Saturday, May 29, 2010

Burner Operating Characteristics

Burners are critical for the successful operation of industrial furnaces. Presented here is a set of equations that can be used to calculate characteristics of burner operation, including flame length, flame diameter, ignitability and flameout conditions. Equations are based on pre-mix burners operating at atmospheric pressure and firing natural gas only. Premix burners create short and compact flames compared to raw gas burners, and are designed to function with fuel-gas mixtures that have consistent specific gravity and composition.

Burner requirements

For direct-fired heaters to function correctly, burners must be capable of providing sufficient heat liberation from the fuel to meet heater processing requirements — based on the lower heating value (LHV) of the fuel. A fuel’s LHV can be defined as the amount of heat produced by combusting a specified volume, and returning the combustion products to 150C. For the heater to operate at the design process flowrate, the burners need to provide the heat necessary to maintain process fluid temperature and meet vaporization requirements at the heating coil outlet.

  • The number, size and placement of burners must allow each coil to operate at the same design outlet temperature
  • Design tube-metal temperature cannot be exceeded at any point on the coils
  • Burner size must allow an outlet velocity that does not result in malfunction over the range of flow conditions
  • Burner flame length should be less than firebox height (for vertical cylindrical heaters) or less than firebox length (for end-wall-fired heaters)
  • Excessive flame height and diameter should be avoided to prevent flame impingement on tubes
  • Burner spacing should be sufficient to allow burner-to-burner, as well as burner-to-tube clearance

The following equations can help establish optimal burner diameter:

Burner clearance

Establishing burner-to-burner clearance and burner spacing should be based on maximum burner flame diameter. Further, burner flame diameter should be evaluated at maximum burner-flame length. Sufficient burner-to-burner, outside diameter clearance should take into account the placement of structural elements between burners.

Sufficient burner-to-burner clearance prevents interference between the flame bodies and unburned fuel cores generated by adjacent burners, which results in the absence of unburned fuel within the burner flame when maximum flame length is reached. Burner center-to-center spacing should be at least one fully combusted flame diameter.

Clearance between the burner-flame (at maximum diameter) and the outside diameter of tubular heating surfaces should be set such that burner-to-tube flame impingement is avoided. Doing so will prevent tube damage due to overheating and will make best use of heating surfaces.


At high burner velocities, flame loss can occur if the heat gain due to burner ignition is less than the heat loss from the burner flame. Burner velocities may be pushed well above that used in normal heater operation in an effort to achieve higher heater capacity. Aside from flame loss while the heater is in operation, flameout can also be characterized by difficulty maintaining a stable flame at startup, or an inability to ignite the burner. The following equations can help predict the circumstances under which flamout conditions might occur:

Flame velocity

The heat generated by combustion is dependent on the flame propagation velocity. In a situation with 0% excess air, the ratio of fuel-to-fuel+air is about 0.1. In that case, evaluation of the flame propagation velocity is straightforward. However, at fuel-to-fuel+air ratios higher or lower than 0.1, it is more difficult. The following equations can help predict flame propagation velocity in those cases:


Qlib heater= Heater liberation, Btu/h

Nb= Number of burners

Db = Burner diameter, ft

Vb= Burner exit velocity, ft/s

Cfuel = Fuel, ft3

LHV = Lower heating value of fuel, Btu/lb

Cair+fuel = Volume of air and fuel mixture, ft3

SVfuel = Specific volume of fuel, ft3/lb

Df max = Maximum flame diameter, ft

Lf = Flame length, ft

SVflame= Specific volume of flame, ft3/lb

Vf = Flame propagation velocity, ft/s

Qgain = Burner heat gain, Btu/h

Qloss = Burner heat loss, Btu/h

As = Flame front area, ft2

(HTC)c (HTC)f, (HTC)r = Natural convective, forced- convective, and radiative heat transfer coefficients, respectively, Btu/h-ft2-F

Tflame = Flame temperature, R

Tsurr = Surrounding temperature, R

Eg = Flame emissivity

Cp = Gas specific heat, Btu/lb-F

A = Frequency factor in the Arrhenius equation

H = Heat of activation, Btu/lb-mol R

R= Gas constant, 1.987 Btu/lb-mol R

T= Gas Temperature, R

dCm/dt= Fuel concentration change, mol per ft3/s

K = Reaction velocity constant, s–1

Wf= Fuel, lb/h

*The text was adapted from the article “Fired-Heater Burner Performance,” by Alan Cross. It appeared in the April 2008 issue of Chemical Engineering.

