[BETA] Fortnightly

To the Editor:

An article in the June 2006 edition of Public Utilities Fortnightly  (”Pondering PJM's Energy Price Run-Up ”) by Howard Spinner of the Virginia State Corporation Commission staff raises the question of whether the observed increase in PJM average system prices in the second half of 2005 was the result of fuel-price increases and increased loads, or the result of market power. The article suggests that the increase in prices in PJM was not the result of higher loads or fuel prices, but was the result of the exercise of market power. It also suggests that the result of this asserted market power was an increase in net revenues to generators.

Spinner’s assertions are based on a series of incorrect assumptions about the types of units on the margin, the frequency of unit types on the margin, and the efficiency of the units on the margin. The results reported in the Spinner article are incorrect and have been previously demonstrated to be incorrect (see “PJM Energy Prices—2005: Response to Howard M. Spinner Paper” available at www.pjm.com/markets/marketmonitor/reports.html). There is no evidence of market power in 2005 or of increased market power in the latter half of 2005. The increase in PJM prices in 2005, particularly in the latter half of 2005, was the result of increased fuel prices and increased demand (see the Market Monitoring Unit (MMU) 2005 State of the Market Report for relevant data and analysis. The report can be found at www.pjm.com/markets/market-monitor/som.html).

In the PJM energy market, prices are set by the marginal unit or units. Under economic dispatch, the lowest-cost available units are dispatched in ascending order of cost to meet the load in real time. The last unit dispatched, the highest-cost unit of the dispatched units, is the marginal unit. When there are binding transmission constraints, multiple marginal units and multiple prices result. Thus, the behavior of marginal units is critical in evaluating the competitiveness of the PJM market.

Based on the significance of marginal units in determining prices, Spinner attempts to estimate the hourly system marginal cost of energy by estimating the costs of marginal units. His estimates are based on a combination of hourly data on the type of fuel on the margin in the Real-Time Market, posted by PJM, and his own assumptions regarding unit type and heat rate. The simple monthly average of the resultant hourly marginal cost is compared to the simple monthly average of PJM hourly prices in the Day-Ahead Market, and the difference between the two is the basis for his assertions regarding market power.

PJM currently posts the time-weighted hourly marginal fuel type for each hour of the year for the Real-Time Market. In order to translate the hourly fuel type into information about marginal costs, Spinner had to make assumptions about the type of unit burning the fuel. The three fuel types that are primarily on the margin in PJM are coal, oil, and natural gas.

While it is generally reasonable to assume that coal is burned in a baseload steam unit,