Revamping the Texas Power Grid: Insights from Rice University’s Electricity Research

Executive Summary

How we generate electricity matters to our air, our climate, our wallets, and our quality of life. Texas power plants burn more coal and natural gas and emit more air pollutants and greenhouse gases than those in any other state. But Texas also leads the nation in wind power, has the fastest growing solar sector, and is home to companies that are driving technological innovations in geothermal energy. Widespread power outages during the February 2021 freeze and price spikes during heat waves have highlighted the vulnerability of our isolated power grid to extreme weather events amid growing demand, an aging power plant fleet, an inadequately winterized natural gas supply system, and increasing reliance on variable wind and solar power.

This report synthesizes research published by Professors Daniel Cohan and James Doss-Gollin and their research groups at Rice University since 2020 and updates those studies to the latest conditions in Texas and the Electric Reliability Council of Texas (ERCOT) market. Though not a comprehensive review of Texas electricity, this report provides important insights as the state works to improve the reliability and sustainability of its power supply amid a changing climate.

The report is organized into four sections, with key highlights presented here:

Section 1 discusses how Texas power plants affect air quality, health, and climate. Although Texas environmental regulators have historically focused on reducing power plant nitrogen oxide (NO_x) emissions to address ongoing nonattainment of ozone standards in Texas cities, this section shows that power plants are even more pivotal as the source of 59% of statewide emissions of the sulfur dioxide (SO₂) that forms sulfate particulate matter (Table E1). Our review of scientific literature finds that sulfate is one of the top two components of fine particulate matter (PM_2.5) across Texas field studies, comprising 21%-57% of PM_2.5 at each site. Thus, control of power plant SO₂ emissions will be crucial to helping several Texas counties attain the more stringent proposed PM_2.5­ standards that they currently exceed and helping other counties attain SO₂ standards that they already violate (Table E2).

The need to attain these air quality standards and protect public health will likely require additional emissions controls at most Texas coal power plants, which could prompt the plants to convert to gas or retire amid competition from cheaper wind, solar, and natural gas power producers. All four remaining Texas coal plants in the Southwest Power Pool already plan to close or switch to gas by 2028. Of the 10 remaining coal plants in ERCOT, only two have announced retirement dates, but others may follow suit due to poor operational performance, challenging market conditions, and more stringent environmental regulations. The Environmental Protection Agency’s (EPA) Good Neighbor Plan will require power plant NO_x emissions to fall by half by 2028, and its proposed Regional Haze Plan for Texas will require SO₂ controls at the state’s largest coal plants — Parish and Martin Lake. Taken together, we estimate that these air quality regulations will prompt at least 8.6 gigawatts (GW) of the remaining 13.6 GW of coal plants in ERCOT to install major emissions control devices, convert to gas, or retire. Given the poor profitability of these plants and the high cost of emission control devices, retirements are the most likely outcome.

Table E1 — Power Plant Emissions in Texas in 2021 and Their Share of Texas Emissions Overall and of US Power Plants

 

Table E2 — Texas Regions That Violate the 70 ppb Ozone Standard, Counties That Would Violate EPA’s Proposed 9 or 10 µg/m³ PM_2.5 Standard (vs. Current 12 µg/m³), and Counties that Violate the 75 ppb SO₂ Standard

 

Section 2 explores the complementary roles that solar and wind can play in replacing the output of closing coal plants and meeting growing power demand in Texas. Although it is not sunny or windy all of the time, our analysis shows that it is either sunny or windy somewhere in Texas over 90% of the time. Solar power that peaks midday, together with wind power that peaks at night in West Texas and on summer afternoons with coastal sea breezes, can provide complementary sources of power most of the time. However, there is substantial variability and occasional periods when it is neither windy nor sunny (Figure E1).

In January 2020, solar and wind capacity in ERCOT were at 6.0 GW and 31.1 GW, respectively. By 2025, these numbers are projected to grow to a respective 34.9 GW and 39.7 GW, a reflection of the increasingly important role of wind and solar in easing the burden on the aging power plant fleet. This requires a paradigm shift from focusing on peak gross load to net load instead, after accounting for variable renewable generation (Figure E2). For example, whereas the burden on dispatchable resources traditionally peaked alongside demand on summer afternoons, peak net load is now shifting to the hours around sunset. We project that the wind and solar farms expected to be added in 2023 and 2024 will provide an average of 8 GW and reduce the burden on dispatchable resources by 4­–8 GW during the hours of peak net load in 2025 (Figure E2). This would more than offset projected growth in electricity demand under typical meteorology and could contribute to offsetting some of the potential loss in output from coal power plants. However, the heat wave of 2023 demonstrates that extreme heat can further raise demand, warranting analysis that is beyond the timeframe of this study.

Figure E1 — Wind and Solar Capacity Factors in ERCOT by Hour and Month from 2017–22

 

Figure E2 — Gross Load and Net Load (Subtracting Wind and Solar Output) in 2020–22 and Projected for 2025

 

Section 3 provides a brief overview of opportunities for geothermal energy in Texas as a prelude to studies that will be issued by the Cohan group later this year.

Section 4 delves into the impact of temperature extremes on heating and cooling demands, which influence peak loads on the power grid. Understanding these risks can inform exercises such as ERCOT’s seasonal assessment of resource adequacy (SARA). Although we do not explicitly link temperature to overall electricity demand, which varies due to factors like population, technology, behavior, and building types, quantifying the risk of temperature extremes is vital for anticipating peak loads and ensuring resilience of the energy system. This section highlights two main analyses.

First, the section compares the February 2021 temperature extremes in Texas to past cold snaps. Using temperature-based proxies for heating and cooling demand (akin to heating degree days or cooling degree days), we demonstrate that there have been several events of similar magnitude since 1950. Although the 2011 cold snap was used as a basis for the extreme scenario in the winter 2020–21 SARA, it was relatively minor compared to the complete historical record.

Second, the section examines how climate change has already impacted heating and cooling demands in Texas. Analyzing the same temperature-based proxies reveals that, on an annual average basis, heating demand is decreasing while cooling demand is increasing, as would be expected from a warming trend. However, weather patterns causing extreme cold are highly variable and show no clear trends. Our findings suggest that although climate change has led to increased cooling demand and decreased heating demand overall, the risk associated with major cold extremes has not diminished.

Collectively, the findings from Section 4 emphasize the inadequacy of existing methods used to assess seasonal resource adequacy, and the imperative of making better use of climate models, the complete observational record, and the changing relationship between extreme temperatures and electricity demand to ensure energy system resilience.

Figure E3 — Annually Aggregated Peak Demand for Heating (Red) and Cooling (Blue) in the Region Served by ERCOT

 

Taken together, the findings of this report show that Texas has a tremendous opportunity to transition away from the coal-fired power plants that are causing a disproportionate share of its climate-warming emissions and air pollution nonattainment challenges and adopt more affordable sources of power. However, intensifying summer heat waves, volatile winter weather, growing power demand, and inadequate transmission within and beyond ERCOT all pose challenges to electric reliability amid this transition. Simultaneous efforts to improve efficiency, expand transmission, add dispatchable resources, and foster continued growth of wind and solar will be needed to overcome these challenges and secure an affordable, sustainable, and reliable power supply for all Texans.

Section 1: Air Quality, Health, and Climate Impacts of Fossil-Fueled Power Plants

Fossil-fueled power plants, especially coal plants, are among the largest emitters of air pollutants and greenhouse gases in Texas, with serious consequences for health and climate. This section provides context on air quality in Texas and the role of power plant emissions. It also discusses how market forces and environmental regulations might lead some coal plants to close or convert to natural gas.

