< Back

Retrofit of Gas Combustion Equipment and Low-NOx Oil Atomizers at a 550-MW Oil-Fired Utility Boiler

Robert C. Carr, Zigmund J. Frompovicz, Coen Company, Inc.
Marco Alberti, Ansaldo Energia
Christopher J. Nagel, ComEd

ABSTRACT:

ComEd is conducting a project to retrofit natural gas combustion equipment and low-NOx oil atomizers at Units 4 and 5 of the Collins Station in Morris, Illinois. Collins Units 4 and 5 are 550-MWg opposed-wall-fired boilers which were originally designed to burn No. 6 fuel oil. Each unit has 28 dual-register burners, overfire air, and flue gas recirculation (FGS) to the windbox.

The retrofit approach consisted of adding gas-firing capability while retaining the major components of the burners, i.e., air registers, burner frame and support assemblies, and oil combustion equipment. Specific hardware furnished included: (1) gas burner assemblies with flame detectors and adjustable pokers; (2) automated sliding sleeve air registers for the main burners and overfire air ports; (3) gas ignitors with flame detector and high-energy spark; (4) oil gun retracts; (5) new flame stabilizers which are compatible with gas and oil firing; and (6) segmented V-Jet oil atomizers designed to achieve low-NOx emissions. Designs of the new stabilizers and oil atomizers were based on improved combustion technology referred to as Reduced Emissions and Advanced Combustion Hardware, or REACH.

The new combustion equipment was installed at Collins Unit 4 during a ten-week outage in the spring of 1996 (much of the gas piping and associated electrical work was conducted pre-outage). Results showed that maximum load of 550-MW was achieved on gas within one week of main burner startup. At maximum load on gas, NOx emissions were 0.16 lb/MBtu with overfire air and 9% FGR. This paper describes the retrofit approach, equipment supplied, and results of combustion optimization tests on gas fuel.

INTRODUCTION:

BOILER DESCRIPTION:

Collins Units 4 and 5 are identically designed, Carolina type, natural circulation, opposed-wall-fired boilers manufactured by the Babcock & Wilcox Company. Each unit has a gross generating capacity of 550 MW, and is equipped with windbox flue gas recirculation (FGR), flue gas recirculation to the furnace hopper, and overfire air (OFA) via manually-operated NOx ports. The boilers were originally designed to burn No. 6 fuel oil (only), and are equipped with 28 dual-register burners. The burner arrangement is two elevations of seven burners each on the front and rear walls. The vertical distance between the burners is fifteen (15) feet, and the horizontal separation is approximately six (6) feet. A dividing plate in the windbox separates the two burner elevations. However, there are no dampers to bias combustion air between the burner elevations. There are seven NOx ports on each firing wall, with one port above each burner column. The vertical distance of the NOx ports above the top burner elevation is eleven (11) feet. Combustion air for the NOx ports is supplied from the windbox for the to burner elevation. The steam generators are comprised of water-cooled hoper-bottom furnace, vertical pendant superheater and reheater sections, and a horizontal bank economizer section. The units are designed to generate 3,800,000 LB/hr of superheated steam at 1,005°F and 2,410 psig. The burners at Collins 4 and 5 are operated in pairs and trios, i.e., burner pairs on the wings and burner trios in the center. New oil-burning trios in the center. New oil burning equipment was retrofitted by EPT and ComEd in 1991 to improve oil atomization and combustion performance. The combustion technology retrofitted is referred to as Reduced Emissions and Advanced Combustion Hardware, or REACH.1,2 The retrofit included: (1) replacement of Y-jet atomizers with internal-mix atomizers, (2) replacement of diffusers with compound-curve-vane swirlers for flame stabilization, and (3) conversion of the atomization steam system from constant steam pressure at 150 psig to constant steam-to-oil differential pressure of 10 psid over the load range.

Each of the twenty-eight burners was originally equipped with mechanically coupled, inner and outer air registers operated by a single electric drive. The out air registers are 54-inch diameter, and the inner registers are 42-inch diameter. There are twelve, 12-inch wide flat blades on each of the inner and outer registers. The air registers have multiple purposes: (1) to induce swirl to the combustion air for aerodynamically stabilizing the flame; (2) to proportion air between the inner and outer air zones of the burner for NOx control; (3) to permit throttling of air flow through achieve uniformity among all the burners; and (4) shutoff of combustion air to the burners. Approximately 45% of the combustion air (not counting what flows through the inner air register and 55% through the outer air register. Binding of the air registers was historically a maintenance problem at Collins. Moreover, when adjusting the registers for optimum flame shape on individual burners, the uniformity of the air flow distribution among the burners was affected.

