Sour water stripping Part 3: waste water treatment

Mass transfer rate-based simulation is used in a case study to examine two of the important factors that determine the performance of a 2-Stage Chevron WWT (waste water treatment) unit.

Nathan A Hatcher, Clayton E Jones and Ralph H Weiland
Optimized Gas Treating, Inc.

Viewed : 14428

Article Summary

The purpose of such a plant is to produce relatively pure ammonia 
and hydrogen sulphide products rather than 
just combusting the ammonia in the 
Claus furnace. It is argued that the performance of the whole plant is controlled almost entirely by the operation of the hydrogen sulphide stripper.

Sour water is generally classified as either phenolic or non-phenolic. Non-phenolic water, also called HDS water because it is produced by hydro-treating in hydrodesulphurisation or HDS units in refineries, contains almost exclusively ammonia, hydrogen sulphide, and possibly a trace of carbon dioxide. Part 1 of this series1 addressed sources of non-phenolic sour water, sour water chemistry, phase equilibrium in sour water systems, and the removal of contaminants in sour water strippers (SWS). Phenolic (or more broadly, non-HDS) water typically contains heat stable salts (HSSs) and HCN, although phenols and caustic may also be present, depending on how the water has been previously used in the refinery. In Part 2, the stripping of phenolic water is discussed2. In the present article, attention is turned to the WWT (waste water treatment) technology originally developed by Chevron and recently acquired by Bechtel. This is generally referenced in the literature as the Chevron WWT process.

WWT technology is a two-stage sour water stripping process whose objective is to separate the hydrogen sulphide and ammonia components of sour water into two separate streams, each relatively free from the other component. Historically there were problems with the handling of ammonia in a SRU, although today these problems can be overcome by using a high enough temperature and a long enough residence time in the SRU furnace to completely destroy ammonia. Nevertheless, when high-nitrogen, high sulphur (heavy) crudes are processed, the amounts of ammonia produced in a refinery can be high enough to make it worth considering ammonia as a saleable product rather than just routing it to the SRU furnace. Removing ammonia from the Claus plant, unloads the Claus plant hydraulically, reduces front-end air requirement, and allows for a thermodynamically higher sulphur recovery.

Sour Water Treating Using WWT — a Case Study

In its simplest form, the Chevron WWT process follows the flowsheet shown in Figure 1. It is the configuration used in this case study. There are numerous embellishments that can be made to this basic flowsheet that will marginally improve the performance of the overall unit. However, for the purpose of this article they would likely obscure the message. Table 1 shows the sour water composition from which dissolved hydrocarbons, phenol, and heat stable salts have been omitted to simplify the study—they do not add materially to the analysis.

Sour water is preheated against hot stripped water in the cross exchanger and enters the H2S stripper onto the top tray of the 40 valve tray stripping section at 190°F. A small (variable) flow of cooled stripped water is fed to the top of a 10-ft bed of random packing where it serves to knock down ammonia from the rising vapour to prevent its escape into the H2S product. The H2S stripper operates at 60 psig. Note that, as is conventional in this process, there is no overhead condenser and the H2S stripper is not internally refluxed. Internally refluxing the stripper would provide insufficient reflux to wash enough ammonia from the hydrogen sulphide. Providing enough reflux by increasing the boilup rate would force even more ammonia into the H2S overhead product, further lowering its purity. Internal reflux is not an option.

The bottoms from the H2S stripper passes directly to the sixth tray from the top of a 35-tray ammonia stripper operating at 18 psig. The NH3 stripper is supplied with enough reboiler steam to produce final stripped water containing <35 ppmw ammonia and <1 ppmw hydrogen sulphide. Thus, the ammonia stripper behaves as a conventional SWS except that most (but not all) the hydrogen sulphide has already been removed.

The fact that the ammonia stripper is nothing more than a normal SWS has an interesting consequence. If the stripped water is to meet reasonable ppm specifications on the residual levels of strippable contaminants, then whatever volatile components (especially H2S) remain in the bottoms from the H2S stripper necessarily appear overhead with the ammonia product from the ammonia stripper. Thus, as far as product purities are concerned, the focus must be entirely on the H2S stripper because its performance alone controls the purity of both the hydrogen sulphide and ammonia products. The ammonia stripper merely serves to produce cleanly stripper water.

In this case study, varying flows of wash water (Stream 20) and varying steam flow rates to the H2S stripper reboiler are examined. Numerous other parameters such as reflux water temperature, tray counts, packing height and size, and sour water composition could have been varied in this study as well, but in the interest of space they were not.

The WWT unit was simulated using the ProTreat mass transfer rate-based simulator. Performance can be discussed in two ways, depending on one’s perspective:
(a) Fractional recovery of sour water feed components in the hydrogen sulphide and ammonia product streams, and
(b) Composition of these product streams on a dry basis.

Compositions of product streams are considered on a dry basis so that differing amounts of water do not obscure the results. As Figure 1 shows, a fraction of the stripped water is recycled back to the top of the H2S stripper as external reflux. The observations and results to be described pertain to the particular conditions of this case study. By and large, they should not be extrapolated to other situations and other sour water feeds.

Recovery of Hydrogen Sulphide and Ammonia in Product Streams
Figures 2(a) and (b) show the respective percentages of the hydrogen sulphide and ammonia in the original sour water that end up in the hydrogen sulphide product stream as a function of H2S stripper reboiler steam flow. The parameter in these plots is the percentage of the stripped water that is used as external reflux in the H2S stripper. There are several worthwhile observations. Firstly, regardless of how high the steam flow is pushed, and no matter what reasonable external reflux flow is used, it appears to be impossible to force more than about 93% of the feed H2S to report to the H2S product stream.

Second, it is the highest steam flows that force the most hydrogen sulphide into the H2S product stream, but these flows also force the most ammonia into that stream. Minimising the ammonia content is done by using lower reboiler steam rates, not higher. Reboiler steam rate has opposite effects as far as H2S recovery and H2S product stream purity are concerned. Third, at first glance, the curves corresponding to 3% recycle rate as external reflux appear to be outliers; however, there is a perfectly good explanation for the seemingly odd behaviour. At recycle flows between 1.5% and 2.5% of the stripped water, the temperature throughout the stripping section is uniformly high and nearly equal to the reboiler temperature. But when the recycle flow reaches 3%, the cold external reflux flow rate is high enough to collapse the temperature throughout most of the stripping section, forcing it to be much closer to the sour water feed temperature. The result of this behaviour is reflected in Figure 3 where it can be seen that corresponding to each level of external reflux there is a reboiler steam rate above which the water content of the H2S product starts to rise rapidly. This also results in much reduced performance vis à vis ammonia contamination of the H2S product stream. It is nevertheless possible to produce a hydrogen sulphide stream containing a very substantial proportion of the original hydrogen sulphide with very little ammonia. For example, with a 2.5% recycle flow and 15,000 lb/h of steam, 92% of the H2S can be recovered with only 0.05% of the original ammonia (100 ppmv, dry basis).

Add your rating:

Current Rating: 4

Your rate: