**4. Discussion**

Air storage systems have been shown to be particularly useful to allow excess energy storage from renewable energies, with air storage in tanks or caverns (in periods of extra energy production) and a successive expansion of the pressurized air in a turbine (in periods of more consistent energy need). Among the different CAES configurations, including diabatic, adiabatic and isothermal mode, isothermal operations appear to be favorable to increase the overall performances, in terms of system efficiency [24]. The developed model for compressed air storage in WWTPs is highly flexible and can be specifically designed for target WWTPs, specifying plant potentiality and wastewater characteristics, both in terms of influent characterization, energy consumption of the biological phase and operating conditions of the activated sludge basin (such as air diffusers efficiency, HRT, SRT, X). It should be remembered that steady-state conditions, such as the ones considered in the present study, are not typically encountered in daily operations in WWTPs. However, this study was aimed at preliminarily evaluating the feasibility of the proposed solution to reduce operating costs for plant managing authorities. Furthermore, data scarcity has been recently recognized as a limiting factor to allow a widespread utilization of mathematical modelling, considering the complicated usage of online sensors and the huge workload required for sampling campaigns [38].

AD has been largely recognized as a significant source of renewable energy from different organic sources, able to provide fully renewable biogas, that has a high potential energy value [36]. Traditionally, AD was known to be beneficial in large-scale WWTPs, where a sufficient amount of sludge is available to continuously feed the digester, avoiding discontinuous operations throughout the year [39]. However, recently, by analyzing a large range (25,000–1,000,000 PE) of WWTP potentialities, it was proved that AD implementation is convenient also in small-scale plants when assessing environmental and economic benefits, particularly if agro-waste addition can be provided to the digester [40]. This can lead to an enhanced biogas generation, due to an augmented organic load, and a consequently favorable energy balance [40]. Regarding the second proposed scenario, further advances could include the co-digestion of sewage sludge with other locally available organic substrates to increase the obtainable methane yield from AD [31] or the application of sludge pre-treatments to increase its biodegradability [41]. Finally, the possibility to upgrade biogas to high-value biomethane, with technical characteristics totally comparable to fossil-derived natural gas, should be considered [41].

A following phase of the work is forecast, where the available commercial items for air storage will be investigated, to allow for an easier implementation of the proposed optimization system in existing plants. An in-depth process analysis of the aerobic (for example, considering the ASM1 model) and anaerobic (such as Anaerobic Digestion Model Number 1, ADM1 [42]) sections of the plant could be performed, better evaluating the performances of the proposed air storage system in dynamic conditions.

The detailed energy consumption data resulting from WWTP modelling through commercially available software's (such as GPSX ®, WEST ® or Biowin ®) could be integrated in the proposed economic saving approach, with further insights in achievable energy saving.

When considering the oxygen requirement of the biological treatment in WWTPs, it is known that the energy consumption for aeration strongly depends on a pool of di fferent parameters (often di fficult to monitor), such as oxygen transfer e fficiency, di ffuser fouling phenomenon, and di ffuser selection [43]. The proposed empirical approach for air supply calculation is extremely simplified and needs to be sustained by an in-depth experimental or modelling campaign. Novel techniques that were recently proposed to save energy in biological wastewater treatment include smart aeration control, consisting of variable frequency drive (VFD), dissolved oxygen (DO) sensors and programmable logic controller (PLC) [44]. When applying the proposed approach to target WWTPs, the experimental data obtained from these novel smart aeration systems could be used to update the input datasets for the proposed economic optimization.
