Continuing where we last left off in Part I, we’ll now describe our simulation methodology and a few of the key takeaways from the simulation study. 

Methodology

The primary objective of our study is to develop a reusable simulation-based framework to robustly model residential V2X value streams in wholesale markets. To accomplish this, we used data inputs for residential driving behavior, residential building loads, and ERCOT market price data. To generate simulated driving profiles, we used Texas-specific residential driving data for commuters and non-commuters from the National Household Travel Survey. This data was used to calibrate parameters such as the number of trips per day, the correlation between travel distance and travel time, and the frequency of trips to a certain location at different hours of the day. These calibrated parameters were used as inputs to a Python-based driving-profile simulator called Emobpy to generate annual driving profiles at 15-min resolution. For load, we used simulated annual time-series profiles from NREL’s ResStock database. For our experiments, we considered a single EV with a battery capacity of 70 kWh and a residential charger with two power capacity options: 11.5 and 7.7 kW. We considered two energy markets (day-ahead and real-time) and four ancillary markets (regulation-up, regulation-down, regulation-reserve, and non-spin). The year 2022 was considered as the target year for our analysis. More than a thousand annual V2X scenarios were simulated that considered different combinations of driving and load profiles, market access, battery cycling constraints, and whether export is allowed. To determine the upper bound on V2X revenue potential, historical market prices in 2022 were used as a substitute for price forecasts.  

Key Takeaways from the Simulation Study

In this section, we discuss the three key takeaways and implications for residential V2X revenue potential in ERCOT.

Question 1: How much annual revenue can be generated by a residential EV owner by participating in the Texas wholesale markets?

In an idealized setting where an EV is always connected to a charger, a stationary vehicle can earn as high as $4700 per year for a typical residential charger, assuming exports are compensated and full market participation is allowed. We ran simulations that considered four market participation scenarios, where an EV can participate in 

• Day-ahead energy market only
• Day-ahead and real-time energy markets only
• Day-ahead energy and ancillary markets only
• Day-ahead, real-time, and ancillary markets (full-market access)

The revenue distribution for the different market scenarios is illustrated in Figure 1. Just participating in the less volatile day-ahead market can earn up to $1935 annually, while access to both day-ahead and ancillary markets increases revenues to $3361 - an 84% increase.   Similarly, additional participation in the more volatile real-time energy market increases annual revenues by over 73% compared to the case with only day-ahead market access. For the full-market access scenario, real-time energy emerged as the most valuable market by contributing 30% of the total revenue, followed by the day-ahead market at 27%, and regulation-up at 24%.

Figure 1: Distribution of annual residential V2X revenue under different ERCOT market participation scenarios for a stationary EV

Question 2: How valuable is the residential V2X market in Texas?

We observed a broad range of annual V2X revenues for scenarios that jointly consider driving behavior and market dynamics, ranging anywhere from $300 to $3000 per charger. Figure 2 depicts the distribution of the annual V2X revenues for all the non-stationary EV cases when full market participation is allowed.

Figure 2: Distribution of annual V2X revenues

Thus, driving behavior does limit value, especially as drivers in Texas average around 15,000 miles annually and many of them, such as daily commuters, do not have their vehicles at home during the day. However, we also observed that revenue value in ERCOT  is concentrated in a few key hours of the year mainly driven by market price trends. This trend is observed in Figure 3, where only a small fraction of hours in the year result in high revenues (observed most distinctly for the volatile real-time market). This highlights the importance of correctly predicting price spikes in a volatile wholesale market and unlocking the true potential of V2X by connecting EVs to the grid during such peak-price events.

Figure 3: Hours of the year sorted according to the hourly prices for each wholesale market

Question 3: How much additional value can be extracted by using a sophisticated V2X energy scheduling platform compared to a basic platform?

From our analysis, we observed that an excellent V2X platform - with optimized forecasting, battery health, customer selection, and market participation capabilities - can generate over 10 times more revenue for a residential customer than a basic V2X platform. We define a basic platform as one that allows participation in the day-ahead market only, uses simple forecasts (e.g., using historical averages as forecasts), considers highly constrained battery management, and does not have the ability to change driving behavior patterns. Figure 4 shows the incremental increase in annual revenue as more sophisticated V2X optimization and forecasting features are added to a basic platform. For ERCOT, the highest marginal revenue increase occurred when cross-market participation and perfect forecasting features were included (313% improvement), followed by optimal customer selection (41%) and driving behavior change (22%). 

Figure 4: Change in revenue as sophisticated features are integrated into V2X software platform 

A V2X platform’s ability to change driving behavior via feedback on future revenue potential is especially pertinent in a volatile market like ERCOT with infrequent but large price spikes. For example, Figure 5, illustrates the revenue potential (annual and per day) as a function of the number of days when the vehicle is always available for charging. The highest revenues are concentrated in a relatively small number of days, where there were 6 days in a year where staying at home would earn an EV owner at least $50 per day. If a sophisticated V2X platform can accurately predict the revenue potential of these high-spike days, it can potentially change driving behavior by incentivizing EV owners to exploit such opportunities.

Figure 5: Annual revenue (on the left) and daily revenue (on the right) as a function of the number of days when a vehicle is available 100% of the time for charging or discharging.

Concluding Remarks

The unique characteristics of the Texas market grid, coupled with high volatility in wholesale prices and increasing storage capacity installations, make it an attractive proposition for V2X. In this two-part blog series, we investigated the revenue potential in wholesale energy and ancillary markets in Texas for residential V2X, subject to uncertain driving behavior, storage, and charger characteristics. Using a simulation-based optimization framework, the results clearly demonstrate the high revenues a residential EV owner can earn by utilizing intelligent EV storage scheduling and market participation features provided by Fermata Energy’s proprietary dispatch optimizer software. Stay tuned for the next post in this series where we will extend our analysis to study the revenue potential in California and highlight the differences in regulations, market trends, and revenue estimates compared to Texas.