Experience exchange on replacing ion membranes

Thread Content

This post was last edited by sunjl1981 on 2013-1-6 23:53. Summary of membrane replacement: On March 5, 2007, the current load of the six electrolyzers was 9 KA. Since it was initially determined that four membranes in cell F (B660#, B618#, B629#, B623#) were leaking, it was decided to increase the current in cells A–E to 10.8 KA while maintaining the total load on the electrolyzer, in order to shut down cell F for inspection and replace the leaking ion membranes. Around 15:50 in the afternoon, after the treatment of slot F was completed, a first membrane leak test was conducted at a pressure of 0.62 mH2O. It was found that three membranes were leaking (no leakage was detected in B629#); specifically, B660# had a leakage rate of 7–10 l/hr, B623# had a leakage rate of 35 l/hr, and B618# also had a leakage rate of 7–10 l/hr. These three membranes were replaced. During the replacement process, it was discovered that the anode frame and anode mesh in the three respective cell slots were corroded, and the gaskets for those anode cells were also corroded and thus unusable. Therefore, while replacing the membranes, the gaskets for the anodes in these three slots were replaced as well. Subsequently, additional electrolyte filling operations were carried out on slot F in accordance with the procedures. When the liquid level on the anode side reached approximately 3/4 of its capacity, it was observed that the outlet hose at B661# on the anode side started to overflow first (no overflow was observed in the other cell slots), and the flow rate was particularly high. The filling operation was immediately stopped; it was determined that the leakage was caused by a membrane defect. Meanwhile, an analysis was conducted on the liquid flowing out of the hose to determine its alkalinity, and the analysis result showed an alkalinity concentration of 20.28%. The electrolyzer was cleaned by draining liquid once again, and the B661# ion membrane was replaced. It was found that there was a tear in the upper part of the removed membrane, approximately 40 cm in length, while the gaskets of the cell were intact. A second membrane leak test was conducted afterwards, using a test pressure of 0.5 mH2O. As a result, leaks were detected in four membranes: B636# had a leakage rate of 10 l/hr, B625# had a leakage rate of 22 l/hr, B620# had a leakage rate of over 100 l/hr, and B714# also had a leakage rate of over 100 l/hr. After these four membranes were replaced, it was found that the upper part of the B620# ion membrane was torn as well, with a length of about 10 cm; similarly, the upper part of the B714# ion membrane was torn, with a length of about 2 cm. It was also found that the anode tank frames and anode grids in cells B625# and B636# were corroded, as well as the gaskets of those anode cells. Finally, a membrane leak test was conducted for the third time on slot F, with a test pressure of 0.45 mH2O. It was observed that B634# and B716# had slight leaks; tests using rotameters showed that the flow rate in both cases was around 4–6 l/hr, which meets the production requirements. Therefore, the personnel on duty were instructed to carry out a liquid injection cycle. When the current in slot F rose to 0.5 KA, personnel were sent to measure the voltage of each slot, and a new issue was identified: the voltages of B716# and B634# were low, at 2.25 V and 2.32 V respectively, while the voltages of the other slots were on average 2.38 V ； B687# and B716# were colored later; when the current was around 3 KA, the hoses were colored normally. B634# has normal coloring. # ， ， &