Share |

Hydrogen Production By Steam Reforming

Management of the gas is critical for petroleum refiners

Ray Elshout Energy, Systems Engineering

Steam reforming of natural gas at petroleum refining facilities is the predominant means of producing hydrogen in the chemical process industries (CPI). Areas where hydrogen is heavily consumed include ammonia production, the cryogenics industry and methanol production (Table 1). Because hydrogen needs within various sectors of the CPI are at their highest levels in history, and are continuing to grow, an understanding of this method of hydrogen production and purification can be useful.

A major percentage of hydrogen used in the CPI goes toward production of ammonia, which continues find greater demand in the chemical fertilizer industry. On the other hand, methanol usage is declining in connection with its use as a feedstock for making methyl tert-butyl ether (MTBE; by reaction of methanol with tertiary butylene). In the U.S., MTBE had been used as a gasoline blend stock until recently, when use of the chemical as a gasoline oxygenate was phased out in favor of ethanol.

In addition to being producers of hydrogen, largely through steam reforming, petroleum refineries are also large consumers of the gas. Consumption of hydrogen by petroleum refineries has increased recently due to clean-fuels programs, which require refiners to produce low-sulfur gasoline and ultralow-sulfur diesel fuel. Management of hydrogen is a critical concern for refiners because various processes require different hydrogen pressure levels and purity.

Hydrogen-using processes that require high pressures and high purity, including hydrocracking, use hydrogen above the 100 kg/cm2 (1,500 psig) level. When a recycle gas system is used, the higher pressures are needed to maintain hydrogen partial pressure at the desired level as methane concentration in the hydrogen feed to a hydrocracker increases. Sufficient hydrogen partial pressure promotes the intended reactions without producing undesirable coke.

If the hydrogen partial pressure cannot be maintained, the recycle gas should be bled. With pressure swing adsorption (PSA) processes producing hydrogen of purity in the range of four-nines (99.99%), this is not a problem.

Other hydrogen users, like those engaging in milder hydrotreating, can use lower-purity hydrogen at lower pressures (600 psig or lower). One approach that makes sense is recovering hydrogen from the users requiring higher pressure and reusing it at the lower pressure levels.

Minimizing the hydrogen bled into the fuel gas can keep the hydrogen production levels manageable. However, the hydrogen plant feed usually includes some hydrogen that goes through for a “free ride,” except for the cost of heating it up to reformer temperature and ultimately cooling it back down to recovery level.

A recently employed practice in the industry is for the hydrogen to be produced for adjacent producers and sold to the user as “over-the-fence” hydrogen. This keeps the production costs off the books from the adjacent user and has found popularity not only in the U.S. but also in Europe.

Figure 1. Steam-methane reforming is still responsible for the bulk of hydrogen production in petroleum refineries

Steam-Methane Reforming

Refinery hydrogen comes primarily from two sources — catalytic reforming of byproduct gas from the dehydrogenation of naphthenes into aromatics and high-octane gasoline blend stocks, as well as from direct hydrogen manufacture. The bulk of direct hydrogen manufacturing in a petroleum refinery is still accomplished via either steam-methane reforming (Figure 1) or steam-naphtha reforming. Partial oxidation of heavier hydrocarbons is also used to a limited extent.

In the overall steam methane reforming (SMR) reaction, methane reacts with steam at high temperatures and moderate pressures in catalyst-filled tubes to generate synthesis gas, a mixture of hydrogen, carbon monoxide and some carbon dioxide.

The reactions for the two simultaneous SMR mechanisms are shown as Equations (1) and (2). Both are endothermic, as shown by the positive heat of reaction. The reaction requires heat transfer to maintain temperatures favorable to the equilibrium reactions.

As the molecular weight of the feedstock increases, such as when heavier hydrocarbons (such as ethane, propane or butane) are included in the feed, the reactions are shown by Equations (3) and (4), with the corresponding heat requirements [2].

Product gas from the steam reforming of the methane and naphtha contains equilibrium amounts of hydrogen, carbon dioxide, carbon monoxide and excess steam. The calculated effluent composition of a reformer always needs to be checked against the equilibrium constant equations to ensure that simulations agree with known values.

Excess steam above the theoretical requirements is maintained to prevent the reforming catalyst from coking. The temperature exiting the reformer furnace tubes is usually about 760oC (1,400oF), a level that provides maximum hydrogen production within the temperature limitation of the reformer tube metallurgy (discussed later).

Water-shift gas reactions

Additional hydrogen can be generated from the carbon monoxide byproduct following the reforming reaction. First, the reformer effluent gas is cooled in two steps to favor the equilibrium toward the right side of the reaction. The first cooling step is followed by the high-temperature shift reactor, and the second cooling step is followed by a low-temperature shift reactor. Shift reactions are promoted as effluent gas flows down through the fixed catalyst reactor containing a ferric oxide catalyst in accordance with the reaction in Equation (5). Note the water-shift reaction is exothermic, which results in a temperature increase across the reactors as water reacts with CO to form CO2 and more H2.

