Mar-2011

Chloride removal in refineries

A review of catalytic removal of chlorides from refinery streams and a critique of current analytical techniques for estimating chloride content

Mani Erfan
Unicat Catalyst Technologies

Article Summary

In recent years, refineries have realised that there is a significant and growing need for the improved removal of chlorinated species from refinery streams. There is a general rise in awareness of the need for a more comprehensive approach to the challenges caused by untreated HCl or organic chlorides within the refinery system.

These challenges are wide ranging and include almost all refining operations. From refineries that are now recognising the need for dechlorination reactors for first-time installation to operations with multiple existing reactors that face operational challenges, there are few that do not have a need for significant improvement in this area of operation.

Most refinery operations lacked a sophisticated and thorough method to deal with chloride removal. If there were chloride removal units in operation, they would usually deal with less critical applications, with no great attention paid to them except at the time of absorbent replacement. In many cases, it was observed that unit engineers did not really track EOR conditions for the reactors and would run the units blind for months at a time.

Evaluation methods for monitoring performance were limited and hinged on occasional Draeger tube testing for a HCl component. Outside laboratory analysis was usually never utilised and refinery laboratory operations lacked a proper methodology for in-house analysis. Proper analysis and performance evaluation remain key areas in need of improvement and standardisation.

It was also observed that many chloride removal operations actually suffered from operational issues such as a short lifecycle, green oil formation, pressure drop increase, and continued downstream corrosion and catalyst deactivation. Yet, the catalyst and absorbent solutions employed remained unchanged and did not meet daily operational challenges.

This article reviews the latest developments in catalyst and absorbent technology to better meet the growing requirement for chloride removal, to discuss common operational challenges — their causes and possible solutions — and to discuss current analytical methodologies to evaluate on-stream performance. Its aims are to assist refinery operations managers to better manage and optimise their current or future operations; to understand causes and employ solutions to current problems and/or avoid future problems; and to establish a framework within which the refinery is able to monitor a unit’s performance.

Performance evaluation and analytical methodology
After reviewing with many refineries in the US their analytical and measuring capabilities for chlorides, the author concluded that almost all refineries did not have adequate equipment and standardised procedures. The author found the normal response to monitoring capabilities was:
• Draeger tube for HCl measurement of gas and or liquid systems
• No measurement method or procedure for measuring and identifying organic chlorides
• No use of outside laboratories that could perform such analytical work.

Overall, refineries do not have the ability to evaluate adequately the performance of their chloride guard beds.
Unicat also researched outside laboratories and equipment manufacturers. In general, the scope of this research was limited to North American. We found a surprisingly limited number of laboratories that claimed the ability to analyse for organic and inorganic halogens in hydrocarbon media.

Ten located laboratories could analyse for organic chlorides (RCl) and HCl in liquid streams to ppm levels, but levels of accuracy were limited to >100 ppm. Only four of these laboratories claimed the ability to analyse RCl and HCl in liquid streams to low ppm levels (less than 10 ppm). Only two of the remaining four could analyse RCl in vapour-phase systems. Only one laboratory claimed the ability to speciate the RCl content in liquid streams to low ppm levels.

While the research did not expand extensively into other regions, it is our understanding that the availability of such laboratories is also very limited across the globe. We also found that there are more laboratories involved in speciation and analytical identification of sulphur species in hydrocarbon streams in North America, rather than halogen species, and the general consensus is that the identification and speciation of organic chlorides is not the highest priority in independent laboratories. The methodologies for identification of sulphur species are more advanced and standardised than for halogen species. Among those that could detect organic and inorganic halogens, none of the three laboratories followed the same analytical methods and procedures.

While there were proven methods for sulphur speciation in the laboratories mentioned, with sensitivity down to 10 ppb, the lowest detection limit for organic chloride species in the one laboratory that could speciate them is 100–500 ppb levels (according to species). A mixture of standardised and in-house methods was used in the laboratories. The standardised methods were EPA 9076 and ASTM D808/D-4327. General feedback from the laboratories was that an effective method for detecting and quantifying chlorides in gas or liquids does not exist at this time.

Challenges for laboratory analysis
• No ASTM or standardised method is available for the detection of organic halogens in gas samples. Analysis is usually by measurement of total chlorides and deducting the inorganic portion from the total value
• Most of the laboratories use flame ionisation chromatography to detect organic chlorides and potentiometric analysis for inorganic chlorides. Such analysis suffers from potential interference from other halogen compounds, such as bromides and fluorides, and from sulphur species. Their presence can give a false higher measurement, since the methods are designed to detect halides in general. Typical values from such analysis are consistently higher than expected in gas analysis
• Using the ASTM method with an EDC detector allows for the identification of other halides and sulphates in liquid streams. This, in turn, provides a clearer picture of the net chloride content in a liquid stream. This is not possible in vapour-phase samples, so accuracy levels for total chloride in gas analysis (organic plus inorganic) are not high