Control of Mercury Emissions by Injecting Powdered Activated Carbon (PAC) Presentation to Utility MACT Working Group May 13, 2002 EPA, RTP, NC Michael D. Durham, Ph.D., MBA ADA Environmental Solutions 8100 SouthPark Way B-2 Littleton, CO 80120 303 734-1727

Outline ADA-ES DOE/NETL Hg Control Program Background on PAC Injection Technology Results from PAC with an ESP Results from PAC with a FF Conclusions and Future Plans

ADA-ES Hg Control Program Full-scale field testing of sorbent-based mercury control on non-scrubbed coal-fired boilers Primary funding from DOE National Energy Technology Laboratory (NETL) Cofunding provided by: Southern Company Wisconsin Electric PG&E NEG EPRI Ontario Power Generation TVA First Energy Kennecott Energy Arch Coal

Project Overview Perform first full-scale evaluations of mercury control on coal-fired boilers (up to 150 MW equivalent). Evaluate effectiveness of sorbent-based Hg control (activated carbon). Test several different power plant configurations. Document all costs associated with Hg control.

DOE/NETL Test Sites Test Site Coal Particulate Test Control Dates Alabama Power Bituminous HS ESP Spring Gaston COHPAC FF 2001 Wisconsin Electric PRB Cold Side ESP Fall Pleasant Prairie 2001 PG&E NEG Bituminous Cold Side ESP Summer Brayton Point 2002 PG&E NEG Bituminous Cold Side ESP Fall Salem Harbor 2002

Coal-Fired Boiler with Sorbent Injection and Spray Cooling Ash and Sorbent ESP or FF Hg CEM Spray Cooling H2O Air

Semi-Continuous Mercury Analyzer Heater Dry Air CVAA Flue Gas Chilled Impingers Gold Trap Mass Flow Controller Micro controller with Display Waste

Sampling Time Required

Comparison of OH and S-CEM*, Long Term Tests (10 lbs/MMacf)

Capture of Vapor Phase Hg by Solid Sorbents Mass Transfer Limits (getting the Hg to the sorbent) Removal increases with particle concentration Produces percentage removal independent of concentration Particle control device (FF vs ESP) is a critical parameter Sorbent Capacity to hold Hg depends upon: Sorbent characteristics Temperature Mercury concentration Concentrations of SO3 and other contaminants

Equilibrium Adsorption Capacities at 250°F Upstream and Downstream of SO3 Injection

WEPCO Pleasant Prairie Testing completed fall of 2001 PRB coal ESP only Spray cooling SO3 conditioning system

Activated Carbon Storage and Feed System

ESP Configuration, PPPP Spray Cooling Carbon Injection

Powdered Activated Carbon Injection System

Baseline Hg Measurements (g/dscm) Location Particle Bound Oxidized, Hg2+ Elemental, Hg0 Total, Hg Inlet ’99 0.16 2.29 6.21 8.65 Inlet ‘01 1.84 2.34 11.39 15.55

Mercury Trends Week 1

Response Time for PAC Injection on an ESP

Carbon Injection Performance on a PRB Coal with an ESP

Long Term Trend Data

Speciated Mercury Measured by Ontario Hydro Method (10 lbs/MMacf) (microgram/dncm) PARTICULATE ELEMENTAL OXIDIZED TOTAL Baseline ESP Inlet 1.97 12.22 2.51 16.71 ESP Outlet 0.01 9.80 6.01 15.82 Removal Efficiency 99.5% 19.8% -139.3 5.3% PAC Injection ESP Inlet 0.98 14.73 1.73 17.44 ESP Outlet 0.00 4.27 0.44 4.71 Removal Efficiency 100.0% 71.0% 74.5% 73.0%

Alabama Power E.C. Gaston Alabama Power Company E.C. Gaston Electric Generating Plant Unit 3, Wilsonville, AL 270 MW Firing a Variety of Low- Sulfur, Washed Eastern Bituminous Coals Particulate Collection System Hot-side ESP, SCA = 274 ft2/1000 acfm COHPAC baghouse supplied by Hamon Research-Cottrell Wet Ash Disposal to Pond

Site Test Configuration with EPRI TOXECON at Alabama Power Plant Gaston Sorbent Injection COHPAC Fly Ash (2%) + PAC Coal Fly Ash (98%) Electrostatic Precipitator

S-CEM Duct Traverse

Example of S-CEM Data

Response Time of PAC Injection with a Fabric Filter

Mercury Removal vs. Injection Rate

Pressure Drop Increase from PAC Injection

Mercury Removal vs. Injection Rate PAC Rate Limit Due to Pressure Drop

5-Day Continuous Injection

Average Mercury Removal Long-Term Tests Gaston, Ontario Hydro (microgram/dncm) PARTICULATE OXIDIZED ELEMENTAL TOTAL Baseline COHPAC Inlet 0.09 9.54 5.97 15.60 COHPAC Outlet 0.01 11.19 3.34 14.54 Removal Efficiency 89.1% -17.3% 44.1% 6.8% PAC Injection COHPAC Inlet 0.23 6.37 4.59 11.19 COHPAC Outlet 0.12 0.91 0.03 1.05 Removal Efficiency 45.6% 85.7% 99.3% 90.6%

Comparison of Sorbent Costs for a Fabric Filter and an ESP

Conclusions (PAC General) PAC injection can effectively capture elemental and oxidized mercury from both bituminous and subbituminous coals Additional field tests and long-term demonstrations are necessary to continue to mature the technology Fabric filters provide better contact between the sorbent and mercury than ESPs resulting in higher removal levels at lower sorbent costs New COHPAC FF’s will have to be designed to handle higher loadings of PAC to insure high (>90%) mercury removal Conventional FF’s should not require any modifications for PAC

Conclusions (Response to Concentration Variations) Response times to changes in inlet concentrations: Feedback data from outlet CEMs—tens of minutes Impact of changes in injection rate: tens of minutes to hours Long averaging times will be required to recover from upsets Injection at somewhat higher rates will make the technology more capable to handle inlet fluctuations PAC injection lends itself to the use of feed rate parameters as a definition of Maximum Achievable Control Technology

Future Plans Short-term testing at additional sites Long-term testing PG&E Brayton Point (Bituminous coal, large ESP) 6/ 2002 PG&E Salem Harbor (Bituminous coal, SNCR, large ESP) 9/2002 * TBD (PRB coal, small ESP) 3/2003 * Southern Company (Bituminous coal, small ESP) 8/ 2003 Long-term testing *Alabama Power (Bituminous coal, COHPAC FF) 2002-2003 *CCPI Program (PRB Coal, COHPAC FF) 2004-2006 *CCPI Program (Bituminous Coal, COHPAC FF) 2004-2006 * Proposed

For More Information www.adaes.com www.adaes.com/mercury.htm Link to other mercury related web sites Publications/reports www.adaes.com/MercuryPublic.htm Public information on DOE/NETL Mercury Control Program www.netl.doe.gov/products/environment/index.html DOE/NETL Website