Real-time control technology for enhancing biofiltration performances and ecosystem functioning of decentralized bioretention cells Open Access

Decentralized bioretention cells, also referred to here as micro-bioretention, biofilters, rain gardens or bioswales are among the most extensively used and promising nature-based solutions (NBS) to achieve at-source urban runoff management (Roy-Poirier et al. 2010; Skorobogatov et al. 2020; Spraakman et al. 2020). Their stormwater buffering capacity as well as runoff treatment rely mainly on the integrated effects of physical, chemical and biological functions of a high permeability vegetated soil ecosystem (Roy-Poirier et al. 2010; Skorobogatov et al. 2020). Most recent research advances highlighted the predominant role of biofilter media composition and depth for particulate-bound pollutant or nutrient retention including soluble contaminant adsorption (Skorobogatov et al. 2020; Zhang et al. 2021). Vegetation type and plant functional traits, as well as soil microbial communities eventually shaped by plant species, are additional drivers of contaminant removal through biosorption, bioaccumulation, immobilization or biodegradation (Read et al. 2009; Lange et al. 2020a; Skorobogatov et al. 2020; Vijayaraghavan et al. 2021). Positive influences of an internal anaerobic water storage zone (altering soil redox potential) have been reported also for improved microbial denitrification (Zhang et al. 2021) and metal removal, especially Cu (Blecken et al. 2009).

Great inconsistencies remain nevertheless in the effectiveness of bioretention systems to remove systematically major inflow nutrients, especially the more labile and soluble forms of reactive phosphorus and nitrogen (e.g. nitrate) (Read et al. 2009; Zhang et al. 2021). Recent works also pointed out that heavy metal distribution between particulate, dissolved colloidal (<0.45 μm) or truly dissolved fractions (≃2–3 nm) may be affected by bioretention systems, depending notably on soil pH and redox modifications in relation to soil moisture regime, plant roots and concomitant microbial activities (Lange et al. 2020a, 2020b). The dissolved and therefore more bioavailable and potentially toxic contaminant fraction may exceed local water quality standards in effluents (Lange et al. 2020a, 2020b). Besides, remobilization of previously captured (and contaminated) particulate organic matter or increased colloid transport during wetting events may favour pollutant leaching out of biofilters. This phenomenon may be enhanced by macropore formation and preferential flow along decaying roots or even soil cracks after extended drying periods, potentially leading to water and pollutant by-passing (Skorobogatov et al. 2020; Técher & Berthier 2022). Climate change involving extreme weather events like prolonged heatwaves or heavy rainfalls, combined with population growth and demands for treatment of a growing number of emerging pollutants in urban runoff, has put an additional increasing strain on bioretention systems (Spahr et al. 2020). Meeting performance goals regarding runoff treatment has become consequently more difficult for conventional biofiltration systems. As an illustration, these latter have not been designed to remove soluble organic pollutants such as ‘hydrophilic trace organic contaminants’ (e.g. plasticizers, pesticides, corrosion inhibitors, flame retardants, personal care products and pharmaceuticals) (Spahr et al. 2020). Recently, Rodgers et al. (2022) reported that persistent and mobile organic compounds (PMOCs) such as the highly soluble perfluorinated alkyl substances (PFAS) possessing soil-organic carbon distribution coefficients (log₁₀ DOC) below 2.7 (e.g. tris(2-chloroethyl) phosphate (TCEP) used as a flame retardant) were not retained by bioretention cells. In fact, many bioretention systems have been designed with highly sandy media to ensure fast infiltration and drainage over years, generally focusing more on hydrology than on pollutant capture or perennial vegetation development (Spraakman et al. 2020; Rodgers et al. 2022). Yet proper plant species establishment may enhance also drainage and rhizosphere-mediated bioremediation processes (Skorobogatov et al. 2020; Vijayaraghavan et al. 2021; Técher & Berthier 2022). Potentially insufficient soluble pollutant removal is worsened by the fact that bioretention systems are usually projected to function passively, without any water control system to regulate dynamically the contaminated flow to be managed, in order to ensure sufficient retention time for high-rate dissolved pollutant adsorption (prior to biological uptake or degradation) or enhanced anaerobic period (for denitrification). Biodegradation of organics generally also requires optimal soil moisture levels (i.e. near field capacity) to promote microbial activity that could lead to complete contaminant mineralization (Barros et al. 1995).

