ERCOT 2019: Market Performance Assessment (Updated March 23, 2020)

ERCOT 2019: Market Performance Assessment

Jesse Schneider and Michael Goggin, Grid Strategies LLC

October 14, 2019 (Updated March 23, 2020)

For the last several summers, a lot of eyes have been on the Electric Reliability Council of Texas (ERCOT) market’s performance, given its relatively low reserve margins and lack of a mandatory forward capacity market. The results from this past summer are in, and with no firm load shed due to supply shortages, reliability was preserved. The next question for electricity market designers is whether prices were at the right level to charge consumers for what they needed while providing accurate signals to attract and retain needed generation. This blog analyzes revenue adequacy and finds that prices were indeed at the right level, so the market design seems to have once again passed with flying colors.

The key distinction between ERCOT and regions with a capacity market or resource adequacy requirement is that generators in ERCOT are solely responsible for assessing the level of supply and demand and for their performance in meeting that demand. ERCOT does not set the required level of capacity demand or determine how much credit different resources receive towards meeting that demand.

One would expect that when the system is low on capacity, as it was this past summer with around an 8 percent reserve margin, that spot energy and reserves prices would occasionally be high. Economic theory predicts that in an efficient market at equilibrium, there would be enough “rent,” or profit earned from prices that exceed generators’ operating costs, over the course of the year to make new generators see enough profit incentive to enter. So the question is, were prices over the last year high enough to attract and retain needed units? Our analysis below indicates the answer is YES.

Figure 1: ERCOT Region

ercot

ERCOT has two relatively new mechanisms to more appropriately reflect scarcity conditions and cause prices to go high enough to contribute to suppliers’ rent.[1] The first is a $9,000/MWh system-wide offer cap based on the Value of Lost Load (VOLL) estimate, which has increased steadily since 2011 and is significantly higher than the caps in other regions.[2] The second is the Operating Reserve Demand Curve (ORDC), which is an adder curve that allows prices to not only reflect Locational Marginal Prices (LMPs), but also the value of online reserve capacity as well as potential price suppression that can occur from actions the RTO takes to maintain system reliability.[3]

Let’s take a look at the prices in 2019. Figure 1 below uses ERCOT historical real-time ORDC data[4] generated during each Security Constrained Economic Dispatch (SCED) interval and compares prices for January-December of each available year. To calculate prices for each interval, we take the LMPs (system lambda) from the SCED display, and add to it the Real-time Online Reserve Price Adder (RTORPA) and the Real-time Online Reliability Deployment Price Adder (RTORDPA). After filtering prices from highest (the $9,000/MWh system-wide offer cap) to lowest, we identify the number of hours that prices have exceeded generators’ operating cost, or $200/MWh.

Figure 1: ERCOT Price Duration Curve (2014-2019)

curve update

As shown above, prices were consistently higher in 2019 than in previous years. In 2019, prices exceeded $200/MWh for 106 hours, with 4 hours and 10 minutes reaching the system-wide offer cap price. September alone, which had the most record-high temperature days since 2011[5], was responsible for 10 minutes worth of prices at the offer cap and 20 hours worth of prices above $200/MWh. For reference, 2018 saw approximately 70 hours over $200/MWh and only 10 minutes at the offer cap. Since the creation of the ORDC in June of 2014, ERCOT only saw prices hit the offer cap one other time in 2016 for 5 minutes.

So prices have been higher, but were they high enough to attract entry? To answer that question we can look at net margin for different units. The Peaker Net Margin (PNM) is used by ERCOT and its market monitor as a measure of generators’ economic rent.[6] ERCOT has estimated that in long-term supply-demand equilibrium, a peaking generator should expect to earn $105/kW-year, more when supply is short and less when the system has excess supply. This reserve margin in 2019 was around 8 percent. For reference, analysis by the Brattle Group indicates that ERCOT’s equilibrium reserve margin is around 10.25 percent,[7] and in other regions regulated reserve margins and capacity market demand levels tend to be set in the 13-15 percent range. So, with last year’s reserve margin at 8 percent, we should expect net margins at or above $105/kW-year.

Figure 2 below displays the ERCOT market monitor’s estimated PNM from 2006-2018.[8] We calculated the numbers for 2019, and added them to the chart as the light blue dotted line. The PNM estimate for 2019 uses price data from each 5-minute SCED interval over the last year.

