Consistency and challenges in replicating slow sand filtration for drinking water production across Mini- and Midi-scales

Abstract

Slow sand filtration (SSF) remains a crucial component in producing safe drinking water. SSF performance depends on the filter and Schmutzdecke maturity, yet determining full maturity remains challenging. The relationship between Schmutzdecke development (biochemical properties and microbial composition) and bacterial removal during ripening is still poorly understood. Laboratory-scale filters of different sizes often yield inconsistent results when used to mimic full-scale systems. This study investigated Schmutzdecke development and bacterial removal during initial SSF ripening under varying sand sizes, inoculum additions, and two scales (Mini/Midi). Over one year, we monitored biochemical indicators (carbohydrates, proteins, biomass) and total/active microbial communities and assessed their correlation with bacterial removal. Age and other operational parameters (i.e. sand size and additional inoculum) showed similar effects on Schmutzdecke biochemical and biomass development across scales. Microbial community composition shifted over time but was not consistent between Mini- and Midi-scales during the first 4 months. Bacterial removal remained low (<0.7 log₁₀) throughout the first year. While the Schmutzdecke developed over time, age alone proved unreliable for predicting filter maturity and performance. Instead, Schmutzdecke biochemical characteristics may be a more reliable indicator of filtration performance. To our knowledge, this is one of the first parallel comparisons of SSF ripening at two laboratory scales, with monitoring up to one year under identical conditions, linking Schmutzdecke biochemical properties and total/active microbial communities to bacterial removal. These findings highlight the limits of age as a maturity indicator and provide new insights for improving SSF design, monitoring, and lab-to-full-scale translation.