PROJECT DESCRIPTION:

Electric Power Technologies, Inc. (EPT) was retained as prime contractor by ComEd to design and furnish retrofittable, gas-burning equipment to convert the existing burners at Collins Units 4 and 5 from oil-only to gas-and-oil operation. EPT subcontracted with Ansaldo Energia s.p.a. (Ansaldo) and Combustion Technologies, Ltd. (CTL) for supply of specific burner components. EPT, Ansaldo, and CTL supplied the following for each unit:

• Gas Burner Assemblies. Gas manifold and six adjustable pokers per burner (designed and supplied by Ansaldo), flame stabilizer, flame detector, and flexible hoses to connect to the main gas header. Figure 1 shows a schematic drawing of the gas burner elements.
• Automated Sliding Sleeve Air Registers for the Main Burners. Sliding sleeves with electric linear drives were retrofitted to permit control of air flow balance among the burners, and positive shut-off of out-of-service burners. The sliding sleeve assembly consisted of a two-sleeve design as shown in Figure 2, one sleeve for the out air register and one for the inner air register. The inner and outer vane air registers were decoupled, and the electric drives were removed. Manual adjustment mechanisms were then installed to permit the positions of the inner and outer registers to be independently adjusted for optimum flame shape.
• Automated Sliding Sleeve Air Register for NOx Ports. Sliding sleeves with electric linear drives were installed to permit automatic adjustment of air flows among the overfire air flows among the overfire air ports used to control NOx emissions. The sliding sleeves were of similar design to the main burners, except only a single sleeve per NOx port was required.

Figure 1, Gas Burner Assembly Retrofit at Collins Unit 4

The primary elements consisted of the gas manifold, pokers with spuds, flame stabilizer, and gas ignitor

Figure 2, Schematic Drawing of the Sliding Sleeve Assembly Retrofitted to the Dual-Register Burners at Collins Unit 4

The sleeves are shown in the closed position

• Gas Ignitors. The existing No. 2 oil ignitors and low-energy spark rods were replaced with gas ignitor assemblies which included flame stabilizer, ignitor flame detector, high-energy spark, and flexible hoses to connect to the ignitor gas header. The ignitors were designed for a maximum heat input of 15 million Btu/hr.
• Ancillary Burner Equipment. Included new oil gun retracts, flexible hoses for oil and atomizing steam, miscellaneous piping, etc.

EPT also furnished installation and startup support, and combustion optimization services. The modified burner were guaranteed to achieve the following:

• NOx emissions £ 0.18 LB/MBtu when burning gas and £ 0.28 LB/MBtu when burning No. 6 oil for all boiler loads.
• Particulate matter mss emissions less than 0.10 LB/MBtu.
• Opacity £ 10% for steady state conditions, and £ 20% for operating transients.

The cost for the combustion equipment, and design, installation, and startup support furnished by EPT for Units 4 and 5 was less than $3/kW.

EQUIPMENT INSTALLATION & STARTUP:

Flame scanners were programmed simultaneous with startup and checkout of the ignitors and main burners. Since new burner elements were installed, and the original ignitor sight pipes were designed for viewing No. 2 oil flames, the ignitor sight pipes were repositioned to insure that the flame scanners had an unobstructed view of the new, gas ignitor flames. Similarly, it was necessary to evaluate (and move as needed) the positions of the scanner sigh pipes for the main burners to achieve good flame discrimination of the gas and oil flames. This is an important consideration for conversion project from oil-only togas- and oil fired operation.

BURNER CHARACTERIZATION TESTS:

Prior to starting the unit up on gas, burner air flow balancing was performed. This was followed by ignitor and main burner startup tests. Presented below are results from the air flow balancing tests, and combustion optimization tests on gas.

BURNER AIR FLOW BALANCING:

Air flow probes were fabricated to permit measurement of relative air flow among the burners. The probes were designed to be inserted down the oil gun guide pipe, and were equipped with a multi-tap annubar section which could be extended into a position perpendicular to the axial flow direction. Cold air flow tests were performed with the unit off line and the forced draft fans operating at an air flow rate equivalent to full load. With the sliding sleeves in the full open position, the relative air flow to each burner was measured to determine the baseline air flow distribution. The sliding sleeves were then adjusted to balance the air flow among the burners. Figures 3 and 4 present burner air flow balance for the baseline and modified condition, respectively. As shown, before balancing the percent deviation from the average burner air flow varied from -34% to +23%. Balancing reduced the spread of the deviation such that all but two burners were within ±10%. Air flow balancing is typically performed with the inner and outer vane-type air registers full open.

