Recovery of UF Performance Through Cleaning Plans Developed From Continuous Monitoring - Part 2 of 2

Read Part 1 of 2 here.

The Miwatek team implemented cleaning and operational regimens aimed at addressing the biological problems. During this period, the UF settings for the MCs and air scours settings were updated to increase the duration of high pH oxidative MCs, extend the duration of air scours, and increase air scour frequency. Additionally, more frequent recovery cleans (RCs) were carried out to control bacterial growth. Although performance on the UF improved slightly, TMP would continue to rise rapidly after the high pH oxidative and low pH recovery cleans.

More frequent recovery cleans were carried out after 3 – 4 days once a TMP of 180?kPa?–?200?kPa was reached for the first half of September. These cleans gradually became less effective as the UF continued to struggle to return to baseline performance, indicating a significant loss of area. A more rigorous approach was adopted wherein "mini-RC 1+2's" were carried out on the UFs every day. This strategy was devised to allow for more cleaning whilst maintaining a high utilization. This period, 3?–?23?September, was characterized by rapid rise and fluctuations in TMP and is highlighted in Figure 5. The operating parameters for this period are tabulated in Table 2. ?No improvement in UF performance was observed during this period, suggesting the problem was not purely biological.

Figure 5: TMPs on the UFs from September to October 2024.

Table 2: Characterization of UF performance for 1st to 23 September 2024.

Analyzing the feed chemistry, the team modified its approach. Instead, we addressed the possibility of inorganic foulants since the addition of the detox stream to the IWP altered the water chemistry, increasing the copper, manganese, nitrite, nitrate, and ammonia levels. Table 3 shows the changes in water chemistry.

Table 3: Feed water components affected by CND treated stream to IWP feed.

Following the analysis of the data, the following changes were implemented:

• A high pH oxidative outside-to-outside recirculatory clean was performed, followed by a low pH outside-to-outside recirculatory clean;
• Each of the cleans was repeated until the cleaning liquor was visibly clear, with the previous liquor dumped and a new cleaning liquor made up;
• The UF settings were adjusted such that the intermittent clean would alternate between a high pH oxidative clean and a low pH clean.

The change carried out proved to be effective, as indicated by the more stable TMP curve generated between the?23rd and 27th of?September in Figure 5. The ability of the low pH cleans to maintain performance confirmed the presence of inorganic fouling. Therefore, the frequency of the low pH cleans was increased to specifically target organic foulants.

Following the positive impact of the previous recirculatory clean, a recirculatory clean with extended soaks and the chelating agent, citric acid, was developed; this clean consisted of:

• A high pH oxidative outside-to-outside recirculatory clean;
• High pH oxidative soak for 3 hours loaded from inside to outside;
• A low pH outside-to-outside recirculatory clean;
• A low pH with 1 wt.% citric acid outside-to-outside recirculatory clean;
• A low pH outside-to-outside recirculatory clean;
• Low pH soak for 3 hours loaded from inside to outside;
• A high pH oxidative outside-to-outside recirculatory clean.

The cleaning procedure started with a high pH oxidative and a low pH clean to increase the contact area for the chelating agent (citric acid) to promote more effective cleaning. Soaking of the inside of the UF lumen was introduced to remove potential foulants contained within the pores of the lumen. As previously, each of the steps was repeated until the cleaning liquor was visibly clear, with the previous liquor dumped and a new cleaning liquor made up. Figure 6 shows the liquor after the citric acid recirculation, indicating a significant build-up of foulants occurred on the membranes that were not removed from previous cleans. The citric acid clean could not be repeated due to the limited availability of citric acid. A significant improvement in UF performance has been observed, and operation remains stable, as indicated in the latter part of Figure 5.

Figure 6: CIP liquor from Citric Acid Recirculation.

The performance of an ultrafiltration system that operated continuously for two years declined due to changes in the feed water's chemistry. By consistently monitoring the system using the Miwatek Monitoring app, an iterative approach was implemented to restore the UF's performance. Adjustments were made to the cleaning regimens, including the addition of alternative cleaning chemicals, and their effectiveness was evaluated through the trends generated by the app. Through continuous monitoring of UF performance, the team successfully returned the system to its baseline performance.

Article (Part 1 & 2) written by Wing-Han Chan & Jivesh Naidu - Miwatek Process Engineers.

Read Part 1 of 2 here.