November 1, 1998, through April 30, 2000

Project Title: CATALYTIC OXIDATION OF NO IN FLUE GAS FOR CAPTURE IN WET SCRUBBERS

ICCI Project Number: 98-1/1.1E-1

Principal Investigator: Gary A. Robbins, CONSOL Inc., Research & Development

Project Manager: Ronald H. Carty, ICCI

ABSTRACT

CONSOL R&D, the research arm of Consolidation Coal Company, which mines about 5 million tons of Illinois coal each year, began development of NOx oxidation catalyst technology that was expected to result in cost-effective NOx removal by wet flue gas desulfurization (FGD) scrubbers. Wet FGD is the preferred technology for removing sulfur dioxide generated during the combustion of high- and medium-sulfur Illinois coal in the production of electric power. In this work, catalysts were tested for the ability to oxidize NO to NO₂ for efficient capture of NO₂ in wet scrubbers. The concept of catalytic in situ oxidation of NO to NO₂ was demonstrated. Several catalysts were identified that oxidized about 50% of NO to NO₂ in the absence of SO₂. Unfortunately, some of these catalysts were unable to oxidize 50% of the NO when SO₂ is present in synthetic flue gas. The catalysts also oxidized a substantial amount of SO₂ to SO₃, which can raise the acid dew point and cause corrosion in power plant ducts and air heaters.

Project background and objectives are provided in the report. The report describes the design and construction of the catalyst test unit, the catalyst test runs, test unit operating difficulties, and suggestions for future development.

Because selective catalytic reduction (SCR) NOx control technology is costly, CONSOL R&D initiated laboratory development of a low-cost alternative NOx control method. This technology was based on the use of existing or new wet flue gas desulfurization (FGD) scrubbers. The overall project goal was to commercialize, by the year 2003, a low-cost alternative NOx control method suitable for use on coal-fired utility boilers that could achieve a 65 µg/MJ (0.15 lb/MMBtu) NOx emission rate. In 2004, additional NOx controls are expected when the recently revised NOx SIP regulations set the framework for NOx emissions limits. The novel NOx control process is based on catalytic oxidation of NO to NO₂. At typical flue gas conditions downstream of the boiler economizer, most of the NOx (>95%) is NO. The residual oxygen (3% to 6%) in the flue gas is the oxygen source. Based on bench-scale tests, (Shen and Rochelle, 1995; Zemansky et al., 1993), NO₂ is more soluble in aqueous solutions than NO. An existing flue gas desulfurization (FGD) wet scrubber can remove the NO₂. The catalyst can be placed upstream of the air heater section, downstream of the air heater but upstream of the particulate collector, or downstream of the particulate collector. A desirable characteristic of the catalyst is that it minimize the oxidation of SO₂ to SO₃. Based on equilibrium calculations, 60% to 95% removal of NOx is possible, but laboratory evaluation is required to determine the actual NOx removal efficiency. The process can be used in combination with combustion modification (i.e., low-NOx burners) to control NOx emissions. The process requires a relatively small capital investment to achieve high NOx removal.

This technology will promote the use of wet FGD scrubbers for combined NOx and SOx control. This will benefit the use of medium- to high-sulfur coals, such as those produced by the Illinois coal industry, by preserving or expanding their markets.

The planned research activities included three tasks. However, because a suitable catalyst was not identified, Task 3 was not completed.

Task 1 - A catalyst screening program to identify a catalyst capable of oxidizing at least 50% of the NO to NO₂ while oxidizing less than 5% of the SO₂ to SO₃ under typical coal-fired flue gas conditions. A catalyst test unit was constructed that uses simulated flue gas. A major catalyst manufacturer provided ten candidate catalysts for evaluation in this program.

Task 2 - Extended catalyst testing to simulate boiler cycling and observe catalyst deactivation. Originally, boiler cycling was to be simulated by multiple start-up/shut-down cycles. However, at the suggestion of the catalyst manufacturer, these tests were performed by conducting a 4-hour lineout without SO₂, followed by an 8-hour lineout with 2000 ppm SO₂, prior to measuring NO and SO₂ conversion.

Task 3 - A process economic evaluation to estimate the process economics, identify significant cost areas for process improvements and cost reductions, and evaluate options for process scale-up. However, because none of the test catalysts satisfied the criteria of oxidizing at least 50% of the NO to NO₂ while oxidizing 5% or less of the SO₂ to SO₃, a process economic evaluation was not justified.

