IntroductionEthylene is typically obtained from the non-catalytic thermal cracking of saturated hydrocarbons such as ethane and propane and alternatively from the thermal or steam cracking of heavier liquids such as naphtha and gas oil. Steam cracking produces a variety of other products, including diolefins and acetylene. The latter are costly to separate from the ethylene, usually by extractive distillation and/or selective hydrogenation of the acetylene back to ethylene. An ethylene plant typically achieves an ethylene selectivity of about 85 percent calculated on a carbon atom basis at an ethane conversion of about 60 percent. In addition, thermal cracking processes for olefin production are highly endothermic. Accordingly, these processes require the construction and maintenance of large, capital intensive and complex cracking furnaces to supply the heat. An alternative is to catalytically crack paraffinic hydrocarbon in the presence of oxygen to form mono-olefins, that is, autothermal partial oxidation of paraffinic hydrocarbons to olefins. The words "partial oxidation" imply that the paraffinic hydrocarbon is not substantially oxidized to deep oxidation products, specifically, carbon oxides. Rather the partial oxidation comprises one or both of oxidative dehydrogenation and cracking to form primarily olefins. The process is conducted under autothermal reaction conditions wherein the feed is partially combusted, and the heat produced during combustion drives the endothermic cracking process. Under these autothermal process conditions there is no external heat source required. However, substantial amounts of carbon oxides are usually formed, and the selectivity to olefins has been low compared to thermal cracking. Dow has recently been awarded a patent for an autothermal process for the production of olefins which shows promise in terms of ethylene yields (US 6,566,573). This process and catalyst for the partial oxidation of paraffinic hydrocarbons such as ethane, propane, naphtha and natural gas condensates to olefins such as ethylene and propylene involves contacting a paraffinic hydrocarbon with oxygen in the presence of hydrogen and a catalyst under autothermal process conditions. Preheating the feed decreases oxygen consumption and increases the net hydrogen balance. The catalyst comprises a Group 8B metal, preferably a platinum group metal, and at least one promoter selected from Groups 1B, 6B, 3A, 4A, and 5A, optionally supported on a catalytic support, such as magnesia or alumina. A fluid bed reactor is proposed as the preferred reactor style. A second patent (US 6,624,116) describes an on-line method of synthesizing or regenerating catalysts for autothermal oxidation processes. The dehydrogenation of ethane to produce ethylene has been covered in a previous PERP report (98/99S9). The present report focuses on the production of ethylene via catalytic ethane oxidation. Dow PatentsThe Dow patent U.S. 6,566,573 involves the partial oxidation of a paraffinic hydrocarbon to form an olefin. Numerous examples are provided in the patent. In the preferred embodiment of the catalyst composition, the platinum group metal is platinum. The promoter is selected from tin, copper, and mixtures thereof. The support is selected from alumina, magnesia and mixtures thereof. Magnesia is the preferred support material because it produces fewer cracking products and less carbon monoxide. Moreover, the hydrocarbon conversion and olefin selectivity tend to be higher with magnesia. The modifier is selected from tin, lanthanum and mixtures thereof. The patent examines various catalysts: platinum/tin, platinum/antimony, platinum/tin/antimony, and platinum/copper, the latter being the preferred based on the test results. A second patent awarded to Dow (U.S. 6,624,116) describes the on-line synthesis and regeneration of the catalyst used in the autothermal oxidation. The invention involves simply feeding a volatile Group 8B metal compound and/or a volatile promoter compound into the oxidation reactor simultaneously with the reactant feedstream under ignition conditions or autothermal process conditions. The regeneration can be beneficially employed on-line to replace metal components of the catalyst, which are lost over time through vaporization. This procedure eliminates the necessity of preparing the catalyst prior to loading the reactor and eliminates the necessity of shutting down the reactor to regenerate or replace the deactivated catalyst. Since the process produces little, if any, coke, burning coke deposits off the catalyst is not necessary. The patent examined several feeds including ethane, propane and natural gas liquids (NGL). Ethane is converted primarily to ethylene. Propane and normal butane are converted primarily to ethylene and propylene, while isobutane is converted primarily to isobutylene and propylene. Naphtha and other higher molecular weight paraffins are converted primarily to ethylene and propylene. One example in the patent evaluated the yields as a function of gas hourly space velocity (GHSV) in a fixed bed reactor (no preheat, pressure = 170 kPa). Over the wide range of space velocities tested, it was found that the ethane conversion and ethylene selectivity did not change significantly at constant temperature and pressure. Also at constant pressure and space velocity, the conversion increases with increasing temperature. The process was evaluated for variations in pressure. It was found that as the pressure of the process increased the ethane conversion increased and the ethylene selectivity