Title | Emission Reduction with Innovative Oxygen Injection |
Creator | Lenhert, D. |
Contributor | Damstedt, B., SaNeto, V., Laux, S. |
Date | 2017-12-11 |
Description | Paper from the AFRC 2017 conference titled Emission Reduction with Innovative Oxygen Injection |
Abstract | The nature of some industrial combustion processes makes the simultaneous reduction of NOx and CO very difficult and often requires very expensive end-of-pipe pollution abatement as more stringent limits have essentially "boxed" operators into a very narrow operating window. The targeted injection of oxygen can be a very effective tool to reduce these air pollutants and provide operational flexibility by addressing CO from the combustion process at the outlet of the system. This presentation will provide three examples of this strategy from very different industries and introduce the equipment that was used.; ; The first example will review the commercial installation of CONOx™ technology to address NOx and CO from a refinery FCC regenerator unit. The unique technology is able to address flue gas CO and NOx simultaneously without converting NOx precursors to additional NOx. A second example will introduce the injection of oxygen into a cement kiln to reduce CO slip from the combustion process. A key success factor in this application is the rapid and effective mixing of a relatively small oxygen flow into a much larger flue gas stream. Lastly, oxygen was injected into an aluminum re-melting furnace to reduce large amounts of CO that are present at the beginning of the melting cycle due to oils and other combustibles on the dirty scrap. Although this approach is known, it was recently enhanced with the introduction of Praxair's OPTIVIEW™ technology. This technology automates the oxygen injection based on real-time flame imaging and analysis to minimize the number of exceedances and reduce demands on operators. |
Type | Event |
Format | application/pdf |
Rights | No copyright issues exist |
OCR Text | Show Emission Reduction with Innovative Oxygen Injection David Lenhert*, Bradley Damstedt, Valmiro Sa Neto, and Stefan Laux, Praxair, Inc. *Presenter Contact Information: David Lenhert, Technology Commercialization Director Praxair, Inc., 1585 Sawdust Road, Spring, TX 77380, USA Phone: 1-716-909-045, Email: David_Lenhert@Praxair.com I. Introduction The classical method to control CO and NOx from combustion processes has been to control the amount of excess oxygen. Sufficient excess oxygen ensures low CO emissions and complete combustion, see operating point 1 in Figure 1. Without equipment changes lower NOx can only be achieved by lowering the excess oxygen. As the amount of excess oxygen decreases in the flue gas, NOx formation is reduced. However this also increases CO emissions, see operating point 2. Figure 1: Typical relationship of excess oxygen on CO and NOx formation 1 Recent years have seen tightening of both CO and NOx emissions limits that have essentially "boxed" operators into a very narrow operating window or require expensive endof-pipe control technologies, such as SNCR, SCR, LOTOx, and others. The nature of some industrial combustion processes makes the simultaneous reduction of NOx and CO very difficult. The targeted injection of oxygen can be a very effective tool to reduce these air pollutants and provide operational flexibility by addressing CO from the combustion process at the outlet of the system. This paper will provide three examples of this strategy from very different industries and introduce the equipment that was used. II. CONOx Technology for FCCU Emission Control The Fluid Catalytic Cracker (FCC) unit is one of the single largest CO and NOx emission sources in a refinery and will continue to be regulated heavily in attempts to minimize the refinery's environmental impact. Praxair has developed a flexible, low capital cost technology that significantly reduces both the CO and NOx emissions from the FCC1,2. Due to the low capital cost of implementing this technology, it can be used in conjunction with other emissions control devices to achieve an even larger reduction in emissions. This technology also may enable more profitable operation through increased feed rate, improved yield, or reduced fuel consumption. Praxair's CONOx technology uses a specialized lance3, which injects heated oxygen into the flue gas duct between the regenerator and the CO boiler or heat recovery boiler. The key to the technology is a high velocity jet Figure 2: Schematic of CONOx Technology of heated oxygen that rapidly mixes with process gases. The lance also creates high concentrations of free radicals, enabling rapid reactions at lower bulk temperatures. Due to the extraordinary jet mixing 1 Reduction of CO and NOx in Regenerator Flue Gas, US Patent No. 7,470,412 2 Reduction of CO and NOx in Full Burn Regenerator Flue Gas, US Patent No. 7,959,892 3 Thermal Nozzle Combustion Method, US Patent No. 5,266,024; EU Patent No. EP0590572. 