The catalyst test unit consists of a tubular quartz reactor that is wrapped with heating tape and temperature controlled. The honeycomb catalyst samples were 1" dia. x 4" long. The catalyst test unit design conditions were 300-600 F, 10,000-60,000 hr^-1 space velocity (std L/hr gas flow per L catalyst bed volume) of simulated flue gas containing 100-1000 ppmv NO, 1000-3000 ppmv SO₂, 3-7 vol % O₂, 10-14 vol % CO₂, 4-8 vol % H₂O, and the balance N₂. For this range of space velocity, and the catalyst bed volume chosen, the combined flow (including water vapor) of simulated flue gas will be 8.6-51.5 SLPM. Gas will be fed from the plant nitrogen supply and mass flow controllers were used to meter the flue gas component gases into the mixing and preheat zone prior to entering the catalyst reaction zone. Exit gas was analyzed using continuous NO/NOx and SO₂ analyzers; SO₃ yield was calculated from changes in the SO₂ concentration. Catalyst test duration was defined by steady-state conditions, as determined by gas analyzer data. Several test runs were 24 hours in duration.

The test strategy was to screen all catalysts at one test condition, and select the best catalysts for more extensive testing. The conditions chosen for the screening test were 600 F, 10,000 hr^-1 space velocity, 1000 ppmv NO, 2000 ppmv SO₂, 5 vol % O₂, 12 vol % CO₂, and 6 vol % H₂O. During the extensive test program, the space velocity and temperature was varied at several levels, two levels were used for the SO₂ and NO concentrations, and the concentrations of H₂O, O₂ and CO₂ were tested at only one level.

The catalyst test unit consists of five subsystems: (1) The gas supply and mixing subsystem provides inert gas (nitrogen from the plant supply) for reactor start-up and shutdown, inert gas (nitrogen from the plant supply) for the balance of the simulated flue gas, and CO₂, O₂, SO₂, and NO from cylinders to produce the simulated flue gas. Special features provide for safety of personnel and equipment. (2) The humidification/gas preheat system consists of metered water injection into the gas stream, followed by evaporation and preheating to approximately 230-260 F with the gas flowing through a coil in an oil bath. (3) The reactor system consists of a custom-built three-sectioned tubular quartz reactor; two sections are heated and one is cooled. In the first section, the simulated flue gas is preheated to reaction temperature, 300-600 F. The second section contains the catalyst, and the third section cools the gas to about 300 F at the reactor outlet. (4) The flue gas conditioning and analysis system provides reactor effluent gas drying, pressure control, and gas analysis for NO/NOx, SO₂, CO₂, and O₂. (5) The data recording system logs data, including mass flow rates of component gases, temperatures of the reactor zones, and the concentrations measured by the gas analyzers.

After construction of the catalyst test unit was completed, several equipment problems were identified and corrected during the integrated system start up. Failure and re-start modes of several components were evaluated as part of the safety considerations. Specific procedures were set up for area monitoring for NO and SO₂ using color indicator tubes. A unit operating and safety manual was written.

In the absence of SO₂, all of the catalysts tested exhibited some degree of activity for oxidation of NO to NO₂ over the temperature range 350-600 F at the conditions tested (10,000 hr^-1 space velocity, 1000 ppmv NO, 2000 ppmv SO₂, 5 vol % O₂, 12 vol % CO₂, and 0-6 vol % H₂O). Several catalysts met the target of 50% or greater conversion of NO at these conditions. Performance with each catalyst depended on temperature.

Most of the catalysts showed moderate to substantial conversion of SO₂ to SO₃ at the test conditions. The SO₂ to SO₃ conversion was dependent on space velocity and temperature. In the absence of NO, oxidation of SO₂ to SO₃ still occurred. To be considered a suitable candidate for large-scale testing, a catalyst must oxidize at least 50% of the NO to NO₂ while oxidizing no more than 5% of the SO₂ to SO₃. None of the catalysts tested satisfied both criteria simultaneously.

Neither the NO oxidation nor the SO₂ oxidation reactions were dependent on the presence of water vapor in the gas stream. With 6% water vapor in the gas, the SO₃ formed was deposited on tubing surfaces downstream of the reactor as sulfuric acid condensation. Without water vapor in the gas (and with NO in the feed gas), the SO₃ formed was deposited as a white, solid, sulfur-nitrogen compound that would eventually plug the system.