Water shift gas equilibrium is not affected by pressure, since there is no volume change. Reduced temperatures favor the conversion of CO to H2, as might be expected by its exothermic nature. A variety of catalysts are available for the service.

Hydrogen Plant Process

Figure 1 shows a schematic of a conventional steam-reforming hydrogen plant [4]. The plant is based on a feed gas with high sulfur content, requiring plant operators to hydrotreat the feed before the zinc oxide removes the sulfur compounds. The H2 purification at the end of the process is based on the removal of CO2 with a pressure swing adsorber (PSA) system shown as the H2 purification block. The reformer is shown as a vertical furnace type with side firing. The reformer furnace design alternatives will be discussed below.

Feed gas — usually a mixture of hydrogen, methane and other light hydrocarbons — is first compressed to about 300 psig. The initial compression has been found to provide product hydrogen at a pressure that can easily reach the desired hydro-processing pressure with a four- or five-stage reciprocating compressor. This equipment is not part of the hydrogen plant.

The feed gas is preheated with reformer effluent gas and hydrotreated to convert the various sulfur compounds (such as mercaptans, carbonyl sulfide and carbon disulfide) to hydrogen sulfide. The gas is then passed through desulfurization reactors, usually containing a zinc oxide catalyst, which adsorbs the hydrogen sulfide. Low-sulfur feeds may not require the hydrotreating step.

Reforming furnace

The sulfur-free gas is mixed with a fixed amount of superheated steam to maintain the desired steam-to-hydrocarbon ratio. The steam-to-hydrocarbon ratio is kept within a range that is high enough to prevent laydown of coke on the reforming catalyst, but low enough to avoid overloading the reformer duty. Typically for a methane feed, the ratio would be three, whereas the theoretical requirement is somewhat less.

The combination of hydrogen and steam is heated to about 760oC (1,400oF). Since all of the reforming reactions are endothermic, additional heat is required to maintain the reaction temperature as the mixture flows down through catalyst-filled reformer tubes.

A critical factor in the reformer heater design is keeping the tube-wall temperature uniform and hot enough to promote the reforming reaction. Two types of heater designs have been employed for this purpose. Figures 2 and 3 show schematic diagrams of the side-firing reforming furnace, and the roof-fired heater design approach is shown in Figures 2 and 4.

Figure 2. Maintaining a tube-wall temperature
that is hot enough for the reforming reaction
is a critical factor in reformer heater design

Figure 3. A typical reformer furnace could
have over 300 burners

Figure 4. Hydrogen plants with single heaters
and capacities up to 100,000 ft3/d have used
a down-firing approach

Side-fired reforming heaters. The coil arrangement in a typical side-fired reformer furnace (Figure 3) consists of two parallel rectangular fire boxes connected at the top with horizontal duct work into the vertical convection stack. Two rows of vertical tubes arranged on a staggered pitch are present in each of the radiant boxes. Several (typically four) rows of burners are used to fire each side of the two radiant sections. This arrangement allows direct radiant fire to reach most of the tube wall. Platforms are provided to access the burners at each of the four burner levels. A typical reformer furnace could have over 300 burners. Reformer tubes typically have diameters of 5 in. (127 mm), walls 0.5-in. (13 mm) thick and about 34 ft (11.5 m) of wall exposed to the burners. The tube metallurgy is usually 25% chrome, 20% nickel or a high-nickel steel such as HL-40.

The inlet manifold at the top of the heater has “pigtails,” which uniformly transfer the feed gas to the top of the tubes. Another manifold at the bottom of the heater connects another set of pigtails to the outlet transfer line. The pigtails provide for thermal expansion as the heater goes from startup temperature to reaction temperature. The objective is to have an equal pressure drop across each tube, which produces uniform flow to each of the tubes. The convection section includes several different coils. The hottest coil is a steam generation coil that protects the other coils from radiant heat. Usually, there is also a steam superheat coil, a feed preheat coil and another steam generation coil. Above these coils, there may be a boiler feed water (BFW) pre-heater and deaerator preheat coil.

Typically an induced draft fan is used to keep the fire box pressure slightly negative. Some reformers also have an air pre-heater and a forced draft fan.

Top-fired reformer. This type of reformer heater is usually a rectangular box. The tubes are still vertical, and inlet and outlet pigtails are used to connect the inlet header and the outlet transfer line, respectively. Figure 4 shows a schematic diagram of a down-fired reformer furnace [9].

The tubes are spaced on a pitch, which allows the burners to fire down between the tubes. The burners have a special “pencil-shaped flame” design. All burners are located in the penthouse above the inlet manifold. The flame and the flow through the tubes travel in the same direction.