2 characteristics and reactivity achieved, the CONOx technology is significantly more effective at initiating low temperature reactions than a standard oxygen jet. The heart of the FCC unit is the reactor and regenerator. A schematic of a typical modern FCC unit is shown in Figure 3. Since the hydrocarbon cracking reactions produce coke that is deposited on the surface of the catalyst, the catalyst must be regenerated by burning off the deposited coke with air or enriched air blown into the regenerator. The regenerator can be operated in one of two ways, either "full burn" or "partial burn", which refers to the extent in which the coke is burned off the catalyst. For full burn regenerators, sufficient oxidant is introduced to burn all of the coke off the catalyst surface. As a result, the flue gas leaving the regenerator has N2, CO2, H2O, and O2, along with trace amounts of CO and NOx. For partial burn regenerators, sufficient oxidant is not introduced to burn all of the coke off the catalyst surface, hence the flue gas contains N2, CO2, H2O, and CO, along with trace amounts of O2, NH3, HCN, and NOx. The CO is then combusted with excess oxygen in a CO Boiler which is used to generate steam/power for the refinery. The CO boiler generates considerable levels of NOx by two basic mechanisms, a) converting the NOx precursors, NH3 and HCN, in the flue gas to NOx and b) burner generated NOx (via Thermal and Prompt NOx formation mechanisms). In either full or partial burn, reducing the amount of excess air in the final combustion step is a good way to reduce NOx. However, reducing the excess air level too much can increase carbon monoxide, which is another regulated pollutant. This results in the refinery having to continually balance excess oxygen to minimize both NOx and CO emissions in a very dynamic process. Figure 3: Schematic of typical FCC Unit. 3 For full burn regenerators, the CONOx technology enables operators to minimize NOx emissions by operating the regenerator at very low excess oxygen levels (e.g., < 1% excess O2). Any corresponding rise in CO in the flue gas duct is effectively oxidized by the CONOx technology without creating any additional NOx. This oxidation occurs at temperatures less than 760oC (1400 °F), which is usually manageable within existing flue gas ductwork without any modifications. Thus, the technology effectively allows the refinery to decouple the stack emissions of NOx and CO. In partial burn regenerators, the CONOx technology operates slightly differently. The oxygen is injected such that the flue gas remains deficient in oxygen after all oxygen from the lance is consumed. This results in a significant reduction of NOx precursors (NH3 and HCN) in the flue gas going to the CO Boiler. By reducing the concentration of NOx precursors in the duct prior to conversion to NOx in the CO boiler, stack NOx can be reduced considerably. Consequently, Praxair's CONOx technology can be used to debottleneck operations due to CO boiler or NOx limits, thus enabling a wider regenerator operating window while still achieving low NOx emissions. The actual amount of NOx reduction in full and partial burn units will vary depending on feed, catalyst properties, regenerator design, CO boiler, and other constraints on the units. However, NOx reductions of up to 60% can be achieved in both full burn and partial burn regenerators. The following section of this paper reviews one example where Praxair's CONOx technology was used to reduce FCC stack emissions. CONOx Technology Application Example The refinery in this application was processing approximately 27,000 BPD of feed thru the FCC unit and was faced with the challenge of meeting a tighter NOx emission limit. The customer was looking for a low-capital NOx solution that did not involve reduced charge rate or more expensive feedstock, and they needed the solution quickly. Praxair's CONOx technology was selected as it could be installed rapidly without requiring a unit shutdown and could achieve the required NOx reduction. 4 The FCC at the refinery uses a two stage regenerator with the first stage operating in partial burn and the second stage operating in full burn. The refinery had identified through stack and flue gas sampling that the partial burn unit and associated CO boilers produced Emission Measurements Flue Gas Regenerated Catalyst Heat Recovery Stack Full Burn Regenerator the majority of the stack NOx CONOx emissions. It was determined Treated Flue Gas that using the CONOx CO Boiler #1 technology on only the partial burn flue gas would require CO Boiler #2 fewer operational changes to the regenerator, provide sufficient Air Partial Burn Regenerator Air Spent Catalyst ValeroFigure Ardmore FCCU Layout 4: Refinery FCCU Layout NOx reduction, and minimize