Evaluating High Temperature Oxy-Natural Gas Retrofit of a Coal-fired Boiler

Paper from the AFRC 2013 conference titled Evaluating High Temperature Oxy-Natural Gas Retrofit of a Coal-fired Boiler by Bradley Adams

AFRC 2013 Industrial Combustion Symposium Evaluating High Temperature Oxy-Natural Gas Retrofit of a Coal-fired Boiler Bradley Adams*, Timothy Shurtz Reaction Engineering International, 77 West 200 South, Suite 210, Salt Lake City, UT 84101, USA Introduction Concern about greenhouse gas emissions has produced increasing pressure on coal-fired power plants to reduce CO2 emissions. Oxy-combustion has been identified as one method for enabling CO2 capture and compression in preparation for transport for enhanced oil recovery (EOR) or sequestration. Numerous studies and several proof-of-concept tests and pilot-scale demonstrations have been conducted to evaluate benefits and challenges associated with oxy-coal combustion. The greatest challenge for oxy-combustion to date has been the large parasitic power requirements for the air separation (oxygen production) and CO2 compression, resulting in reduced plant efficiency and higher cost of electricity. To address this, a high flame temperature oxy-combustion concept has been developed to increase plant efficiencies. This technology has been tested at pilot-scale as part of a DOE R&D program. During this same period of oxy-combustion research, development of U.S. shale gas reserves has resulted in wide-spread availability of lower-cost natural gas. This raises the possibility that the use of natural gas with oxy-combustion may provide a beneficial approach to reducing CO2 emissions and improving plant efficiency. Additional benefits of retrofitting a coal plant to oxy-natural gas include reduced fuel cost, reduced emissions, operational flexibility, improved plant efficiency (relative to other carbon capture options), and potential to convert back to solid fuel firing if needed (as opposed to repowering). This presentation reviews an evaluation of heat transfer impacts when retrofitting a 350 MWe coal-fired plant for oxy-natural gas combustion. Approach A 350 MWe coal-fired plant sited near an oxygen pipeline was identified as a potential candidate for retrofit. Key challenges to retrofitting were identified, including impacts on the firing system (e.g., burners, pressure parts, piping and windbox, control systems), radiant furnace heat transfer, convective pass heat transfer, and equipment impacts (e.g., fans, air heater, APCDs). The focus of this study was on the radiant furnace and convective pass heat transfer. Relative to air-coal systems, oxy-natural gas firing provides flexible flame temperatures, potentially higher thermal efficiency (through higher heat transfer), lower emissions, and the potential for easier CO2 capture. In this work, REIs combustion-based CFD code, Glacier, and process simulation code, SGE, were used to model the air-coal plant and the potential oxy-natural gas retrofit. CFD simulations were used to assess the heat transfer in the radiant furnace for different flue gas recycle (FGR) conditions. FGR absorbs energy and adds mass flow, therefore the amount and injection location of FGR in the furnace impacts flame temperature and heat flux profiles. FGR also impacts the distribution of heat transfer between the radiant and convective sections of the boiler. Managing this heat transfer split is one of the key challenges in a retrofit design. The SGE process tool was used to model the steam circuit of the boiler and quantify the trade-offs between heat transfer in the radiant and convective sections. * Corresponding author. Tel.: [phone]; fax: [phone]. E-mail address: [e-mail] 2013 Industrial Combustion Symposium 2 The boiler combustion and heat transfer were first modeled for the air-coal case. Then the natural gas-oxy case was simulated and the results compared. The furnace operating conditions are described in Table 1. Table 1. Furnace Operating Conditions for Air-coal and Oxy-NG Firing. Furnace Operating Conditions Air-Coal Oxy-NG Total Furnace Firing Rate (MBtu/hr) 940 940 Coal/Natural Gas Firing Rate (MBtu/hr) 857 857 Process Gas (PG) Firing Rate (MBtu/hr) 83 83 Total Air or O2+FGR Flow Rate (klb/hr) 924 549 Theoretical Excess O2 (%, wet) 4.7% 1.2% Theoretical Excess O2 (%, dry) 5.1% 3.0% Overall Furnace Stoichiometric Ratio 1.31 1.09 Coal Burner SR 1.26 NG Burner SR 0.984 PG Burner SR 1.76 0.984 Figure 1 shows a schematic of the furnace geometry, including labels for the natural gas burners, process gas burners, recycled flue gas inlets and a region of air in-leakage. The furnace modeled was the reheat furnace for the split furnace boiler, hence the firing rate shown in Table 1 is much lower than that for the full unit firing rate with combined furnaces. Some of the key assumptions for the retrofit boiler analysis were: • Furnace firing rate was maintained when changing from air-coal to oxy-natural gas firing. • Process gas (PG) was fired at same rate as the baseline coal case. • Air in-leakage was estimated at 3% of the total flow rate. • FGR composition and oxygen injection rate were based on maintaining 3% excess O2 (dry) at exit. • There was 26.7% O2 in the O2+FGR mixture, which was estimated to match heat transfer in the air-coal case. • A generic natural gas burner design was used. Flue gas was injected in burner zone but not mixed directly with natural gas-oxy streams. This approach was based on Jupiter Oxygen Corp's high flame temperature concept. Figure 1. Schematic of furnace geometry.AFRC 2013 Industrial Combustion Symposium 3 Results The key results compared in the air-coal and oxy-natural gas cases were the flame temperature, furnace exit gas temperature, radiant furnace heat transfer, peak furnace wall temperatures, flue gas flow rates, and convective section heat transfer. These results are summarized in Table 2 and will be discussed in more detail below. Table 2. Modeling Results for Air-coal and Oxy-NG Firing Configurations. Result Baseline Coal-Air NG-Oxy w/ FGR Furnace Exit Gas Temperature (°F) 1850 1925 Exit CO Concentration, wet (ppm) 138 2044 Exit O2 Concentration, wet (%) 4.7 1.3 Peak Wall Temperature (°F) 840 852 Radiant Furnace Heat Transfer (MBtu/hr) 459 497 Back Pass Flue Gas Flow Rate (klb/hr) 1011 582 Heat Transfer to Superheater (MBtu/hr) 255 247 Heat Transfer to Economizer (MBtu/hr) 110 66 Flue Gas Temperature Leaving Economizer (°F) 626 538 Figure 2 shows the predicted gas temperature profiles for the air-fired coal combustion case and the oxygen-fired natural gas combustion case. For the natural gas-oxygen case, there was a significant amount of flame surface over 4500 °F. This was by design as the recycled flue gas was injected adjacent to the burners rather than directly through the burners or pre-mixed with the oxygen. Radiation from the high-temperature flames to the furnace water walls is mitigated by the high absorptivity (high H2O and CO2 concentration) flue gas. The location of the flue gas recycle (FGR) inlets impacts the local flame temperatures as well as the exit gas temperatures. Figure 2. Predicted gas temperature profiles for air-coal and oxygen-natural gas simulations. Figure 3 shows the CO concentration profiles for the coal and natural gas simulations. The CO concentrations in the natural gas case are very high in the lower furnace because the process gas and natural gas burners are fired at sub-stoichiometric levels (0.984), a result of the FGR and air in-leakage AFRC 2013 Industrial Combustion Symposium 4 being assumed to enter in and above the burner zone. These high CO concentrations can be reduced by optimizing the locations of the FGR injection. The O2 concentration profiles shown in Figure 4 again indicate the fuel-rich state of the lower furnace, and also reflect the overall higher levels of oxygen concentration in the air-coal case (see Table 2). Figure 3. Predicted CO concentrations profiles for air-coal and oxygen-natural gas simulations. Figure 4. Predicted O2 concentrations profiles for air-coal and oxygen-natural gas simulations. Figure 5 plots the predicted net wall heat flux for both cases. Although the peak flame temperatures are much higher in the oxygen-natural gas case, the net heat flux values are similar in magnitude and location for the two configurations. This is due to three factors - flame emissivity, flame temperature, and flue gas absorptivity. The air-coal flame has a higher emissivity (due to particles) and a lower flame temperature. Since flame emissive power is the product of flame emissivity and the flame temperature raised to the fourth power, the flame radiative emission of the two flames is roughly similar, with the oxygen-natural gas flame being somewhat higher due to its higher flame temperature. The oxygen-natural gas case also AFRC 2013 Industrial Combustion Symposium 5 has a higher flue gas absorptivity due