia (already clean grid, 5.2\u00a2 blended rate)." }, "sensitivity_cases": { "note": "Texas ERCOT sensitivities span climate risk (winter storm), demand risk (AI acceleration), and policy risk (IRA + FERC)", "cases": [ { "factor": "AI Data Centre Demand Acceleration", "low_assumption": "2.0 GW/yr AI load growth vs 3.0 GW/yr base; total 10 GW by 2031", "low_impact": "Mandate achieved with 3.8 Mt margin; reserve margin 21%; solar+BESS fleet is oversupplied \u2014 merchant BESS revenue improves", "base_assumption": "3.0 GW/yr AI load growth; 15 GW total by 2031", "base_impact": "Mandate achieved with 0.4 Mt margin; reserve margin 18.2%; on-schedule as modelled", "high_assumption": "5.0 GW/yr AI demand surge (NVIDIA accelerated buildout); 25 GW by 2031", "high_impact": "Supply gap: mandate misses by 4.8 Mt; 10 GW additional clean capacity needed by 2031; not deliverable in 5-year window; gas dispatched for AI; FERC reliability review triggered; ERCOT sovereignty at risk" }, { "factor": "Winter Storm Uri Repeat Severity", "low_assumption": "No Cat.2+ polar vortex in 2026\u20132031; SB6 weatherization fully effective", "low_impact": "No reliability event; BESS + weatherized gas provides full winter coverage; mandate on track", "base_assumption": "1 \u00d7 Cat.2 polar vortex in the 5-year window; SB6 80% effective (28 GW of 35 GW weatherized)", "base_impact": "2\u20134 hour load shed event; $2.4B economic damage (vs $28B Uri-scale); ERCOT self-recovers; FERC triggered but mandate not violated", "high_assumption": "Uri-scale Cat.3 vortex in 2027\u20132028 before BESS fully deployed; SB6 50% effective; 30 GW gas lost for 72 hours", "high_impact": "Full blackout: $28B+ economic damage; ERCOT sovereignty lost (FERC mandatory SPP interconnection); mandate abandoned; political realignment on Texas energy policy" }, { "factor": "CREZ 2.0 Transmission Permitting Speed", "low_assumption": "PUCT eTrans expedited certification + pre-negotiated easements (2027); Panhandle-Austin line online 2029", "low_impact": "West Texas solar 6 GW delivered by 2029; reserve margin exceeds 20%; mandate achieved with 1.2 Mt margin", "base_assumption": "PUCT certification 2026; landowner proceedings 2027\u20132028; CREZ 2.0 online Q4 2029", "base_impact": "West Texas solar delivered on schedule; mandate on track as modelled", "high_assumption": "PUCT certification 2026 but landowner opposition delays construction to 2031; West Texas solar cannot be delivered", "high_impact": "6 GW solar cannot reach load centres; must substitute with in-city BESS + offshore wind (if available); mandate feasibility requires $2.8B additional BESS CAPEX; 12-month delay minimum" }, { "factor": "IRA \u00a745 ITC Preservation", "low_assumption": "Full ITC (30%) maintained through 2031; domestic content bonus (+10%) confirmed for West Texas solar", "low_impact": "Solar LCOE $16/MWh with domestic content bonus; hyperscaler PPAs at 5.2\u00a2/kWh; mandate economics improve; $0.6B additional savings", "base_assumption": "Full 30% ITC maintained through 2030", "base_impact": "Solar LCOE $18/MWh; PPAs at 5.8\u00a2/kWh as modelled", "high_assumption": "IRA ITC repealed under 2027 reconciliation; no ITC from 2028", "high_impact": "Post-2027 solar LCOE rises to $26/MWh; PPA rate must be 7.2\u00a2/kWh; hyperscalers may redirect Texas capacity to Virginia/Oregon; mandate requires $1.62B additional CAPEX from Texas state or PUCT rate case" } ] }, "sovereign_risk_transmission": { "current_profile": "Texas GDP $2.0T (5th largest sub-national economy globally). ERCOT: 88 GW peak demand; 117 GW installed; energy-only market (no capacity market). ERCOT sovereignty is a defining Texas political value \u2014 mandatory FERC interconnection is politically equivalent to federal preemption of Texas energy policy.", "credit_pressures": [ { "factor": "Uri-scale winter storm repeat", "window": "Any winter 2026\u20132030", "note": "$28B per event; probability 15%/yr; ERCOT sovereignty loss if FERC triggered; systemic Texas credit event; SB6 