The Electricity Strategy Game Explorer is a visual tool for experimenting with different auction scenarios in the Electricity Strategy Game, a class exercise run in Energy and Environmental Markets (MBA212) at UC Berkeley's Haas School of Business. To summarize for those not familiar with the game, I'll quote professor Severin Borenstein's introduction from the instructions:
The Electricity Strategy Game is designed to replicate the world of strategic trading in wholesale electricity markets. Players utilize generation portfolios to compete in a sequence of daily electricity spot markets. Spot market conditions will vary from hour to hour and day to day. Players must develop strategies to deploy their assets over this sequence of spot markets while accounting for the cost structure of their portfolio, varying levels of hourly demand, and the strategies of other players
The Electricity Strategy Game (ESG) Explorer assumes that you have already purchased a generating portfolio, and are interested in optimizing your bids in the hourly spot market auctions. As such, it shows you a supply curve including all generators with their potential bids along with the demand forecast for the hour you have selected. You can explore various bidding strategies that you and your opponents might use, and see the effect of your choices on profitability.
Please note that since this is a class project and time is limited, I have not tested this visualization extensively for cross-browser compatibility. It seems to work well in Chrome 9 (where it was developed), Firefox 3.6 and Safari 5 on my OS X 10.6.6 machine. I expect that it will work well in these browsers on Windows and Linux as well; it may not fare as well in Microsoft Internet Explorer (IE), although it was designed to standards so hopefully if current versions of IE don't support it, future ones will.
ESG Explorer supports three modes of interaction: adjusting bids, adjusting the confidence interval around the demand forecast, and changing the day and hour (thus changing the demand forecast).
The current economic disposition of each power plant is represented by a colored box. The color indicates which portfolio the plant belongs to. The width of the box corresponds to the capacity of the plant, in MW. The height of the bottom edge of the box shows the variable costs associated with running the plant, in $/MWh. The thicker top edge of the box shows the plant's current bid price in $/MWh. Thus, the area inside the box represents the plant's expected profits in a pay-as-bid auction, if its bid is accepted. The box will be shaded if the bid is actually accepted (i.e. the bid is below the current clearing price, which is determined by intersecting the supply and demand curves) or hollow if the bid was not accepted. If the plant is one of the marginal producers in the current scenario, its box will be partially shaded, according to the proportion of power it is able to sell at its current bid price. If you let your mouse rest on a box, a tooltip will appear containing numeric values for the plant's bid, marginal cost (MC) and expected profit.
If you have selected a day with a uniform price auction, the shaded profit box will extend above the bid price for plants whose bids clear, indicating that they are paid the clearing price rather than the bid price.
To adjust a plant's bid, simply drag the thick line representing the bid to the new bid value. When you release the drag, the plant will re-sort according to its new position in the supply curve, and a new clearing price and quantity will be calculated if applicable. A "bid editor" will appear above the plant whose bid you just changed, allowing you type in a more precise bid if you like. When you have entered your adjustment (or if you don't wish to adjust) press enter or click the ✔ button. The bid editor will disappear, applying any changes you made.
You can reset all bids to their initial values by clicking the "Reset Bids" button at the lower left corner of the page. Note that the default bids are a conservative "optimal" bid governed by the rules:
This bid is "optimal" in the sense that it provides the most profit possible while keeping the risk of not having a bid clear near zero. In practice, this is about the most conservative bid that would be reasonable, and in many situations higher bids will be better.
The demand forecast for the hour you have chosen is represented by the slightly oblique vertical black line. However, in the ESG the actual demand may differ somewhat from the forecast. According to the game rules, the standard error of the forecast is 3%. The gray lines to the left and right of the demand forecast represent the edges of a confidence interval around the forecast demand. The interval is initially ±3% (one standard deviation). You can adjust it to whatever level of certainty you're comfortable with (up to 6 standard deviations) by dragging the gray lines. The status bar in the lower-right will show you numerically the size of your confidence interval.
Note that the clearing price and quantity displayed on the chart and the shading of the individual power plants are driven solely by the median demand forecast and are not affected by the confidence interval. The confidence interval only affects the range of predicted profits, discussed below.
To explore a different day or hour, use the day and hour selection boxes in the bottom bar, or use the arrow buttons to more forward and backward one hour at a time. Note that the ESG Explorer is currently limited to ESG year 1, as year 2 has more complex rules that I have not modeled.
The panel above the supply and demand chart shows the expected total profit for each portfolio for the hour. The profit value is obtained by adding up all capacity sold at the accepted price, subtracting marginal costs of generation (fuel and variable operations & maintenance), and subtracting ¼ of each plant's daily operation and maintenance cost. Since there are four hours in a game day, this last subtraction accounts for the fixed operation costs of the plant, divided equally throughout the day.
The bold vertical line with the dollar figure label shows the predicted profits if the demand comes in exactly as forecast. The lighter shaded area shows the range of possible profits within the demand confidence interval you have selected. Note that the shaded bar may be "missing" if a portfolio's profits are not affected by changes in demand because the supply curve near the forecast is very flat, etc.
If I have more time to invest in this project, I would like to:
If you have other ideas or thoughts to share, feel free to e-mail me.