Knowledge-based system for automatic MBR control.

MBR technology is currently challenging traditional wastewater treatment systems and is increasingly selected for WWTP upgrading. MBR systems typically are constructed on a smaller footprint, and provide superior treated water quality. However, the main drawback of MBR technology is that the permeability of membranes declines during filtration due to membrane fouling, which for a large part causes the high aeration requirements of an MBR to counteract this fouling phenomenon. Due to the complex and still unknown mechanisms of membrane fouling it is neither possible to describe clearly its development by means of a deterministic model, nor to control it with a purely mathematical law. Consequently the majority of MBR applications are controlled in an "open-loop" way i.e. with predefined and fixed air scour and filtration/relaxation or backwashing cycles, and scheduled inline or offline chemical cleaning as a preventive measure, without taking into account the real needs of membrane cleaning based on its filtration performance. However, existing theoretical and empirical knowledge about potential cause-effect relations between a number of factors (influent characteristics, biomass characteristics and operational conditions) and MBR operation can be used to build a knowledge-based decision support system (KB-DSS) for the automatic control of MBRs. This KB-DSS contains a knowledge-based control module, which, based on real time comparison of the current permeability trend with "reference trends", aims at optimizing the operation and energy costs and decreasing fouling rates. In practice the automatic control system proposed regulates the set points of the key operational variables controlled in MBR systems (permeate flux, relaxation and backwash times, backwash flows and times, aeration flow rates, chemical cleaning frequency, waste sludge flow rate and recycle flow rates) and identifies its optimal value. This paper describes the concepts and the 3-level architecture of the knowledge-based DSS and details the knowledge-based control module. Preliminary results of the application of the control module to regulate the air flow rate of an MBR working with variable flux demonstrates the usefulness of this approach.

Carbon monoxide conversion by thermophilic sulfate-reducing bacteria in pure culture and in co-culture with Carboxydothermus hydrogenoformans.

Biological sulfate (SO(4)) reduction with carbon monoxide (CO) as electron donor was investigated. Four thermophilic SO(4)-reducing bacteria, Desulfotomaculum thermoacetoxidans (DSM 5813), Thermodesulfovibrio yellowstonii (ATCC 51303), Desulfotomaculum kuznetsovii (DSM 6115; VKM B-1805), and Desulfotomaculum thermobenzoicum subsp. thermosyntrophicum (DSM 14055), were studied in pure culture and in co-culture with the thermophilic carboxydotrophic bacterium Carboxydothermus hydrogenoformans (DSM 6008). D. thermoacetoxidans and T. yellowstonii were extremely sensitive to CO: their growth on pyruvate was completely inhibited at CO concentrations above 2% in the gas phase. D. kuznetsovii and D. thermobenzoicum subsp. thermosyntrophicum were less sensitive to CO. In pure culture, D. kuznetsovii and D. thermobenzoicum subsp. thermosyntrophicum were able to grow on CO as the only electron donor and, in particular in the presence of hydrogen/carbon dioxide, at CO concentrations as high as 50-70%. The latter SO(4) reducers coupled CO oxidation to SO(4) reduction, but a large part of the CO was converted to acetate. In co-culture with C. hydrogenoformans, D. kuznetsovii and D. thermobenzoicum subsp. thermosyntrophicum could even grow with 100% CO (P(CO) = 120 kPa).

Effect of carbon monoxide, hydrogen and sulfate on thermophilic (55 degrees C) hydrogenogenic carbon monoxide conversion in two anaerobic bioreactor sludges.

The conversion routes of carbon monoxide (CO) at 55 degrees C by full-scale grown anaerobic sludges treating paper mill and distillery wastewater were elucidated. Inhibition experiments with 2-bromoethanesulfonate (BES) and vancomycin showed that CO conversion was performed by a hydrogenogenic population and that its products, i.e. hydrogen and CO2, were subsequently used by methanogens, homo-acetogens or sulfate reducers depending on the sludge source and inhibitors supplied. Direct methanogenic CO conversion occurred only at low CO concentrations [partial pressure of CO (PCO) <0.5 bar (1 bar=10(5) Pa)] with the paper mill sludge. The presence of hydrogen decreased the CO conversion rates, but did not prevent the depletion of CO to undetectable levels (<400 ppm). Both sludges showed interesting potential for hydrogen production from CO, especially since after 30 min exposure to 95 degrees C, the production of CH4 at 55 degrees C was negligible. The paper mill sludge was capable of sulfate reduction with hydrogen, tolerating and using high CO concentrations (PCO>1.6 bar), indicating that CO-rich synthesis gas can be used efficiently as an electron donor for biological sulfate reduction.