Optimizing Reverse Osmosis Membrane Life case study
An Independent District Heating and Cooling Facility used a municipal Lake Michigan water supply to operate a reverse osmosis system with three 200- gpm trains and was experiencing highpressure differentials and short runs between cleanings. The water had an SDI of 5–6 prior to the prefilters and an SDI of 3–4 after the prefilters. The system does not require continuous operation of all three trains and frequently only uses one or two trains, which can amplify potential biological problems.
Frequent cleanings, after only 1–4 million gallons of throughput monthly, were dramatically impacting membrane life and performance. When the cleaning frequency exceeds four cleanings per year, the reliability of the membranes is diminished and cost of operation escalates. The equipment supplier initially suggested additional pretreatment equipment and ultimately offered an extended service contract to remedy the poor performance of their system.
ChemTreat and the local engineering team worked together to optimize the operation and cleaning procedures. The increasing delta pressure indicated a high potential for biological or organic fouling. The membrane autopsies revealed biological fouling and membrane telescoping clearly indicated pressure surges. Multiple theories were investigated including biological or polymeric fouling, the rate of pressure increase at startup, and antifoulant interactions with the flocculant used by the municipal water source. The problem was resolved by a combination changes to the operation and cleaning procedures.
Decreased the rate-of-pressure increase at start-up to prevent membrane damage.
Altered biocide feed. Prior to system layup, ChemTreat CL207 biocide was fed to the system feed pump intake prior to shutdown. In addition to the biocide feed at shutdown, biocide was also fed every 12 hours during operation prior to the prefilters.
The system has been operating for over four months without cleaning and with no increase in differential pressure. The chart below illustrates the dramatic change in differential pressure for the first train with an increase in product recovery.