AERO Radiant Wall Burner for Ethylene Cracking and Steam Methane Reforming

Paper from the AFRC 2016 conference titled AERO Radiant Wall Burner for Ethylene Cracking and Steam Methane Reforming

A new radiant wall burner design for ethylene production is presented. The burner design was optimized utilizing computational fluid dynamics (CFD) and physical prototype testing. Due to the unique design of this burner, flame is consistently fired against the firing wall preventing impingement on heater process tubes while requiring minimal burner maintenance. These improvements utilize a simplistic design to deliver long-life to the burner and process equipment as well as increased performance

Honeywell UOP Callidus Technologies Callidus Technologies, L.L.C. 7130 South Lewis Avenue, Suite 335 Tulsa, Oklahoma 74136 Tel: [phone] www.callidus.com AERO™ Radiant Wall Burner for Ethylene Cracking and Steam Methane Reforming Testing Supervised by Tim Lowery Report by Richard Martin (July 2016) Report Presentation by Kurt Kraus (September 2016) Contents Introduction .................................................................................................................................................. 2 Figure 1 - Furnace Fired With Pre-Mixed Radiant Wall Burners ......................................................... 2 Traditional Operational Issues for Pre-mixed Burners ............................................................................. 3 Figure 2 -Conventional Radiant Wall Tips ................................................................................................ 3 Figure 3 - Pre-Mixed Burner Components ............................................................................................ 4 Figure 4 - Aero Tips and Mixers................................................................................................................ 5 Conventional ......................................................................................................................................... 5 Low NOx (staged fuel) ........................................................................................................................... 5 Aero Pre-Mixed Burner Design Features ...................................................................................................... 5 Mixer (Venturi).......................................................................................................................................... 6 Burner Tip ................................................................................................................................................. 6 Flashback Discussion and Test Procedure .................................................................................................... 7 NOx Emissions ............................................................................................................................................... 9 Figure 5 - Estimated NOx Emissions ......................................................................................................... 9 1 Introduction The patented (US 9,194,579) AERO™ radiant wall pre-mixed burner is carefully designed to overcome historical operating problems occuring for both conventional and "low NOx" pre-mixed radiant wall burners. Typically radiant wall, pre-mixed fuel gas burners h a v e b e e n used i n fired applications such as steam methane reforming furnaces and ethylene cracking furnaces. They provide a re lative ly high heat release in a small disk-shaped volume adjacent to a refractory wall while providing re lative ly low pollutant gas combustion emissions. Incomplete combustion products such as carbon monoxide and unburned hydrocarbons are minimized because of the pre-mixing of the combustion air and fuel prior to the initiation of combustion. The nitrogen oxide emissions (NOx) are relatively low when compared two different types of conventional burners firing in the same application because of the relatively high surface area of the flame per unit of heat released. Typical NOx emission levels for the traditional pre-mixed burners for steam methane reforming applications is in the range of 65 to 75 ppm corrected to 3% oxygen concentration in the dry combustion products. For ethylene cracking furnaces NOx emission levels of 90 to 100 ppm corrected to 3% oxygen concentration in the dry combustion products are typical. Figure 4 illustrates the configurations for a conventional pre-mixed radiant wall burner and a staged fuel "low NOx" burner. Following (figure 1) is an illustration of a typical furnace that is fired with pre-mixed radiant wall burners. Figure 1 - Furnace Fired With Pre-Mixed Radiant Wall Burners 2 Traditional Operational Issues for Pre-mixed Burners While pre-mixed burners used for heating "radiant walls" offer the advantages of small well-defined compact flames and at least partial "automatic" control of excess air as the burner firing rate changes, depending on the composition of the fuel, flashback can be a potential problem for the burner. Flashback is the combustion of a pre-mix of fuel and combustion air inside the radiant wall burner tip and mixing chamber (venturi). It can occur when the flame propagation speed exceeds the discharge velocity of the fuel and air mixture exiting the tip and can cause thermal damage to the burner tip and the mixing chamber (venturi) in addition to creating various operational problems. The burner should be designed so that the discharge speed of the fuel and air mixture leaving the burner tip is uniform and exceeds the flame propagation speed. Current state of the art pre-mix burners feature geometries that do not necessarily provide a uniform flow of the fuel and air mixture through the assembly and gives way to acceleration and deceleration of the fuel and air mixture, causing a non-uniform flow exiting the tip. As a result of such non-uniform flow, turbulence is created resulting in excess pressure loss which will reduce the quantity of air entrained by the fuel gas. Commonly in the art, burner tips feature a cylindrical tip design with multiple discharge openings or a multiple