To calculate cumulative net profit, we subtract a peaker’s marginal operating cost and variable operations and maintenance (VOM) costs from the energy price and sum the calculated profits across all intervals in which price is greater than marginal cost. The marginal cost is calculated by taking the monthly average price of natural gas multiplied by the market monitor’s assumed heat rate for a new combustion turbine of 10 MMBtu/MWh. Similar to the market monitor’s PNM analysis, we also assume a VOM of $4/MWh and an annual total outage rate (including planned and forced outages and derates) of 10 percent.[9]

Figure 2: ERCOT Peaker Net Margin (2006-2019)

updated PNM

According to our analysis, 2019 brought in the highest peaker net revenue seen in the last eight years. Prior to that, only 2008 and 2011 were able to reach the break-even level of $105,000/MW-year, and 2018 came close to crossing that threshold. Using 2019 data, we estimate net revenue to be $145,620/MW-year, or 38 percent above the target level. Table 2 below displays the estimated cumulative monthly peaker net revenue for 2019:

Table 2: Estimated Cumulative Monthly Peaker Net Revenue (2019)

Month

Cumulative Monthly Net Revenue (Cumulative $/MW)

Jan

1,893

Feb

4,814

Mar

9,627

Apr

14,376

May

18,045

Jun

22,723

Jul

29,113

Aug

109,205

Sep

130,077

Oct

139,808

Nov

144,584

Dec

145,620

 

A close look at the net revenues of a new combined cycle natural gas units tells a similar story to the peaker analysis. While 2011 was the last year that these units exceeded their target net revenue level, which lies between approximately $110-125/kW-year, recent years have seen increases in net revenue, as shown in Figure 3. A similar analysis to the peaker net revenue estimation above indicates that 2019 exceeded the cost of new entry for a gas combined cycle generator. Using the same methodology as the above but instead assuming a heat rate of 7 MMBtu/MWh, the market monitor’s assumed heat rate of a hypothetical new combined cycle unit, we estimate net revenues over the last year to be approximately $160/kW-year, or approximately 28 percent above their target level.

Table 3 below displays the estimated cumulative monthly combined cycle unit net revenue for 2019:[10] 

Table 3: Estimated Cumulative Monthly Combined Cycle Unit Net Revenue (2019)

Month

Cumulative Net Revenue (Cumulative $/MW)

Jan

3,283

Feb

6,865

Mar

13,221

Apr

18,879

May

24,130

Jun

29,636

Jul

37,338

Aug

118,995

Sep

141,596

Oct

152,571

Nov

158,755

Dec

160,050

 

Figure 3: Combined Cycle Net Revenues[11]

CC net revenue

Signaling investment

In ERCOT, high spot prices signal to Retail Electric Providers to go out and sign more contracts with generators so they can shield themselves from high spot market prices. Those long-term power purchase agreements (PPAs) then are used by prospective generators to finance their new plants. An influx of 4,000 MW of solar and 5,000 MW of wind plants expected by this coming summer will likely take care of much of this need.[12] Loads need to also make sure they will have enough supply to cover themselves during periods of low renewable output. Demand response and storage, as well as conventional sources, are options for that need. The retirement of a 650 MW coal plant and the strong renewable growth will continue to reduce emissions in Texas.

Conclusion

2019 may have been the best test to date of the ERCOT market design. ERCOT now has two critical tools to ensure prices are at an efficient level and attract and retain resources when they are needed: a bid cap in the range of the value of lost load of $9,000/MWh, and an Operating Reserve Demand Curve (ORDC). With these features in place, and with a hot summer and low reserve margin, 2019 was a great test of this market design. The results indicate that no firm load was shed due to supply shortages, while the system did provide sufficient price signals to attract and retain needed resources. Results indicate that two types of units earned 38 and 28 percent above their target level, respectively. That, as well as the level of market response through new entry over the coming years, should answer a lot of questions about whether ERCOT’s unique market design works.

###

Appendix A: References for Peaker Net Margin and Combined Cycle Net Revenue Analysis

Potomac Economics (2019a), ERCOT Wholesale Electricity Market Monthly Report, https://www.potomaceconomics.com/wp-content/uploads/2019/09/Nodal_Monthly_Report_2019-08.pdf.

