Oct-2010

Utilising the checkerwall as a 
static mixer to improve SRU 
process performance

A little more than a decade ago, as Blasch Precision Ceramics began to steadily grow their hex head ferrule business, they began to field requests from some of their SRU-operating customers to develop a checkerwall that would be easy to install, mechanically stable under all conditions, allow for design with varying degrees of open area, and could incorporate the use of an integral manway, if desired, Fig. 1.

J J Bolebruch, E l Collins and N Teator
Blasch Precision Ceramics Inc

Article Summary

At that time, concerns were strictly mechanical, and there were no requests for mixing or flow management. The goal for many operators was to put in a checkerwall that would survive a campaign intact.

The result was the Blasch HexWall checkerwall. The initial checkerwall blocks were designed to be 9” wide, the same as a row of bricks in the hotface. That way, the wall could be erected into a slot in the existing lining, thus further serving to hold it in place. The blocks are mechanically engaged through a series of tabs and slots, so the wall is quite stable, even when 3-4 metres in diametre. The openings are round, so that any portion of the block not directly supported by other blocks is in the form of an arch, and, as such, has greater mechanical stability than a flat span. As larger furnaces were encountered, additional designs utilizing longer blocks were developed so as to retain a reasonable aspect ratio of wall width to height, Fig 2.

Over the next few years, as the stability and durability of the Blasch-designed checkerwalls was proven out, some operators, who previously spent their time rebuilding walls, began to question what they were actually designed to do. After considerable consultation with process licensors, it was determined that there were a variety of ideas as to what made each licensor’s process more effective, but it also became apparent that most everyone agreed that time, temperature, and turbulence were the variables that played the largest roles on the process constituents. The difference was in how each licensor chose to influence these factors including the use and design of checkerwalls and baffle walls in the reaction chamber.

Initial mixing efforts
Initial efforts centred on process flow within the checkerwall feature, and, based on observations with spray nozzles, focused on a variety of inserts that were designed to introduce turbulence into flow through the block itself, Fig. 3. Like spray nozzles, they did an excellent job of mixing the gas flow within the confines of that individual block, but at a cost of a high pressure drop, and still worse, they did little to improve conditions within the furnace as a whole, as the hundreds of narrow, now thoroughly mixed streams did not seem communicate with one another, and as a result, this did not significantly improve overall mixing, reduce separation within the furnace, or improve homogeneity if there were multiple injection points or feedstocks.

A different approach would be needed, and a return trip to the drawing board was indicated.

Wider focus
Improvement in mixing of a gas stream where large scale non-homogeneities are distributed across the entire cross section of the reaction chamber requires the use of the full reaction chamber. It follows, then, that if it is possible to redirect the flow downstream of the checkerwall, it will allow the full use of the reaction chamber volume in the improvement of mixing and overall homogeneity. By stepping away from the narrow focus within the checkerwall block itself and looking at how to influence the flow downstream it should be possible to engineer a variety of flow fields to influence different mixing and reaction shortcomings of current technology.

Initial vortex work
The first flow field examined was a vortex, Fig 4. The mixing properties and commercial uses of vortex mixing are well documented. One such example is the use in fuel injectors for high performance engines. A simple, publicly available definition from Wikipedia states, “A vortex (plural: vortices) is a spinning, often turbulent, flow of fluid... The speed and rate of rotation of the fluid are greatest at the centre, and decrease progressively with distance from the centre. The fluid pressure in a vortex is lowest in the centre where the speed is greatest, and rises progressively with distance from the centre”.

Modelling was performed to determine what the flow out of each individual block should look like in order to contribute to the overall desired flow. As a result of this, a relatively simple vectoring hood called a vector tile was developed that would fit into the outlet ends of the existing blocks, and could be oriented and secured in the field in such a way as to redirect, or “vector”, the downstream flow allowing each individual block assembly to contribute to the creation of the overall desired flow, Fig 5. This variation of the HexWall checkerwall was christened the VectorWall.

This design was modelled via a commercially available computational fluid dynamics software program using process details from a customer inquiry at the time, Fig 6.

The CFD results seemed to indicate that the desired flow field was developing, and as a result, tooling was fabricated that allowed for the manufacture of the vector tiles in the same material as the checkerwall via the Blasch process, and existing block tooling was modified to make blocks that could accept them with a smooth flow path transition.

One of Blasch’s hex ferrule customers that had expressed dissatisfaction with the process efficiency of their reaction furnace was identified as a prospective VectorWall trial site. A system was installed in the spring of 2009 and monitoring began, Fig 7.

Feedback from the field
It was immediately apparent in this first trial that there was a substantial improvement in the process dynamics as a result of the induced vortex flow. Product yield increased, excess oxygen required was cut by 50%, and the number of process upsets related to equipment fouling decreased to near zero. This customer is exceptionally pleased with the performance achieved and has adopted this as a company-wide best practice and begun to field these walls in other plants.

Additional VectorWall mixing checkerwall units were fielded by Blasch at locations unrelated to this site, and within months it became clear that the advantages extended to them as well. The positive field results encouraged wider thinking regarding the VectorWall’s ability to affect mixing and turbulence. How could other configurations influence the process flow within the reaction chamber and what would it mean to process effectiveness.