What did LMPs tell us this summer in PJM's new neighborhood?
With one summer under its belt as a member of PJM, ComEd has been called a complete success by some, boring by others. On the whole, prices in the region were quite flat, and spreads between the 1,281 price nodes in the region were small to non-existent. In fact, the best opportunity for a generator to make money in this market was to export power to a neighboring market.
Some of the lack of action this summer surely is due to the fact that it was a cool summer. But from a transmission perspective, it's clear that ComEd was a well-oiled machine. Using data from Energy Velocity's Market Ops and Weather products, we can see the impact of transmission and weather on ComEd this summer.
ComEd's LMP Market
ComEd became part of PJM last May, adding 1,281 new pricing points to the region. To simplify the analysis, we will focus only on peak prices from the real-time market between June and August.
First, the average ComEd zonal price of $40.24, which is a weighted average price for the entire ComEd zone, was 30 percent lower than the equivalent zonal average in PJM of $57.24, shown in Figure 1. Neither market was particularly volatile, but ComEd was as much as 30 percent more volatile than PJM.
Second, ComEd as a region demonstrated extremely low transmission congestion. Figure 2 presents the average peak prices for the summer by price node. Areas shaded red represent the highest prices and those shaded blue represent the lowest. This figure is useful in presenting the spatial dispersion of prices in the ComEd region. However, a close look at the legend reveals that the regions shaded red are only about $2/MWh higher than those areas shaded blue.
Figure 3 tells a similar story by comparing the spread of congestion prices in ComEd across all 1,281 price nodes and the spread across all price nodes in PJM and PJM West combined. As is clear in the figure, the two regions can't even be compared on the same scale because 99 percent of the average on peak ComEd congestion prices fell between $41/MWh and $44/MWh.
Knowing that the summer as a whole was flat, we decided to search all 2,200 hours of the summer for any hour where congestion might have registered in the LMPs. Significant price differences arose on the grid only from June 6 to June 8. These first days of the summer also brought the first heat wave of the season, which likely caught the market off-guard. While this caused the largest spread of the summer, and some fairly high prices, 93 percent of the LMP locations still demonstrated prices in the relatively narrow band of $80/MWh to $90/MWh.
The spatial pricing trends have implications for the value of location in the ComEd region. It was surprising to see that the highest prices during the summer of 2004 were in the outskirts of Chicago and in the much smaller urban area of Rockford, Ill. But even in these areas, the average on-peak premium was only $2/MWh more than the value of power at the lowest value locations.
In fact, the highest prices in ComEd this past summer were not in the ComEd region at all, but at ComEd's interfaces. The highest-priced interface was the export interface with the New York ISO at $58.36/MWh. ComEd's interfaces with Wisconsin Electric and Illinois Power followed close behind at just over $55/MWh. While exporting power to New York is not a likely transaction, exporting to Wisconsin and Southern Illinois is common. Transmission costs to these regions were higher this summer. This might be expected given that ComEd has a reserve margin of just about 30 percent, and neighboring regions, especially Wisconsin, have lower reserve margins. Many of ComEd's neighbors are net importers of power.
A Cool Summer
Part of the explanation lies in an unusually cool summer for 2004. Using the standard measure of cooling-degree days (CDD) for the period June through August we see total CDD of 442. By comparison, 2002 and 2003 had CDD totals of 845 and 602, respectively. In fact, one needs to go back to 1992 to find a cooler summer. Typical summers in Chicago have 671 CDD.
Not only was the summer of 2004 cool on average, but there was a noticeable lack of extreme days, too. Daily high temperatures extended into the low 90s only 3 times, none of which occurred during peak hours. Over the same period in 2002, temperatures reached the 90s for 21 days including several extended hot spells.
What If It Were Warmer?
Because 2004 was such a cool year in the ComEd region, we estimated what prices would have looked like in the summer of 2002. As shown in Figure 4 (), the number of CDD in 2004 was 442 compared with 845 in 2002. The difference can also be seen in Table 1, which shows the distribution of temperatures at Chicago O'Hare in 2002 and 2004. Note that for each temperature range greater than 74 degrees, 2002 had a higher percentage of on-peak hours in the range compared with 2004. Note also that more than 20 percent of the on-peak hours in 2002 had temperatures at 85 degrees or above, compared with less than 2 percent in 2004.
We wondered how the ComEd market might heat up during a summer like 2002. To estimate the impact on prices, we calculated, by temperature range, the average and standard deviation of the real-time, on-peak LMP for the ComEd zone as shown in Table 2 (). Assuming that during a summer like 2002 the average price would be similar within each temperature band, we calculated a weighted average and standard deviation of the price (weighted by the percentage of hours in each temperature range). In 2004, the weighted average price and standard deviation were $39.98 and 19.7 respectively. For a year like 2002, we estimated an average on peak price of $48.63, or about 20 percent higher. Likewise, we estimated that prices would be more variable, as indicated by the higher standard deviation estimate of 20.4.
A 20-percent increase in prices would have a dramatic impact on revenues. Consider unit 8 at the Joliet 29 plant in Will County, for example. The coal-fired steam turbine has a summer capacity of approximately 518 MW. From 2000 through 2003, this unit had an average capacity factor of 76.8 percent during on-peak hours. A 518-MW base-load unit generating at 76.8 percent capacity for the summer on-peak hours would realize approximately $16.5 million in revenues during the on-peak period but would realize over $20.1 million based on our hotter weather scenario if the capacity factor remained unchanged.
Similarly, consider a peaking unit such as Elgin Energy Center CT03. This gas turbine unit has a summer capacity of 117 MW and a marginal cost of approximately $68/ MWh. Based on the average prices in Table 2, it would be profitable to operate this unit any time the temperature was greater than 85 degrees. A 117-MW peaking unit that usually generates in the on-peak period any time the temperature is above 85 degrees would have operated for 19 hours in 2004 and approximately 209 hours in a hotter summer like 2002. Assuming average prices by temperature band in 2004 and an 80-percent capacity factor when running, this same unit would realize just over $100,000 in energy revenues in 2004, compared with about $1.3 million in a summer like 2002.
ComEd's new LMP market is supposed to use prices to highlight significant load pockets and transmission congestion points. Looking at the past summer it appears that there isn't much congestion in ComEd right now. Of course, neither the grid nor the generating capacities of the region were challenged by the weather of 2004. For the most part, the market provided no clear incentive to either site generation in a particular location, or to invest in a particular transmission infrastructure. Generators in Rockford were the best off, with an average $2 on-peak location premium over the entire summer. The best opportunities to make money in the ComEd region were at the export interfaces, indicating that to fetch the highest prices, generators would be best off to locate in another market altogether.
A warmer year, or even a typical year, likely would result in higher overall energy prices. Our simple analysis estimated a 20-percent increase from a summer like 2004 to a summer like 2002, assuming a linear relationship between CDD and market price. It is likely that the relationship between weather and market price is nonlinear and would result in even higher prices.
During a cool summer like 2004, a transmission trader would find ComEd boring. But ComEd is likely well prepared for hotter summers to come. In addition, ComEd appears to be a good market for generators who lack the ability to analyze more lively transmission markets like those in PJM proper. However, independent generators in the region may want to find a deal on firm transportation to New York if they want to make a buck during a summer like 2004.
Articles found on this page are available to subscribers only. For more information about obtaining a username and password, please call our Customer Service Department at 1-800-368-5001.