1.1 Texas Air Pollution Challenges: Ozone, Fine Particulate Matter, and SO₂

Air quality management in Texas has long focused mostly on ground-level ozone, since multiple regions have for decades violated federal standards for this potent respiratory irritant. This has led the state to prioritize emissions controls for nitrogen oxides (NO_x), an ozone precursor, rather than sulfur dioxide (SO₂), an air pollutant and precursor of fine particulate matter (PM_2.5 denotes particles smaller than 2.5 microns in aerodynamic diameter). State implementation plans issued by the Texas Commission on Environmental Quality (TCEQ) for ozone nonattainment regions required power plants to install control technologies such as low-NO_x burners and selective catalytic reduction to reduce their emissions of NO_x, but they neglected requirements for SO₂ controls. However, several Texas regions may soon face nonattainment of proposed new EPA standards for PM_2.5, an air pollutant that is estimated to cause far more deaths and illnesses than all other air pollutants combined. Several other counties violate SO₂ standards near coal-fired power plants.

Ozone

The Houston-Galveston-Brazoria[2] and Dallas-Fort Worth[3] regions have been in nonattainment of various federal ozone standards since at least 1990. Although peak ozone concentrations have declined since then, both regions — along with El Paso and San Antonio — have failed to attain the more stringent 70 part per billion (ppb) eight-hour ozone standard issued in 2015 (Table 1.1).[4] The design values for the regions based on 2020-22 data are 77 ppb in Dallas-Fort Worth, 81 ppb in El Paso-Las Cruces, 78 ppb in Houston, and 75 ppb in San Antonio.[5] Thus, substantial further reductions in ozone and its precursor emissions are needed. Ozone forms from atmospheric reactions involving nitrogen oxides (NO_x) and hydrocarbons, and tends to be most sensitive to NO_x emissions on the hot afternoons that determine attainment of peak ozone standards.[6] Power plants emit 10% of Texas NO_x emissions, mostly outside of nonattainment regions, and are thus a contributing but not leading cause of ozone nonattainment in Texas.

Table 1.1 — Ozone Design Values and Nonattainment Status in Texas Regions Based on 2020–22 Data

 

Fine Particulate Matter (PM_2.5)

Meanwhile, fine particulate matter (i.e., PM_2.5, denoting particles smaller than 2.5 microns in diameter) is the deadliest air pollutant[8] and causes substantial health effects — even at levels below the EPA’s 12 microgram per cubic meter (μg/m³) annual standard.[9] However, ongoing attainment of that standard by narrow margins across Texas regions has led Texas regulators to focus less on PM_2.5 and its precursors (e.g., SO₂ and ammonia) than on ozone-forming NO_x. With the World Health Organization recommending[10] a PM_2.5 standard of 5 μg/m³ and the EPA’s Clean Air Science Advisory Committee recommending 8–10 μg/m³,[11] further reductions in PM_2.5 and its precursors would help protect public health.

In January 2023, the EPA proposed tightening the annual PM_2.5 standard to a level between 9 and 10 μg/m³.[12] Based on PM_2.5 measurements reported to the EPA by TCEQ in 2020–22, Harris, Hidalgo, Kleberg, and Webb counties would violate a 10 g/m³ standard, and Bowie, Cameron, Dallas, El Paso, Tarrant, and Travis counties would also violate the standard if a 9 g/m³ limit is chosen (Table 1.2 and Figure 1.1). If those pollution levels persist, TCEQ would — for the first time — need to develop PM_2.5 attainment plans for the Houston, McAllen, Kingsville, and Laredo regions, as well as the Dallas-Fort Worth and Austin regions under a 9 g/m³ standard. Nonattainment regions typically extend beyond the county with the violating monitor, to ensure that upwind sources are adequately controlled. Thus, substantial portions of the Texas population and industrial facilities could fall within PM_2.5 nonattainment regions for the first time.

Table 1.2 — Annual PM_2.5 Design Values in Texas Counties, Based on 2020–22 Data

 

Figure 1.1 — Map of Regions That Would Have Exceeded EPA’s Proposed 9 μg/m³ or 10 μg/m³ Annual PM_2.5 Standard Based on 2019–21 Data[13]

 

It should be noted that TCEQ measures PM_2.5 in only 20 of the state’s 254 counties (Table 1.2) and at far fewer monitors than it operates for ozone. Hybrid analysis synthesizing satellite observations with modeling that the EPA included in its regulatory impact analysis (Figure 1.2) indicates that PM_2.5 may exceed 9 μg/m³ across many of the state’s most populous regions. This could lead to calls to expand the state’s PM_2.5 monitoring network if the EPA tightens the national standard.

Figure 1.2 — Annual Average PM_2.5 Concentrations Based on a Hybrid Approach Assimilating Satellite Data with Modeling[14]

 

Sulfur Dioxide (SO₂)

In 2010, the EPA tightened federal standards for SO₂, a potent respiratory irritant that is also a precursor of PM_2.5. Several Texas counties violate SO₂ standards, mostly in the vicinity of coal-fired power plants (Figure 1.3).

Figure 1.3 — Map of Texas SO₂ Nonattainment Areas as of March 2023

 

1.2 The Role of Power Plant Emissions in Texas Air Pollution

Because they burn more coal and natural gas than those in any other state, Texas power plants lead the nation in emissions of NO_x, SO₂, and CO₂­ (Table 1.3).[15] Power plants emit the majority of SO₂ in Texas and large shares of NOx and CO₂.

Table 1.3 — Power Plant Emissions in Texas in 2021 and Their Share of Texas Emissions Overall and of U.S. Power Plants[16]

 

The leading role of power plants in SO₂ emissions will make them pivotal to the attainment of the more stringent PM_2.5 standards that the EPA is expected to issue in 2023. Sulfate ranks alongside organic carbon as the top two components of PM2.5 in many regions.[17] Texas lacks sufficient regulatory monitors to adequately speciate the composition of its PM_2.5. However, scientific field studies have found sulfate to be the largest or second-largest component of particulate matter in several parts of Texas. In eight field studies across various sites in Southeast Texas, sulfate comprised 21%–57% of fine particulate matter (Table 1.4). Since SO₂ is the main precursor of sulfate particulate matter and 59% of Texas SO₂ emissions come from coal-fired power plants,[18] reductions of power plant SO₂ emissions could be pivotal to bringing Texas regions into attainment of tightened PM_2.5 standards.

 

Table 1.4 — Sulfate Contribution to Particulate Matter in Texas Field Studies

Study	Sampling Time	Location	Method	Type	Sulfate (%)
Chiou et al.[19]	July 2003– August 2005	Beaumont (Hamshire)	Filter sample and ICa	PM2.5	57%
Karnae and John[20]	July 2003–December 2008	Corpus Christi (CAMS635)	Filter sample and ICa	PM2.5	30%
Cleveland et al.[21]	August 15–September 28, 2006	Houston (University of Houston)	Q-AMSb	PM1*	38%
Leong et al.[22]	September 6–September 29, 2013	Houston (NW and SE Houston)	ACSMc and SP-AMSd	PM1*	24% (NW) 30% (SE)
Bean et al.[23]	August 24–October 1, 2013	Conroe	ACSMc	PM1*	27%
Dai et al.[24]	Winter: February 3–17, 2014 Summer: May 1–31 2014	Sugar Land (UH Sugar Land)	HR-ToF-AMSe	PM1*	23% (Winter) 36% (Summer)
Wallace et al.[25]	February 7–27, 2015	Houston (Ship Channel)	HR-ToF-AMSe	PM1*	23%
Naiema et al.[26]	May 13–29, 2015	Houston (Clinton Drive)	HR-ToF-AMSe Filter sample and ICa	PM1* PM2.5	45% 21%
Schulze et al.[27]	May 24–June 14, 2016	Galveston	HR-ToF-AMSe	PM1*	44%

Notes

^a Ion chromatography

^b Quadrupole aerosol mass spectrometer

^c Aerosol chemical speciation monitor

^d Soot particle aerosol mass spectrometer

^e High-resolution time-of-flight aerosol mass spectrometer

* Nonrefractory particles with aerodynamic diameter smaller than 1 micron

 

All of the coal-fired power plants in ERCOT are located in the eastern half of the state, putting them near the state’s most populous regions and most of the regions that violate the current ozone standard or would exceed a tightened PM_2.5 standard (Figure 1.3 and Tables 1.1 and 1.2). Four additional coal-fired power plants — Harrington, Pirkey, Tolk, and Welsh — operate outside ERCOT in the Southwest Power Pool; all are scheduled to cease coal burning between 2023 and 2028.

