Constellation Power Source Generation, Inc.

Results of Test Burns Conducted to Evaluate Coal-Based Synthetic Fuel at Brandon Shores Station

December 3, 2002

A.Introduction

Section C of the October 24, 2002 test burn protocol for the evaluation of coal-based synthetic fuel to be used at Constellation Power Source Generation’s (CPSG’s) Brandon Shores station specifies the results to be reported on the test burn conducted. The protocol indicates that CPSG is to submit a data report to the Maryland Department of the Environment (“MDE”) Air and Radiation Management Administration (“ARMA”) and Department of Natural Resources (“DNR”) Power Plant Research Program (“PPRP”) providing the block hourly average SO2 and NOx emission rates (in lb/mmBtu) for the unit tested (in this case Unit 1) recorded by the continuous emission monitoring system (CEMS) during the 24-hour synfuel test burn period. Results of proximate and ultimate analyses of the synfuel and visible emission readings are also to be provided.

CPSG is to compare the CEM emission data and visible emissions readings during the test burn of synfuel to applicable Federal and State emission limits, representative historical emission to show no significant increase in emissions or the values that served as the basis of a CPCN, as applicable. The test plan established 1st Quarter 2002 NOx emissions and the SO2 emission limit established by the Brandon Shores CPCN and federal New Source Performance Standard as baseline emissions to compare synfuel emissions for both Brandon Shores units. More specifically, the test burn protocol established baseline values of 0.48 lb of NOx/mmBtu for Unit 1 and 0.52 lb of NOx/mmBtu for Unit 2 and 1.2 lb of SO2/mmBtu for both units.

Fuel analyses results are also to be included in the data report comparing the as-fired proximate, ultimate and trace element (metals) of the synfuel to values expected from Brandon Shores base coal.

Detailed discussions of the test burn procedures and results are provided in Sections B and C that follow.

B.Test Burn Procedure

A twenty-four hour long synfuel test burn began on Brandon Shores Unit 1 at noon on November 12, 2002.

The synfuel used for the test burn was manufactured by Sempra Energy at its synfuel facility currently located in southeastern Virginia. The synfuel tested was produced from coal meeting the specifications required for its use at Brandon Shores along with the Accretion binder described in the test protocol.

UNIT 1 TEST BURN COAL HANDLING

Test burn synfuel was unloaded from the barge unloader and directed to the north side of the coal yard stacker-reclaimer on November 8, 2002, thus allowing the test synfuel to be separated from coal usually used at Brandon Shores. Plant fuel handlers began sending synfuel to the Unit 1 bunker on Monday night, Nov. 11, at midnight (0000 hours on Tuesday, Nov. 12). Unit 1 continued to operate normally while synfuel was fed to the coal bunkers and by the late morning of November 12, Unit 1 was burning 100% synfuel - the Unit 1 coal bunker typically holds about 7 hours of coal. To ensure that there would no question that synfuel was being fired, the test period officially began at noon on November 12 and ended at 1159 AM on November 13. Synfuel samples were collected every 6 hours during the test burn; the samples were composite samples of all five of the unit’s coal feeders. Ultimate, proximate and trace element analyses were completed on the synfuel samples for the unit as indicated in the test protocol.

C.Test Burn Results

C1.CEMS Data

Table 1 provides the 24 hourly average values and summary statistics of gross load, SO2 emission rate (lb/mmBtu), NOx emission rate (lb/mmBtu) and percent opacity values recorded by unit CEMS during the Brandon Shores Unit 1 synfuel test burns. In addition, summary statistics of SO2 and NOx emission rates recorded over the 1st Quarter of 2002 are provided for comparison against the test burn data for each unit.

Figure 1 is a time series graph of the 24 hourly test burn NOx emission rate values for Unit 1 compared to the average and the upper control limit (calculated as the average plus two standard deviation unit values) calculated from the 1st Quarter 2002 baseline data. Figure 2 is a time series graph of the 24 hourly test burn SO2 emission rate values for Unit 1 compared to the 1.2 lb of SO2/mmBtu emission limit established in the Brandon shore CPCN and federal New Source Performance Standard (NSPS) applicable to Brandon Shores.

Figure 3 includes time series graphs of hourly SO2 and NOx emission rate values for Unit 1 from the 1st Quarter 2002 baseline data and are intended to demonstrate the long-term variability of both SO2 and NOx emission rates with time.

