Feb-2010

Reducing sulphur oxide emissions from the FCC unit

Application of a SOx-reduction additive enabled a refiner to meet a pressing requirement to reduce emissions from its FCCU

Colin Baillie, Renaud Kieffer and Maria Luisa Sargenti, Grace Davison Refining Technologies Europe

Article Summary

Although the amount of sulphur oxide (SOx) emitted from FCC units (FCCUs) is relatively small, it still accounts for most of the SOx released from oil refineries. The major options for reducing SOx emissions from the FCCU include flue gas scrubbing (FGS), hydrodesulphurisation (HDS) and the use of a SOx-
reduction additive. FGS requires a significant capital investment and the operating costs are high. In addition, discarding the spent treating compounds such as lime, caustic soda or other compounds can be problematic. HDS requires the highest capital investment, but also provides improved yields as well as lower SOx emissions. Therefore, refiners often choose SOx-reduction additives, as they require very little capital investment and are extremely effective in reducing SOx emissions.

Due to new, more stringent, local environmental legislations, Refinery A, located in western Europe, was confronted with the challenge of reducing SOx emissions from its FCC regenerator by 30%. This is highlighted in Table 1, which shows daily and yearly SOx bubble limits for the refinery before and after 1 January 2009.

Faced with the various options for tackling SOx reduction, the refinery chose the additive route and began using Super Desox in December 2008. The results of this trial will be discussed in this article.

Super Desox
Super Desox was commercialised in 2003 and is based on a patented magnesium-alumina spinel structure, which has been successfully used by more than 70 refiners worldwide. To understand why this magnesium-alumina spinel structure works so well, it is necessary to examine the mechanism for the catalytic reduction of SOx emissions in the FCCU. A schematic diagram of the oxidation and reduction reactions of sulphur compounds in the FCCU is shown in Figure 1.

In the regenerator, sulphur contained in the coke is oxidised to SO2 and SO3. Both of these sulphur oxides are absorbed on the SOx additive, where SO2 is further oxidised to SO3 in the presence of oxygen and an oxidation catalyst --in the additive. In the regenerator, the SO3 reacts with magnesium oxide in the alumina spinel and is converted to magnesium sulphate. Compared to free magnesium oxide, the spinel in Super Desox is more effective for this sulphate formation and a stable sulphate is formed under regenerator conditions.

In the reactor, the magnesium sulphate is reduced by hydrogen to form magnesium sulphide. Super Desox plays an important role here too because the magnesium in the alumina spinel is less basic than free magnesium oxide, thereby making the sulphate easier to reduce. In addition, vanadium on the spinel structure further decreases the stability of the magnesium sulphate, especially under reducing conditions and in the presence of steam. The magnesium sulphide is then hydrolysed to a magnesium oxide in the stripper, which returns to the regenerator to complete the catalytic cycle.

Establishing a baseline for uncontrolled SOx emissions
Several operating variables have been identified as having significant effects on the performance of SOx-reduction additives.1-3 Some of these include the presence of combustion promoters, the ratio of catalyst circulation rate to unit catalyst inventory, unit temperatures, availability of oxygen in the regenerator, feed sulphur content and SOx concentration.

Various scientific studies have shown that the fraction of thiophenic sulphur in the feed has a direct impact on the coke sulphur content deposited on spent catalyst and thus on SOx emissions. Therefore, changes in uncontrolled SOx emissions can be accurately estimated by measur-ing thiophenes in the feed. However, it is uncommon for refiners to be able to measure thiophenic feed content on a regular basis. It is therefore recommended to estimate uncontrolled SOx emissions by measuring feed sulphur or even better with slurry sulphur (the slurry being the most aromatic product of the FCCU).

Due to the operating variables that can affect SOx additive performance, it is necessary to establish a baseline of uncontrolled SOx emissions to be able to calculate the true SOx reduction caused by the additive. Such a baseline is calculated by performing correl-ations for every relevant parameter based on pre-trial unit data. To establish a baseline for Refinery A, three months’ unit data were used.

It was observed that, in general, parameters such as slurry density and slurry sulphur content correlated well with SOx emissions. However, due to a lack of slurry measurements, it was decided to use feed properties instead. In this respect, the main parameters turned out to be feed sulphur content (S.charge), vacuum residue throughput (RSV) and flue gas oxygen content, as shown in Figure 2. Using the data from the refinery, the correlation shown in Figure 3 was established using the raw SO2 values from the pre-trial period (from 21 October 2008 to 12 January 2009). The high value of R2 (84.2) signifies the good correlation obtained.

As would be expected, it can be seen that increased feed sulphur plays a major role in the above correlation; SOx emissions increase with increasing feed sulphur. In addition, increased residue through-put can be expected to increase SOx emissions, which is indeed observed in the correlation. The incorporation of more residue results in increased SOx emissions due to the more refractive nature of its sulphur-containing compounds, which will thus end up more readily in the coke on catalyst. The other important variable in the correlation is flue gas oxygen content. As oxygen increases, the coke yield decreases and, therefore, less sulphur reaches the regenerator with coke.