to the higher H2O (primary) and CO2 (secondary) concentrations in the recycled and combustion flue gases. This tends to absorb more of the radiant energy from the flame, in effect using some of the radiative energy emitted from the flame to heat up the flue gases (as opposed to transferring the radiant energy to the water walls). This process is consistent with the higher furnace exit gas temperature (FEGT) in the oxygen-natural gas case. The process also illustrates one of the advantages of injecting the flue gas separately from the oxygen-natural gas burners, which is that the flame temperature and wall heat transfer can be controlled somewhat independently, and that the localized use of recycled flue gas can be used to adjust the flame temperature and wall heat transfer. Note that in the case modelled here, the flue gas is assumed to have no removal of moisture during the recycle process. Table 3 compares heat transfer totals on a zone-by-zone basis for the two cases. Figure 5. Predicted net wall heat fluxes for air-coal and oxygen-natural gas simulations. Table 3. Comparison of Zone-by-zone Furnace Heat Transfer. Baseline Air-Coal Oxy-NG (w/ FGR) Zone Heat Transfer (MBtu/hr) % of Total Heat Transfer (MBtu/hr) % of Total 1 42 9.1 54 10.9 2 154 33.6 188 37.9 3 189 41.2 170 34.1 4 74 16.1 84 17.0 Total 459 497 Consistent with the net heat flux results, the wall temperature profiles for the air-coal and oxygen-natural gas cases are similar as shown in Figure 6. The maximum peak wall temperature for the air-coal case was 840 F and for the oxygen-natural gas case was 852 F, emphasizing the similar wall temperature and heat transfer properties of the two cases.AFRC 2013 Industrial Combustion Symposium 6 Figure 6. Predicted wall temperatures for air-coal and oxygen-natural gas simulations. Beyond the radiant furnace, Table 2 summarizes the convective pass differences in heat transfer between the two cases. The much lower flue gas flow rate for the oxygen-natural gas case (582 kpph vs 1011 kpph) is partially compensated by the higher exit gas temperature such that the heat transfer to the superheater is similar for the two cases (247 MBtu/hr vs 255 MBtu/hr). However, the impact of the lower energy available in the oxygen-natural gas flue gas is illustrated by the reduced heat transfer and exit gas temperature of the economizer. These lower values may or may not be significant depending on the design of the overall boiler steam circuit and the preheat requirements of the gas-gas preheater used in the flue gas recycle loop. Summary The reheat furnace of a 350 MWe split-furnace boiler was modelled for air-coal firing and oxygen-natural gas firing to compare flame and wall temperatures and radiant furnace and convective pass heat transfer. Both CFD and process modelling were used in the furnace evaluations. Modelling results showed that oxygen-natural gas firing is feasible for mimicking air-coal fired furnace heat transfer, but firing system will need to be optimized for specific furnace design to control CO concentrations and furnace heat transfer. Results also showed that: • Oxy-NG burners produced high flame temperatures (>4800 °F) which aided furnace heat transfer (higher than air-NG flames). • 25-27% oxygen in FGR+O2 inlet mixture gave similar radiant furnace heat transfer to the air-coal case. • The chosen firing system design produced similar radiant furnace heat flux distribution and peak wall temperatures as the air-coal design. • Burner design and FGR layout can be used to manipulate furnace heat transfer and CO. • Reduced flue gas flow slightly decreased heat transfer in backpass superheater and significantly decreased heat transfer in economizer. The use of both CFD modelling and SGE process modelling illustrated that modelling tools can help evaluate conversion of boilers from coal to natural gas firing and can particularly help evaluate trade-offs between radiant and convective heat transfer for different fuels, oxidizers, burner designs, firing rates and flue gas recycle rates in retrofit boiler designs.AFRC 2013 Industrial Combustion Symposium 7 Acknowledgements This research was sponsored in part by the US Department of Energy National Energy Technology Laboratory under contract DE-0005288, Andrew Jones program manager. This work was also sponsored in part by Jupiter Oxygen Corporation. CFD graphics were produced using Fieldview software from Intelligent Light (http://www.ilight.com)