Figure 1.4 — Coal Power Plants Operating in ERCOT in 2019

 

1.3 Implications for Coal Power Plant Retirements

Several coal power plants have closed in Texas since 2017, including Big Brown, Monticello, JT Deely, and Oklaunion. Market forces and environmental regulations are likely to prompt several more coal plants to retire or convert to natural gas by the end of the decade. All 4 GW of the remaining coal power plants in the Texas portion of the Southwest Power Pool (Harrington, Pirkey, Tolk, and Welsh) along with Coleto Creek in ERCOT have announced plans to retire or convert to natural gas by 2028 or sooner.[29] Thus, we will focus on how recent environmental regulations may accelerate the retirements or conversions of remaining coal-fired power plants in ERCOT.

In March 2023, the EPA issued the Good Neighbor Plan, aimed at reducing interstate formation of ozone smog. The plan will shrink ozone-season NO_x emissions budgets for each state in cap-and-trade markets, including a nearly 50% cut to the Texas budget from 2023 to 2029 (Figure 1.5).[30] Modeling by the EPA showed that the plan would likely prompt Martin Lake to convert from coal to gas.[31] Additional measures at other plants will be needed to further cut NO_x emissions.

Figure 1.5 — Ozone Season NO_x Emissions at Texas Power Plants Subject to the Cross-State Air Pollution Rule in 2021 and 2022, and The Emissions Budgets Under EPA’s Good Neighbor Plan for 2023–29[32] 

 

In April 2023, the EPA proposed a Regional Haze Plan for Texas that would mandate steep SO₂ emissions reductions at the Parish, Martin Lake, and Coleto Creek coal plants in ERCOT and the Harrington and Welsh power plants in the Southwest Power Pool.[33] The EPA expects Texas power plant SO₂ emissions will fall by more than half under the plan. Since Coleto Creek,[34] Harrington,[35] and Welsh[36] have already announced plans to cease coal operations between 2025 and 2028, the Regional Haze Plan would primarily affect Parish and Martin Lake. Those two plants emitted more SO₂ in 2021 than all other Texas power plants combined.[37] Three units at each plant began operation between 1977 and 1980, just before SO₂ scrubbers were required at all new coal-fired power plants. The units at Martin Lake now have scrubbers that are far less effective than modern designs, whereas the Parish units lack scrubbers altogether. Emissions from Martin Lake are responsible for the SO₂ nonattainment area in Rusk and Panola Counties. TCEQ does not operate any SO₂ monitors near Parish in Fort Bend County.

Table 1.5 summarizes factors that could affect the retirement of ERCOT coal plants. Only Coleto Creek[38] and Limestone[39] have announced retirement dates. The EPA’s proposed Regional Haze Plan for Texas[40] would require Martin Lake to upgrade its scrubbers and two units at Parish to add scrubbers, convert to gas, or retire. The EPA’s Good Neighbor Plan sets Texas’ ozone-season NO_x budget at 21,623 tons in 2028, a 54% cut below 2022 emissions at affected sources (Figure 1.5).[41] That cap-and-trade rule will add urgency to cutting NO_x emission rates, which averaged 1.24 lb/MWh at Texas coal plants in 2021.[42] Furthermore, in April 2023, the EPA proposed cutting the mercury emissions limit for lignite-burning plants such as San Miguel and Twin Oaks by 70%.[43] Taken together, 8.6 GW of the 13.6 GW of coal power plants in ERCOT would likely need substantial SO₂ and/or NO_x emissions controls or a conversion to natural gas to continue operations (Table 1.5). Additional steps may be needed to comply with new effluent limitations on wastewater discharges from steam power plants, which were issued by the EPA in March 2023.[44]

Table 1.5 — Coal Power Plants in ERCOT and Pending Retirements and Emissions Control Needs[45]

 

Section 2: Complementarity of Wind and Solar Power in Texas

Electricity has historically been produced mainly from sources that can be turned on and off —natural gas, coal, nuclear, and hydropower. These resources are known as “dispatchable” sources of electricity. By contrast, the fastest growing sources of electricity — wind and solar — are “non-dispatchable” because they can generate power only when the wind blows or the sun shines. That poses challenges to electric grid operators. However, if the wind blows and the sun shines at different times, the output of wind and solar farms could complement each other and lessen the number of hours when neither resource is available.

2.1 Prior Publications by the Cohan Research Group

The Cohan research group at Rice University has pioneered research quantifying the complementarity of wind and solar power in ERCOT, most notably with studies published by Joanna H. Slusarewicz and Daniel S. Cohan (2018) and Richard Morse et al. (2022).[46]

The Slusarewicz/Cohan study considered hypothetical wind farms and solar farms in two parts of ERCOT — West Texas, where both wind and solar farms were historically most prevalent, and South Texas, where a growing number of wind farms were being deployed near the coast at the time of the study. Slusarewicz and Cohan showed that, on average, winds blow most strongly in West Texas at night, and along the southern Texas coast on summer afternoons and evenings with the sea breezes. These findings suggest that wind farms from different regions can be complementary with each other and with solar power, which peaks midday when winds tend to be slow.

The Morse study, which was funded by the Energy Foundation, extended beyond the Slusarewicz/Cohan study by considering all wind and solar farms in the ERCOT interconnection queue as of June 2020 (Figure 2.1). Using the NREL System Advisor Model and WIND Toolkit to simulate the hypothetical output of each project under 2009–11 meteorological conditions, Morse et al. found that during the winter, spring, and fall, winds across ERCOT blow most strongly at night, complementing the daytime output of solar farms (Figure 2.2). During summer months, a more spatially heterogeneous pattern emerges, with West Texas winds continuing to peak at night, but winds along the southern Texas coast peaking with afternoon and evening sea breezes, consistent with the Slusarewicz findings.

Figure 2.1 — Wind (Top) and Solar (Bottom) Projects in the ERCOT Interconnection Queue as of June 2020, Aggregated by County[47]

 

Figure 2.2 — Average Regional Capacity Factor of Wind (Left) and Solar (Right) Sites in ERCOT Interconnection Queue in Peak Demand Months of January (Top), July (Middle), and September (Bottom) 

 

The complementary output modeled by Morse can also be observed in actual wind and solar generation data in ERCOT in recent years (Figure 2.3). Wind output peaks at night throughout the years, while solar follows expected patterns. Close inspection of the figure shows that in recent years it has been relatively rare for wind capacity factors to drop below 20% at night when solar power is unavailable. As will be shown in Table 2.3, a substantial number of wind and solar projects are now in ERCOT’s interconnection queue in all of these regions.

Figure 2.3 — Wind and Solar Capacity Factors in ERCOT by Hour and Month from 2017–22

 

Morse et al. conducted optimization modeling to identify combinations of wind and solar projects that could together replace most of the output that ERCOT received from coal-fired power plants. They found that it would be impossible to fully replace coal with wind and solar alone without the use of battery storage or other backup resources. This is due to the fact that there are occasional times when it is neither windy nor sunny anywhere in ERCOT. Instead, the optimization modeling identified a least-cost ensemble of 15,798 MW of wind farms and 10,156 MW of solar farms from the June 2020 queue that would leave just 10% of the output from coal plants uncovered (what the study termed “slack” to be covered by natural gas, storage, or other options), producing surplus output at other times. The output from that ensemble would have fallen below a 20% capacity factor during 8.8% of all half-hour periods in the 2009–11 meteorological conditions considered in that study (Table 2.1). Such low output would have occurred mainly on fall and spring nights, when power demand was relatively low. Surpluses would have occurred during most daylight hours, thanks to the output from solar farms.