Rodgers et al. (2022) showed that increasing the proportion of evapotranspiration or decreasing the saturated hydraulic conductivity of bioretention media could help retaining the soluble part of mobile pollutants. With this in mind, future directions for bioretention design with more efficient water depuration capacity may consider trade-offs between lower permeability biofilter media to enhance pollutant capture and proper plant choice to ensure adequate ecohydrological functioning (involving both water infiltration and evapotranspiration fluxes) and rhizosphere-mediated remediation processes (Rodgers et al. 2022; Técher & Berthier 2022). Similarly to other agricultural (Griffiths et al. 2022) or forest ecosystem research domains (where the ‘bioretention concept’ was initially derived (Roy-Poirier et al. 2010)), there is a need for further research addressing the interactive effects of plant functional traits, microbial communities and media to optimize biofiltration performances (Lange et al. 2020a; Skorobogatov et al. 2020). For this purpose, plant trait-based approaches better anchoring phytoremediation into the whole plant ecophysiology show great promises (Gervais-Bergeron et al. 2021). However, more studies are required to take better advantages of the ‘fast-slow’ plant economics spectrum (Reich 2014) of species used in biofilters to propose species assemblages enhancing water transfer and contaminant removal while being resilient to environmental stress. Moreover, an improved understanding of coupled water, carbon and energy fluxes considering the central role played by vegetation in bioretention ecosystem considered as a whole could lead to the design of systems more adapted to local contexts, potentially resulting in enhanced ecosystem services.

The broader application of real-time control (RTC) technology to bioretention systems would have great potential to help address all of these aforementioned issues regarding soluble pollutant capture and improved bioretention ecosystem health and activity, while improving sustainable urban water management through enhanced stormwater treatment and possible reuse. Indeed, RTC relies on (wireless) sensors to enable real-time water or soil monitoring and actuators (e.g. valves, water gates and so on) in order to actively and automatically operate components of the water flow (Xu et al. 2021). It has been increasingly investigated to enhance successfully conventional wastewater treatment system performances. By contrast, only a few studies have been published so far that dealt explicitly with experimental RTC application (through vegetated bioretention column experiments) to improve stormwater runoff quality (Persaud et al. 2019; Shen et al. 2020). As ‘proofs of concept’, they showed that simple inflow or outflow active control valve opening could be effective to adapt stormwater detention time or soil moisture level to help removing pathogens (Shen et al. 2020) or reducing slightly ammonium effluent concentration (Persaud et al. 2019), respectively. Accordingly, more research investigations should be conducted, taking advantage of RTC and the ‘toolbox’ of environmental monitoring and ecological engineering sciences (i.e. related to the use of low-cost environmental sensors, the knowledge on water and contaminant transport, their interactions with ecosystem components including the impacts of plant and microbial functional diversity and activity (Cherqui et al. 2019; Lange et al. 2020a, 2020b)). That could help enhance the beneficial effects of the real-time active control of critical bioretention operational variables (e.g. water inflow/outflow rates, water detention times, soil moisture level control) for consistent removal of a broader range of soluble runoff-conveyed nutrients (N, P) and pollutants (heavy metals, metalloids and emergent hydrophilic trace organic contaminants). As suggested by Persaud et al. (2019) and Shen et al. (2020), different hybrid RTC strategies (e.g. combined real-time and forecast control and operational schemes (e.g. dynamic soil moisture/internal water storage level controls, mixed anaerobic/aerobic functioning, detention times) should help to cope better with the stochastic nature of climatic events (drought, heavy rainfalls) and variable runoff characteristics (volume, pollutant load).

