Fouling of the heater tubes determined the timing of the shutdown. The heater tubes themselves were cleaned by the operators, using a process called steam-air decoking. The coke was burnt off in a controlled manner by the use of air for combustion and steam to regulate its speed. Steam also
Figure 37.4 View of Soaker Overhead Line
helped to keep the tubes from getting overheated and to blow away the unburnt coke pieces. The decoking process was fairly fast; it took about three days to decoke both the heater cells.
There were two 70' long, 10" outlet pipes, one from each cell, flanged at the heater end and on the 16" soaker inlet pipe, as shown in Figure 37.2. The location of the flanges at the Soaker end made it difficult to clean the horizontal section of these pipes. Similarly, the location of the flange on the 16" pipe leading to the Soaker made that section unwieldy to handle.
The planner and I were confident that the shutdown duration could be reduced significantly. We analyzed the main activities described above and formulated some ideas, as follows:
1. Soaker—Use of HP steam would improve the safety of the cleaning activities significantly. A longer steam-out period would speed up the cleaning process and help reduce overall duration.
There was no simple solution to minimize the access time for scaffolding inside the skirt of the Soaker. A few months later, the maintenance supervisor came up with an interesting idea. While this would not solve the access time problem, it was still worth pursuing; more about this later.
2. Fractionating Column—In another part of the refinery, nitrogen cooling was used in the reactors of the Hydro-Cracker Unit. Nitrogen was supplied from a portable liquid nitrogen plant. These units liquefy air and fractionate it to get liquid nitrogen. The contractor who supplied the unit confirmed it was possible to supply cold air, using the same unit. We decided to hire the unit for the next TCU shutdown to cool the column quickly, and to provide a comfortable working environment for the people. With ambient temperatures in the Middle East over 100°F for most of the year, this would be a welcome change. The concept was a somewhat revolutionary welfare measure, and nobody had so far considered air conditioning a column. I was thus at the receiving end of jokes.
3. Pump piping—We discussed the problem relating to the cleaning of the discharge piping of the residue pumps with the process technologist. One possibility was to introduce a 2" nozzle at the tee junction, as illustrated in Figure 37.5. We evaluated the risks and initiated a Plant Change.
4. Fractionating Column strainer—It was evident that the coke build-up in the strainer was contributing to the pump failures. The loss of pump capacity was so severe that it was not feasible to operate the unit beyond six months. The process technologist and I came to the conclusion that the strainer design had to be modified to overcome this problem. He designed a strainer with a larger number of holes, so that the pumps had sufficient net positive suction head (NPSH). This was another Plant Change.
Figure 37.5 Cleaning Access Nozzle
5. Soaker overhead pipe—The unit manager and I discussed the logic of lowering the 18" Soaker-to-Column transfer line for cleaning. After seeing the evidence, he agreed that this line need not be cleaned regularly in future. This item was deleted from the work list, clipping at least eight hours from the duration.
6. Heater outlet lines—Cleaning of the heater outlet lines would be much simpler if we could relocate two pairs of flanges. Similarly, relocating one pair of 16" Soaker inlet pipe flanges would speed up access to the Soaker. The process technologist agreed to raise a Plant Change, illustrated in Figure 37.6.