Yes Energy News and Insights

Four Ways to Enhance Virtual Trading

Power trading is a nuanced, complex, and constantly-evolving industry. Monitoring 60-second power plant production and transmission line flow data enables real-time traders to stay a step ahead of market moves.

The Situation 

Since virtual traders only submit bids once a day, there’s a 14-38-hour lag between bids and settlement, depending on the market. As such, some virtual traders consider minute-by-minute monitoring optional for day-ahead trading. However, several of our virtual trading clients have an opposing view. Here are four ways Yes Energy® clients use Live Power® data to enhance their trading. 

1. Understanding Current Power Generation Activity and Transmission Constraints

In order for virtual traders to predict what’s going to happen, they need to understand current market dynamics. Without a strong foundation, day-ahead trading assumptions can be inaccurate. 

Seeing inter-zone transmission flow in order to spot real-time congestion is critical to that foundation. For example, when a new constraint appears, traders need to know how much power nearby plants are generating, as they may be exacerbating or relieving the congestion. Viewing changes in power plant output before, during, and after congestion occurs allows traders to better understand the current picture, informing their decision making for day-ahead trading. 

Image 1 -1
Image 1.5

Congestion shift factors and Live Power monitored plants, as seen through Yes Energy's tools.

Understanding the larger relationships behind plant behavior (e.g., when different plants turn on and off) strengthens that foundational understanding, as does being able to see constraints in the market and how they affect each node. Traders also need to know what type of fuel (i.e., gas, wind or solar) is affecting the constraint.

In the Constraint Profile module, traders can use Live Power data to create histograms showing the correlation between generation output and congestion. For example, if we look at last month’s data and divide a plant’s output into three categories (i.e.: 0-100 MW, 100-200 MW, 200-300 MW), we can analyze how many times we saw congestion on a particular constraint in each of those categories. From there, we can establish the larger patterns of congestion in order to diagnose what’s going on.

In addition, the Time Series Analysis module allows traders to view Live Power data alongside price, load, and generation forecast data in order to generate scatter plots. This information grants traders the ability to survey the magnitude of congestion as well as correlations within the congestion in order to provide clear visual queues between market fundamentals and outcomes.

Virtual traders have reported these tools help them create a solid foundation for bridging the gap between what’s happening right now and the day-ahead market, helping to test and validate their assumptions about future market conditions. 

Image 3-2

Live Power generation and line flow data viewed alongside nodal prices and transmission constraints.

2. Validating Reported Generation Outages 

Another popular use of the Live Power data is to confirm or disprove reported outages. Being able to see online/offline dispatchable capacity and unit-by-unit output helps traders track unit commitment and evaluate day-ahead spreads.

For example, if sources report that a plant will be offline from January 1st to January 10th, virtual traders have an advantage if they’re able to verify that information before placing their bids. Live Power allows traders to check in on January 1st to see if the plant turned off or if it skipped the outage. We may see that the generation dropped when we expected it to, it dropped a bit later or perhaps it’s still running strong, information which allows traders to adjust their bids accordingly.

If Live Power data confirms the initial report of generation dropping off as forecasted, the tool increases confidence that the forecast was indeed accurate, and the planned outage as prescribed will likely be valid. Because real-time data is delivered every 60 seconds, Live Power allows continual monitoring in case power generation comes back online. This visibility into the supply side of the equation allows traders the confidence to adjust their bidding strategy when activity doesn’t align with forecasted reports.

3. Leading Indicator of Wind Forecast Error 

Live Power tools also serve as a leading indicator of wind forecast error. With the rise in renewable penetration, wind has become a dominant force in pricing in many ISOs. When wind is forecasted to be low but comes in much higher (or vice versa), it creates a tremendous amount of volatility. Live Power helps traders anticipate that volatility.

If a trader sees a wind farm drop off in the real-time market, they can quickly adjust their pricing. Live Power enables traders to view renewable production on the geographic “edges” of the market, allowing them to make similar adjustments for the next day. For example, if a trader sees wind farms in West Texas (i.e., Amarillo, Lubbock, and Odessa) drop off more than expected as they’re placing their bids for the next day, they recognize that tomorrow’s wind will come in lower than expected and can adjust their bidding accordingly. In another likely scenario, they may see forecasted wind drop off a day early, helping them update their day-ahead bids to account for that risk. Viewing renewable production on the periphery of the market changes in ISO-wide fleet performance is important information for traders to track.  

The second leading indicator for forecast error is wind farms that turn on but generate less power than anticipated (or none at all) due to self-scheduled curtailments. Since wind energy facilities can claim tax credit on every kilowatt-hour of electricity sold for 10 years after a facility is placed into service, when those tax credits expire, financial incentives may cause the farm to run at lower levels when prices drop below a certain point. By analyzing historical data, Live Power allows traders to view curtailment activity at various wind farms. For example, if a trader sees that a wind farm has a capacity of 200 MW, but is only producing 100 MW even with blustery wind, they’re likely curtailing their output. Traders can then combine that knowledge with Yes Energy’s pricing information to discern the cutoff price, and more accurately forecast that wind farm’s future output levels under similar conditions.

4. Forecast Generation for the Next Day 

The best way for virtual traders to price their bids is to have a generation forecast for the next day. To build that forecast, they need to have a solid set of historical information. In one such example, combining Live Power's historical generation data with fuel cost information allows traders to see how much it costs a coal or gas plant to generate power. If a trader expects power prices to be greater than that amount, they expect the plant to run, and vice versa. They can then use those historical outputs to build a model to predict likely output.  

Image 4-2

 Live Power data from Barney M. Davis and gas prices which could be used to train a model. 

For wind, traders could use historical data alongside the National Weather Service forecast and wind speed data to create an initial model for a wind farm. Once they train their model on those aspects, they can clarify the relationship between those variables, then add future variables to predict generation output.

Creating a complex generation forecast takes time and skill; traders without an R or Python programming background can create forecasts based on different scenarios and/or assumptions. For example, if a trader is analyzing the relationship between a gas plant and prices at a particular node and has used Live Power historical data to identify full output, 50% output and 0% output, they can then run models at each setpoint to determine price expectations, blending potential outcomes to calculate a final price forecast.

Stay Ahead of Market Moves

With today’s rapidly changing power grid, virtual traders need instant visibility into what’s happening to stay a step ahead of market moves. Understanding what’s happening right now as well as what’s happened historically is the foundation for virtual trading. To determine a price forecast, traders need a generation forecast, and to create a generation forecast, they need historical generation data.

Live Power delivers fastest-in-industry reporting of generation and transmission flow data to support your trading. Real-time data is delivered every 60 seconds and fully integrated within Yes Energy power analytic tools, providing unparalleled visibility into the supply side of the equation. 

Want to learn more?