By Randy Smith, Fossil Energy Research Corp.
Regulatory and legislative requirements such as the Clean Air Interstate Rule (CAIR) NOX emission caps have driven the need to develop NOX control technologies for existing coal-fired power plants.
In November 2006, Power Engineering magazine reported on the Department of Energy, National Energy Technology Laboratory’s (DOE/NETL) advanced NOX emissions control research and development (R&D) efforts to provide more cost-effective options for coal-fired power plants to comply with ever more stringent emission limits.
Included in the article, titled “DOE/NETL’s Advanced NOX Emissions Control Technology R&D Program,” was a summary of the development and testing of an in situ device for real-time catalyst deactivation measurements. Fossil Energy Research Corp. (FERCo) developed the device to reduce selective catalytic reduction (SCR) operating costs through optimized catalyst management. It was tested at Southern Co.’s Plant Gorgas Unit 10. Now, five years of testing data is available, as shown in Figure 1.
FERCo’s KnoxCheck in situ measurement device collects real-time SCR performance data by continuously measuring catalyst activity. As the data is collected, it is analyzed by a catalyst management software program, providing information on boiler operating conditions that negatively impact catalyst activity. This information can then be used to optimize boiler operation with respect to catalyst deactivation rate and the catalyst replacement schedule.
FERCo conducted tests on the SCR installed at Plant Gorgas Unit 10, which is owned by Southern Co. subsidiary Georgia Power. The 700 MW tangentially fired plant burns bituminous coal. The SCR started operation in 2002 and is designed with two parallel SCR reactors containing three extruded honeycomb catalyst layers plus a spare. FERCo installed KNOXCheck on each of the three catalyst layers in one of the reactors prior to the 2005 ozone season and took six sets of catalyst activity measurements throughout the summer at four-week intervals.
The KNOXCheck uses a self-contained ammonia feed system to control ammonia concentration and extracts upstream and downstream flue gas samples to analyze the inlet and outlet NOX concentration. The in situ measurement technique is similar to the laboratory measurement, whereby a small auxiliary ammonia injection grid (AIG) is located above the section of catalyst to be tested and ammonia is added such that the local NH3/NOX ratio exceeds 1.0 and the maximum NOX reduction across the small test section is measured.
Catalyst activity is assessed using a metric known as reactor potential (RP), which provides a measure of the overall potential of the SCR reactor to reduce NOX by accounting for both catalyst deactivation and catalyst layer blockage. The reduction in reactor potential can be characterized by the ratio RP/RPo, where RP is the current measurement and RPo is the value for the initial fresh, unexposed catalyst layer. RP/RPo results for the individual layers illustrate the accelerated deterioration of catalyst performance that occurs over time in the first layer compared to the second layer and similarly the second layer compared to the third. The RP/RPo ratio for the overall reactor dropped from around 0.7 to 0.6.
KNOXCheck operated during the 2005 to 2008 ozone seasons and for the entire 2009 year at Plant Gorgas Unit 10. While multiple test locations exist across each layer of catalyst, all the gas samples and ammonia injection controls were centralized into two cabinets. The probes for inlet and outlet NOX sampling and supplemental NH3 injection were installed in standard 4-inch ports at the inlet and outlet of each catalyst layer.
Five years of testing results are shown in Figure 1 as a function of RP/RPo, where RP is the current measurement of reactor potential and RPo is the value for the initial fresh, unexposed catalyst layer when the SCR started up in 2002.
The RP/RPo results for the individual layers reflect the effect of catalyst layer replacements (Layer 1 before the 2006 ozone season and Layer 3 during 2009) and also illustrate the deterioration of catalyst performance that occurs over time.
Replacing Layer 1 with new catalyst prior to the 2006 ozone season provided an opportunity to quantitatively compare the results of the in situ KNOXCheck measurements to both the activity provided by the catalyst vendor and an activity value provided by a third-party catalyst testing laboratory that tested an unused sample of catalyst. The results of this comparison testing showed good agreement of the in situ KNOXCheck activity measurements to the vendor’s value of activity, as well as the third-party laboratory’s measurements. The KNOXCheck measurements were 6 percent higher than the value from the vendor, while the third-party laboratory’s value was 3 percent lower than the vendor’s reported activity.
FERCo plans to continue the KNOXCheck measurements at Gorgas Unit 10 in 2010. The in situ testing there was conducted on-site by FERCo personnel, but a similar system installed at a FERCo client in Texas is configured for remote access via a secured virtual private network (VPN) connection to the plant network. In the near term, plans call for incorporating the KNOXCheck system into several new refinery SCR installations, as well as some gas turbine and biomass SCR applications.
KNOXCheck provides a direct measurement of reactor potential, accounting for the actual flue gas flow rate and blockage values, not a calculated quantity based on the design area velocity and an estimate of the blockage. A true reactor potential at actual operating conditions can help to make informed catalyst management decisions.