In general, NOx emissions rates for Unit 1 during the test burn were comparable to the emission rates of the 1st Quarter 2002 baseline. The average NOx rate for the Brandon Shores synfuel test period of 0.44 lb/mmBtu is less than the average rate of 0.48 lb/mmBtu for Unit 1 during the baseline period. Figure 2 illustrates that the Brandon Shores Unit 1 test burn NOx rate values were consistently less than the average value and well below the upper control limit value from the baseline data. Unit 1 NOx values were slightly greater than the average baseline value during the synfuel test for (3) one-hour periods (0200 AM to 0400 AM on 11/13/02). These somewhat elevated NOx rates (about 0.5 lb/mmBtu) can be attributed to the reduction in load level as a result of typical late night reduced system demands.

Figure 4 is a scatter plot of NOx emissions as a function of gross load from the 1st Quarter of 2002 baseline period for Unit 1 with synfuel test burns data superimposed for comparison. Although not specifically called for in the test protocol, Figure 5 offers a similar scatter plot of SO2 emissions as a function heat input. These figures illustrate two key points in evaluating emission data:

  1. There is substantial “natural” variability in hourly emission rate values, even for a given operating level, and
  2. The superimposed synfuel test data is largely indistinguishable from the baseline data (i.e., is easily within the upper and lower boundaries of variability of the baseline data).

Hourly opacity readings obtained from the Unit 1 CEMS as summarized in Table 1 were generally below the applicable 10% opacity limitation and were typical of values normally achieved by Unit 1. However, Unit 1 experienced elevated opacity during the 0100 PM and 0200 PM hours on 11/12/02 as a result of a malfunction of its electrostatic precipitator (ESP) particulate matter control equipment. The elevated opacity was unrelated to the synfuel being burned and opacity was reduced to normal levels after the ESP problem was corrected. Additional discussion of this event is provided in Item C3. Visible Emission Observations, below.

C2.Fuel Analyses

Tables 2 through 4 summarize the results of the proximate, ultimate and metals analyses specified in the test protocol. Laboratory analyses of the composite samples gathered during the test burn periods were performed by Commercial Testing and Engineering, Inc. of Baltimore, MD. As indicated in these tables, the results of the proximate, ultimate and metals analyses were consistent with values typically found for coal. The corresponding values of the undiluted binder are also presented in these tables for comparison to the synfuel test values.

C3.Visible Emission Observations

Although the overcast and rainy weather conditions were not favorable for observations, an hour of visible emissions observations was made on Unit 1 on 11/12/02 beginning at 0100 PM. As indicated above, Unit 1 experienced elevated opacity in this time period because of a malfunction of Unit 1’s ESP. More specifically, power was lost to an electrical cabinet that energizes the precipitator’s “rappers” (devices that mechanically clean dust collection surfaces in the precipitator). As a result, the precipitator dust collection surfaces were not cleaned as they normally are, leading to an excess accumulation of flyash on them. Flyash continued to accumulate until it then began falling off of the precipitator collection plates under its own weight, becoming partially re-entrained the exhaust gas stream and resulting in the opacity that was observed. Plant operators restored power to the rapping system and precipitator operation and opacity levels returned to normal.

An additional hour of visible emissions observations was made on Unit 1 on 11/13/02 beginning at 0900 AM under more favorable weather conditions and normal ESP operations; no visible emissions or condensing plumes were observed and normal opacity monitor readings were recorded during this observation period. Both sets of observations were performed by a certified observer according to EPA Reference Method 9 (see Appendix for the report of these readings).

D.Conclusion

The NOx and SO2 CEMS data, visible emissions observations and fuel analyses called for the test protocol and presented in this report clearly demonstrate that the use of a coal-based synthetic fuel manufactured using the Accretion binder and tested at Brandon Shores Unit 1 did not result in increased emissions. In addition, no changes or modifications were made to existing fuel handling, storage, preparation or delivery equipment or to methods of operation of Unit 1 in order to accommodate the use of synfuel during the test burns. For these reasons, CPSG concludes that the use of Accretion binder based synfuel at the Brandon Shores station is not a modification of an electric generating unit as defined under PSC law. As indicated in the October 24, 2002 test plan, since both Brandon Shores units are essentially identical, any effects from the synfuel would be the same on both units; therefore the results and conclusions from synfuel testing conducted on Unit 1 also apply to Unit 2.

1