Hydrogen plants with single reformer heaters and capacities up to 100 million ft3/d have used the vertical, down-firing approach. Each burner’s radiant flame covers one-quarter of four adjacent vertical tubes (except for the outside burners, which cover half of the two adjacent tubes).

The radiant gases exit the box horizontally through a horizontal convection section. The horizontal convection section is located about 3 m above grade to allow enough height for passage. The horizontal convection provides for a simpler support structure than that of the side-fired unit.

Transfer-line steam generator

The outlet transfer line from the reformer is used to generate high-pressure (usually 650 psig) steam. The reformer effluent gas exits through the transfer line at about 1,400oF and enters the tube side of a single-pass steam generator. BFW is fed through the shell side and becomes 650 psig steam. Depending on the size of the reformer, there may be two transfer lines exiting opposite ends of the reformer and feeding two steam generators. Figure 3 shows the two transfer line steam generators.

Feed preheat exchanger.Gas is cooled to about 650oF and is moved out of the steam generator. It then enters the tube side of the feed preheat exchanger. Feed gas is preheated to about 600F using heat from the effluent gas. This temperature can be controlled by partial bypass of the effluent side to maintain the desired hot-shift gas reactor temperature.

Hot shift-gas reactor.Effluent gas containing carbon monoxide and steam is passed over the hot gas-shift catalyst, where the water-shift gas reaction shown in Equation (5) occurs. This reaction is slightly exothermic, resulting in a temperature rise across the reactor.

More steam generation.Additional medium-pressure steam is generated, reducing the hot-shift reactor effluent to a temperature of about 500oF, which shifts the reaction equilibrium toward more hydrogen production.

Cold shift-gas reaction.Additional hydrogen is produced by the gas-shift reaction at the lower temperature. The shift reaction is exothermic, which results in a temperature rise across the reactor.

Condensate removal. Cold gas-shift effluent is cooled by heat exchange with BFW, deaerator feedwater, and cooling water to about 34oC (100oF). Condensate is separated from the gas in a vertical knockout drum.

Hydrogen purification

Hydrogen purification is generally carried out using one of two approaches — solvent-based systems or pressure-swing adsorption (PSA) processes.

Solvent systems Most older units remove carbon dioxide from the hydrogen rich gas using a solvent, such as Catacarb or amines, in a typical acid gas separation unit (Figure 5).

Figure 5. Most older units remove carbon
dioxide from the hydrogen-rich gas with
a solvent

Remaining carbon oxides (primarily carbon monoxide) are reacted with hydrogen in a methanator reactor to convert them to methane. Methane is an undesirable component in the makeup gas to a hydrocracker because it builds up in the recycle gas, requiring bleeding of the recycle gas to maintain the desired hydrogen partial pressure in the hydrocracker.

Most solution-type carbon dioxide removal systems are similar. Gas enters the bottom of the absorber, where it contacts lean solution. The carbon dioxide is absorbed from the gas, leaving the rest of the contaminants and hydrogen relatively untouched.

The rich solution is then heat-exchanged with lean solution and enters the top of the stripper. The stripper uses a steam reboiler to regenerate the solvent, stripping out the absorbed carbon dioxide. The overhead from the stripper goes through a condenser to condense solvent and then to an overhead drum, where the carbon dioxide is separated from the stripper reflux.

PSA unit.The newer PSA process produces a hydrogen stream of four-nines (99.99%) purity. It separates carbon monoxide, carbon dioxide and unconverted hydrocarbons. A bank of adsorbers operates in a cycle where the adsorbers are rotated through a higher-pressure adsorption portion, followed by a pressure reduction, which allows the contaminants to be released from the adsorber. The hydrogen gas passes through the adsorber as almost-pure hydrogen. The contaminants flow into a fuelgas surge drum.

Figure 6 shows a schematic diagram of such a system. The valve openings and closings are all controlled by the central processing unit.

Figure 6. A PSA unit separates carbon monoxide, carbon dioxide and unconverted hydrocarbons from hydrogen. Adsorbers operate in a high-pressure to low-pressure cycle to adsorb and then release contaminants

The fuel gas is relatively low-BTU carbon oxides. It is supplemented with natural gas or other fuels as feed to the reformer furnace burners.

Pre- and post-reforming

These are two techniques used to expand the capacity of exisiting plants where the reformer furnace is heat-transfer-limiting.

Pre-reforming Pre-reforming is used when spiking the feed with liquified petroleum gas, which is used to increase the capacity of the existing unit. Examining the reforming Equations (1), (2) and (4) reveals the advantage of a heavier feed that yields more hydrogen per feed mole. The pre-reformer reaction breaks down the heavier hydrocarbons (propane and butane) to methane ahead of the heat-intensive reforming reactions, essentially shifting part of the load upstream of the reformer heater as shown in Figure 7 [8].