Table 2.1 — Frequency of Low Output From a Cost-Optimized Wind and Solar Deployment Scenario Under 2009–11 Meteorological Conditions, and the Average and Maximum Load During Those Times[49]

 

Four important caveats must be noted that limit the applicability of the Morse study. First, the study did not consider battery storage. Second, the study assumed ample transmission capacity and did not quantify the additional high-voltage transmission lines that would be needed to transmit power from the windiest and sunniest regions to the cities and industrial centers that need it most. Third, the study assumed that wind farms and solar farms perform as well as wind speeds and solar irradiance allow. However, freezing precipitation and condensation during the February 2021 and February 2023 winter storms severely impaired output from many wind farms and some solar farms.[50] Most Texas wind turbines lack de-icing equipment, and solar panels can be covered with snow. Finally, wintertime peak demand has grown dramatically since 2011, as the population has grown and the proportion of homes heated with electric resistance heating or heat pumps has reached 60%.[51] That heightens the importance of studying how all generation resources perform under the freezing precipitation, shifting wind speeds, and frigid temperatures that can accompany winter storms. It should be noted that coal and gas power plants and the natural gas supply suffered severe impairments during the February 2021 freeze,[52] and that coal plants have failed during various other severe weather events.[53]

2.2 Updated Analyses

The Slusarewicz/Cohan and Morse studies were written amid a time of rapid growth in wind and solar in Texas. Wind and solar outputs rose sharply from 2017 through 2022 (Figure 2.4) and are projected to continue to grow.

Figure 2.4 — Wind and Solar Output in ERCOT by Month From 2017–22

 

ERCOT’s interconnection queue has greatly expanded since the June 2020 queue considered by Morse et al. (Table 2.2). Numerous solar, natural gas, and battery projects have been added to the queue, though wind projects have dipped. Meanwhile, more than 5 GW of wind, 8 GW of solar, and 2 GW of batteries have been built in ERCOT since 2020 (Table 2.2). Proposed projects span each region of ERCOT (Table 2.3). Thus, any update to the Morse study would start from a larger baseline of existing renewable resources and find greater opportunities to add solar projects and complement their variable output with new natural gas plants and batteries.

Table 2.2 — Total Capacity (in MW) of Projects in the ERCOT Generator Interconnection Status Report in June 2020 and January 2023, Compared with Operational Capacity Reported by ERCOT Resource Capacity Trend Charts

 

Table 2.3 — Regional Distribution of the Wind and Solar Capacity Online in the Summer 2023 Seasonal Assessment of Resource Adequacy, and of the Wind and Solar Projects in the June 2023 ERCOT Generator Interconnection Status Report That Have Requested a Full Interconnection Study

 

ERCOT expects that solar will continue to lead all other sources of new generation capacity (Figure 2.5). That could ease, rather than worsen, strains on dispatchable resources. Solar power tends to be abundant during summer afternoons when air conditioning use peaks and on days when winds are slow. Thus, it can help satisfy summertime peaks in demand and ease the burden on thermal power plants when winds are slow during daytime hours.

Figure 2.5 — Trends in Generating Capacity from Solar (Upper Left), Wind (Upper Right), Gas Combined Cycle (Lower Left), and Other Gas Power Plants (Lower right), in ERCOT Resource Capacity Trend Charts Issued in January 2023

 

A closer look at periods of record summer and winter peak demand in 2022 and 2023 (Figure 2.6) suggests that a new paradigm is needed for assessing resource adequacy. When summer demand hit an all-time high of 80 GW on July 20, 2022,[54] wind and solar provided over 15 GW of power (Figure 2.6a). Wind and solar provided 20 GW of power when a new record of 85 GW was set on August 10, 2023 (Figure 2.6b). Peak net load came on August 25, 2023, when gross load was 78 GW, but wind and solar output fell below 8 GW amid slow winds around sunset (Figure 2.6c), prompting a voluntary conservation notice.

When winter demand hit an all-time high of 74 GW on December 23, 2022, wind provided 23 GW of power during nighttime conditions (Figure 2.6d). Tighter conditions and higher prices came later that week, when winds slowed and net load on thermal resources peaked (Figure 2.6d). Seasonal assessments of resource adequacy reports focus on the hours of peak demand, but it is now the hours of peak net load (demand minus variable renewable output), as well as times of unexpectedly high-power plant outages, that pose the greatest risks of shortfalls.

Figure 2.6 — Electricity Generation by Energy Source in ERCOT During the Weeks of Peak Gross Load During (a) Summer of 2022 and (b) Summer of 2023; (c) Peak Net Load in Summer of 2023; and (d) Peak Gross Load in Winter of 2022–23

(a)

(b)

(c)

(d)

 

Those times of peak net load deserve closer scrutiny because they must be satisfied by an aging fleet of thermal power plants. More than 30 GW of power plants in ERCOT are now older than 30 years, including more than 20 GW of power plants that are more than 40 years old (Figure 2.7). Retirements and outages at those plants can threaten reliability as power demand continues to grow.

Figure 2.7 — Dispatchable Resources Sorted by Year Entering Service, as Reported by ERCOT’s Seasonal Assessment of Resource Adequacy for Summer 2022

 

To examine trends and projections of gross load and net load, we projected gross load to 2025 based on the growth ratios from ERCOT’s 2023 Long-Term Load Forecast Report (Figure 2.8). The forecast anticipates 2.1% annual average growth rates from 2023–32, compared to the 2.6% growth rates that were observed from 2013–22.

Figure 2.8 — Annual Energy Trends and Projections from ERCOT’s 2023 Long-Term Load Forecast

 

We then plotted cumulative distribution functions of gross load and net load in each hour of 2020–22, where net load was computed by subtracting wind and solar generation from gross load (Figure 2.9a-c). Across those three years, growth in wind and solar output (Figure 2.4) nearly kept pace with growth in load, allowing net load to grow far more slowly than gross load (Figure 2.9a-c) and thus mitigating the burden placed on dispatchable resources. Applying the growth rates from the Long-Term Load Forecast to the three base years of meteorological conditions, we created a synthetic projection of gross load for 2025 (Figure 2.9d). We note that this projection is below the 85 GW record load observed in August 2023 amid a record heat wave, illustrating the potential for extreme weather to drive demand above historical norms.

We then computed wind and solar output based on either the capacity that existed at the end of 2022 (brown swath in Figure 2.9d) or the capacity that is expected in 2025 as wind and solar farms proliferate (brown plus purple swaths). As can be seen in the figure, projected growth in wind and solar is sufficient to nearly eliminate hours when more than 60 GW is needed from dispatchable resources under historical (2020–22) meteorology. However, if adverse policies halt the growth in wind and solar, far greater burdens would be placed on the aging fleet of dispatchable power plants. Projected growth in wind and solar by 2025 would reduce peak net load during the top 100 hours of the year by an average of 4 GW, and by 6 GW during the top 500 hours, while producing an annual average of nearly 8 GW (Figure 2.10). Thus, despite the variability of solar and wind power, their complementary nature and strong performance during heat waves and cold fronts allow them to substantially reduce peak demands on thermal power plants. Still, the 60 GW remaining projected net load in 2025 (Figure 2.9d) (or more under extreme weather conditions) is far greater than the 51 GW that natural gas power plants provided during peak net load in summer 2023 (Figure 2.6d). This indicates that additional dispatchable generation and storage resources will be needed if most coal plants close.