Apart from abiotic retention processes, all of these operational variables may eventually greatly influence and enhance on a short (i.e. few hours) to a middle-time scale (i.e. few days to weeks) the microbial and vegetation-mediated biofiltration processes, which would thus deserve further research investigations, too. Of particular interest, additional artificial irrigation of bioretention cells thanks to the connexion with greywater systems (Barron et al. 2020) may help to overcome the negative impacts of prolonged drought on vegetation (and microbial activity) and resulting soil structure modifications, meanwhile allowing for the purification (and potential reuse) of the injected greywater volumes. Similarly, some kinds of cisterns or rainwater tanks may be envisioned to be used as additional stormwater storage layouts acting against flooding risks and allowing also for real-time controlled artificial irrigation of bioretention cells during the dry season, with potential water reuse. Eventually, other co-benefits regarding several ecosystem services (related to greenhouse gases reduction due to enhanced stormwater reuse instead of direct groundwater pumping (Batalini de Macedo et al. 2021), biodiversity support and cultural services (Rippy et al. 2021)) may arise from healthier bioretention ecosystems that would have been less prone to environmental stress.

We acknowledge that upgrading existent bioretention systems with RTC may not be always an easy (and feasible) task from a technical viewpoint, strongly depending on the bioretention construction specificities, as well as the local implementation context. Yet for existing systems that are lined at the bottom, for instance, just applying an outflow valve actively controlled may help to avoid important soil moisture level variations by preventing total bioretention cell drainage and maintaining optimal soil water content, thus creating more efficient soil microenvironments for vegetation-assisted rhizoremediation. Field trials and research efforts to adapt RTC to bioretention systems for a variety of contexts are still needed (Vijayaraghavan et al. 2021). However, it could be expected, in a near future, that the increasing availability of low-cost and low-power sensors (sometimes ready to use for Internet-of-Things (IoT) applications connecting together several stormwater control measures (Cherqui et al. 2019; Xu et al. 2021)) could help to expand the implementation of RTC technologies at broader scales. That could be reinforced by their combination with innovative and sustainable bioenergy production such as for instance plant microbial fuel cells. Some studies about soil moisture sensors fuelled by microbial fuel cells have already illustrated the potential of such applications (Tapia et al. 2018).

Then, the next step could be the integration of real-time controlled vegetated bioretention systems with other stormwater control measures at the district or maybe city scale, relying on smart process controls of the different kinds of connected stormwater management units thanks to optimization algorithms (Liang et al. 2021). Indeed, based on the work of Liang et al. (2021), more resilience to extreme events like flooding may be expected also by considering bioretention cells as transient stormwater storage layouts with optimization of their respective outflow rates at the catchment scale.

Developing efficient RTC and active strategies to interlink more closely hydrological and ecological functioning of bioretention systems may be a step further in the research agenda of sustainable urban water management. That should help to enhance the water treatment performances of bioretention ecosystems, their resilience to environmental disturbances (especially climate variations) and, last, but not least, the potential for enhanced stormwater reuse regarding non-potable water demands (e.g. irrigation, urban agriculture, domestic purposes and so on) and possibly other ecosystem services such as urban heat island mitigation (through evapotranspiration and air cooling) and city amenity. Integrating RTC systems directly into future biofiltration device design may eventually offer a unique opportunity for urban planners and researchers to combine more stringent stormwater treatment process with sustainable developmental goals (SDG) achievement (SDG 6 – clean water and sanitation, SDG 11 – sustainable cities and communities, SDG 13 – climate action) within the water–energy–food or greenhouse gases nexus in cities (Batalini de Macedo et al. 2021).

All relevant data are included in the paper or its Supplementary Information.

The authors declare there is no conflict.