March 12, 2008
This post onwards, my postings are not in actual order of occurence. The next job I did was the problem solving of higher loop pressure in ammonia plant synthesis loop. The loop pressure was running by 10% higher value i.e. ~220 Ata compared to 200 Ata in other similar size, identical, same technology plants of same vintage and company was unable to identify the reason for last 12 years as it was a problem since commissioning.
When I evaluated this system I used the ammonia sysnthesis converter simulation program developed in C++. The major first difference identified was that the overall conversion of reactants into ammonia was higher than other companies (We compared our data with identical plants). The reason of this difference was identified that the second bed in the converter was giving mor econversion & first bed was giving less compared to others.
Ammonia converter S-200 radial flow type Haldor Topsoe make has two beds with one interbed exchanger. Feed gas first goes to the shell annular space rises up providing cooling to the sheel & then enters to interbed exchanger & is preheated to the desired temp range. Then it goes to first bed radially exits from it at higher temp & goes to interbed exchanger shell side for cooling. Relatively cold gas enters to second bed & exits from the bottom.
Now on identification of above reason of higher overall conversion, less conversion in first bed, higher conversion in second bed was sufficient for identifying the problem area.
On simulation again with the actual conditions, we found that some amount of additional gas is required to enter to second bed directly wihtout passing thru first bed. At this point came the idea of some internal bypass or leakge in the exchanger first which was ruled out as it will result opposite to the actual phenomenon. It will lead to higher conversion in first bed.
So finally upon detailed analysis of reactor internals drawing for few days, the only possible way of gas flow to second bed was identified thry bottom support plate sealing.
Based on this it was decided to open the reactor in next turnaround but it was openend earlier due to basket failure which was also an unique phenomenon in this particular reactor. Such failures are not common but it was happening in my company for 4-5 th time.
Surprisingly it was found on inspection of plate that tightening bolts were loose & they were only hand fitted & probably engineer missed the step. This expalined both the reasons of high loop pressure & repeated failure of converter internal pipe. SO COSTLY..........MISTAKE..........A good learning for all of us.....................