API PUBL 932-A:2002 pdf download

API PUBL 932-A:2002 pdf download.A Study of Corrosion in Hydroprocess Reactor Effluent Air Cooler Systems.
A plot of the data is shown in Figure I. The four horizontal hands represent the four levels of corrosion but it can be seen that there is considerable overlap in the range of K factors at each level. The best conclusion that can be drawn is that there is a trend showing that the severity of corrosion increases with increase of K,,. confirming Piehl’s observation. The imprecise nature of the correlation however does not permit a useful guideline to be developed. Similar plots were developed for corrosion severity versus:
I. Downstream separator bisullide concentration (see
Figure 2).
2. Maximum air cooler tube velocity (see Figure 3).
Note: Where values were not reported by the plants, ulucs calculated by UOP were used. These calculated values are valid only for balanced systems with assumed unitimt (low distribution. They may be inaccurate where less than full condensation has occurred. As presented, the data were so scattered that no correlation or int’erence could be drawn. It was esident that occurrences of corrosion do not correlate well with individual parameters partly because of the maccuracies just discussed and in addition, because of the interdependency of sonic of the par.uneters. By simultaneously plotting the level of corrosion against both bisulfide concentration and velocity. only a very slight improvement was obtaáncd (see Appendix A. Figure A-i). However, when the influence of piping symmetry on the distribution of flow through the air coolers was introduced a’. a fourth parameter. more striking relationships were apparent (see Appendix A. Figures A-2 through A-5).
The air cooler piping system consists of a single inlet pipe connected to a branched manifold system called the inlet header, which distributes the flow equally to each cell of the air cooler. The outlet flow from each cell is gathered by a similar manifold arrangement called the outlet header, which reduces to a single outlet pipe leading to the separator. With the piping headers inlet and outlet) hydraulically balanced (see Figure A-3). no corrosion or low corrosion of the air cooler tubes is apparent. Several moderate corrosion pinis appeared. hut two of these were associated with very high bisu tide concentrations. Similarly, where the inlet piping header is balanced, corrosion tends to be low (see Figure A-4). On the other hand, when both headers were unbalanced (see Figure A-5). corrosion of the tubes was predominantly severe or moderate, It is clear that uneven distribution of flow through the air cooler creates conditions of either low or high velocities where corrosion can occur. Unfortunately. iii most cases reported. neither the hisultide concentration nor the magnitude of the velocity at the location where corrosion occurs is known, so that a more refined correlation is not possible. The data were screened to exclude air coolers with return bend tubes and those with ferrules. All of the plots are included in Appendix A (see Figures A-I to A-5).