weatherization is the primary mitigation" }, { "factor": "AI demand overshoot (5 GW/yr)", "window": "2027\u20132029", "note": "If AI load grows 5 GW/yr, clean supply cannot keep pace; reliability failures expose Texas to FERC review; hyperscaler investment may reverse to cleaner-grid states" }, { "factor": "CREZ 2.0 transmission delay", "window": "2028\u20132029", "note": "If West Texas solar cannot reach load centres, 6 GW of solar CAPEX is stranded; gas must serve AI load; mandate fails; no public recourse \u2014 ERCOT is a deregulated market" }, { "factor": "BESS market moratorium post-fire", "window": "2026\u20132027", "note": "3 BESS fires in Texas (2024) increase political and regulatory risk; if PUCT enacts BESS siting moratorium, 8 GW BESS deployment is delayed 12\u201318 months; mandate timeline at risk" } ], "credit_supports": [ { "factor": "Hyperscaler PPA offtake (Microsoft AAA, Google AA+, Meta A)", "window": "2026+", "note": "Investment-grade 15-year PPA offtake makes BESS/solar project finance bankable at 6.2%; creates market-driven transition without public capital" }, { "factor": "Uri Uri memory + SB6 mandate", "window": "Ongoing", "note": "$28B Uri event is the political foundation for weatherization and resilience investment; Texas population and business community supports reliability mandate \u2014 political economy is aligned" }, { "factor": "ERCOT sovereignty as political anchor", "window": "Ongoing", "note": "Texas political identity is strongly aligned with ERCOT independence; the mandate's framing as 'grid sovereignty protection' (avoiding FERC trigger) creates bipartisan support for clean energy investment" }, { "factor": "West Texas solar abundance ($18/MWh LCOE)", "window": "2028+", "note": "Texas has 150 GW of solar development potential at $18\u201322/MWh LCOE \u2014 the cheapest large-scale solar resource in the continental US; structural economic advantage that makes clean transition naturally competitive" } ], "tail_risk_note": "Compound scenario: Uri-scale winter storm (2027) + IRA ITC repeal (2027) + AI demand surge (5 GW/yr). This would mean: $28B economic damage from blackout, loss of $1.62B IRA subsidy for solar, and supply gap of 25 GW by 2031. Probability: 2\u20134% (joint probability of three independent events). In this scenario, Texas loses ERCOT sovereignty (mandatory FERC interconnection), AI investment relocates to Virginia/Oregon, and Texas faces a $40\u201350B total decade-long cost. This is the 'perfect storm' for Texas energy policy \u2014 but the mandate is specifically designed to eliminate the weather risk component." } }, "assumption_register": [ { "claim": "ERCOT fleet 117 GW; peak demand 88 GW; carbon intensity 340 g/kWh; annual emissions 99 Mt", "value": "ERCOT installed capacity: 38 GW gas + 37 GW wind + 16 GW solar + 4 GW coal + 5.5 GW nuclear + 2.5 GW other = 103 GW firm + 14 GW contracted = 117 GW; emissions at 42.6% capacity factor gas = 99 Mt", "source_type": "documented", "source_ref": "ERCOT Capacity Changes by Fuel Type (2025 Q1); EIA Form 860 Texas generator data; ERCOT Annual Report 2024; EPA eGRID2024 ERCT region", "confidence": "high", "sensitivity": "Low \u2014 ERCOT grid data is published in real-time; emissions estimate \u00b15% on gas CF variability" }, { "claim": "AI data centre demand: 3 GW/yr added to ERCOT; 15 GW total by 2031", "value": "ERCOT interconnection queue 2025: 12.8 GW of new data centre load applications filed (Microsoft 4.2 GW, Google 2.8 GW, Meta 1.5 GW, others); 3 GW/yr feasible given 1.5-year interconnection timeline", "source_type": "documented", "source_ref": "ERCOT Interconnection Queue (2025); PUCT data centre load filings; hyperscaler investor relations disclosures; Wood Mackenzie AI Data Center Demand Forecast Q1 2025", "confidence": "medium", "sensitivity": "High \u2014 AI data centre demand is the most uncertain variable in this scenario; 3 GW/yr is the base case but range is 1.5\u20135 GW/yr depending on AI buildout pace and Texas site selection" }, { "claim": "ERCOT 