Apr-2014

Reliability in recovery

Reliability is a buzz word used a lot these days. You hear it a lot and it means different things to different people based on their job description.

Matthew Allen, Total Petrochemical and Refining USA
Jeffrey Bolebruch, Blasch Ceramics, USA

Article Summary

I think today one thing most refiners could agree on is that improving and maintaining the mechanical reliability of the Sulfur Recovery Unit (SRU) is critical. When one is down, depending on treating capacity, you might be backing down unit charge rates based on your plants SRU capacity. So while the SRU might not make any money it can sure lose a lot when things go wrong or unexpected delays are experienced in an outage or during startup.

In the mid-1990s Total’s Port Arthur facility was approached by Blasch to look at their hex head ferrule design that they had just implemented at a competitor’s facility across town. It appeared to offer a number of advantages, namely getting rid of castable refractory that was always full of cracks.

The SRUs were on the same outage schedule as the hydrotreaters so every cat change and then a major turnaround every 5-6 years. Every time the SRU was down for a planned event, inspection took a quick look at the shape of the refractory and ferrules. This was usually followed by a discussion to determine which cracks could be tolerated and which needed repair. It seemed there was always some level of repair that took place.

Inspection
The Claus type units went online in 1984 and 1993. They had the standard round ceramic sleeves, or ferrules, which were inserted into the tube ends, and packed with a high temperature castable refractory material, to a depth that would drive the temperature down to an acceptable level, given the thermal conductivity of the material used, typically 4in.

For many years this worked well to protect the tube ends and the tubesheet from the hot corrosive gas path. The drawbacks were in the quality of the installation, proper startup (refractory dryout) and the nature of the service. Integrity of the castable installation is affected by the skill of the installer and proper mixing of the castable, too wet or dry will cause issues so having a good inspection follow up is essential.

A large amount of new castable makes proper heat up critical. If the castable heats up too quickly the trapped water vaporises to steam at a rate that causes failure in the refractory. This can include anything from large cracks to spalling entire areas off leaving the tubesheet unprotected. Therefore, unit startup must follow a heat up schedule (typically provided by the castable vendor in terms of ËšF/hr.) and one concern is that the main burner inputs too much heat and is therefore uncontrollable. Most of the time, as long as everything goes well this does not pose an issue but mistakes can happen. In some cases to reduce risk a third party is often used to cure the refractory using a smaller burner delaying startup.

Another problem that most are familiar with is cracking in castable between the ferrules. Cracking was a concern due to a through thickness crack that allows hot gasses to bypass to the tubesheet hot face resulting in corrosion (since the tubesheets are normally carbon steel). The refractory is expanding and contracting as it heats up and cools down. Large fluctuations in temperature are problematic because they create gradients in the body where differences in the rate of expansion or contraction create enough stress to lead to the aforementioned crack formation. The larger the body, or the more rapid the thermal cycle, the greater the stress, and the more damage is done during that cycle.
Cracking and spalling aside another issue is tubesheet inspection and mechanical cleaning. Either of these requires the ferrules to be removed. Some may be better at it than others, but all of the ferrules always ended up being replaced because removing the refractory more often than not cracked the ferrules.

Also, the fact that the design of the Blasch ferrule offered advantages in a lower dP was considered. The standard round ferrule design has a ‘high’ dP (tubesheet dP is not the limiting factor in a design) associated with it. While not quite sharp edge (in terms of K factor) it does impose some amount of pressure drop. This is typically taken into account during design but due to the fact these units typically run at a low pressure (below 15 psig) any pressure increase on the front end imposes some limit to unit throughput. Any reduction in dP, however small, is directionally good for throughput.

New design solutions
When Blasch design was examined, the following improvements over original design were observed:
•    Minimised use of castable so reduced risk of castable failure at startup (degassing).
•    Easy to remove/reinstall for inspection and cleaning (so they are reusable).
•    Lower dP than traditional design.

Based on these findings, the decision was made to try the Hex head ferrules out in 2001. Since the tubesheet was square pitch the ferrules are actually square, but the concept is the same. One of the problems encountered in the field was that the tubesheet was not drilled by a modern computer numerical control (CNC) machine so we were in some cases up to 0.25in off on the centrelines. Ultimately, a small number of round ferrules had to be installed at the top and bottom because the new square head ferrules were not properly centred in the tubes. So if one is retrofitting an older design it is best to take a dimension check (perhaps during a cat change outage) before the order is placed. Another thing that needs to be taken into account is removing any refractory anchors one might have.

In Figure 3 it can be seen where round ferrules had to be utilised in some locations due to the relaxed tolerances on tubesheet drilling. Also, one can observe that for a new design one must ensure that enough room exists at the periphery for castable. This means that the barrel diameter must be checked and one might have to add a little to the OD. It was a little tight, in this example.

Installing the ferrules is very easy and the installation goes very fast. It is prudent to order extra in case of breakage though, 5% or so is safe. One should make sure to band the outer circumference beforehand packing in castable, otherwise ferrules could vibrate out on the run.

Positive results
Experience since then has been positive and tubesheet inspections to date have not revealed any tubesheet corrosion (which would be expected if bypassing hot gasses past the ferrule heads). One or two of the ferrules that were at the outer edges and were in too tight cracked just behind the head but this is due to the tubesheet drilling and not the ferrule design. In the end the major benefit is being able to quickly perform a tubesheet inspection.