criterion for cleaning of RO membrane elements?

Published on by

Hello, I wish to know what is a good criterion for cleaning of RO membrane elements? Thanks in advance.

3 Answers

  2. Hi Vishakha, I agree with Ashwin but need to clarify his statement. The 10% increase in Normalised Feed Pressure is against the base line of the system with brand new membranes. Also the 10% decrease in Normalised Permeate Flow is also base lined against system with new membranes. So from the logged data you should see when the system has new membranes, the normalised feed pressure is lower, and the normailised feed pressure is higher. These are your base line numbers, when they change by 10% - schedule a CIP. Also recommend FT Norm - it is good (although I have doubts over the correction it has for temperature - I find on some plants at either high or low extremes, the temperature correction in the formule gives me strange data). But the above criteria are what is recommended by manufacturers, and I find tends to be overly conservative (results in frequent CIP, and ultimately reduced membrane life from too frequent exposure to pH extremes, especially if high pH CIP is required fopr bio/organic fouling). Also, in real life situations I find that operators don't always know how to use excel (and there fore can't calculate the normalised data), so in the abscence of process engineer support, the trigger points for CIP can be missed. Hence we generally work directly off of the differential pressures. As the operators can read the DP's directly from the gauges and the plant display they are aw3are of each stage DP at all times - but they don't have visibility to the normalised data. We are beginning to program in the normalisation calculations directly in to the control system for this reason (but have only done this for a few plants). So in general, we would expect the DP of each stage (with new membranes) to be approx 0.8 - 1.2 bar. We generally allow a 1.0 - 1.5 bar rise in DP before a CIP is performed. When the DP rises above 1.0 barm the operators schedule the CIP, then it should be performed before the DP rises above 1.5 bar. Of course using this method, relies on constant plant flowrates (if the feed flowrate to the plant varies, then the DP's on the stages will also vary). So this needs to be taken into account. So although it is not recommended todo it like this (and we easily have the capability to use the recomended method), we have found that simply using the DP achieves very good results. Generally this results in CIP once every 3-4 months, whereas using the normailsed trigger could be double this rate (or even more frequent). Essentially this is a balance, you want to maintain good quality permeate, and resuce your plant energy costs. But you also want to minimise downtime and extend the asset life (membrane). We've found this strategy works for us almost universally, and is a good balance between acceptable permeate quality and increased energy costs, and good membrane life with maximum time in service. We generally achieve 5-7 year membrane life (but this is highly dependant on the plant design feed water quality and plant operators). Another strategy that is sometimes used by some of our plant managers (and we try to discourage this), is to run the plant until the permeate quality is so bad it is out of spec, then perform CIP to brine the permeate back in spec. Please don't do this, as will result in excessively high DP's, and eventual reduction in permeate flows, and replacement of membranes in 2-3 years. Please dont do this. Hope this helps, Anthony

  3. You should consider normalized data: 10 % increase in pressure requirement and salt passage or 10% decrease in flow is indicative of cleaning requirement. Normally every membrane manufacturer provides excel work sheet to calculate normalized data. you may refer to DOW's FTNORM. This is downoadable. Ashwin Doshi