Mar-2007

Key factors in selecting refractories for the hydrocarbon processing industry

Longer campaigns, higher temperatures, and more oxygen all add up to increased wear and tear on process vessels and equipment and make proper material selection even more critical.

Jeffrey J Bolebruch
Blasch Precision Ceramics, Inc

Article Summary

Unanticipated failure in refractory lined vessels or failure of critical pieces of equipment, whatever the material of construction can have expensive and tragic consequences.

The same refractory lining designs that have performed admirably in service for years may not be suited to the new realities of the process industry. Higher temperatures can require higher duty compositions, and at the very least, exacerbate the differences in thermal expansion between refractory and steel as the insulated vessel walls remain the same temperature, while the hot face of the refractory can be seeing considerably higher temperatures than it was prior. Also, processes utilising oxygen enrichment can operate more unstably, causing large swings in temperature, which in turn stresses the refractory as it is forced to expand and contract repeatedly.

This highlights the point that refractory linings are dynamic bodies, expanding and contracting with swings in operating temperature, and, as such, should be treated with careful consideration. Large refractory bodies will thermal fatigue and crack over time, as the number of thermal cycles they see increases.

This type of fatigue can be mitigated through material selection – choosing refractory compositions that have lower coefficients of thermal expansion, or by physical means such as segmenting linings in order to discourage the transmission of stresses across entire bodies.

Most refractory linings in process vessels are composed of a single or double layer of castable refractory. These linings are typically secured to the vessel wall via a series of metallic hangers that are welded to the shell, forming fields of metal fingers protruding from the wall in a variety of shapes and lengths. It is these lengths of steel that serve to “knit” together the castable, providing a physical means of holding it together and keeping it attached to the steel shell of the process vessel.

Unfortunately, metal has a much higher coefficient of thermal expansion than ceramic, and in particular, in applications where temperatures are very high and the hangers come close to the surface, there is potential for the hangers to actually force the refractory to crack as they expand at nearly twice the rate of the refractory for a given operating temperature. In addition to expansion from a purely temperature related mechanism, many process vessels contain corrosive or oxidising atmospheres that can penetrate far enough into refractory linings to attack the metallic hangers, cause swelling and further exacerbating the cracking problem.

Why then are monolithic refractories used so extensively? They are relatively inexpensive and easy to install for experienced refractory contractors. They also adapt to nearly any shape or size without a requirement for tooling up front. The process industry is in business to make money, and to expect them to spend any more money on refractory materials than is absolutely necessary is just not realistic. 

What then is the purpose of this article? Simple. There are a number of processes that are pushing the envelope to the point where the cost of unexpected downtime and damaged equipment can and do greatly outweigh the incremental up front cost of an engineered lining.

Take for example, the Claus sulphur recovery process. In the Claus process, hydrogen sulphide (H2S) is combusted in a refractory lined thermal reactor at temperatures in the low to mid two thousand degrees Fahrenheit. H2S is extremely corrosive in and of itself, but increasingly, oxygen enrichment is being used, which forces the temperatures up to, and sometimes above, 3,000°F. Further, oxygen enrichment tends to lead to greater swings in temperature as air, with its large percentage of inert nitrogen is replaced with much more combustible oxygen and smaller changes to process parameters have greater impact on the business end.

Consequently, there have been a number of cases over the last few years, particularly in Claus units with a combination or large diameter tubesheets and high operating temperatures/oxygen enrichment, which have had failures of the tubesheet refractory, in some cases, leading to the destruction of the entire waste heat boiler. In almost every case, the mode of failure was cracking of the monolithic refractory, caused either by repeated thermal cycles or corrosion of the hangers buried in the refractory, and the subsequent shearing off of the tube ferrules that protect the waste heat boiler tube ends, leading to significant corrosion of large areas of the tubesheet and the tube-to-tubesheet welds, and immediate failure.

There have been enough installations of precast hex head ferrules over the last few years to show, that when properly designed, a segmented lining can not only safely protect a tubesheet in these difficult environments, it can also extend the length of the campaign.

There are a couple of reasons for this. A segmented lining, by definition, is comprised of a series of smaller precast shapes, which are designed to expand at operating temperature and form what is in effect a monolithic face, but also to contract upon cool down, returning to their status as individual components, thus preventing the accumulation and transmission of stresses across the larger body. There is a good reason why sidewalks have expansion joints, and the temperature swings there amount to maybe a hundred degrees!

Additionally, the use of a precast, pre-fired lining, allows for the incorporation of refractory fibre insulation, which cannot tolerate the moisture present when monolithic refractories are applied. A precast liner, designed as part of a two piece system including a fibre component, will have much greater insulating value than a two piece monolithic castable lining consisting of a lower density insulating castable and a high duty castable on the face. 

This permits the design of thinner, more dynamic linings, with much less mass than cast in place systems. Less refractory mass means less energy used to heat the lining and more going directly to the process.

Inspection of vessel walls is simplified, as parts or all of a segmented system may be removed and re-installed without having to jackhammer an entire lining out and re-cast it, allowing days for subsequent cure-out. There are other applications where this approach has shown merit.