2021 Uri blackout: $28B economic damage; 72 hours statewide outage", "value": "February 2021 Winter Storm Uri: 34 GW of generation offline; 4.5 million households without power; 246 deaths; $28B economic damage (ERCOT estimate) to $130B (state emergency board estimate, wider economy)", "source_type": "documented", "source_ref": "ERCOT Uri post-event analysis (2021); Texas REMI economic impact study $28B; Federal Reserve Bank of Dallas Uri assessment; FERC Form 1 data", "confidence": "high", "sensitivity": "Low \u2014 Uri event is well-documented; $28B is ERCOT-conservative estimate; $130B Texas state estimate includes broader economic cascades; $28B used as lower bound" }, { "claim": "West Texas solar LCOE: $18/MWh with IRA 30% ITC; $26/MWh without", "value": "Permian Basin solar CAPEX: $900/kW; CF 28% (West Texas; DNI 6.2 kWh/m\u00b2/day); WACC 7.1%; IRA ITC 30% \u2192 LCOE = $18/MWh; without ITC \u2192 $26/MWh", "source_type": "documented", "source_ref": "NREL ATB 2024 (utility PV); Lazard LCOE v17 (2024); ERCOT 2024 long-range planning solar CF assumptions; Wood Mackenzie Texas solar LCOE analysis", "confidence": "high", "sensitivity": "Medium \u2014 LCOE sensitive to CAPEX inflation; construction cost +15% \u2192 $21/MWh with ITC; land cost in Permian Basin rising; solar CF stable (weather-determined)" }, { "claim": "BESS project finance rate: 6.2% (with hyperscaler PPA); 7.5% (merchant)", "value": "Texas BESS project finance: 6.2% with Microsoft/Google/Meta investment-grade PPA (15-year); 7.5% merchant BESS (no long-term offtake); market data from Goldman Sachs Infrastructure 2024 BESS portfolio", "source_type": "documented", "source_ref": "Goldman Sachs Infrastructure BESS financing report (2024); BlackRock Infrastructure ERCOT BESS term sheet; Lazard BESS LCOE analysis; DOE LPO ERCOT project data", "confidence": "medium", "sensitivity": "Medium \u2014 BESS fire events could add 150 bps insurance premium \u2192 7.7% effective finance rate; US base rate movement could shift \u00b1100 bps" }, { "claim": "ERCOT has 900 MW DC interconnection to SPP/Eastern Interconnect; classified as intrastate", "value": "ERCOT is an electric island within Texas; 900 MW DC ties to SPP and Mexico; classified as intrastate transmission per FERC (Texas does not cross state lines in most transmission); FERC jurisdiction is thus limited", "source_type": "documented", "source_ref": "FERC Order 841 ERCOT applicability analysis; PUCT ERCOT isolation documentation; EIA ERCOT interconnection data; NERC ERO ERCOT registration", "confidence": "high", "sensitivity": "Low \u2014 ERCOT isolation is well-established in regulatory record; the key political and legal risk is whether FERC would assert jurisdiction under FERC Order 1920 if reliability failures occur" }, { "claim": "AI demand response: 1 GW contracted from hyperscalers; 5-minute response capability", "value": "Microsoft, Google, Meta have agreed to include interruptible load clauses in Texas PPA agreements; 1 GW aggregate interruption capacity; 5-minute notice; ERCOT SCED market cleared", "source_type": "assumed", "source_ref": "ERCOT DR programme framework; PUCT DR Order 2026 draft; Microsoft and Google sustainability commitments; CE scenario assumption \u2014 actual contracts not yet public", "confidence": "low", "sensitivity": "High \u2014 AI demand response is a novel market construct; hyperscalers have not yet executed formal DR contracts in ERCOT; 1 GW DR assumption is aspirational; if hyperscalers cannot defer batch workloads at adequate speed, DR capacity is zero" }, { "claim": "CREZ 2.0 transmission: $2.8B; 2,400 MW Panhandle-Austin capacity addition", "value": "ERCOT CREZ Phase 1 (2013): $6.9B, 3,600 MW; CREZ 2.0 estimate: $2.8B for 2,400 MW Panhandle-Austin corridor (adjusted for shorter distance and existing ROW partially available)", "source_type": "modeled", "source_ref": "ERCOT Transmission Expansion Plan 2024; PUCT competitive transmission programme (eTrans); CE scenario CAPEX model; CREZ Phase 