CCRU Recycle Gas Compressor Steam Quality 1

[recklesskid]

Hi, I'm a new process engineer in a refining industry and I'm currently handling the CCRU.

A few months ago, we experienced deterioration on the performance of our recycle gas compressor (600# steam turbine-driven). We observed continuous decrease in recycle gas flow and we also experienced difficulty in meeting the set speed of the compressor. We even opened the hand valve (provision for additional two nozzles for the steam turbine) of the of the compressor to help increase the recycle gas flow, and it increased. But the deterioration still persisted.

The convection section of our CCRU normally produces 26 MT/hr of steam and these are diverted to the steam reboilers of the debutanizer and reformate splitter of the unit (sums to 23 MT/hr consumption). THe recycle gas compressor consumes about 10 MT/hr of steam (the balance of the CCRU heaters plus the steam from the HP steam header [from utilities]).

The deterioration has been experienced when the generation has increased to about 29-30 MT/hr. We had the steam from the CCR heaters tested by a third party and it showed high sodium (2000 - 8000 ppb) and silica content (500 - 5000 ppb) and conductivity of 29 - 60 uS/cm. From this and the observed decrease steam consumption, we identified that the problem is deposition of steam impurities to the turbine nozzles and/or blades. Question: will there be other possible causes of the problem?

With this, we maintained the steam generation at 27 MT/hr max and we noticed gradual increase in the recycle gas flow and we noticed that the set speed can now be met like before the deterioration has experienced. Deterioration started last July and it continued until August. The RG flow has already decreased for about (12,000 sm3/hr) during the period. The highest set speed that the compressor can meet was only 9300 rpm. Normally, it can achieve to about 10300 rpm. Starting September when we closely monitored the steam generation and maintained it below 27 MT/hr, we have observed improvement in the RG flow and compressor performance. Currently, our RG flow has gotten back to its normal RG flow and we can already achieve 10,000 rpm of compressor speed. Question: Is it possible that the if the nozzles have been plugged, it has been cleaned through time? How is this possible? We have been locally testing steam quality from CCRU at generation <= 27 MT/hr and the conductivity has been <= 15 uS/cm and silica content <= 0.8 ppm.

For our analysis, since we cannot easily replace our units steam disengaging drum, which (based on our evaluation) is undersized, we would divert the steam of CCRU to other units (which are using steam reboilers) . With this, I also want to know if the steam quality has significant effect on the performance of reboilers. If we cannot prevent steam generation of 29-30 MT/hr, this steam would still be used in steam reboilers. Would this also deteriorate the performance of the equipment?

your inputs would be highly appreciated. Thanks

 

[JJPellin]

My background is rotating machinery, not process. So, I am going to comment on issues more closely related to the turbine. Hopefully, someone else can help with the process issues.

We have seen a similar problem with the drive turbine on our FCC main air blower in the past. I am wondering if you have considered performing an on-line water wash of the turbine to restore performance while you work out the process issues needed to keep it from happening again? We were able to set up a condition based water wash of our big turbine and restore the performance completely with the unit on-line.

This was done with the turbine at minimum governor which obviously required a unit rate cut. After it was over we were able to confirm that the turbine performance came back up to previous levels. The cost of the rate cut as compared with the cost of a unit shutdown and full turbine overhaul was considerable.

If you wanted to consider an on-line wash, you should work closely with the turbine manufacturer to be sure that your method is compatible with their design.

At subsequent overhauls we were able to verify that the build-up in our turbine was occurring in the blades of the first two or three stages. It was not building up in the nozzles. If you have already seen some recovery, it is understandable that some of this build-up has eroded away on its own. If you look closely at the vibration data, you may be able to confirm a slight change in run speed vibration phase angle if not amplitude to prove that the build-up was rotating.


Johnny Pellin