leaf design with slots separating the leaves through which the fuel and air mixture is discharged into the furnace. Figure 3 illustrates to commonly used tip configurations. Figure 2 -Conventional Radiant Wall Tips The designs inherently do not provide uniform mixture flow at the tip exit simply because the fuel/air mixture enters the tip at relatively high velocity at the entry to the tip with a decrease in mixture velocity along the length of the tip. Typically both of these designs provide increased flow through the ports or slots at the downstream end of the tip. The nature of the design creates a situation where the flow is decelerated and then re-accelerated as it approaches the discharge openings. The resulting turbulence and differing velocities create non-uniform flow exiting the tip. In some locations the velocity can be extremely high, greatly exceeding the flame propagation speed, while in other locations the exit velocity can be extremely low, and in some cases even negative creating "reverse" flow back into the tip. Flashback may occur in the low o r r e v e r s e velocity regions. The flashback problems are usually associated with fuels with a relatively high hydrogen concentration. The traditionally acceptable maximum fuel hydrogen concentration is in the range of 65 to 75% hydrogen by volume at the "design" burner firing rate for the conventional radiant wall pre-mixed burner. The range of acceptable hydrogen concentrations is most likely somewhat lower for the staged fuel version 3 because there is a lower energy level to entrain the air resulting in a lower tip exit velocity. Also, it is very difficult to prevent flashback with these hydrogen concentrations as the burner firing rate is decreased. It resent years the typical fuel hydrogen concentrations have increased and it is not unusual to receive specifications stating hydrogen concentrations greater than 85% by volume. The key to firing these fuels with higher hydrogen concentrations is achieving the maximum possible uniform velocity as the fuel/air mixture exits the burner tip. The AERO Pre-Mixed Burner has been specifically designed to maximize the use of the energy of the fuel to entrain combustion air and eliminate low velocity regions at the tip exit. The AERO pre-mixed radiant wall burner are designed to maintain a constant fuel/air mixture velocity to a point just upstream of the tip discharge ports. There is a slight acceleration of the fuel air mixture as it exits through the tip ports. Figure 3 - Pre-Mixed Burner Components Burner Mounting Plate Tip Muffler Secondary Air Control Tile Air Inlet The conventional radiant wall burner uses 100% of the fuel for air entrainment and normally provides 100% of the required combustion air as pre-mixed air. The staged fuel Low NOx version may require some secondary air depending on the specified NOx emissions and quantity of staged fuel required to achieve the NOx emissions. 4 Figure 4 - AERO Tips and Mixers Conventional Aero Equal Velocity Tip Mixer (venturi) Fuel Orifice Low NOx (staged fuel) AERO Equal Velocity Tip Mixer (venturi) Primary Fuel Orifice Staged Fuel Orifice AERO Pre-Mixed Burner Design Features The AERO Pre-Mixed Burner design improvements include: 1. An improved burner tip design that maintains a constant velocity from the discharge of the mixer to the discharge ports in the tip. 2. An improved mixer (Venturi) design that increases the quantity of combustion air entrained per unit of the fuel flow. 3. An improved method for more precise secondary air flow control. 4. A simplified of cast construction was easily replaceable components. 5 The following discussion will address only the first two items listed above. Mixer (Venturi) The ability to prevent flashback in a pre-mixed burner is dependent upon maintaining a fuel air mixture velocity that is greater than the flame propagation speed for all points located in the tip discharge openings. Obviously, the higher the combustion air entrainment rate the easier it is to achieve appropriate velocities at the tip discharge openings. The first design step was to evaluate typical existing Venturi designs using CFD analysis. Next was the optimization of the Venturi design. It was determined that a venturi could be produced that is physically smaller for a given heat release than current designs with an increased pre-mix air side capacity by as much as 20%. If as suggested earlier a pre-mixed burner is capable of operating with 65% to 75% by volume hydrogen in the fuel mixture then logically it would appear that with the improved mixer the hydrogen concentration that could be handled may be significantly higher. Burner Tip As previously discussed, flashback occurs when the discharge velocity from the burner tip is less than the flame propagation speed. For all radiant wall burner designs that were reviewed (Callidus or other vendors) the calculated average velocity of the fuel/air mixture from the tip is always much higher than the flame propagation speed. The problem occurs when there is non-uniform flow from the tip openings. Traditionally, burner vendors used burner tip pressure loss to solve this problem. Many years ago radiant wall tips were designed based on a tip discharge area of approximately 5 square inches per million BTU fired for fuels with essentially no hydrogen and 4 square inches per million BTU fired when the hydrogen content of the fuel exceeded approximately 20%. Unfortunately, when this criterion is used the internal pressure in the tip is relatively high and as a result it is very difficult to achieve 100% of the required combustion air as pre-mix air. Because of this, burner vendors began to increase the tip port area to achieve additional pre-mix air capacity. In other words, the limitation on port area to ensure relatively uniform flow through the ports resulted in a decreased capacity. As the port areas per unit heat release were increased flashback became more of a problem. In fact, if the slot width was maintained for a traditional pre-mix radiant wall burner and the length of the slot was increased to provide the additional area (and hopefully increased capacity) then reverse flow (flow from the furnace back into the tip) occurred with flashback resulting (see Figure 2). This was demonstrated in physical testing 30 years ago and can now be shown using CFD modeling. It was necessary to develop four tip configurations. They are: 1. Conventional radiant wall burner for "normal" hydrogen concentrations. 