Radiant coil

Commercially and industrially most important, ketene itself, H2C—C—O, is produced by pyrolysis of acetic acid [64-19-7]. In this process, high quahty acetic acid is evaporated and the vapor passed continuously through a radiant coil under reduced pressure at 740—760°C. 

The materials of constmction of the radiant coil are highly heat-resistant steel alloys, such as Sicromal containing 25% Cr, 20% Ni, and 2% Si. Triethyi phosphate [78-40-0] catalyst is injected into the acetic acid vapor. Ammonia [7664-41-7] is added to the gas mixture leaving the furnace to neutralize the catalyst and thus prevent ketene and water from recombining. The cmde ketene obtained from this process contains water, acetic acid, acetic anhydride, and 7 vol % other gases (mainly carbon monoxide [630-08-0][124-38-9] ethylene /74-< 3 -/7, and methane /74-< 2-<7/). The gas mixture is chilled to less than 100°C to remove water, unconverted acetic acid, and the acetic anhydride formed as a Hquid phase (52,53). 

Larger-si2ed heaters are usually hori2ontal box heaters. The radiant coils can be located either on the side walls so that the units are fired from underneath, or in a center row of tubes in which the heater is fired from both sides to provide a higher heat flux for reducing the radiant surface. An access area at one end of the box is required in order to remove the tubes. Sometimes multiple coils are included in the same box, which may require access to both ends of the box. 

Over 25 years ago the coking factor of the radiant coil was empirically correlated to operating conditions (48). It has been assumed that the mass transfer of coke precursors from the bulk of the gas to the walls was controlling the rate of deposition (39). Kinetic models (24,49,50) were developed based on the chemical reaction at the wall as a controlling step. Bench-scale data (51—53) appear to indicate that a chemical reaction controls. However, flow regimes of bench-scale reactors are so different from the commercial furnaces that scale-up of bench-scale results caimot be confidently appHed to commercial furnaces. For example. Figure 3 shows the coke deposited on a controlled cylindrical specimen in a continuous stirred tank reactor (CSTR) and the rate of coke deposition. The deposition rate decreases with time and attains a pseudo steady value. Though this is achieved in a matter of rninutes in bench-scale reactors, it takes a few days in a commercial furnace. 

The reactant mixture then enters the tubular reactor or the radiant coil at the cross-over temperature generally above 1000° F. It is rapidly heated to the cracking temperature by radiant heat supplied by burners in the combustion chamber. The gas leaving the reactor enters the transfer line exchanger where it is rapidly quenched to avoid decomposition of valuable products. 

The following discussion is confined to the radiant coil where pyrolysis reactions occur. 

Furnace Simulation. The purpose of this example is to demonstrate the capability of the PF60 system to predict yield structure and the tubeskin temperature in commercial operating furnaces. Table II summarizes the data from a commercial reactor processing primarily propane as feedstock. At the time the data were taken, the furnace had been on stream less than four days and hence an unfouled radiant coil condition could be assumed. The yields were recorded by on-line chromatographs. The tubeskin temperatures were measured in 15 locations by calibrated infrared pyrometers. 

During decoking of the radiant coils, the TLXs can also be partially decoked. For complete decoking, the furnace is usually cooled down and the TLXs are separated from the coils and hydrojetted with high pressure water. In some cases, the coke in the TLXs can also be burnt off, and hence no mechanical cleaning is required. 

The research programme was initiated through an investigation of radiant coil alloys and deposited coke. Radiant coils are constructed in high alloy steels (eg Incoloy 800, HK 40, 34 CT), with iron, nickel and chromium as major components and a range of minor constituents such as silicon, titanium and manganese. 

Soft X-ray appearance potential spectroscopy (SXAPS) studies of radiant coil alloys were carried out to determine surface compositions resulting from exposure to a range of environments. 

Coke deposits from a commercial ESC plant were also examined. Bulk metals concentrations were determined by ashing at 900°C prior to emission spectroscopy analysis. High levels of metals were found at the radiant coil-coke interface due, presumably, to oxide spallation, but concentrations of metals (primarily chromium, iron and titanium,totalling ca200 ppm) were also found into the bulk of the coke and at the coke-process stream interface (Table I). SXAPS examination of coke heated under vacuum for several hours at 900°C also revealed the appearance of up to 1% concentration of metals at the surface of some coke samples, and significantly chromium, iron and titanium were again the metals observed. 

Several studies (7,8) have indicated that the propensity of transition metals for catalysing coke formation increases with decreasing stability of the metal carbide - ie increasing propensity along the series Ti < Cr < Fe < Ni. However, the examination of the ESC radiant coil revealed only low levels of nickel in the surface oxide and more particularly virtually no nickel in the coke samples. Therefore, iron and possibly chromium, which are both present at the cracker coil surface and within the coke, are more likely to be responsible for catalysed coke formation. This is supported by research of Albright and coworkers (5) on the pyrolysis of hydrocarbons over Incoloy 800 surfaces, which found iron to be the predominant metal in the coke. 

In summary, these investigations suggested that coke deposition in ESC plant is enhanced by catalytic interaction of metals from the radiant coil walls, a view later supported by high resolution TEM studies. This led to an examination of means of passivating alloy surfaces. 

These experiments demonstrated that the passivating qualities of steam pretreatment may be partially associated with the generation of a silica-rich oxide. However, brief attempts to increase the surface concentration to approach a continuous silica layer excluding all the alloying metals (by variations in preoxidation conditions) met with failure, exhibiting bulk inhomogeneity and oxide spallation problems. Further, many ESC plants are constructed with radiant coil alloys low in silicon content, and so preoxidation can only be considered a partial palliative. Hence, attention turned to the prospect of coating the internal surfaces of radiant coils with a thin, continuous layer of silica. 

Mindful of the difficulty of coating the internal surfaces of long ESC radiant coils, a chemical vapour deposition (CVD) route was seen as presenting the best approach. The deposition of silica by CVD onto silicon is well known in the electronics industry (11), but had not been extensively applied to metals. 

Run Length, Coke is produced as a side product that deposits on the radiant tube walls. This limits the heat transfer to the tubes, and increases the pressure drop across the coil. The coke deposition not only limits the heat transfer, but also reduces the olefin selectivity. Periodically, the heater has to be shut down and cleaned. Typical mn lengdis are 40 to 100 days between decokings. Prediction of mn length of a commercial furnace is stiU an art, and various mechanisms are postulated in the literature. Coke also deposits in transfedine exchangers. Mechanisms for coking in radiant coils and transfedine exchangers appears to be different for different feeds. 

The feed system is fairly conventional. Light feeds are pressured into the unit and liquid feeds are pumped from weighed feed tanks. The radiant coil is divided into four sections of small diameter stainless tubing. The outlet bulk gas temperature of each zone... 

---