Federal and state limits on emissions of mercury (Hg) from power plants typically require periodic or continuous measurement of Hg levels by approved reference methods (RMs) and continuous emission monitoring systems (CEMs). For example, the U.S. Environmental Protection Agency's Clean Air Mercury Rule calls for the installation by January 1, 2009, of certified Hg monitors on all stationary sources that emit more than 29 pounds of the toxic pollutant annually. Reporting of data for compliance monitoring will begin on January 1, 2010. This allows less than 18 months for installation and certification of Hg CEMs and sorbent trap monitoring systems, and 2.5 years until mandatory reporting of their readings.
In July 2006, under a program organized by Lehigh University's Energy Research Center, the performance of four commercial continuous and semi-continuous Hg CEMs and three sorbent trap (ST) systems from different manufacturers was compared in the field at Allegheny Energy's Armstrong Power Station. The results were compared to the U.S. and EU reference methods. The project was conducted with the support of the U.S. EPA, EPRI, U.S. electrical utilities (particularly Allegheny Energy), Italy's Ministry of Economic Development, and the Institute for Environment and Sustainability of the EU's Joint Research Centre.
Based on the number of tests, the Armstrong program represents one of the most comprehensive examinations of Hg field measurement conducted in the U.S. Of the 294 total test samples (Table 1), 72 used the Ontario Hydro Method (OHM), 36 used samples obtained by the EU Reference Method, and 186 tests used ST samples.
Table 1. Tests performed at Armstrong Power Station. Source: Lehigh University Energy Research Center
The perfect host
Allegheny Energy's Armstrong Power Station (Figure 1) has two units rated at about 190 MW (gross) each. The station receives coal by truck from a number of local mines, as well as Virginia coal delivered by rail. Two electrostatic precipitators (ESPs) in series control particulate emissions. As a result, Armstrong's stack opacity is very low, typically in the 3% range. Although the plant normally runs in baseload mode, it occasionally is used to help control grid frequency.
1. The tests' host. Armstrong Power Station in Adrian, Pa., has two coal-fired units, each with a gross rating of 190 MW. Courtesy: Allegheny Energy Inc.
Units 1 and 2 at Armstrong discharge flue gas to the atmosphere through two 1,000-ft steel stacks enclosed in a common concrete liner. The 14.5-ft diameter of each stack makes the average flue gas velocity about 75 ft/sec at full load. The main CEM platform, about 350 ft above ground level, is accessible by an elevator.
Field testing of the CEMs was performed on Unit 2 in July 2006 by a joint team of U.S. and EU researchers. During the first 10 of 18 total test runs, the plant fired locally mined bituminous coals with high and variable Hg content. For Tests 11 through 18, it burned a low-Hg coal from Virginia.
The continuous Hg monitors were located at ground level. The flue gas samples, extracted from the stack and conditioned by sample extraction probes, were delivered to the Hg analyzers by 450-ft-long heated umbilical lines. The samples were also sent to a semicontinuous Hg monitor located at the main CEM platform via a short, heated umbilical cord.
Two paired Ontario Hydro Method trains and two EU computer-controlled automated isokinetic trains obtained samples in accordance with the OHM and EN-13211 reference methods for Hg measurement. The samples were analyzed on-site and overnight in a Western Kentucky University (WKU) mobile chemical analysis laboratory.
Samples of flyash and coal were collected at least three times per day to make them more representative. The coal samples were collected from the coal mills, and the flyash samples from the ESP hoppers.
Coal composition measurements indicated that the properties of the coal being burned changed after Test 10 (Figure 2). During Tests 1 through 10, firing local coal produced Hg emissions in excess of 20 µg/wsm3. During Tests 11 through 18, a low-Hg Virginia coal was burned, lowering mercury emissions below 10 µg/wsm3. This change in coal properties allowed test data to be collected over a wider range of mercury concentrations, from 7 to 23 µg/wsm3. The Virginia coal had a 50% lower Hg content, as well as a lower chlorine (Cl-) content, than the local coals. The Cl-/Hg ratio for the Virginia coal was more than 50% higher than that of the local coals.
2. Half and half. The mercury and chlorine content of as-received coal used for the tests at Armstrong Power Plant Source: Lehigh University Energy Research Center
OHM: The gold standard
The Ontario Hydro Method is the oldest and most-proven RM. It is, however, manpower-intensive and time-consuming, and therefore expensive to use. The OHM was initially developed to support the U.S. EPA's effort to characterize and inventory Hg emissions from the nation's coal-fired power plants. After several revisions by the American Society for Testing and Materials (ASTM), it was finalized as standard test method ASTM Method D 6784-02. Since then, the OHM has been used by the EPA and other research organizations as both a research and potential compliance tool. The OHM is the EPA's RM for measuring total (HgT) and speciated mercury.
At Armstrong Power Station, speciated mercury samples were collected from sampling ports at the CEM elevation using EPA Method 17. Two sets of field blanks and reagent blanks were taken at each location and analyzed for quality control purposes. Operators had to manually adjust procedures to maintain an isokinetic sampling rate. It took about two hours to collect a sufficient sample volume, confirming that OHM is a manpower-intensive and expensive Hg measurement method.
To comply with EPA Part 75 Relative Accuracy Test Audit criteria, OHM sampling must be performed in paired trains, and the paired samples must be within 10% of the relative standard deviation (RSD) of results. The precision of the OHM when used as an RM is a critical parameter. The OHM precision data, collected from various reports on pilot-scale and field measurement of gas-phase Hg by the OHM, indicated that RSD is in the 5% to 15% range (Figure 3).
3. OHM precision. Precision of the Ontario Hydro Method. Source: Lehigh University Energy Research Center
The OHM results were analyzed to determine their precision and other performance parameters. Despite changes in Hg concentration caused by changes in fuel quality and unit load, the HgT values, measured simultaneously by four individual OHM sampling trains located in two stack ports, were very close to each other.
Email
Comments
Print
Reuse