2. Conventional radiant wall burner for "high" hydrogen concentrations. 3. Low NOx radiant wall burner for "normal" hydrogen concentrations. 4. Low NOx radiant wall burner for "high" hydrogen concentrations. 6 Flashback Discussion and Test Procedure As previously discussed radiant wall burners can be demonstrated at the maximum firing rate operating with relatively high hydrogen concentration, but as the burner is "turned down" flashback can occur. The flashback testing for the AERO Burner was conducted at 120% of normal firing, 100% of normal firing and 80% of normal firing. The operation of the burner was considered successful when the burner could be turned down to 64% of the design firing rate without flashback. The above described testing was first conducted with a furnace temperature of approximately 1900°F. All of the tips tested were able to fire with higher fuel hydrogen content than is believed to have been possible with the conventional type tips as shown in figure 4. To determine the impact of a higher furnace temperature on flashback, insulation was added to the furnace to increase the temperature to approximately 2300° and addition testing was performed. All tests were conducted with approximately 3% oxygen in the combustion products. The following paragraphs provide a brief description for some sample test points. The test points described are representative, but certainly are only a small sample of the tests conducted. Test 1- Conventional Radiant Wall Burner Tip For "Normal" Hydrogen concentrations. It was determined that the burner would operate at 0.8 million BTU per hour with a fuel hydrogen content of 65% hydrogen. It was then determined that the burner would operate at the design heat release of 1.2 million BTU per hour with 80% hydrogen concentration in the fuel. The hydrogen content was slowly raised until the burner flashed back with 82% hydrogen in the fuel. The design rate (1.2 million BTU per hour) pressure was approximately 25 Psig. Test 2- Conventional Radiant Wall Burner Tip For "High" Hydrogen Concentrations. The burner was operated at design heat release of 1.2 million BTU per hour with a fuel hydrogen content of 80%. When the hydrogen content was raised to 85%, the burner flashed back. The design rate pressure was approximately 40 Psig. This burner configuration was then fired at 1.0 million BTU per hour with a fuel hydrogen content of 75%. The fuel hydrogen content was raised to 80% with no flashback. The fuel hydrogen content was then slowly raised to 85% and flashback occurred. The hydrogen content was reduced to 82% with no flashback. The burner heat release was then reduced to 0.8 million BTU per hour with a fuel hydrogen content of 80%. A second "run" was conducted for the same conditions with a duration time of 10 minutes. The fuel hydrogen content was then raised and the burner flashed back when the fuel hydrogen content reached approximately 89%. Test 3 - Low NOx Radiant Wall Burner Tip For "Normal" Hydrogen Concentrations. Approximately 25% of the fuel was staged. As a result, some secondary air was required (the conventional burners previously described were essentially 100% pre-mix). The burner was fired with heat releases of both 0.8 million BTU per hour (0.6 million primary) and 1.0 million BTU per hour (0.75 million primary) with 80% fuel hydrogen content. No attempt was made to increase the hydrogen content. The burner was then fired at the design rate of 1.2 million BTU per hour (0.9 million primary) with 100% hydrogen. The burner was "barking" but did not flashback. The hydrogen content was not reduced to eliminate the barking. 7 Test 4 - Low NOx Radiant Wall Burner Tip For "High" Hydrogen Concentrations. Approximate 25% of the fuel was staged. The burner was fired at the design rate of 1.2 million BTU per hour with a fuel hydrogen content of 90%. The fuel hydrogen content was then raised to 100% with no flashback occurring. The design fuel pressure was approximately 40 Psig. The burner was then fired with heat releases of both 0.8 million BTU per hour (0.6 million primary) and 1.0 million BTU per hour (0.75 million primary) with 90% fuel hydrogen content. Flashback occurred for the 1.0 million BTU per hour firing rate as the fuel hydrogen content was increased to 100%. Flashback occurred for the 0.8 million BTU per hour case when the hydrogen content was increased to approximately 97%. Test 5 - Conventional Radiant Wall Burner Tip For "High" Hydrogen Concentrations (High Furnace Temperature) In an effort to determine the impact of increased furnace temperature on the flashback characteristics, installation was added to the furnace to increase the temperature to approximately 2300°F. The burner was operated at a design heat release of 1.2 million BTU per hour with a fuel hydrogen content of 80%. When the hydrogen content was raised to 93%, the burner flashed back. The design rate pressure was approximately 40 Psig. 8 NOx Emissions For all configurations tested, the NOx emissions were considerably less than one might based on the emissions for the traditional radiant wall burner designs. Following is a graphical representation of the test facility measured values for the NOx emissions. It should be noted that the values are for a specific set of operating conditions and will of course vary depending on a number of factors including fuel composition, furnace operating temperature and excess air. It should be noted that for all of the testing the fuel was composed of natural gas and hydrogen. So, for any given percent hydrogen listed the remaining fuel components is natural gas. Figure 5 - Estimated NOx Emissions NOx Vs Hydrogen Content Staged Fuel AERO Burner 45 40 Corrected NOx (PPMV) R² = 0.932 35 30 25 20 15 10 5 0 0 20 40 60 80 % Hydrogen *NOx Corrected to 3% O2, and 2000°F 9 100 120