(75a) Additives to Prevent Coking in Ethylene Furnaces

Coke formation in ethylene furnaces during naphtha cracking is produced by a mechanism in which the iron and nickel in the metal reactor walls catalyze the formation of graphitic carbon in the shape of long filaments. The formation of coke on the metal tube walls is thought to be a three step process where 1) hydrocarbons decompose on the hot metal surface, 2) carbon atoms diffuse into the metal lattice where they react with Fe and Ni forming metastable iron and nickel carbides, which then 3) decompose into graphitic carbon and metal according to C2H2n+2 --> C + Fe + Ni --> Fe3C + Ni3C --> Fe + Ni + graphite. The formation of graphitic carbon is a very strong thermodynamic driving force for the reaction. Once graphitic carbon forms in the bulk metal, it creates pressure within crystal structure, causing some small metal particles to become dislodged. In the pyrolytic environment of the naphtha cracking process, the dislodged crystallites continue to catalyze carbon formation, sometimes forming a carbon filament on the back side of the metal particle. Understanding this mechanism for coke deposition is critical to identifying approaches for minimizing coking during naphtha cracking for ethylene production. TDA Research Inc. has identified several additives that block the formation of the metastable carbide intermediates by forming strong, thermally stable surface compounds upon which filamentous coke cannot form. These compounds are stable at temperatures in excess of 1000C (1832F), and in the case of one of our more recent formulations, also resists attack by sulfur. We have conducted laboratory experiments using hot metal tubes to determine how well our additives suppressed coke formation and found that under conditions that would otherwise result in rapid coking, our additives completely blocked filamentous coke formation when used at concentrations as low as 10 ppm (Figure 1). In tests done using stainless steel tube sections, 761 mg of coke was formed in 8 hours when our additive was not used. In contrast, only 58 mg coke was formed in 16 hours when 10 ppm of our additive was used. In addition, microscopic examination showed that the coke that did form was not filamentous (Figure 2). We also found that pretreating the stainless steel tube sections with 10 ppm of our additive for four hours prior to n-heptane cracking, filamentous coke formation was suppressed for up to 40 hours after the flow of additive was stopped. Finally, we tested Inconel 625 and Incoloy 800 HT tube sections at 900C, with and without one of our coke suppression additives and for both alloys, the presence of our additive effectively eliminated the formation of filamentous coke.