Wastewater treatment plants (WwTPs) remove microplastics (MPs) from municipal sewage flow, with the resulting bulk of MPs being concentrated within generated sewage sludge which is frequently recycled back onto agricultural land as accepted practice in many European countries as a sustainable fertiliser resource. This circular process means that MPs successfully removed from WwTPs are deposited into the soil and able to return into the natural watercourse by means of run-off or infiltration to groundwater. This study quantifies the removal efficiency of MPs with size ranging between 1000 and 5000 μm in a primary settlement tank (PST) at a WwTP serving a population equivalent of 300,000 and provides MP concentrations in the generated sewage sludge. Our study revealed that the proportion of MPs partitioning in a PST to settled sludge, floating scum and effluent was 96%, 4% and 0% respectively, implying 100% removal of MPs of 1000-5000 μm in size. The generated sewage sludge was estimated to contain concentrations of approximately 0.01 g of MPs or 24.7 MP particles per g of dry sewage sludge solid, equivalent to ∼1% of the sewage sludge weight. Using these figures and data from the European Commission and Eurostat, the potential yearly MP contamination onto soils throughout European nations is estimated to be equivalent to a mass of MPs ranging between 31,000 and 42,000 tonnes (considering MPs 1000-5000 μm in size) or 8.6×10¹³-7.1×10¹⁴ MP particles (considering MPs 25-5000 μm in size). An estimated maximum application rate of 4.8 g of MP/m²/yr or 11,489 MP particles/m²/yr, suggests that the practice of spreading sludge on agricultural land could potentially make them one of the largest global reservoirs of MP pollution. Hence, recycling raw sewage sludge onto agricultural soils should be reviewed to avoid introducing extreme MP pollution into the environment.

To simulate the large-scale impacts of wind farms, wind turbines are parameterized within mesoscale models in which grid sizes are typically much larger than turbine scales. Five wind-farm parameterizations were implemented in the Weather Research and Forecasting (WRF) Model v4.3.3 to simulate multiple operational wind farms in the North Sea, which were verified against a satellite image, airborne measurements, and the FINO-1 meteorological mast data on 14 October 2017. The parameterization by Volker et al. underestimated the turbulence and wind speed deficit compared to measurements and to the parameterization of Fitch et al., which is the default in WRF. The Abkar and Porté-Agel parameterization gave close predictions of wind speed to that of Fitch et al. with a lower magnitude of predicted turbulence, although the parameterization was sensitive to a tunable constant. The parameterization by Pan and Archer resulted in turbine-induced thrust and turbulence that were slightly less than that of Fitch et al., but resulted in a substantial drop in power generation due to the magnification of wind speed differences in the power calculation. The parameterization by Redfern et al. was not substantially different from Fitch et al. in the absence of conditions such as strong wind veer. The simulations indicated the need for a turbine-induced turbulence source within a wind-farm parameterization for improved prediction of near-surface wind speed, near-surface temperature, and turbulence. The induced turbulence was responsible for enhancing turbulent momentum flux near the surface, causing a local speed-up of near-surface wind speed inside a wind farm. Our findings highlighted that wakes from large offshore wind farms could extend 100 km downwind, reducing downwind power production as in the case of the 400-MW Bard Offshore 1 wind farm whose power output was reduced by the wakes of the 402-MW Veja Mate wind farm for this case study. Significance Statement: Because wind farms are smaller than the common grid spacing of numerical weather prediction models, the impacts of wind farms on the weather have to be indirectly incorporated through parameterizations. Several approaches to parameterization are available and the most appropriate scheme is not always clear. The absence of a turbulence source in a parameterization leads to substantial inaccuracies in predicting near-surface wind speed and turbulence over a wind farm. The impact of large clusters of offshore wind turbines in the wind field can exceed 100 km downwind, resulting in a substantial loss of power for downwind turbines. The prediction of this power loss can be sensitive to the chosen parameterization, contributing to uncertainty in wind-farm economic planning.

A large-eddy simulation (LES) of a laboratory-scale horizontal axis tidal stream turbine operating over an irregular bathymetry in the form of dunes is performed. The Reynolds number based on the approach velocity and the chord length of the turbine blades is approximately 60,000. The simulated turbine is a 1:30 scale model of a full-scale prototype and both turbines operate at very similar tip-speed ratio of λ ≈ 3. The simulations provide quantitative evidence of the effect of seabed-induced turbulence on the instantaneous performance and structural loadings of the turbine revealing how large-scale, energetic turbulence structures affect turbine performance and bending moments of the rotor blades. The data analysis shows that wake recovery is notably enhanced in comparison to the same turbine operating above a flat-bed and this is due to the higher turbulence levels generated by the dune. The results demonstrate the need for studying in detail the flow and turbulence characteristics at potential tidal turbine deployment sites and to incorporate observed large-scale velocity and pressure fluctuations into the structural design of the turbines.

The efficiency of tidal stream turbines in a large array depends on the balance between negative effects of turbine-wake interactions and positive effects of bypass-flow acceleration due to local blockage, both of which are functions of the layout of turbines. In this study we investigate the hydrodynamics of turbines in an infinitely large array with aligned or staggered layouts for a range of streamwise and lateral turbine spacing. First, we present a theoretical analysis based on an extension of the linear momentum actuator disc theory for perfectly aligned and staggered layouts, employing a hybrid inviscid-viscous approach to account for the local blockage effect within each row of turbines and the viscous (turbulent) wake mixing behind each row in a coupled manner. We then perform large-eddy simulation (LES) of open-channel flow for 28 layouts of tidal turbines using an actuator line method with doubly periodic boundary conditions. Both theoretical and LES results show that the efficiency of turbines (or the power of turbines for a given bulk velocity) in an aligned array decreases as we reduce the streamwise turbine spacing, whereas that in a staggered array remains high and may even increase due to the positive local blockage effect (causing the local flow velocity upstream of each turbine to exceed the bulk velocity) if the lateral turbine spacing is sufficiently small. The LES results further reveal that the amplitude of wake meandering tends to decrease as we reduce the lateral turbine spacing, which leads to a lower wake recovery rate in the near-wake region. These results will help to understand and improve the efficiency of tidal turbines in future large arrays, even though the performance of real tidal arrays may depend not only on turbine-to-turbine interactions within the array but also on macro-scale interactions between the array and natural tidal currents, the latter of which are outside the scope of this study.