COMBUSTION OPTIMIZATION - GAS TEST DESCRIPTION:

The objectives of the testing were to define operating modes that maintained NOx emissions below 0.18/MBtu over the load range when firing natural gas. Recommendations for excess oxygen, position of the NOx ports, and windbox flue gas recirculation (FGR) rates were developed from tests conducted at six load conditions from 100-MW to 548-MW

MEASUREMENT METHODS:

A multipoint extractive system was used to obtain flue gas samples for analysis of NOx, CO, CO2, and O2. Samples were extracted from probes installed in the "A" (north) and the "B" (south) flue gas ducts immediately upstream of the air heaters. Gaseous samples were obtained from probes installed in four ports in each duct. Three sampling probes were installed in each port (long-, intermediate-, and short-length probes). Thus, samples were obtained from 24 points in the flue gas duct, ensuring a representative sample of the boiler exhaust gases. Twenty-four sample lines were strung from the probes to a mobile emissions monitoring laboratory that was located at the ground level adjacent to Unit 4. The gas samples were conditioned to remove water, and were then directed to emission monitors. Kilkelly Environmental Associates (KEA) supplied and operated the emissions monitoring laboratory under the direction of EPT.

NOx, CO, CO2, and O2 were measured using EPA continuous monitoring procedures, and EPA certified calibration gases. O2 was measured with a Servomex Model 570A instrument, CO2 with a ACS Model 3300 instrument, CO with a TECO Model 48 instrument, and NOx with a TECO Model 10A chemiluminescence instrument.

Figure 3, Baseline Air Flow Balance for the Burners at Collins Unit 4

Figure 4, Modified Air Flow Balance for the Burners at Collins Unit 4

Data from the plant Yokogawa zirconium oxide O2 probes installed in the windbox were used to calculate the windbox flue gas recirculation rate using relationships developed by EPT between windbox O2, economizer excess O2, and windbox FGR. At each test condition, the boiler control room data listed below were recorded.

• Steam temperatures, pressures, and attemperation requirements.
• Windbox pressure and temperature, and furnace draft.
• Air heater inlet and outlet air/gas temperatures.
• Plant excess O2 from six Yokogawa zirconium oxide probes.
• Plant windbox O2 from four Yokogawa zirconium oxide probes.
• NOx emissions and opacity from the plant CEMS.
• FD/ID fan amps, boiler control settings, fuel oil flow rate, and temperature.
• Natural gas flow rate to the boiler (ignitors and main burners).

EMISSIONS TEST RESULTS:

NOx emissions as a function of load for gas fuel are presented in Figure 5 for three scenarios: (1) uncontrolled (no NOx controls), (2) 100% overfire air, and (3) 100% overfire air and 10% FGR. Also shown for comparison purposes are uncontrolled NOx emissions data on No. 6 fuel oil obtained before the gas conversion (fuel nitrogen content = 0.22%).

The data show that uncontrolled NOx emissions on gas increased sharply with load, i.e., from 0.14 LB/MBtu at 100-MWg to 0.58 LB/MBtu at 500-MWg (it was interesting to note that NOx emissions on gas were comparable to oil before the gas conversion project). Opening the NOx ports (100% OFA) reduced NOx emissions at 500-MWg by 53% from 0.58 to 0.27 LB/MBtu, and adding 10% FGR further reduced NOx emissions by 48% to .14 LB/MBtu (10% FGR was the maximum because only one of the two GR fans was operable). At 540-MWg, Nox emissions were easily maintained below the NOx emissions limit of 0.20 LB/MBtu by using OFA and FGR. Carbon monoxide emissions were less than 130 ppm for all the test conditions.

Figure 6 illustrates the effect of FGR on NOx emissions at 500-MWg. All data were obtained with only one GR fan in service. The top curve shows NOx emissions without OFA, whereas the lower curve shows the effect of FGR with the NOx ports 100% open. As expected when firing natural gas, FGR was extremely effective in reducing NOx emissions for both operating modes. With 100% OF AND 10% FGR, NOx emissions were reduced 75% below uncontrolled levels to 0.14 LB/MBtu. If two GR fans were available, an FGR rate close to 15% would be possible, which would be expected to further decrease NOx emissions to 0.10 LB/MBtu, or less. However, the use of higher levels of FGR at maximum load must be balanced against the need to maintain acceptable superheater metal temperatures.

The effect of excess O2 on NOx emissions is presented in Figure 7. With 100% OFA and 10% FGR, NOx emissions increased approximately.

Figure 5, NOx Emissions vs. Load at Collins Unit 4.

Data presented for natural gas include uncontrolled, wit OFA ad no FGR, and with OFA and FGR. Pre-retrofit uncontrolled emissions data for oil firing are also shown for comparison purposes.

Figure 6, NOx Emissions vs. FGR at Collins Unit 4 for Natural Gas Firing

The FGR data were obtained using one GR fan.

