Effective operation of RO systems is a data-intensive, multivariate, statistical, and preventative maintenance problem. Operators must monitor system’s key performance indicators (KPIs) such as pressures, conductivities, flow rates, permeate and concentrate streams of a given RO membrane unit, let alone an entire system. Data from individual RO membrane units must be collected and normalized to calculate a given system’s KPIs, criteria that help govern when preventative action should occur to minimize costs. Due to the high volume of data being monitored, calculating these KPIs can be a lengthy process. Currently, many facilities still perform manual data collection and review maintenance logs infrequently. At a RO plant, engineering and operations teams manually collect SCADA data, typically monthly, and then analyze the data in Excel or PowerBI against set KPIs which may take a week. Consequently, by the time operators receive the KPIs, sometimes more than a month will have passed, whereby the window of opportunity to preventatively take action is lost. Understanding the challenges RO facilities face when attempting to hasten the data collection process, calculate KPIs, and inform stakeholders in real-time are crucial for an effective RO facility’s operation and maintenance strategy. In addition, data quality checks are more cumbersome and usually are subjective based on what the reviewer considers to be erroneous data using professional judgment. Adjustments to treatment, maintenance, and membrane operations are considered and implemented as needed. These adjustments are not usually designed for evaluating data quality and filtering out erroneous values in real-time.
This presentation is available to AMTA Members only.
- Javad Roostaei
- Hazen and Sawyer
- AMTA/AWWA Membrane Technology Conference, Las Vegas
- AMTA/AWWA Membrane Technology Conference
- Reverse Osmosis, Data Collection, Analytics