capital cost. The system required very limited process modifications for integration to the existing FCC, allowing rapid online implementation of the technology. The refinery and Praxair agreed to an aggressive implementation schedule of six months, following a three month design and approval period. Significant schedule challenges included: Completing CFD modeling to select an optimal lance location. Performing rigorous design and safety reviews. Finalizing and programming the control logic. Designing, fabricating and testing the equipment before delivery. Hot tapping the flue gas line to allow the lance to be inserted while the FCC is in operation. The design of the specialized lance started with identifying the range of required operating conditions which allowed Praxair to set critical internal dimensions and ensured the required flexibility during operation. Next, CFD modeling was used to evaluate the flow in the flue gas line and its impact on the mixing characteristics of the hot oxygen jet. This allowed us to evaluate potential injection points to optimize jet trajectory and mixing. Once the optimum location of the lance was identified, a hot tap of the flue gas line was required to install the lance without shutting down the FCC. 5 Praxair designed and constructed the lance, an insertion/retraction mechanism, and the gas flow control system. High temperature seals designed into the insertion/retraction system allow repeated lance insertion into the process while maintaining a safe Figure 5: Photo of oxygen injection lance working environment for operators. Once built, the entire CONOx system underwent thorough factory acceptance testing at Praxair's facility in Tonawanda, New York prior to installation. This ensured a smooth and rapid startup at the refinery. Within two weeks of the equipment delivery to the refinery, the CONOx system was achieving the required NOx reductions. NOx from the CO boilers had to be decreased by greater than 15% to achieve the goal for NOx reduction in the combined full burn and partial burn stack effluent. Plus, the CONOx technology also had to operate over a wide range of FCC charge rates and regenerator operating conditions. Stack CEMs were used to measure NOx with the CONOx system operating at various flow rates and also turned off to check the baseline levels. Although many other factors impact the NOx leaving the combined stack, the CONOx system provided the ability to rapidly respond to changing operating conditions and was able to keep NOx below the desired target. Using CONOx in the partial burn unit also improved CO boiler operation due to the higher inlet temperatures and conversion of some of the CO in the flue gas line. With the CONOx lance, the refinery operated at a lower CO level in the stack along with a lower stack NOx, as shown in Figure 6. NOx reduction was proportional to oxygen flow rate up to the point where the system became limited by maximum duct temperature and pressure drop through the orifice box. The effect of changes in the CONOx oxygen flow was immediately seen in downstream flue duct temperature and stack NOx response, allowing straightforward control through oxygen flow. 6 Figure 6: Simultaneous reduction in both stack CO and NOx achieved. III. CO Emission Control for Cement Kilns Cement plants process a blend of limestone, sand and clay minerals by grinding, preheating, calcining, clinkering, cooling and grinding the material to the final product. Cement making consists of two main chemical conversions, the calcining of limestone (CaCO3) to lime (CaO) and the subsequent clinkering of the material at high temperatures to cement clinker consisting mainly of calcium silicates, calcium aluminates and calcium ferrites. The ground clinker is then blended as needed with gypsum and sometimes fly ash for the final product. The clinkering step takes place in large rotary furnaces that are slightly inclined. The calcined limestone, sand and alumino-silicate materials fed into the kiln and are subsequently sintered together at very high temperature of 1400 to 1500 °C (2500 to 2700 °F) to form pebble-sized nodules that are dropped into a cooler. A single large burner at the discharge end of the kiln provides the heat for the process in the rotary kiln resulting in a countercurrent flow of the gas versus the raw materials. The raw materials entering the kiln pipe are preheated and calcined with the hot flue gas that leaves the rotary kiln at the temperature level of 1000 to 1050 °C. To reduce operating costs and to provide a very luminous hot flame most cement kilns are using coal, petcoke or refuse derived fuels from the recycling industry. 