Login / Signup

Modeling of Hydraulic Performance in Disks and Full-Scale Ceramic Water Filters.

Camille HeylenEbenezer AnnanKyle MonahanGabrielle M StringDaniele S Lantagne
Published in: Environmental science & technology (2021)
Ceramic filters for household water treatment can improve water quality and reduce diarrheal disease. Hydraulic performance is critical for quality control and user acceptability, and hydraulic models have previously been developed and tested with experimental full-scale filters. As filters are cumbersome, there is interest in using disks instead of filters in laboratory efficacy studies. To assess the validity of disk use, we collected experimental volume from three sets of full-scale frustum-shaped filters and matching disks with different burn-out material sieve sizes and firing temperatures. We compared the experimental and fitted data by calibrating hydraulic conductivities from filters and disks. Hydraulic conductivities increased with larger burn-out material and higher firing temperatures but were comparable between filters and disks (2.00-6.15 × 10-7m·s-1 and 2.69-6.32 × 10-7m·s-1, respectively). We found that previously described hydraulic models successfully predicted cumulative volumes for filters and disks with rRMSE ranging from 2.1 to 9.6% (filters) and 3.4 to 4.7% (disks). The error increased slightly (rRMSE: 5.0-15%) when predicting hydraulic parameters for filters from the hydraulic conductivity of disks. Our results validate a method to predict full-scale filter hydraulic performance from hydraulic conductivity of disks and can be used to simplify and increase testing capacity, resulting in higher quality, more acceptable filters that improve household drinking water quality.
Keyphrases
  • water quality
  • quality control
  • machine learning
  • quality improvement