Figure 2.9 — The Cumulative Number of Hours that the ERCOT System has More Than Each Amount of Load Per Year

 

Figure 2.10 — Extent to Which the New Wind and Solar Farms (9.5GW and 28.6 GW, Respectively) Anticipated to be Added in 2023 and 2024 are Expected to Reduce Net Load in 2025, With Hours Ranked by Net Load

 

2.3 Energy Efficiency

Since very little new thermal capacity is likely to be added in ERCOT before 2025 (Table 2.2 and Figure 2.3) and most coal plants face environmental challenges (Table 1.5), it will be crucial to consider the demand side of the equation to keep supply and demand in balance. Energy efficiency and demand response are crucial to maintaining the reliability of the grid. The American Council for an Energy Efficient Economy ranks Texas 29th among states for its energy efficiency policies and finds substantial opportunities to improve building efficiency.[55] Texas also lags in its implementation of demand response measures, which specifically aim to reduce peaks in demand.[56]

Although energy efficiency and demand response are beyond the scope of this study, they are likely to be pivotal to efforts to make Texas electricity reliable, affordable, and clean.

Section 3: Opportunities for Geothermal Energy in Texas

Geothermal power plants harness heat from within the Earth to generate electricity. That heat can also be used directly for district heating or certain industries that require low-temperature heat. Life-cycle greenhouse gas emissions from this renewable resource are similar to those from solar photovoltaics and are more than an order of magnitude lower than fossil fuels.[57]

Historically, most geothermal electricity has been generated from hydrothermal resources, where hot water or steam is available at shallow depths. High-quality hydrothermal resources are available at only a limited number of accessible sites nationwide, none of them in Texas (Figure 3.1). Thus, geothermal power plants generated only 17 gigawatt hours (0.4%) of U.S. electricity and virtually none in Texas in 2022, according to data from the U.S. Energy Information Administration.

Figure 3.1 — Geothermal Resources in the United States

 

Technologies developed by the oil and gas industry can make it possible to access higher temperature fluids at depth in locations that have too little permeability for conventional technologies. Enhanced (or, “engineered”) geothermal systems (EGS) inject fluids deep underground to create flows of hot water that can be used for direct heat or to generate electricity. The Department of Energy (DOE) Frontier Observatory for Research in Geothermal Energy’s (FORGE) site in Utah has enabled companies and researchers to test emerging technologies for EGS. In 2022, DOE established an Enhanced Geothermal Earthshot that aims to bring the cost of EGS down to $45 per megawatt hour by 2035. Researchers are also exploring how geothermal power could be dispatched flexibly to balance fluctuations in load and wind and solar output.[58]

In 2023, researchers from five Texas universities and their collaborators issued a major study on opportunities for geothermal power in Texas.[59] The study noted that EGS or related technologies for closed loop geothermal systems or hybrid approaches could be applicable in Texas if costs come down sufficiently. The study noted that most of the best geothermal resources in Texas are located in the eastern half of the state, including locations in close proximity to major population centers (Figure 3.2).

Ongoing research in the Cohan group, funded by Project InnerSpace, is exploring the extent to which geothermal resources could be deployed in Texas and nationally if costs decline to near the levels sought by the DOE Earthshot. That research is comparing cost estimates from the Geothermal Electricity Technology Evaluation Model (GETEM) and GEOPHIRES model,[60] and using NREL’s Regional Energy Deployment Systems (ReEDS) capacity expansion model to simulate electricity supply, demand, and transmission. Results from that study will be available in late 2023.

Figure 3.2 — Temperatures at 6.5 km Depth in Texas, Based on SMU Geothermal Laboratory Estimates[61]

 

Because hotter geothermal resources are available in western states and initial deployments are being spurred by regulatory policies in California that require clean firm power,[62] Texas is unlikely to be a hotbed for geothermal deployments this decade. However, Texas companies are taking a leading role in developing advanced geothermal technologies. At least 12 geothermal startups are headquartered in Texas,[63] and various incumbent oil and gas companies are exploring opportunities to invest in geothermal companies or pivot their technologies to geothermal applications. If learning by doing elsewhere is able to bring down the costs of advanced technologies, geothermal could become viable in Texas as a firm and reliable complement to variable wind and solar power.

Section 4: Climatology of Texas Temperature Extremes and Electricity Demand

In recent years Texas has experienced both hot and cold temperature extremes that have strained the energy and electricity systems, highlighting the limitations of existing methods to assess resource adequacy. For example, ERCOT estimates that peak demand without load shedding during the February 2021 cold snap would have been 76,819 GW,[64] which dramatically surpassed ERCOT’s “extreme winter forecast” of 67,208 MW in its seasonal assessment of resource adequacy.[65]

Three issues must be addressed. First, because temperature extremes, particularly cold extremes, are driven by unique weather patterns that occur infrequently, short observational records are inadequate for risk assessments. Instead, extended records and model simulations should be used. Second, climate change has altered and will further increase the frequency and intensity of hot extremes; no trend in cold extremes is significant within the historical record. Third, the relationship between temperature and demand changes rapidly from one year to the next. Collectively, these challenges highlight the limitations of existing approaches to assess resource adequacy.

4.1 Short Observational Records are Inadequate for Risk Assessment

The seasonal assessment of resource adequacy published by ERCOT for the 2020–21 winter[66] used the February 2011 cold snap as a basis for assessing risk. February 2011 saw the coldest weather in several decades across Texas. However, when contextualized within the full historical record, the 2011 cold snap stands out as relatively unexceptional. For example, Figure 4.1 shows temperature anomalies across Texas during historical cold snaps. The most severe one-day, three-day, and five-day periods are shown. Qualitatively, the February 2021 cold snap looks like the February 1951, December 1983, and December 1989 events. The February 2011 cold snap is qualitatively less extreme than the earlier storms. This qualitative comparison suggests that the 2011 cold snap was not particularly extreme when contextualized within the full historical record.

To make these qualitative insights more quantitative, James Doss-Gollin et al.[67] computed the “inferred demand for heating” by i) calculating the amount of heating needed at each grid cell to reach a comfortable indoor temperature and then ii) aggregating across each grid cell in the region served by ERCOT, weighting each cell by its 2020 population. Then, the annual maximum series of the aggregated inferred demand for heating were computed. Figure 4.2(a) shows the time series of the peak six-hour and two-day demands for heating for each year in the dataset. At the six-hour duration, the 1989 cold snap was colder than the February 2021 storm, and others were nearly as cold at both durations. Figure 4.2(b) shows the estimated return periods for the 2021, 1989, and 2011 storms. While the specific values (2021 is estimated to be a 50-to-100-year event, depending on duration) are sensitive to methodological assumptions and should be interpreted with care, the overall finding that the February 2021 cold snap had multiple precedents within the historical record is robust.

Figure 4.1 — Temperature Anomalies During the February 2021 Cold Snap Were Qualitatively Similar to Those Experienced in Previous Storms

 

Figure 4.2 (a) — Time Series of the Annual Maximum Inferred Demand for Heating at the Six-Hour and Two-Day Durations; (b) — Estimated Return Periods for the 1989, 2011, and 2021 Storms at Different Durations

 

Although the February 2021 cold snap had substantial precedent when aggregated across the state, there were specific locations that experienced more severe cold. Figure 4.3 shows the estimated local return periods. Some areas (bright yellow) experienced cold in February 2021 that they would have expected to exceed with 1% probability in a given year. In most areas, however, including locations where specific infrastructure components failed, the return period was between 20 and 50 years. As for Figure 4.2b, the specific values are sensitive to methodology and should be interpreted with caution.