7 The arrangement of the burner firing into the narrow kiln pipe and the counter-current flow of material can lead to significant stratifications at the feed end of the kiln. Flue gas traveling at the bottom of the kiln may have much higher CO concentrations than the flue gas in the upper part of the kiln pipe. This is exasperated by the use of solid fuels, specifically coarse refuse derived fuels or alternate fuels. Although the residence time in the flame is quite large the trajectory of these large fuel pieces brings them to the bottom half of the kiln and they may be trapped in the clinker bed where they burn much more slowly. Low oxygen availability near the clinker bed combined with limited cross-mixing in the long and narrow kiln results in high CO concentrations in this region. If this CO stratification is not mixed with oxygen at sufficiently high temperatures for reaction (>1000 °C or 1800 °F) then high CO emissions from the kiln can be the consequence. CO emissions from cement kilns have been traditionally high in the range of thousands of ppm, especially with high use of alternate fuels. Recently, the permitting process has focused more on CO in addition to NOx and producers are increasingly "boxed in" by NOx, CO and in case of SNCR and SCR end-of-pipe solutions by ammonia slip in the flue gas. Customers may be required to reduce CO or will see higher CO when they try to further reduce NOx with combustion control methods. The combined emission requirements present a difficult choice as retrofit NOx control technologies cannot address high CO or aggressive combustion air staging leads to higher CO emissions. Increasing excess oxygen has a detrimental effect on NOx emissions and efficiency. However, oxygen injection at the right locations can offer a solution to reduce CO from the kiln process and meet emission requirements. To improve the CO emissions from rotary kilns, Praxair has adapted the hot oxygen technology developed for the CONOx technology, discussed above, for the process conditions present in these kilns. The key to this technology is to break up the CO stratification and oxidize the CO by providing additional oxygen. The approach is utilizing Praxair's OPTILANCE™ Hot Oxygen Technology which combines providing highly reactive hot oxygen with superior flue gas mixing4. The operating principle of the lance is shown in Figure 2 above. The high momentum of the hot oxygen effectively mixes with the unburned gas and provides additional oxygen for combustion, Figure 7. Tests that compare cold oxygen injection to OPTILANCE technology show the very high momentum of the 4 Cement Clinker Production with Reduced Emissions, US Patent No. 7,682,447 8 lance improves CO reduction as higher jet momentum drives improved mixing of flue gas with the highly reactive oxygen. Figure 7: Schematic of a Cement Kiln with OPTILANCE Technology for CO Reduction The application of hot oxygen to cement kilns also offers the opportunity to apply staged combustion to the entire rotary kiln and reduce NOx from the combustion. The addition of combustion air in stages to reduce primary NOx is a well-established technology (power plants, low-NOx calciners, etc). However, the confinement of a rotary kiln makes it difficult to add hot burnout air in the later stages of combustion after the main flame. Oxygen injection at the kiln feed end offers a low cost opportunity to reduce the overall air admitted to the main flame and reduce NOx through combustion staging. This combination of technologies can be used as an initial reduction to an SNCR system that needs to further reduce NOx beyond the typical 75% NOx reduction possible for SNCR. 9 OPTILANCE™ Technology Application Example OPTILANCE Technology was successfully implemented at a cement plant with a preheater kiln to reduce the CO emissions from the kiln in response to new local emission limits. This cement producer uses waste, plastics, lignite and anthracite as fuels and utilizes oxygen injection into the main burner flame to support combustion of hard to burn fuels without negative impact on clinker quality or emissions. The plant has an SNCR system that is not able to reliably meet a future NOx emission limit. CO emissions are required to be reduced by 20% at the same time. This presented an ideal opportunity to utilize OPTILANCE hot oxygen injection as a complementing approach to reduce NOx and CO emissions. Oxygen is injected at the riser duct and through optimization of the lance location and direction the bulk of the CO is mixed with the hot oxygen. At the same time the secondary air at the main burner was reduced through combustion optimization to lower NOx emissions from the combustion process. The increase of CO due to excess air reduction was also effectively addressed through the oxygen injection. Figure 8 shows the response of CO at the flue stack to changes of Oxygen flow at the lance. Reducing lance O2 flow causes an immediate increase in stack CO, restoring lance O2 flow reduces the CO. Figure 9 compares daily averages of stack CO and NOx for cold and hot O2, normalized to