Use of the MABR fingerprint soft sensor to control biofilm thickness and optimize hybrid membrane aerated biofilm reactor (MABR) performance

Abstract

Installing commercial membrane-aerated biofilm reactor (MABR) units into existing systems to form hybrid MABRs us ab emerging approach to increase wastewater treatment capacity and enhance nutrient removal due to tightened discharge regulations and growing populations. Biofilm thickness is identified as a critical parameter affecting MABR performance. To monitor biofilm thickness and develop a biofilm thickness control strategy for practical full-scale MABR units, this study develops the MABR fingerprint soft sensor as the linear relationship between off-gas oxygen content (% O₂) and bulk ammonia concentration (mg-N/L) under diurnally varying loading conditions. Simulations were conducted to test the implementation of the MABR fingerprint soft sensor in MABR units in a completely mixed tank and MABR units arranged in series. Results indicate that relatively thin biofilms, quantified by a higher numerical value of the MABR fingerprint soft sensor metric, should be maintained for the single-stage hybrid MABRs and in the initial stage for MABRs in series to optimize nitrification and denitrification simultaneously. The MABR fingerprint soft sensor metric can serve as a proactive signal to indicate biofilm thickness changes before noticeable deviations in performance occur. The numerical value of the MABR fingerprint soft sensor metric associated with optimal process performance was found to be relatively consistent over a range of process loadings and operation conditions, including temperature variations, wet weather flow conditions, and varying wastewater characteristics. A general procedure to implement the MABR fingerprint soft sensor to optimize the performance in the large-scale hybrid MABRs is provided.