Figure 4.3 (a) — Population Density in 2020; (b, c, e, f) — Estimated Return Period of the Cold Observed in February 2021 at Each Grid Cell (Contours Show 50- and 100-year Levels); (d) — Location of Electricity Generation Infrastructure

 

4.2 Effects of Climate Change on Temperature Extremes

While most electricity systems are designed to handle peak demand during summer months, pathways to deep decarbonization generally electrify building heating, thus increasing electricity demand during winter. A key question is how climate variability and change will affect peak heating and cooling demand in an electrified future.

A collaborative analysis between the Doss-Gollin lab at Rice University and a research team from Columbia University assessed trends in temperature-based proxies for electricity demand over the past 70 years. In Texas, demand for heating (cooling) decreases (increases) over most of the contiguous U.S. However, while climate change drives robust increases in peak cooling demand, trends in peak heating demand are generally smaller and less robust. Because the distribution of temperature exhibits a fat left tail, severe cold snaps dominate the extremes of thermal demand. As building heating electrifies, system operators such as ERCOT must account for these events to ensure reliability.

Total annual demand for heating (Figure 4.4, red line) has been slowly decreasing since 1980, while total demand for cooling has been increasing (Figure 4.4, blue line). These effects are broadly consistent with the first-order warming effects of climate change. Combined, the total demand for heating and cooling (calculated in degrees Fahrenheit as in Section 4.1) is roughly constant (black line), though it is important to note that because this demand does not account for the efficiency of heating, the energy source considered, or building characteristics, it does not correspond directly to electricity demand (see Section 4.3).

Figure 4.4 — Annually Aggregated Mean Inferred Demand for Heating (Red) and Cooling (Blue), and Total (Black) in the Region Served by ERCOT

 

While trends in demand for heating and cooling are important for understanding total energy requirements, understanding trends in peaks is important for assessing system risks. Figure 4.5 shows the trends in annual peak demands for heating and cooling. Peak demand for cooling, i.e. the hottest 24 hours of the summer, shows a small, possibly zero, upward trend. Taken with Figure 4.4, this suggests that the effect of warming on Texas demand for cooling has been primarily to increase the number of hot days rather than a large increase in how hot these days are. Peak demand for heating (red line) shows much larger interannual variability than demand for heating. No trend is readily apparent. As noted in Section 4.1, the large interannual variability of cold extremes means that the recent past is insufficient for risk assessment.

An important point is that most U.S. electricity systems, including ERCOT, are designed to meet peak loads during summer months, when demand for cooling peaks. However, infrequent but severe cold temperatures may lead the largest electricity demands to occur during rare winter peaks, particularly as the electrification of home heating continues (see Section 4.3).

Figure 4.5 — Annually Aggregated Peak Demand for Heating (Red) and Cooling (Blue) in the Region Served by ERCOT

 

4.3 Response of Demand for Electricity to Peak Temperatures

Assessing resource adequacy involves mapping scenarios of possible temperatures onto electricity demand. This is a challenging and imprecise task because the relationship between temperature and electricity demand evolves rapidly over time due to changing demographics, technologies, industrial needs, and other factors. This is illustrated in Figure 4.6, which shows how the relationship between temperature and daily average power demand has changed from 1996 to 2020 using data from ERCOT.[70]

Figure 4.6 — Relationship Between Temperature and Electricity Demand in 1996 Versus 2020

 

Different studies have used different methods to model the response of electricity to temperature. For example, Jangho Lee and Andrew Dessler use a polynomial regression model while Blake Shaffer, Daniel Quintero, and Joshua Rhodes use a temperature fixed effects model.[71] Both identify two challenges. The first is that there is, by definition, a lack of data at the extremes, meaning that estimates of electricity demand at temperatures that have been experienced rarely or not at all in the recent historical record are subjective and difficult to verify a priori. The second is that the relationship evolves over time, as shown in Figure 4.6, and because this change depends on factors like the rate of adoption of electric vehicles, the price of heat pumps, and the speed of population growth, it is difficult to model accurately.

Section 5: Conclusion

The research and data synthesized here provide important insights as Texas works to improve the reliability and sustainability of its power supply amid a changing climate. The power grid of today leaves Texans vulnerable to outages under extreme weather conditions even as emissions from coal plants damage our air quality, health, and climate. Environmental regulations and economic factors are likely to drive the closures of some coal-fired power plants, heightening the need to add other resources to meet the needs of a growing economy.

Continued growth in wind and solar generation is crucial to easing the burden on other resources. Our research has demonstrated that wind and solar power production tend to be complementary in Texas, including during times of peak demand. Nevertheless, wind and solar resources leave key gaps, especially around sunset and during winter freezes, that must be met by dispatchable resources. Energy efficiency and demand response can curb growth in electricity demand and make it more flexible. Still, there will be a need for construction of more dispatchable generation and storage resources along with transmission to enhance the reliability and resilience of electricity in ERCOT.

Taken together, our research shows that Texas has a tremendous opportunity to transition away from the coal plants that are causing a disproportionate share of its climate-warming and health-damaging emissions. However, intensifying summer heat waves, volatile winter weather, growing power demand, and inadequate transmission within and beyond ERCOT all pose challenges to electric reliability amid this transition. Simultaneous efforts to improve efficiency, expand transmission, add dispatchable resources, and foster continued growth of wind and solar will be needed to overcome these challenges and secure an affordable, sustainable, and reliable power supply for all Texans.

Acknowledgments

Funding for this study was provided by the Energy Foundation. The authors acknowledge Chen Chen for analysis of wind, solar, and load data and Chun-Ying Chao for a literature review of particulate matter measurements in Texas.

Notes

---

[1] U.S. Environmental Protection Agency (EPA), eGRID Data Explorer, accessed March 11, 2023, https://www.epa.gov/egrid/data-explorer; EPA, “Air Pollutant Emissions Trends Data,” accessed March 11, 2023, https://www.epa.gov/air-emissions-inventories/air-pollutant-emissions-trends-data.

[2] Texas Commission on Environmental Quality (TCEQ), “Houston-Galveston-Brazoria: Ozone History,” accessed March 11, 2023, https://www.tceq.texas.gov/airquality/sip/hgb/hgb-ozone-history.

[3] TCEQ, “Dallas-Fort Worth: Ozone History,” accessed March 11, 2023, https://www.tceq.texas.gov/airquality/sip/dfw/dfw-ozone-history.

[4] EPA, “8-Hour Ozone (2015) Designated Area State/Area/County Report,” Green Book, accessed March 11, 2023, https://www3.epa.gov/airquality/greenbook/jbcs.html#TX.

[5] EPA, “8-Hour Ozone (2015) Designated Area Design Values,” Green Book, accessed March 11, 2023, https://www3.epa.gov/airquality/greenbook/jdtc.html.

[6] Xue Xiao et al., “Highly Nonlinear Ozone Formation in the Houston Region and Implications for Emission Controls,” Journal of Geophysical Research: Atmospheres 115, no. D23 (2010), https://doi.org/10.1029/2010JD014435; Daniel S. Cohan et al., “Nonlinear Response of Ozone to Emissions: Source Apportionment and Sensitivity Analysis,” Environmental Science and Technology 39, no.17 (2005): 6739–48, https://doi.org/10.1021/es048664m.

[7] EPA, “8-Hour Ozone (2015) Designated Area Design Values”; TCEQ, “Ozone Data,” accessed March 11, 2023, https://www.tceq.texas.gov/agency/data/ozone_data.html.

[8] Richard Burnett et al., “Global Estimates of Mortality Associated with Long-Term Exposure to Outdoor Fine Particulate Matter,” Proceedings of the National Academy of Science 115, no. 38 (2018): 9592–97, https://doi.org/10.1073/pnas.1803222115; J. Lelieveld et al., “The Contribution of Outdoor Air Pollution Sources to Premature Mortality on a Global Scale,” Nature 525, no. 7569 (2015): 367–71, https://doi.org/10.1038/nature15371.