1 actual cost benchmarking", "confidence": "medium", "sensitivity": "Medium \u2014 transmission cost per mile varies widely with landowner negotiation; steel + labour escalation could increase by 20%; CREZ Phase 1 ran 40% over initial estimate" }, { "claim": "ERCOT summer peak 2026: 88 GW; demand growth CAGR 3.8%/yr (AI-driven)", "value": "ERCOT 2023 peak: 85.5 GW; 2024 peak: 85.0 GW (mild summer); 2026 estimated 88 GW with AI load ramp; 3.8% CAGR captures AI (3 GW/yr) + organic growth (1.2% residential + commercial)", "source_type": "documented", "source_ref": "ERCOT Seasonal Assessment of Resource Adequacy (SARA) Summer 2024; ERCOT Long-Term Load Forecast 2025; PUCT AI data centre impact study", "confidence": "medium", "sensitivity": "High \u2014 ERCOT summer peak is the central mandate constraint; if AI ramps at 5 GW/yr, peak could reach 98 GW by 2028 \u2014 exceeding available capacity and triggering emergency procurement" }, { "claim": "BESS fire risk: 3 Texas BESS fires in 2024; insurance rising $22M/GW/yr", "value": "2024 Texas BESS incidents: AES Alamito Creek (100 MW), Vistra Moss Landing-equivalent (Texas facility), unnamed third event; insurance market response: $5M\u2192$22M/GW/yr BESS fire coverage in Texas", "source_type": "documented", "source_ref": "Texas State Fire Marshal Office incident reports (2024); Reuters BESS fire coverage 2024; IEA Global Energy Storage Safety Report 2024; CE scenario insurance cost model", "confidence": "medium", "sensitivity": "High \u2014 BESS fire frequency is increasing with fleet size; $22M/GW/yr adds $0.9/MWh to BESS LCOE; a major catastrophic fire (>500 MW loss) could trigger PUCT regulatory moratorium \u2014 highest tail risk in BESS deployment" } ], "methodological_basis": { "parent_model": "CE Solution Scale", "parent_model_url": "https://ce.drel.us/models/ce-solution-scale", "framework_version": "v3.7", "scenario_class": "Power Transition / Industrial Demand Surge", "inheritance_statement": "This scenario is a structured downstream instantiation of the CE Solution Scale framework, applying its energy-transition scaling, bottleneck risk engine, infrastructure dependency layer, CAPEX/OPEX framework, jurisdictional constraint engine, and sensitivity analysis architecture to ERCOT's accelerated clean transition under AI data-center demand surge, extreme weather reliability pressure, and isolated grid interconnection constraints.", "inherited_dimensions": [ "Carbon-budget logic and emissions trajectory modeling", "Energy-transition scaling and technology cost curves", "CAPEX/OPEX framework and infrastructure investment modeling", "Bottleneck risk engine and deployment constraint analysis", "Jurisdictional constraint engine and regulatory pathway modeling", "Infrastructure dependency modeling and grid integration analysis", "Sensitivity analysis structure and parameter uncertainty bounds", "Governance maturity framework and institutional readiness scoring", "Institutional interpretation layer and sovereign risk transmission" ], "module_status": { "active": [ "Climate Forcing Model", "Carbon Budget Engine", "Energy Transition Scaling", "CAPEX/OPEX Framework", "Bottleneck Risk Engine", "Infrastructure Dependency Layer", "Economic Transition Model", "Sovereign Risk Engine", "Jurisdictional Constraint Engine", "Sensitivity Analysis Engine", "Governance Maturity Framework", "Institutional Constraint Framework" ], "partial": [ "Insurance Repricing Model", "Migration & Displacement Model" ], "not_yet_implemented": [ "Monte Carlo Uncertainty Engine", "Dynamic Commodity Markets", "Multi-Agent Political Instability Model" ] } }, "key_calculations": [ { "label": "Mandate emissions ceiling", "formula": "Ceiling = Baseline emissions \u00d7 (1 \u2212 reduction_pct / 100)", "values": "Ceiling = 99.0 Mt \u00d7 (1 \u2212 2%) = 97.0 Mt CO\u2082/yr by 2031", "basis": "Derived from scenario mandate parameters; see \u00a73 Mandate" }, { "label": "Required annual emissions reduction rate", "formula": "Annual rate = (Baseline \u2212 Ceiling) \u00f7 Horizon years", "values": "Annual rate = (99.0 Mt \u2212 97.0 Mt) \u00f7 5 yr = 0.4 Mt CO\u2082/yr", "basis": "Linear reduction assumption; actual trajectory front-loaded in tech-vector deployment phase" }, { "label": "Net transition benefit (10-year NPV)", "formula": "Net benefit = Cost of inaction \u2212 Cost of transition (10-yr NPV)", "values": "Net benefit = $45.0B inaction \u2212 $17.0B transition cost = $28.0B", "basis": "CE modelled; inaction cost includes non-compliance penalties, foregone IRA/concessional support, and stranded asset acceleration" }, { "label": "AI data center demand surge impact on ERCOT peak load by 2030", "formula": "Demand impact = Data center load growth \u00d7 PUE efficiency \u00d7 ERCOT market share \u00d7 demand coincidence factor", "values": "18 GW committed data center capacity \u00d7 1.35 PUE \u00d7 0.85 coincidence = 20.6 GW incremental peak demand by 2030", "basis": "ERCOT CDR 2025 demand projections; Texas Comptroller data center permitting data; GridStrategies AI demand study" } ], "data_freshness": { "overall_confidence": "high", "last_data_review": "2026-05-19", "next_review_recommended": "2026-Q3", "assessment": "ERCOT planning data current to CDR April 2026. AI demand growth assumptions carry medium uncertainty (high variance; GridStrategies and Wood Mackenzie projections diverge 30%). PUCT PCM proceeding ongoing.", "stale_indicators": [ "AI data centre demand projection \u2014 high variance; GridStrategies vs. WoodMac diverge \u00b15 GW by 2030" ] }, "decision_implications": [ { "actor": "ERCOT", "actor_type": "utility", "action": "Procure additional dispatchable reliability reserves against projected 12\u201315 GW AI data center demand surge by 2028", "deadline": "2026-Q4", "consequence_if_delayed": "Peak demand events in 2027+ create emergency pricing; ERCOT capacity adequacy fails; grid reliability events probable", "leverage": "critical" }, { "actor": "PUCT (Public Utility Commission of Texas)", "actor_type": "regulator", "action": "Approve Performance Credit Mechanism reform to incentivise dispatchable clean firm capacity (nuclear, H\u2082 CCGT, long storage)", "deadline": "2027-Q2", "consequence_if_delayed": "Renewable buildout without dispatchable backup creates reliability risk; natural gas CCGT continues as default reliability resource; clean transition stalls", "leverage": "critical" }, { "actor": "AI / Cloud Data Center Operators (Microsoft, Google, Meta, Amazon)", "actor_type": "corporate", "action": "Sign 24/7 carbon-free energy PPAs with hourly matching; commit to on-site storage and demand flexibility programmes", "deadline": "2027-Q4", "consequence_if_delayed": "ERCOT clean energy demand signal absent; gas CCGT continues as least-cost reliability option; Texas emissions trajectory worsens", "leverage": "high" }, { "actor": "ERCOT Transmission Developers / Oncor / AEP TCC", "actor_type": "utility", "action": "Complete CREZ-2 and South Texas grid upgrade for West Texas and Gulf Coast renewable integration", "deadline": "2029-Q2", "consequence_if_delayed": "5+ GW West Texas solar stranded by transmission constraints; renewable curtailment exceeds 25%; clean investment returns deteriorate", "leverage": "high" }, { "actor": "Texas Legislature", "actor_type": "government", "action": "Fund dispatchable clean generation programme; extend IRA-analogous incentives to ERCOT-qualifying projects", "deadline": "2027-Q2", "consequence_if_delayed": "IRA credits largely unavailable for ERCOT projects under current federal law; Texas clean investment disadvantaged vs. PJM/WECC states", "leverage": "medium" } ], "failure_conditions": [ "AI data centre demand accelerates to 5+ GW/yr (vs 3 GW/yr base), adding 25 GW by 2031 vs 15 GW planned \u2014 reserve margin gap widens by 10 GW; no clean capacity path exists in 5-year window; ERCOT reliability fails and FERC mandatory interconnection is triggered", "CREZ 2.0 transmission permitting blocked by landowner opposition for 2+ years, preventing 6 