Figure 7. A pre-reformer breaks down heavier hydrocarbons into methane ahead of the reforming reactions

Feed at 950oF passes down through the pre-reformer reactor, where the breakdown reactions occur. Then the pre-reformed feed passes through another convection coil to reheat it to about 1,100oF before entering the reformer.

Adding the pre-reformer as a retrofit to an existing facility presents two problems — one of space and one of compatibility. Physical space contraints may not allow adding a feed reheat coil within the convection section. Also, the metallurgy of the inlet pigtails may not be able to handle the higher feed temperature.

Post-reforming. Post-reforming is an attempt to provide additional reforming catalyst outside the reformer heater. A down-flow reactor is added between the outlet transfer line and the waste heat steam generator. This can present a space and piping problem. The additional post-reformer catalyst reduces the overall total space velocity of the combined reformer and post reformer, thus achieving additional reaction. This reduces the downstream shift-reaction requirements. Edited by Scott Jenkins

Share |

Thursday, April 29, 2010

Relief Device Design

Pressure relief valves or other relieving devices are used to protect piping and equipment against excessive over-pressure. Proper selection, use, location, and maintenance of relief devices are essential to protect personnel and equipment as well as to comply with codes and laws. Determination of the maximum relief required may be difficult. Loads for complex systems are determined by conservative assumptions and detailed analysis. By general assumption, two unrelated emergency conditions caused by unrelated equipment failures or operator error will not occur simultaneously (no double jeopardy). The sequence of events must be considered. The development of relief loads requires the engineer to be familiar with overall process design, including the type of pump drives used, cooling water source, spares provided, plant layout, instrumentation, and emergency shutdown philosophy.

This section suggests methods to calculate relief capacity for most emergency conditions, including fire. A common reference for determining individual relieving rates is contained in Section 3 of API RP 521. The design of the proper relieving device must take into consideration all of the following upset conditions for the individual equipment item if such upset can occur. Each upset condition must be carefully evaluated to determine the "worst case" condition which will dictate the relieving device capacity.
Blocked Discharge
The outlet of almost any vessel, pump, compressor, fired heater, or other equipment item can be blocked by mechanical failure or human error. In this case, the relief load is usually the maximum flow which the pump, compressor, or other flow source produces at relief conditions.
Fire Exposure
Fire is one of the least predictable events which may occur in a gas processing facility, but is a condition that may create the greatest relieving requirements. If fire can occur on a plant-wide basis, this condition may dictate the sizing of the entire relief system; however, since equipment may be dispersed geographically, the effect of fire exposure on the relief system may be limited to a specific plot area. Vapor generation will be higher in any area which contains a large number of uninsulated vessels. Various empirical equations have been developed to determine relief loads from vessels exposed to fire. Formula selection varies with the system and fluid considered. Fire conditions may overpressure vapor-filled, liquidfilled, or mixed-phase systems.
Tube Rupture
When a large difference exists between the design pressure of the shell and tube sides of an exchanger (usually a ratio of 1.5 to 1 or greater), provisions are required for relieving the low pressure side. Normally, for design, only one tube is considered to rupture. Relief volume for one tube rupture can be calculated using appropriate sizing equations in this section. When a cool media contacts a hot stream, the effects of flashing should be considered. Also the possibility of a transient overpressure caused by the sudden release of vapor into an all-liquid system should be considered.
Control Valve Failure
The failure positions of instruments and control valves must be carefully evaluated. In practice, the control valve may not fail in the desired position. A valve may stick in the wrong position, or a control loop may fail. Relief protection for these factors must be provided. Relief valve sizing requirements for these conditions should be based on flow coefficients (manufacturer data) and pressure differentials for the specific control valves and the facility involved.
Thermal Expansion
If isolation of a process line on the cold side of an exchanger can result in excess pressure due to heat input from the warm side, then the line or cold side of the exchanger should be protected by a relief valve. If any equipment item or line can be isolated while full of liquid, a relief valve should be provided for thermal expansion of the contained liquid. Low process temperatures, solar radiation, or changes in atmospheric temperature can necessitate thermal protection. Flashing across the relief valve needs to be considered.
Utility Failure
Loss of cooling water may occur on an area-wide or plantwide basis. Affected are fractionating columns and other equipment utilizing water cooling. Cooling water failure is often the governing case in sizing flare systems. Electric power failure, similar to cooling water failure, may occur on an area-wide or plant-wide basis and may have a variety of effects. Since electric pump and air cooler fan drives are often employed in process units, a power failure may cause the immediate loss of reflux to fractionators. Motor driven compressors will also shut down. Power failures may result in major relief loads.
Instrument air system failure, whether related to electric power failure or not, must be considered in sizing of the flare system since pneumatic control loops will be interrupted. Also control valves will assume the position as specified on "loss of air" and the resulting effect on the flare system must be considered.
Share |

