Published April 2012
While methanol carbonylation dominates the current era of acetic acid production, two routes employing the oxidation of hydrocarbons (ethylene and ethane) vie for recognition. These routes have previously been evaluated in PEP Report 37C [PEP037C] at reduced scale, and the analysis concluded that the routes might serve isolated local demands but could not compete on the commercial scale. Since methanol carbonylation supplanted the two-step oxidation route via isolated acetaldehyde, the prospect of a single-step feedstock to acetic acid route might recapture economic efficiency from carbonylation. However, the findings from the previous analyses suggested otherwise, which raises several questions. Is the reduced scale the sole aspect that rendered the previous analyses suboptimal? Or is the oxidation technology inherently inferior? This review addresses those questions by updating the previous evaluations to the present and extrapolating the plant size to the current maximum world scale set by Celanese, namely 600 thousand metric tons per year (kta).
The Showa Denko (SD) route oxidizes ethylene to acetic acid in a single-step conversion, which should generally be more cost effective than the conventional two-step ethylene oxidation process (via acetaldehyde). Therefore, the initial expectation is that the SD route should be closer to economic equivalence with carbonylation at the same economy of scale.
SABIC's ethane-based oxidation route to acetic acid functions as a single-step oxidation using surplus ethane, but also generates by-product ethylene and CO2. The single reaction step with no intermediate isolation suggests a similar impact as the SD route, and the use of surplus feedstock that might otherwise be flared hints at a potential economic advantage. However, significant by-product production requires more of a trade-off assessment depending on the marketability of the by-products. The CO2 by-product presents a particular concern since it becomes an environmental problem and an economic liability for this process if no end use is available. Unsatisfactory economics shown in the previous analysis [PEP037C] at reduced scale again suggests that a world-scale configuration might provide economic feasibility.