[9] WHO (World Health Organization), “WHO Global Air Quality Guidelines: Particulate Matter (‎PM_2.5 and PM₁₀)‎, Ozone, Nitrogen Dioxide, Sulfur Dioxide and Carbon Monoxide,” accessed March 11, 2023, https://www.who.int/publications-detail-redirect/9789240034228.

[10] WHO, “WHO Global Air Quality Guidelines.”

[11] Sean Reilly, “EPA science advisers unanimously back tighter soot limits,” E&E News, March 4, 2022, https://www.eenews.net/articles/epa-science-advisers-unanimously-back-tighter-soot-limits/.

[12] EPA, “EPA Proposes to Strengthen Air Quality Standards to Protect the Public from Harmful Effects of Soot,” news release, January 6, 2023, https://www.epa.gov/newsreleases/epa-proposes-strengthen-air-quality-standards-protect-public-harmful-effects-soot.

[13] K. Kaufmann, “EPA Proposes Lowering Limit for Small Particle Pollution,” RTO Insider, January 8, 2023, https://www.rtoinsider.com/articles/31405-epa-proposes-lowering-limit-small-particle-pollution.

[14] EPA, Regulatory Impact Analysis for the Proposed Reconsideration of the National Ambient Air Quality Standards for Particulate Matter, December 2022, https://www.epa.gov/system/files/documents/2023-01/naaqs-pm_ria_proposed_2022-12.pdf; Aaron van Donkelaar et al., “Monthly Global Estimates of Fine Particulate Matter and Their Uncertainty,” Environmental Science and Technology 55, no. 22 (2021): 15287–300, https://doi.org/10.1021/acs.est.1c05309.

[15] TCEQ, “Ozone Data.”

[16] EPA, eGRID Data Explorer; EPA, “Air Pollutant Emissions Trends Data.”

[17] van Donkelaar et al., “Regional Estimates of Chemical Composition of Fine Particulate Matter Using a Combined Geoscience-Statistical Method with Information from Satellites, Models, and Monitors,” Environmental Science and Technology 53, no. 5 (2019): 2595–611, https://doi.org/10.1021/acs.est.8b06392.

[18] EPA, “Air Pollutant Emissions Trends Data.”

[19] Paul Chiou et al., “Atmospheric Aerosol over a Southeastern Region of Texas: Chemical Composition and Possible Sources,” Environmental Modeling and Assessment 14, no. 3 (2009): 333–50, https://doi.org/10.1007/s10666-007-9120-8.

[20] Saritha Karnae and Kuruvilla John, “Source Apportionment of Fine Particulate Matter Measured in an Industrialized Coastal Urban Area of South Texas,” Atmospheric Environment 45, no. 23 (2011): 3769–76, https://doi.org/10.1016/j.atmosenv.2011.04.040.

[21] M. J. Cleveland et al., “Characterization of Urban Aerosol Using Aerosol Mass Spectrometry and Proton Nuclear Magnetic Resonance Spectroscopy,” Atmospheric Environment 54 (2012): 511–18, https://doi.org/10.1016/j.atmosenv.2012.02.074.

[22] Y. J. Leong et al., “Overview of Surface Measurements and Spatial Characterization of Submicrometer Particulate Matter during the DISCOVER-AQ 2013 Campaign in Houston, TX,” Journal of the Air & Waste Management Association 67, no. 8 (2017): 854–72, https://doi.org/10.1080/10962247.2017.1296502.

[23] Jeffrey K. Bean et al., “Composition and Sources of Particulate Matter Measured near Houston, TX: Anthropogenic-Biogenic Interactions,” Atmosphere 7, no. 5 (2016), https://doi.org/10.3390/atmos7050073.

[24] Qili Dai et al., “Seasonal Differences in Formation Processes of Oxidized Organic Aerosol near Houston, TX,” Atmospheric Chemistry and Physics 19, no. 14 (2019): 9641–61, https://doi.org/10.5194/acp-19-9641-2019.

[25] Henry W. Wallace et al., “Source Apportionment of Particulate Matter and Trace Gases near a Major Refinery near the Houston Ship Channel,” Atmospheric Environment 173 (2018): 16-29, https://doi.org/10.1016/j.atmosenv.2017.10.049.

[26] Ibrahim M. Al-Naiema et al., “Source Apportionment of Fine Particulate Matter in Houston, Texas: Insights to Secondary Organic Aerosols,” Atmospheric Chemistry and Physics 18, no. 21 (2018): 15601–22, https://doi.org/10.5194/acp-18-15601-2018.

[27] Benjamin C. Schulze et al., “The Impacts of Regional Shipping Emissions on the Chemical Characteristics of Coastal Submicron Aerosols near Houston, TX,” Atmospheric Chemistry and Physics 18, no. 19 (2018): 14217–41, https://doi.org/10.5194/acp-18-14217-2018.

[28] Richard Morse et al., “Can Wind and Solar Replace Coal in Texas?” Renewables: Wind, Water, and Solar 9, no. 1 (2022): 1, https://doi.org/10.1186/s40807-022-00069-2.

[29] KyLeah Frazier, “Xcel Energy’s Harrington Power Plant to convert to natural gas by 2025,” News Channel 10, KFDA, February 24, 2023, https://www.newschannel10.com/2023/02/23/xcel-energys-harrington-power-plant-convert-natural-gas-by-2025/; Jeff Della Rosa, “SWEPCO to retire 1 coal-fired plant, cease coal operations at another,” Talk Business & Politics, November 5, 2020, https://talkbusiness.net/2020/11/swepco-to-retire-1-coal-fired-plant-cease-coal-operations-at-another/; Caden Keenan, “Xcel Energy aims to retire coal early at Tolk Generating Station,” KAMR, MyHighPlains.com, accessed April 27, 2023, https://www.myhighplains.com/news/local-news/xcel-energy-aims-to-retire-coal-early-at-tolk-generating-station/.

[30] EPA, “State Budgets under the Good Neighbor Plan for the 2015 Ozone NAAQS,” accessed April 27, 2023, https://www.epa.gov/csapr/state-budgets-under-good-neighbor-plan-2015-ozone-naaqs.

[31] EPA, “State Budgets under the Good Neighbor Plan for the 2015 Ozone NAAQS.”

[32] EPA, “Clean Air Power Sector Programs: Ozone Season Data 2021 vs. 2022,” accessed April 27, 2023, https://www.epa.gov/power-sector/facility-level-comparisons#OzoneSeason; EPA, “State Budgets under the Good Neighbor Plan for the 2015 Ozone NAAQS.”

[33] EPA, “Revision and Promulgation of Air Quality Implementation Plans; Texas; Regional Haze Federal Implementation Plan; Disapproval and Need for Error Correction; Denial of Reconsideration of Provisions Governing Alternative to Source-Specific Best Available Retrofit Technology (BART) Determinations,” Code of Federal Regulations, Title 40, Parts 52, 78, and 97 (2023): 1–216, https://www.epa.gov/system/files/documents/2023-04/Texas%20BART%20FIP_NPRM_2023-04-19_ADMIN_signed%20pre-publication%20copy.pdf.

[34] “Mark Rosenberg, “Coleto Creek Power Plant shutting down by 2027,” Victoria Advocate, updated November 4, 2022, https://www.victoriaadvocate.com/counties/goliad/coleto-creek-power-plant-shutting-down-by-%202027/article_261596c8-342b-11eb-92e8-0f9c2d927a2b.html.

[35] Frazier, “Xcel Energy’s Harrington Power Plant to convert to natural gas by 2025.”

[36] Darren Sweeney, Taylor Kuykendall, and Anna Duquiatan, “More than 23 GW of coal capacity to retire in 2028 as plant closures accelerate,” S&P Global Market Intelligence, February 10, 2022, https://www.spglobal.com/marketintelligence/en/news-insights/latest-news-headlines/more-than-23-gw-of-coal-capacity-to-retire-in-2028-as-plant-closures-accelerate-68709205.