Tuesday, April 27, 2010

Gasoline and LP-Gas Treating

Most gasoline and LP-gas streams contain sulfur in various forms and LP-gas and raw NGL streams also can contain carbon dioxide. Especially objectionable are hydrogen sulfide, mercaptans, and elemental sulfur. Gasoline containing hydrogen sulfide has objectionable odor and is corrosive. Mercaptans give an objectionable odor to gasoline and elemental sulfur makes the gasoline corrosive. Carbonyl sulfide in the LP-gas can hydrolyze and cause the product to become corrosive. A product containing the objectionable materials can be treated to remove the hydrogen sulfide, carbonyl sulfide, and elemental sulfur and to either remove the mercaptans or convert them to less objectionable compounds.

Hydrogen Sulfide and Carbon Dioxide Removal
Hydrogen sulfide and carbon dioxide can be removed from LP-gas and gasoline by liquid-liquid contacting processes using a caustic solution, aqueous alkanolamines, or solid KOH. The caustic wash processes are described later. The alkanolamine processes were described earlier in the gas treating section. For application, both a coalescer and a full flow carbon filter should be used to minimize the introduction of hydrocarbons into the regeneration section of the amine unit. When treating hydrocarbon liquids with amine, contacting is generally accomplished in a liquid-liquid contactor, though stirred reactors can be used. The tower should have a minimum of 20 feet of packing. The design flow rates for packed towers should not exceed 20 gpm liquid per square foot of cross sectional area.
Sulfur Removal
Elemental sulfur is removed from the gasoline by contacting it with a polysulfide wash solution. The solution is made up by using the following amounts of chemicals per 1,000 gallons of water: 1,000 lb. of caustic (NaOH), 800 lb. of commercial Na2S, and 20 lb. of sulfur. The sodium sulfide (Na2S) is melted in a vat by use of a steam lance. Add the sulfur to the melted Na2S. The sulfur must be completely dissolved in the liquid sulfide, and then this mixture is added to the 10% (1,000 gal. water and 1,000 lb. NaOH) caustic solution. Protective clothing and goggles should be worn when handling these chemicals.
Mercaptan Treating
Mercaptans can be converted to disulfides or removed from liquid hydrocarbons by several methods. The method or combination of methods that can be used depends on the mercaptan content of the product to be treated and the specification that must be met.
Carbonyl Sulfide Removal
Because carbonyl sulfide (COS) can hydrolyze and cause sweet LP-gas to become corrosive, and as the concern for minimizing total sulfur emissions has increased, there has been a growth of interest in removing COS from propane and LP-gas streams. Several alternative processes can be used. COS can, of course, be removed sacrificially by MEA. Mick has reported successful use of a combination of potassium hydroxide and methyl alcohol. This process is also sacrificial. The Malaprop process uses diglycolamine but requires unspecified modifications in the process flow from that used for gas treating. The ADIP process utilizes aqueous diisopropanolamine. Molecular sieves can be used for removing COS. The Malaprop, ADIP and molecular sieve processes are regenerative.
Share |

Tuesday, April 20, 2010

Burner Retrofit Considerations

Flame Envelope

1. Conventional raw gas and premix burners have luminous flames. The combustion reaction occurs within the visible flame boundaries. The flame envelope is defined as the visible combustion length and diameter. Ultra-Low NOx and latest generation burners have non-luminous flames. Much of the combustion reaction is not visible. The flame length and diameter are often determined by inserting a CO probe into the firebox and defining the flame envelope as CO concentrations greater than 50 ppm.

2. Ultra-Low NOx and latest generation burners have longer flame lengths than conventional burners. Longer flame lengths change the heat transfer profile in the firebox. Longer flame lengths can result in flame impingement on the tubes and mechanical supports.

3. The flame diameter is often defined in terms of ratios of the burner tile outside dimension. Many burners have flame diameters that are1 to1-1/2 times the diameter of the burner tile. Since the tile diameters are often larger for Ultra-Low NOx and latest generation burners, the flame diameters at the base of the flame many be slightly larger. The flame diameter often necks outs, giving a wider flame at the top.

Physical Dimensions of Firebox

1. Optimized designs have burner spacing that is designed to have gaps between the flame envelopes. Since the tile diameters are often larger for Ultra-Low NOx and latest generation burners, retofits can result in closer Burner-burner spacing and flame interaction. Flame interaction can produce longer flames and higher NOx values. Flame interaction and congealing can interupts the flue gas convection currents in the firebox, reducing the amount of entrained flue in the flame envelope. This condition increases the NOx levels. Ultra-Low NOx and latest generation burners should be spaced far enough apart to allow even flue gas recirculation currents to the burners.