Figure 7, Effect of Excess O2 on NOx Emissions at Collins Unit 4 for Gas Firing

Note the significant effect of ignitors on NOx emissions.

0.05 LB/MBtu as excess O2 was changed from 1.5 to 2.5%. An interesting observation was the effect of ignitors on NOx emissions. As shown, operation with all the ignitors in service increased NOx emissions by approximately 0.02 LB/MBtu. This effect was documented consistently during the testing.

BURNER AND BOILER OPERATION:

Flame stability of the new ignitors and gas burners was excellent. The main flames were well anchored to the flame stabilizers, and flame shape varied little with changes in excess O2, FGR, or load. At full load with 100% OFA and maximum FGR, the furnace was clear with the exception of the individual flames. Each flame exhibited a distinct blue-color at the base, followed by a yellow luminous envelope. There was no flame carryover above the OFA port elevation.

Prior to the retrofit there was concern that maximum load when burning gas might be restricted because of excessive tube metal temperatures in the convective sections of the boiler (a common concern with gas conversion projects). In this regard, heat transfer model calculations predicted that maximum load would be restricted to less than 500-MW. Fortunately, the model results were overly pessimistic and maximum load of 500-MW was achieved.

BOILER VIBRATION:

A common concern when burning gas fuel in utility boilers is combustion-induced boiler vibration (i.e., "rumble"). Therefore, during the combustion optimization test efforts were directed to identify any operating where boiler vibration could occur. In this regard, no boiler vibration occurred when Collins Unit 4 was operated over the load range with maximum FGR and overfire air. However, it was possible to induce boiler vibration at high loads (>525-MWg) with the unit at conditions outside the normal operating envelope, i.e., at very low excess O2 (<0.5%), or with a single burner trio out of service (air slides closed) in the top elevation. Avoiding these non-typical operating conditions is accomplished without difficulty by plant operators.

The operating conditions which induced vibration on gas were not previously investigated with oil fuel. Therefore, it is not certain if the vibration observed was combustion induced, or was caused by other factors such as vibrations propagating to the boiler from the combustion air and FGR ductwork (both which can vibrate significantly under some flow conditions). Tests planned for October-November on oil fuel will further investigate this possibility to determine if the vibration is unique to gas, or has origins elsewhere.

SUMMARY:

ComEd is conducting a project to retrofit natural gas combustion equipment and low-NOx oil atomizers equipment and low-NOx oil atomizers at Units 4 and 5 of the Collins Station in Morris, Illinois. Electric power Technologies, Inc. (EPT) was retained as prime contractor by ComEd to design and furnish retrofittable, gas-burning equipment to convert the existing burners at Collins Units 4 and 5 from oil only to gas- and oil-operation. Equipment supplied by EPT for each unit included: (1) gas burner assemblies, including gas manifold and adjustable pokers (Ansaldo design), flame detector, and flexible hoses to connect to the main gas header; (2) automated sliding sleeve air registers for the main burners; (3) automated sliding sleeve air registers for the NOx ports; (5) oil gun retracts; (6) new flame stabilizers which are compatible with gas and oil firing; and (7) segmented V-Jet oil atomizers designed to achieve low-NOx emissions.

The equipment was installed during a ten-week outage, and testing of ignitor fires was begun in May 1996. Lightoff of main fires on gas was conducted in June, and the unit reached maximum load on gas by the Fourth of July.

Combustion optimization tests showed that application of 100% OFA and 10% FGR reduced NOx emissions 75% below uncontrolled levels. At 540-MWg, NOx emissions of 0.16 LB/MBtu were achieved with 100% OFA and 9% FGR using only one GR fan. If two GR fans were available, an FGR rate close to 15% would be possible, which would further decrease NOx emissions.

Flame stability of the new ignitors and gas burners was excellent. At full load with 100% OFA and maximum FGR, the furnace was clear and each flame exhibited a distinct blue-color at the base, followed by a yellow luminous envelope. There was no flame carryover above the OFA port elevation.

The approach used for this gas conversion project is recommended for utility companies who are interested in retrofitting gas-firing capability to oil-fired boilers, but do not wish to replace the burners. Instead, the major components of the burners can be retained, which can significantly reduce the cost to install the new combustion equipment.

REFERENCES:

1. Kerho, S. E. ad D. V. Giovanni. "Atomizer and Swirler Design for Reduced NOx and Particulate Emissions." NOx Controls for Utility Boilers Workshop. July 1992 EPRI Conference, Boston, Mass.

2. Giovanni, D. V., R. C. Carr, and S. E. Kerho. "Reduction in NOx Emissions by Retrofit of Low-NOx Atomizers on a 550-MW Oil-Fired Boiler." Third International Conference for a Clean Environment. Portugal, July 3-6, 1995.

Back to top