[37] EPA, eGRID Data Explorer.

[38] Rosenberg, “Coleto Creek Power Plant shutting down by 2027.”

[39] “Limestone Generating Station,” Global Energy Monitor Wiki, accessed April 27, 2023, https://www.gem.wiki/Limestone_Generating_Station.

[40] EPA, “Revision and Promulgation of Air Quality Implementation Plans,” Code of Federal Regulations, Title 40, Parts 52, 78 and 97.

[41] EPA, eGRID Data Explorer.

[42] EPA, eGRID Data Explorer.

[43] EPA, “Biden-Harris Administration Proposes to Strengthen the Mercury and Air Toxics Standards for Power Plants,” news release, April 5, 2023, https://www.epa.gov/newsreleases/biden-harris-administration-proposes-strengthen-mercury-and-air-toxics-standards-power.

[44] EPA, “Effluent Limitations Guidelines and Standards for the Steam Electric Power Generating Point Source Category-Initial Notification Date Extension,” Federal Register 88, no. 60 (March 29, 2023): 18440, https://www.federalregister.gov/documents/2023/03/29/2023-04985/effluent-limitations-guidelines-and-standards-for-the-steam-electric-power-generating-point-source.

[45] EPA, eGRID Data Explorer.

[46] Joanna H. Slusarewicz and Daniel S. Cohan, “Assessing Solar and Wind Complementarity in Texas,” Renewables: Wind, Water & Solar 5, no. 1 (2018): 7, https://doi.org/10.1186/s40807-018-0054-3; Morse et al., “Can Wind and Solar Replace Coal in Texas?”

[47] Morse et al., “Can Wind and Solar Replace Coal in Texas?”

[48] Morse et al., “Can Wind and Solar Replace Coal in Texas?”

[49] Morse et al., “Can Wind and Solar Replace Coal in Texas?”

[50] Electric Reliability Council of Texas (ERCOT), “Update to April 6, 2021 Preliminary Report on Causes of Generator Outages and Derates During the February 2021 Extreme Cold Weather Event,” 2021, https://bit.ly/3SxF9pU; “The Timeline and Events of the February 2021 Texas Electric Grid Blackouts,” The University of Texas at Austin Energy Institute, July 2021, https://bit.ly/3QRLdZ3; Diana DiGangi, “Texas Wind Energy Freeze-out Shows Need for Better Resource Adequacy, Says NRG VP,” Utility Dive, February 6, 2023, https://www.utilitydive.com/news/texas-wind-freezout-renewables-resource-adequacy/642031/.

[51] Blake Shaffer, Daniel Quintero, and Joshua Rhodes, “Changing Sensitivity to Cold Weather in Texas Power Demand,” iScience 25, no. 4 (2022): 104173, https://doi.org/10.1016/j.isci.2022.104173.

[52] ERCOT, “Update to April 6, 2021 Preliminary Report on Causes of Generator Outages and Derates During the February 2021 Extreme Cold Weather Event.”

[53] Benjamin Storrow, “How Coal Failed in the Texas Deep Freeze,” E&E News, March 18, 2021, https://www.eenews.net/articles/how-coal-failed-in-the-texas-deep-freeze/.

[54] ERCOT, “Summer 2022 Operational and Market Review,” 2022, https://bit.ly/40shFEg.

[55] Sagarika Subramanian et al., “2022 State Energy Efficiency Scorecard,” ACEEE, December 2022, https://www.aceee.org/sites/default/files/pdfs/u2206.pdf.

[56] Federal Energy Regulatory Commission (FERC), “2021 Assessment of Demand Response and Advanced Metering,” December 2021, https://bit.ly/47pvaH0.

[57] National Renewable Energy Laboratory, “Life Cycle Assessment Harmonization,” accessed July 26, 2020, https://www.nrel.gov/analysis/life-cycle-assessment.html.

[58] Dev Millstein, Patrick Dobson, and Seongeun Jeong, “The Potential to Improve the Value of U.S. Geothermal Electricity Generation Through Flexible Operations,” Journal of Energy Resources Technology 143, no. 1 (2021), https://doi.org/10.1115/1.4048981; Wilson Ricks, Jack Norbeck, and Jesse Jenkins, “The Value of In-Reservoir Energy Storage for Flexible Dispatch of Geothermal Power,” Applied Energy 313 (2022): 118807, https://doi.org/10.1016/j.apenergy.2022.118807.

[59] Jamie Beard et al., The Future of Geothermal in Texas: The Coming Century of Growth & Prosperity in the Lone Star State, accessed March 11, 2023, https://energy.utexas.edu/research/geothermal-texas.

[60] Gregory L. Mines, “GETEM User Manual,” Idaho National Laboratory, July 2016, https://workingincaes.inl.gov/SiteAssets/CAES%20Files/FORGE/inl_ext-16-38751%20GETEM%20User%20Manual%20Final.pdf; Koenraad F. Beckers and Kevin McCabe, “GEOPHIRES v2.0: Updated Geothermal Techno-Economic Simulation Tool,” Geothermal Energy 7, no. 1 (2019): 5, https://doi.org/10.1186/s40517-019-0119-6.

[61] Beard et al., The Future of Geothermal in Texas.

[62] California Public Utilities Commission Energy Division (CPUC), “CPUC Orders Historic Clean Energy Procurement To Ensure Electric Grid Reliability and Meet Climate Goals,” June 24, 2021, https://www.cpuc.ca.gov/news-and-updates/all-news/cpuc-orders-clean-energy-procurement-to-ensure-electric-grid-reliability.

[63] Michelle Lewis, “Here’s Why Texas Is a Geothermal Energy ‘Sleeping Giant,’” Electrek, January 25, 2023, https://electrek.co/2023/01/25/texas-geothermal-energy/.

[64] B. Magness, “Review of February 2021 Extreme Cold Weather Event,” ERCOT, 2021, accessed May 2, 2021, https://bit.ly/3u7rg7H.

[65] ERCOT, “Final Seasonal Assessment of Resource Adequacy for the ERCOT Region (SARA): Winter 2020/2021,” 2020, accessed May 2, 2021, http://www.ercot.com/content/wcm/lists/197378/SARA-FinalWinter2020-2021.pdf.

[66] ERCOT, “Final Seasonal Assessment of Resource Adequacy for the ERCOT Region (SARA): Winter 2020/2021.”

[67] James Doss-Gollin et al., “How Unprecedented Was the February 2021 Texas Cold Snap?” Environmental Research Letters 16, no. 6 (2021), https://doi.org/10.1088/1748-9326/ac0278.

[68] Hans Hersbach et al., “The ERA5 Global Reanalysis,” Quarterly Journal of the Royal Meteorology Society 146, no. 730 (2020): 1999–2049, https://doi.org/10.1002/qj.3803.

[69] Doss-Gollin et al., “How Unprecedented Was the February 2021 Texas Cold Snap?”

[70] Jangho Lee and Andrew E. Dessler, “The Impact of Neglecting Climate Change and Variability on ERCOT’s Forecasts of Electricity Demand in Texas,” Weather, Climate, and Society 14, no. 2 (2022): 499–505, https://doi.org/10.1175/WCAS-D-21-0140.1.

[71] Lee and Dessler, “The Impact of Neglecting Climate Change and Variability on ERCOT’s Forecasts of Electricity Demand in Texas”; Shaffer, Quintero, and Rhodes, “Changing Sensitivity to Cold Weather in Texas Power Demand.”

 

This material may be quoted or reproduced without prior permission, provided appropriate credit is given to the author and Rice University’s Baker Institute for Public Policy. The views expressed herein are those of the individual author(s), and do not necessarily represent the views of Rice University’s Baker Institute for Public Policy.