2. The burner centerline to burner centerline dimension is one of the most important dimensions in the firebox. Many tube failures are causes by flame and hot gas impingement. When Ultra-Low NOx and latest generation burners are being retrofitted, the larger size of the flame envelope must be evaluated. Firebox convection currents can push the slow burning flames into the tubes.

3. Flame impingement on refractory often causes damage. When Ultra-Low NOx and latest generation burners are being retrofitted, the larger burner diameter may result in the burners being spaced closer to the refractory. Unshielded refractory may require hot face protection.

4. Many heaters are designed for flame lengths that are 1/3 to ½ the firebox height. Ultra-Low NOx and latest generation burners typically have flame heights of 2-2.5 ft/million Btu (2-2.5 m/MW). Longer flame heights from Ultra-Low NOx and latest generation burners may change the heat transfer profile in the firebox. The longer flames may result in flame or hot gas impingement on the roof and shock tubes. These tubes may require protection to prevent failures. Protection may include metallurgical upgrades, increased tube thickness or tube shielding. Some older heaters have very short firebox heights and may not be suitable for retrofits to Ultra-Low NOx and latest generation burners.

5. When retrofitting burners, may companies test prototype burners in a test furnace with similar geometry and fuels as the heater to be retrofitted.

Fuel Treatment

1. While many conventional burners have orifices 1/8”(3 mm). and larger, Ultra-Low NOx and latest generation burners often have tip drillings of 1/16”(1.5 mm). These small orifices are extremely prone to plugging and require special protection. Most fuel systems are designed with carbon steel piping. Pipe scale forms from corrosion products and plugs the burner tips. Tip plugging is unacceptable for any burner, but it is even more important not to have plugged tips on Ultra-Low NOx and latest generation burners because plugged tips can result in stability problems and higher emissions. Many companies have installed austenitic piping downstream of the fuel coalescer/filter to prevent scale plugging problems.

2. Coalescers or fuel filters are required on all Ultra-Low NOx and latest generation burner installations to prevent tip plugging problems. The coalescers are often designed to remove liquid aerosol particles down to 0.3 to 0.6 microns. Some companies install pipe strainers upstream of the coalescer to prevent particulate fouling of the coalescing elements.

3. Piping insulation and tracing are required on fuel piping downstream of the coalescer/fuel filter to prevent condensation in fuel piping. Some companies have used a fuel gas heater to superheat the fuel gas in place of pipe tracing.

4. Unsaturated hydrocarbons can quickly plug the smaller burner tip holes on Ultra-Low NOx and latest generation burners.

5. Higher hydrogen content in the fuel gas results in higher NOx production. It also increases the stability of the flame.

6. Natural gas fuel has often produced unstable flames and flameouts on Ultra-Low NOx and latest generation burners. Its use should be avoided if possible.

Air Control

1. Ultra-Low NOx and latest generation burners must be operated at design excess air levels to control NOx emissions.

2. Most refinery general service heaters are natural draft heaters. It is important to control the draft at the design value, usually 0.1”(3 mm) H2O at the top of the radiant section. High drafts increase tramp air ingress and often result in higher excess air levels at the burners. This condition results in higher NOx levels on Ultra-Low NOx and latest generation burners. Automated draft control has been installed on retrofits to obtain better excess air control.

3. Ultra-Low NOx and latest generation burners are usually supplied with individual plenum boxes and individual damper controls. Because excess air control is so important on these burners, some companies have installed individual actuators on each burner damper for better control. Others have connected all the burner dampers on a jack shaft to control the excess air levels.

4. New heaters are designed with seal welded construction to prevent tramp air ingress. Many older heaters have bolted panel design. High temperature silicon and foil tape have been used on these heaters to reduce tramp air. Observation openings should be designed to minimize excess air ingress. Observation openings should be closed when not in use.

5. Ultra-Low NOx and latest generation burners are usually supplied with individual plenum boxes. Many of these burners are supplied with mufflers to control noise emissions. The mufflers are often an effective devise to eliminate excess air fluctuations due to wind. Windscreens are often installed to eliminate wind effects when burner mufflers are not used. A 15 mph wind can cause a ± 0.11” H2O draft variation at the burner, resulting in a ± 15%change in excess air level for a burner designed at 0.4” H2O draft.

6. Forced draft systems should be considered for Ultra-Low NOx and latest generation burner retrofits. The forced draft system provides better excess air control, eliminates wind effects, and the increased burner pressure drop often results in a smaller flame envelope.

7. Ultra-Low NOx and latest generation burners may be installed in a common air plenum. Internal baffles may be required to obtain even air distribution.