performance of an optimized cold O2 jet, as a function of the lance O2 flowrate. Cold O2 emissions data are relatively constant across the range of O2 flows, which is due to the O2 optimization of both the cold O2 lances with the primary burner O2 injection. Hot O2 data, by contrast, are preliminary with no optimization having been performed. Despite this, benefits of OPTILANCE technology are easily discerned. As hot O2 flow increases, CO and NOx simultaneously reduce, trending below 75% of the average cold O2 emission rate. It is expected that further reductions in CO and NOx are achievable with increased control tuning and operating experience. To be successful with the combined reduction approach it is necessary to understand the CO/NOx production characteristics from each kiln, to evaluate the limits required and to systematically optimize the system. A CFD study can greatly help with identifying the success of the mixing of oxygen with the flue gas. 10 Figure 8: Stack emission response with respect to Oxygen Injection Figure 9: Simultaneous CO and NOx reduction using OPTILANCE Hot Oxygen Technology The described combustion optimization approach with targeted oxygen injection using highly reactive hot oxygen OPTILANCE technology can be generally applied to many industries that require reduction of CO or unburned hydrocarbons. Other processing kilns (lime, calcining, etc), small boilers, waste combustion systems and thermal oxidizers can all benefit, if the existing emission control system is unable to reach the required limits. 11 IV. OPTIVIEW™ for Emission Control Many industrial combustion processes are producing highly variable CO and VOC emissions due to rapidly changing conditions in the furnace. These spikes are not acceptable from an environmental point of view. Examples for processes with high emissions of unburned matter are batch melting of metals or incineration of wastes that have a high degree of fuel heating value and therefore combustion oxygen requirement variations. In aluminum remelting furnaces the aluminum scrap is charged into a refractory lined batch melting furnace. The scrap can contain oil from cutting operations, grease or paint which burn off at the beginning of the heating cycle as the temperatures in the furnace increase. Insufficient oxygen availability at this point results in soot (smoke) and high CO and VOC emissions. The measurement of fuel and air or oxygen for the burners cannot be used to calculate and ensure an appropriate combustion stoichiometry and excess oxygen level because there is temporarily an unknown amount of additional fuel in the furnace until all hydrocarbons have been volatilized from the scrap. One simple solution is to increase air flow to the furnace at the beginning of the batch, but this reduces the combustion temperatures and also decreases furnace efficiency, often resulting in a prolonged period of high unburned matter and lower productivity. Oxygen injection into the furnace has been used for many years to combat the initial emissions of hydrocarbons after charging the furnace with "dirty" scrap. However, often these solutions were just simple timed openings of the valve controlling the oxygen flow to the furnace, performed by the furnace operator. This type of control is highly dependent on the operator's experience to visually identify changes on the exhaust flame intensity that would indicate the presence of excess unburnt fuel in the flue gas. More sophisticated solutions have employed in-situ measurements of oxygen or CO to control the excess oxygen. Due to the high amount of soot in this phase of the process and salt used as fluxing agent, these process monitoring tools are costly, difficult to maintain and they usually introduce a delay that makes timely and accurate action impossible. Praxair has developed a control technology to automatically initiate and control the oxygen and fuel flows in the beginning of batch melting processes without an in-situ flue gas measurement. The OPTIVIEW™ Image Analysis uses a video image of the flame in the furnace or the flame formed by excess hydrocarbons in the exhaust to adjust the appropriate time and amount of additional oxygen temporarily injected into the furnace as well as the 12 ideal burner firing rate. The technology was first employed at a tilting rotary furnace for aluminum melting in Europe. This furnace had already an oxy-fuel combustion system and the OPTIVIEW system was retrofitted in January 2017. Figure 10 shows a sketch of the process control scheme. Figure 10: OPTIVIEW™ System Schematic A commercial video camera records an image of the flame emerging from the furnace exhaust if hydrocarbons are present in the furnace. The images are analyzed in real time and the image parameters matched to a program that predicts the amount of hydrocarbon present to generate the flame and how much