Plant-wide modelling and control of nitrous oxide emissions from wastewater treatment plants

Nitrous oxide (N2O) is a greenhouses gas with a global warming potential three hundred times stronger than carbon dioxide (CO2). The IPCC report released in 2014 shows that the CO2 equivalents emitted from the wastewater systems are increasing in the last decades. It was also estimated that 14% of those CO2 equivalents comes from N2O emissions.

It becomes therefore relevant, within the context of reducing the carbon footprint of wastewater treatment (WWT) systems, to develop control strategies aimed at the minimization of the emissions of this gas. Till now, few operation strategies have been developed to reduce the amount of N2O emitted from WWT plants.

However, these strategies have been employed for mainly sequencing-batch systems, where mere regulations of the cycle frequency and/or of the length of aeration and anoxic phases are enough to drastically reduce the amount of N2O emissions. However, in full-scale continuously-aerated wastewater treatment systems such control strategies cannot be implemented.

Furthermore, the available control strategies developed for N2O emissions are not online, namely they do not change the operating conditions automatically as a function of on-line measurements. All of this makes the technologies proposed till now too case-specific and quite a number of adaptations would be needed if the system is changed.

During the present work, a generic control strategy for N2O emission minimization is developed. More specifically, the control strategy is designed in order to prevent the typical biological mechanisms triggering N2O production. Furthermore, for thorough and comprehensive evaluation of such a control strategy prior to its application in real full-scale WWT systems, the developed control strategy is implemented and simulated in different model environments and a multi-criteria evaluation, taking into account not only the N2O emissions but also the effluent quality and the operational costs, is carried out.

This is because the reduction of the carbon footprint of WWT plants cannot be achieved at the expense of worse effluent quality and unreasonably-high operational costs. To build simulation environments where N2O controller could be benchmarked against a reference scenario, three different benchmark simulation models are developed by including N2O-producing processes in the Benchmark Simulation Model No2.

As an outcome, three different benchmark simulation models - the BSM2Na, the BSM2Nb and the BSM2Nc – are available. A scenario analysis showed discrepancies among the N2O predictions by the three models. Since there is at the moment no consensus model considered to describe reliably N2O emissions from WWT plants, all the three models are used for testing the N2O control strategy.

In a second step, a comprehensive sensitivity analysis on the BSM2Na was carried out at the aim of extrapolating the main biological mechanisms responsible for N2O emissions. It was found that the ratio between NOB and AOB activity could indicate the accumulation of those nitrification intermediates, like nitrite and hydroxylamine, which trigger the N2O production via AOB denitrification.

Given the interactive nature and multiple objectives typically required in biological systems, fuzzy-logic approach was chosen as a control technique for the implementation of the strategy. To avoid poor performance behaviour due to intuitive design, a systematic procedure for the design of fuzzy-logic controllers is developed using a partial nitritation/Anammox system as application case.

The same systematic methodology is then adopted to tune the fuzzy-logic controller for low N2O emissions. The ratio between measured nitrate produced and ammonium consumed in the aerobic zone (RNatAmm) is used as controlled variable and oxygen supply is regulated accordingly. The results coming from the benchmarking of the control strategy in the three simulation models showed that, by controlling the ratio RNatAmm, N2O emissions were able to be drastically reduced within reasonable aeration energy consumptions.

To cope with the increased COD demand by heterotrophic denitrifiers, additional control actions regulating the flow rate for carbon addition in the anoxic compartment were implemented. The results of the controller evaluated under comprehensive simulation tests indicate a promising potential for full-scale applications in order to reduce N2O emission from WWTPs.

In addition, implementation of the control concept requires minimum investment (only relevant sensors required and adaptation of aeration control algorithm of the plants) is expected to encourage its take up by WWT plant operators for managing CO2 footprints of WWTPs.