ERCOT (2019), Historical Real-Time ORDC and Reliability Deployment Price Adders ad Reserves, http://mis.ercot.com/misapp/GetReports.do?reportTypeId=13231&reportTitle=Historical%20Real-Time%20ORDC%20and%20Reliability%20Deployment%20Price%20Adders%20and%20Reserves&showHTMLView=&mimicKey

Potomac Economics (2019b), 2018 State of the Market Report for the ERCOT Electricity Markets, https://www.potomaceconomics.com/wp-content/uploads/2019/06/2018-State-of-the-Market-Report.pdf

References:

[1] Potomac Economics (2019), 2018 State of the Market Report for the ERCOT Electricity Markets, June 2019, p. 120, https://www.potomaceconomics.com/wp-content/uploads/2019/06/2018-State-of-the-Market-Report.pdf.

[2] Surendran, R., Hogan, W.W., Hui, H., and Yu, C. (2016), “Scarcity Pricing in ERCOT, June 2016, slide 4, https://www.ferc.gov/CalendarFiles/20160629114652-3%20-%20FERC2016_Scarcity%20Pricing_ERCOT_Resmi%20Surendran.pdf.

[3] The effects from “reliability deployments” include the “commitment of Resources through Reliability Unit Commitment (RUC),” the “deployment of Load Resources other than Controllable Load Resources (Increase demand by deployed amount in second Pricing run of RT SCED),” and the “deployment of Emergency Response Service (ERS) (Increase demand by deployed amount in second Pricing run of RT SCED).” Also note that the Real-Time On-Line Reliability Deployment Adder (RTORDPA) was added to ORDC calculations beginning in July, 2015. See Maggio, D., Moorty, S., and Shaw, P. (2018), “Scarcity Pricing Using ORDC for Reserves and Pricing Run for Out-of-market Actions, slide 28, https://www.pjm.com/-/media/committees-groups/task-forces/epfstf/20180329/20180329-item-07-scarcity-pricing-using-ordc-for-reserves-and-pricing-run-for-out-of-market-actions-ercot.ashx.

[4] Data available here: http://mis.ercot.com/misapp/GetReports.do?reportTypeId=13231&reportTitle=Historical%20Real-Time%20ORDC%20and%20Reliability%20Deployment%20Price%20Adders%20and%20Reserves&showHTMLView=&mimicKey.

[5] Gogo, N., and Eubank, B. (2019), “After Sizzling Hot September, Heat Lingers in Central Texas on First Day of Fall, September 23, 2019, https://www.kvue.com/article/weather/fall-heat-central-texas-hot-september/269-6c184b75-7c80-4a47-83bc-f7a3eb2ad27b.

[6] Potomac Economics (2019), p. 121.

[7] Newell et al. (2018), Estimation of the Market Equilibrium and Economically Optimal Reserve Margins for the ERCOT Region, October 12, 2018, p. vi, http://www.ercot.com/content/wcm/lists/143980/10.12.2018_ERCOT_MERM_Report_Final_Draft.pdf.

[8] Potomac Economics (2019), p. 122.

[9] Potomac Economics (2019), p. 112.

[10] Note: we divide $/MW values by 1,000 to convert to $/kW.

[11] Potomac Economics (2019), p. 113.

[12] Tomlinson, C. (2019), “The Competitive Texas Electric Grid Proves Critics Wrong Again, October 2, 2019, https://www.houstonchronicle.com/business/columnists/tomlinson/amp/The-competitive-Texas-electric-grid-proves-14482970.php?utm_campaign=CMS%20Sharing%20Tools%20(Premium)&utm_source=t.co&utm_medium=referral&__twitter_impression=true.

–Jesse Schneider is a Research Analyst at Grid Strategies and Michael Goggin is the Vice President of Grid Strategies.

 

Leave a Reply

Fill in your details below or click an icon to log in:

WordPress.com Logo

You are commenting using your WordPress.com account. Log Out /  Change )

Google photo

You are commenting using your Google account. Log Out /  Change )

Twitter picture

You are commenting using your Twitter account. Log Out /  Change )

Facebook photo

You are commenting using your Facebook account. Log Out /  Change )

Connecting to %s