Redundancy and Reliability – The 5th Generation of Membrane Bioreactor Design

Membrane bioreactor (MBR) facility designs have progressed through four generations of development (Crawford et al, 2001). The four generations are generally described as evolving from long Solids Retention Time (SRT) package plants, to the inclusion of nutrient removal (total nitrogen and/or phosphorus), to equipment and facility design optimization by reducing both the SRT and the Mixed Liquor Suspended Solids (MLSS) concentrations, and finally to the fourth generation characterized by larger plant designs with the Owner/Engineer assuming increased risk and design responsibility. The small MBR installations of the first three generations, as well as most fourth generation plants, were not required to operate with full reliability and redundancy – most MBR plants to date have included very large flow equalization or flow diversion basin capabilities, or withdraw their influent wastewater from a collection sewer that continues to a larger downstream plant. At these plants, reliability and redundancy is desirable but not essential – system shutdowns can occur without causing a raw wastewater bypass. Some recent MBR designs however are of the 5th generation of design; characterized by the inclusion of fully reliable and fully redundant design features that are essential for end-of-pipe treatment facility design. Detailed design information from operating MBR installations showing the unique reliability and redundancy requirements of end-of-pipe MBR system designs will be presented.Nine key design issues and concepts for 5th generation MBR designs include: 1) Programmable Logic Controller (PLC) System Design, 2) Network Communications System Design, 3) Operator Interface Modules for Monitoring and Control, 4) Alarm and Response Management, 5) Standby Power, 6) Membrane Train Redundancy, 7) Spare Capacity per Train Provisions, 8) Biological Reactors Redundancy, and 9) Flow Equalization Provisions.Manufacturers of membrane systems typically use PLCs to manage and sequence the complex operations of their systems. For small systems, reliability can be enhanced by utilizing the capabilities of modern PLCs to accommodate redundancy in key components, including a hot standby PLC to take over upon failure of the primary PLC. Larger MBR systems require multiple PLCs, and configuration of the PLCs, I/Os and control logic within each is important.The reliability of network communications within the membrane PLC system and between this system and the overall plant SCADA system, and peripheral devices such as Operator Interfaces is also critical. Continuous integrity of the communication allows the membrane control system to initiate and coordinate auxiliary functions essential to maintaining membrane operation. Fully looped communications networks are required, to ensure that there cannot be a single point of failure that would disrupt communications between master and train PLCs. A reliability audit of the communications network can identify weak links in the system.Operator interface (OI) modules for system monitoring and control are typically provided by the membrane equipment vendor, with the graphical displays customized for the equipment control requirements. The OI is the means for operator intervention to compensate for changing operating conditions and/or difficulties with system operations. It is therefore an essential system component to be considered when evaluating system reliability. Issues to be evaluated include local technician capabilities and response time, compatibility of graphical software with other plant systems, software licensing issues and hardware sparing.Alarm and response management includes consideration of all alarms or events that could cause a membrane train to shutdown or to be temporarily removed from service. With water treatment applications, the safe response to most alarms is to stop production from the affected train. With wastewater applications, however, most alarms need to be managed while the affected trains continue to operate, in order to reliably produce treated effluent at desired rates. A reliability audit can sort out critical from non-critical alarms and identify appropriate responses to manage those alarms. Examples of primary and secondary alarms, and the need to manage their automated responses, will be presented.Standby power is required to allow MBRs to continue to produce effluent. Key equipment that requires standby power during power failure events so that the MBR facility can continue to operate will be identified.Membrane train redundancy is one of the more difficult areas of MBR reliability and redundancy to design. Ideally, membrane equipment capacity is designed to satisfy all possible peak flow requirements. In the absence of accurate flow information, the Owner and designer may need to make additional provisions for redundancy of trains to allow for uncertainty.It is common for designers to provide spare capacity within each train for the potential installation of additional membranes, should that be required. Proper reliability and redundancy analysis of that spare capacity should include an understanding of the purposes to which that spare capacity may be applied.The evaluation and provision of biological reactor redundancy is similar to that for conventional treatment plant designs. Since MBR designs are often limited by their ability to transfer sufficient oxygen to meet carbonaceous BOD and nitrification requirements, it is prudent to review the oxygenation capacity carefully in any reliability analysis.Uncertainties may occur relating to the accurate prediction of short duration peak flows, therefore some equalization provisions are needed within the plant design, even if only within the bioreactor tankage, to accommodate short term train shutdowns and restarts, and control system variability. A reliability audit, or design, should include this consideration, and the provisions for inplant equalization of this type for two facilities will be compared.