Structural Considerations

1. Hole for Hole Replacement is the optimum situation for retrofits. However, since many Ultra-Low NOx and latest generation burners have larger burner tiles and larger burner cutouts, hole for hole replacement cannot occur. It is often more economical to replace the floor when hole for hole replacement is not an option.

2. Ultra-Low NOx and latest generation burners often weight more than conventional burners. Retrofits may require additional structural bracing. Floor Levelness/ Refractory Thickness

3. Heaters floor steel should be level. Bowed sections should be repaired or replaced.

4. The floor refractory thickness should be checked to ensure the heater floor steel is an acceptable temperature.

5. Physical constraints below the firebox floor should be checked. There should be sufficient space underneath the burner plenum for tip removal.

Process Related Parameters

1. Ultra-Low NOx and latest generation burners often have longer flames that change the heat flux profile. This is especially important on cracking heaters such as cokers and visbreakers. The longer flames may increase the bridgewall temperature and change the duty split between the radiant section and convection section.

2. When the heat flux profile changes, the location of the maximum tube metal temperature changes. Retrofitting Ultra-Low NOx and latest generation burners in short fireboxes can result in high roof and shock tube metal temperatures.

3. Ultra-Low NOx and latest generation burners may have less turndown capability than conventional burners. High CO levels can occur when firebox temperatures are below 1240ºF. Flame instability and flameout can occurred when firebox temperatures are below 1200ºF.

4. Conventional raw gas burners can handle a wide variation in fuel gas composition. Since Ultra-Low NOx and latest generation burners are often designed at the limit of stability, a fuel composition change may cause a stability problem. Since methane fuel is the hardest fuel to burn, many companies specify burn test using methane as the test fuel.

5. The proper design basis for the burner retrofit is extremely important. Sometimes the process requirements have changed significantly since the furnace was designed. Important design basis items include:

1) Emission Requirements

2) Process Duty Requirements

3) Heater General Arrangement Drawings

4) Turndown Requirements

5) Fuel Composition Ranges

6) Fuel Pressure

7) Startup Considerations

6. It is important to review existing plant data accurately. Heater tube fouling may result in high bridgewall temperatures. Fouled convection sections may result in higher firing rates. Tramp air may result in high excess air levels. Stack dampers are often frozen in place.


1. When retrofitting Ultra-Low NOx and latest generation burners, addition heater instrumentation is often required. Most companies will install analyzers to determineO2 and NOx levels. Some companies will install CO and combustibles analyzers.

2. When retrofitting burners, may companies install the firebox draft indication and damper control on the DCS system to obtain better excess air control.

3. When retrofitting Ultra-Low NOx and latest generation burners, Flame Scanners may be require to protect against flameouts during turndown and startup situations.

4. When retrofitting burners, minimum fuel gas pressure instrumentation may be require to protect against flameouts during turndown and startup situations.

5. The heater should have a bridgewall temperature indicator on the DCS system.


1. Many operators have been trained to observe luminous conventional flames. After retrofitting Ultra-Low NOx and latest generation burners, the operators will have to be trained to observe non-luminous flames. A bright yellow flame on an Ultra-Low NOx burner may be an indication of a burner setup problem.

2. Ultra-Low NOx and latest generation burners have non-luminous flames that are hare to detect. Bright burner tile or flame holder color is often an indication that the primary tips are operating properly.

3. Special startup procedures may be required for Ultra-Low NOx and latest generation burners.

4. Special premix gun inserts and pilot designs may provide additional stability during startup and turndown conditions.

5. Ultra-Low NOx and latest generation burners may have less turndown capability than conventional burners. High CO levels can occur when firebox temperatures are below 1240ºF. Flame instability and flameout can occurred when firebox temperatures are below 1200ºF.

5. Ultra-Low NOx and latest generation burners are designed to operate within closely controlled oxygen levels to obtain the lowest levels of NOx. This may require more operator attention and interaction.

Installation Checkout

1. Correct burner installation is extremely important on Ultra-Low NOx and latest generation burners. It is often beneficial to have burner company representatives assist in checkout before initial operation.

2. Tip orientation and tip height should be checked

3. The burner tile must be installed properly. Check the diameter in different locations to ensure proper diameter dimensions and concentricity.

4. The damper should be checked to ensure freedom of movement through the entire range of operation.

Instability Issues Specific with Low NOx Burners

CFD Modeling

1. CFD modeling (Computational Fluid Dynamic modeling) is a useful tool in modeling firebox conditions. It has been used in multi-burner systems to analyze problems such as flame interactions, firebox currents, and localized high heat fluxes.

2. CFD modeling capability is relatively expensive and should be limited to special situations.

3. CFD modeling capability is improving as companies gain experience. However, field results may vary significantly from the model results.

Share |


Twitter Delicious Facebook Digg Stumbleupon Favorites