decreased. While running in an autothermal mode, the feed does not have to be preheated, so long as the feed contains hydrogen or the catalyst supports combustion beyond the normal fuel-rich limit of flammability. Preheating the feedstream, however, has certain advantages. The advantages comprise a decrease in oxygen and hydrogen consumed, an increase in the paraffin concentration in the feed, an increase in the operating paraffin to oxygen molar ratio, and a net increase in recycle hydrogen in the product stream. In addition, catalysts can be used which do not support combustion beyond the normal fuel-rich limit of flammability. The process was evaluated for the addition of preheat using a platinum and copper on tin and lanthanum-modified alumina monolith supported catalyst. The feed was preheated to temperature ranging from 281ºC to 389ºC (538ºF to 1,092ºF). It was found that by preheating the feed to temperatures about 400ºC substantially the same ethane conversion and product selectivities were obtained at higher hydrocarbon to oxygen molar ratio, as were obtained at lower preheat and lower hydrocarbon to oxygen ratios. Thus, a preheat temperature of 275ºC (527ºF) was selected as the optimum for further evaluation. Examples are presented in the patent with and without hydrogen present with the feed. The examples with hydrogen present showed significantly higher yields of ethylene and lower yields of carbon oxides. In fact, the presence of hydrogen reduces the carbon dioxide selectivity to less than 0.5 percent in some examples and improves the selectivity to ethylene by over 10 percent in all examples. This patent also shows that the presence of tin or antimony or a combination of both also improves the selectivity to ethylene by almost 10 percent. Most of the patent examples employ a fixed bed type reactor. For the fluid bed example, alumina beads were used to prepare a catalyst containing platinum, copper and tin. The preheat temperature was 275ºC and the operating pressure was 135 kPa. It was seen that a reactor operating at slightly above minimal fluidization could be used for the autothermal oxidation of ethane to ethylene to achieve high selectivity to ethylene and low selectivities to methane and carbon oxides. Conversions of over 70 percent and ethylene selectivity of about 82 percent were achieved. In comparing similar results, Dow concluded that although both the fixed bed and the modified fluidized-bed reactors were suitable, the selectivities were more favorable in the modified fluidized-bed reactor. Less methane and carbon oxides were obtained, and more ethylene was produced at closely similar conversions. Catalyst attrition is not discussed but would need to be examined before any commercial demonstration is considered. Attrition plays an important role in the development of a fluid-bed catalyst. By employing a fluid-bed reactor there is a noticeable difference in the capital investment due to the elimination of the large, expensive and complex furnaces used in conventional steam cracking processes. Process DescriptionA conceptual flow diagram has been developed for the partial oxidation of ethane to ethylene using a fluid-bed reactor. Except for the reactor, the process flow follows conventional recovery techniques. Only a small facility is required for the hydrogenation of acetylene, due to reduced quantities made relative to conventional steam crackers. Capital InvestmentThe ethane POX unit has a simpler reactor system and fewer fractionation columns than the corresponding ethane cracker, which results in about a 15 percent lower investment. The OSBL investment for the ethane POX process compared to the conventional route is about 10 percent lower because there are fewer by-products expected with the POX route, and thus the investment in storage facilities is reduced significantly. Production CostsBased on a 620 thousand metric ton per year (1.37 billion pound per year) production facility located on the U.S. Gulf Coast for the second quarter of 2003, the POX route to ethylene has a somewhat higher cash cost of production. The higher cash cost of production is due to the slightly lower efficiency in converting ethane to ethylene and the additional cost of oxygen. In addition because the patent did not define the by-products very well, fuel value was used to determine the by-product credits for the POX route to ethylene. Better by-product definition could justify higher credits. Commercial AnalysisThe report presents a detailed global picture of ethylene supply and demand. Ethylene capacity is provided by region, by producer share, and by feedstock. Capacity is estimated for 2005 and 2010. Essentially no new crackers are planned for Japan, Korea, the United States, or Western Europe. Most new crackers will be based on ethane feedstock and built by oil companies in joint ventures with government entities. This is in contrast to the current 50 percent share for naphtha feedstock versus 30 percent for ethane alone or ethane/propane mix. Presently, the top 10 producers account for about one-third of global ethylene capacity. The report also includes regional analysis of supply and demand for the United States, Western Europe, and Asia Pacific. ConclusionWhile our speculative analysis reveals this POX process it not economically favored, the POX process does represent an interesting development. Low capital investment is an attraction of the POX process. By-products, catalyst performance (activity, regeneration, and durability), and the fluid bed reactor design require further definition and refinement. 
| 