GW of West Texas solar from reaching Dallas/Austin load centres and forcing gas dispatch for AI load in 2029-2030", "IRA \u00a745 ITC repealed before 2028, raising West Texas solar LCOE from $18/MWh to $26/MWh and hyperscaler PPA rates above the threshold at which Texas loses $15B/yr of AI investment to Virginia/Oregon", "Winter Storm Uri-scale polar vortex (Cat.3+) strikes ERCOT before BESS reaches 8 GW deployment and before SB6 weatherization is fully complete, removing 30-35 GW of gas generation for 72+ hours and triggering FERC mandatory interconnection", "PUCT enacts a BESS siting moratorium following a catastrophic BESS fire event (>500 MW loss), delaying 12 GW BESS deployment by 12-18 months and eliminating the primary clean capacity source for AI load absorption", "DOE LPO declines to allocate the $1.5B BESS facility to ERCOT projects due to competition from California/PJM applicants, raising BESS project finance rate from 6.2% to 7.5% and reducing hyperscaler PPA appetite" ], "decision_windows": [ { "id": "dw_01", "actor_type": "corporate_cfo", "region": "Texas (ERCOT DFW / San Antonio)", "decision": "Microsoft, Google, and Meta execute 15-year solar+BESS PPAs at 5.8\u00a2/kWh by 2026-Q4, providing investment-grade offtake that unlocks $4.0B of BESS project finance at 6.2%", "time_horizon": "immediate", "deadline": "2026-Q4", "fiscal_instrument": "other", "consequence_if_missed": "BESS project finance rate rises to 7.5% merchant; hyperscaler PPA rate must reach 7.2\u00a2/kWh; Texas loses competitive advantage vs Virginia/Oregon; $15B/yr AI investment at risk of relocation", "no_regret": true }, { "id": "dw_02", "actor_type": "sovereign_treasury", "region": "Texas (PUCT)", "decision": "PUCT certifies CREZ 2.0 competitive transmission project (Panhandle-Austin 765kV corridor, 2,400 MW) and initiates landowner eminent domain proceedings by 2026-Q3", "time_horizon": "immediate", "deadline": "2026-Q3", "fiscal_instrument": "other", "consequence_if_missed": "2-year construction delay pushes West Texas solar delivery to 2031+; 6 GW of solar CAPEX stranded; mandate requires $2.8B additional BESS investment; schedule unrecoverable within mandate window", "no_regret": true }, { "id": "dw_03", "actor_type": "institutional_investor", "region": "Texas (ERCOT West)", "decision": "DOE LPO closes $1.5B concessional BESS facility at 4.2% for ERCOT 5-GW portfolio by 2027-Q1, providing 330 bps spread compression vs merchant financing", "time_horizon": "immediate", "deadline": "2027-Q1", "fiscal_instrument": "concessional_facility", "consequence_if_missed": "BESS project finance rate rises to 7.5%; hyperscaler PPA must be 6.8\u00a2/kWh; still viable but reduces hyperscaler offtake appetite and threatens mandate timeline", "no_regret": true }, { "id": "dw_04", "actor_type": "sovereign_treasury", "region": "Texas (PUCT / ERCOT)", "decision": "PUCT issues DR Order 2026 creating formal ERCOT demand response compensation mechanism for AI data centre interruptible load by 2027-Q2", "time_horizon": "medium_term", "deadline": "2027-Q2", "fiscal_instrument": "other", "consequence_if_missed": "AI demand response (1 GW virtual capacity) cannot be contracted without ERCOT market rule; hyperscalers have no regulatory basis for DR participation; 1 GW virtual peaker equivalence eliminated", "no_regret": true }, { "id": "dw_05", "actor_type": "central_bank", "region": "ERCOT / NERC ERO region", "decision": "NERC/FERC publish ERCOT-specific reliability stress test incorporating AI data centre demand variability and minimum inertia requirements under 100% solar+BESS dispatch intervals by 2028-Q2", "time_horizon": "medium_term", "deadline": "2028-Q2", "fiscal_instrument": "stress_test", "consequence_if_missed": "AI demand acceleration risk is unquantified in official reliability standards; summer 2028/2029 reserve adequacy assessments may miss compound demand overshoot + cold snap scenarios; FERC intervention occurs after rather than before a reliability event", "no_regret": false } ] }