A summary of the 2018 Hydrocarbon Engineering article on ceramics at both ends of the refinery.
Hydrogen production at one end, sulfur recovery at the other.
Two furnaces where the refractory decides the uptime.
Engineered ceramics raise performance and reliability in the steam methane reformer and the sulfur recovery unit by replacing brick structures that crack, collapse or mix poorly. In the reformer that means mortar-free flue gas tunnels that grow with the furnace. In the Claus reaction furnace it means a checkerwall that stays standing and steers the flow, which in one refinery revamp raised the zone 1 temperature by about 100 °C and ended furnace vibration.
| Publication | Hydrocarbon Engineering, August 2018 |
| Title | Boosting performance and reliability |
| Author | Uday Parekh, Blasch Precision Ceramics |
| Topic | Ceramic systems for hydrogen production and sulfur recovery in refineries |
Refineries need more hydrogen as crude gets heavier and sulfur limits tighter, and most of it comes from steam methane reforming. The hydrotreaters that use the hydrogen produce sour gas, which ends up as hydrogen sulfide in the Claus plant. The author calls the two units the bookends of a refinery: one makes a feedstock, the other turns a waste stream into sulfur.
Down-fired reformers collect their flue gas in tunnels on the furnace floor. Brick tunnels fail after repeated thermal cycles: the lids grow more than the base and the cross section turns trapezoidal, or the whole tunnel grows longer than its joints allow and snakes. The article describes the mortar-free alternative with interlocking base, wall, tie rod and lid components, usable in a revamp, for lids alone or in a new furnace.
A checkerwall that only survives the campaign is half a solution: the same blocks can decide how long the gas stays hot enough to destroy ammonia.
About two thirds of the sulfur conversion happens in the thermal stage, and the reaction furnace also has to destroy hydrocarbons, ammonia and BTEX. It needs residence time, temperature and turbulence, the three Ts of combustion.
Brick checkerwalls, cylindrical or matrix style with mortar, fail through thermal cycling, mortar failure and vibration. A choke ring reflects part of the flow back, but much of it jets through the center and leaves before the minimum residence time is reached.
The article traces the development from the HexWall checkerwall, with dry stacked blocks 9 to 18 inches deep held by tabs and slots, to the VectorWall, where hoods at the block outlets direct the flow into a long path with a narrow residence time distribution. By 2018 more than 100 of these systems were installed, including in SRUs above 1,000 tpd.
Field result: a refinery with ammonia destruction problems replaced its checkerwall with a VectorWall. Zone 1 ran about 100 °C hotter, sour water stripper gas feed rose by about 30%, and the furnace stopped vibrating.
CFD result: Porter McGuffie Inc. modeled an Asian refinery SRU before and after a choke ring was replaced. With the choke ring, 22.5% of the flow missed the residence time needed for ammonia destruction. With the VectorWall, all zone 1 flow met it, and the pressure drop was about 900 Pa lower.
Energy: for a 500 tpd SRU with lean feed and fuel gas at 3 USD per million Btu, the author estimates savings above 250,000 USD a year from less co-firing for BTEX destruction at about 1,050 to 1,100 °C.
If your reformer and your SRU are both on the turnaround list, the same question applies to each: which refractory structure fails or limits the unit, and can a dry stacked, engineered part replace it.
For the reformer, see the StaBlox flue gas tunnel system. For the Claus furnace, see the VectorWall mixing checkerwall and the case where VectorWall ended furnace vibration. Both units are covered on the refining and chemical processing page.
A choke ring is an open ring that reflects part of the flow. A checkerwall is a full wall of blocks with openings. The article shows that a choke ring lets a central jet pass with too little residence time.
Ammonia and BTEX need high temperature for long enough to be destroyed. If zone 1 runs too cool, ammonium salts deposit downstream and can force a shutdown.
No. The CFD study cited in the article found roughly 900 Pa less pressure drop across the VectorWall than across the old choke ring.
Yes. The article names furnace revamps, replacement of the lids alone and new furnaces.
Describe what fails or limits the unit today, and a Blasch engineer will tell you which part addresses it.