additional oxygen is necessary to burn these hydrocarbons along with the respective burner firing rate required to keep the optimum heat input to the smelting process. Alternatively, the combustion system could be enriched with oxygen. As an example, Figure 11 shows the flame present at the beginning of the process without oxygen injection. After installation and initial "teaching" of the image analysis to match the correct combustion requirements to the captured video, the OPTIVIEW system operates automatically. System characterization is done with extractive sampling of flue gas in the exhaust. Figure 12 shows the results of the correlation between the image parameters used in the calculation of the oxygen flow and the percentage of CO in the gas from the furnace. For an industrial furnace with a high degree of variability the quality of the correlation is very good and shows that using image data for the control is an excellent real-time approach to resolve the emission issues of aluminum scrap melting. The video camera is protected against the temperatures and dirt present in metal processing facilities and requires no maintenance. 13 This robust system is now in commercial operation at the customer. New applications of OPTIVIEW Image Processing Technology for combustion control in the area of batch processing and incineration are in preparation. Figure 11: Aluminum Tilting Rotary Furnace Photograph Figure 12: Image Analysis Parameter Correlation to CO from the Process In addition to the desired reduction of smoke, CO and hydrocarbons, the OPTIVIEW system has distinct operating advantages. By injecting oxygen and adjusting the burner firing 14 rate at the right time and in the right amount, the heat of the hydrocarbons emanating from the scrap can be captured in the furnace and used very efficiently to reduce the cycle time for increased production from the furnace. Figure 13 shows the impact of the OPTIVIEW technology on the total melting duration, i.e. the time between first scrap charge to the tapping of the aluminum from the furnace into a holding furnace prior to casting. Analyzing data from several months before and after the installation of the system shows that the average melting time was reduced by 0.6 hours or 12%. This is a significant improvement in productivity. Interval Plot of Melt Duration (h) Bars are One Standard Error from the Mean 5.4 5.3 5.20515 Melt Duration (h) 5.2 5.1 5.0 4.9 4.8 4.7 4.6 4.60014 4.5 After Before OPTIVIEW Figure 13: Impact of OPTIVIEW on Melt Duration The change in combustion control strategy can be seen in Figure 14. Before the system was installed the average oxygen to natural gas ratio used over the entire cycle was 3.25, which is a significant increase of the oxygen flow to combust the natural gas. The stoichiometric ratio of oxygen to natural gas is close to 2. The increased ratio before the OPTIVIEW installation shows the need to significantly increase the oxygen flow to the entire furnace to manage emissions. After the technology implementation the average ratio could be decreased to 2.70, which is a 17% improvement and represents a large reduction in oxygen required to melt a batch of aluminum. The payback of the OPTIVIEW technology is less than one year at this location. 15 Interval Plot of O2/NG Ratio Bars are One Standard Error from the Mean 3.4 3.3 3.24519 O2/NG Ratio 3.2 3.1 3.0 2.9 2.8 2.7 2.69095 2.6 After Before OPTIVIEW Figure 14: Impact of OPTIVIEW on Oxygen Usage V. Conclusions The three technologies discussed in this paper highlight innovative use of oxygen combustion technology to address emission control challenges. These are examples that are readily transferred to other emission reduction needs of similar nature. The application of CONOx™ to reduce emissions in refineries is a successful example for NOx and CO mitigation in a complex operating environment using a unique hot oxygen lance. The installation was conducted following a demanding safety process and detailed commissioning and installation planning. The second application example of hot oxygen technology is the installation of the OPTILANCE™ technology in a cement kiln. Here a unique and robust solution enabled the control of CO and NOx to meet new emission requirements without a significant capital investment. The third example focusses on the control of oxygen flow in highly variable combustion process conditions in aluminum melting using OPTIVIEW™ Image Analysis. By using characteristic image information the oxygen flow can be effectively tailored to the needs of the furnace. This resulted in the desired reduction of CO and hydrocarbon emissions from the aluminum melting furnace coupled with optimized oxygen flow at much improved production capacity. 16 |
ARK | ark:/87278/s69d17fj |
Setname | uu_afrc |
ID | 1388796 |
Reference URL | https://collections.lib.utah.edu/ark:/87278/s69d17fj |