Upscaled Filter Boxes with Iron-coated Sand for Effective Phosphorus Removal from Open Wells and Collector Drain Water

03 May 2023 - Published by Charlotte Boeckaert
Previously, a filter box with a filter material volume of 40L Iron-Coated Sand (ICS) developed by UGent was successful in removing reactive phosphorus (P) from individual drains in agricultural fields (over 80% P removal efficiency). However, the removal efficiency of this relatively small P filter box tended to be lower with larger water flows exceeding 5 m3/day. To upscale the filter, UGent and Inagro constructed and tested a large concentric cylindrical filter element.


140 liter ICS is wire meshed cylinders

The upscaled filter element was placed in a cubic container and consisted of two cylindrically shaped wire meshes filled with 140 L ICS in between them.

Water was pumped into the central cylindrical cavity of the filter element and forced to pass through the concentric cylindrical filter element, which had a thickness of 14 cm from the center towards the outside.

The P in the water passing through the ICS is (partially) sorbed before it is discharged from the cubic container into the ditch. Two such upscaled P filters were installed and tested during two drainage seasons, one on a collector drain (at a farm in Middelkerke, Belgium) and one at an open water pond (at the Godelieve fields of Inagro, Belgium).


 Upscaled P filter installed on a collector drain (left) at a farm in Middelkerke, Belgium, and at an open pond (right) at the Godelieve fields of Inagro, Belgium.


Two drainage seasons

During the first drainage season, with a water flow of 17-36 m3/day, the filter at Middelkerke removed 18.2% of total phosphorus (TP) and 21.3% of dissolved reactive phosphorus (DRP). Although the upscaled P filter showed the potential to treat a large water flow up to 40 m3/day, the P removal efficiency was thus limited, and considerably lower than the performance of the small P filters for individual drainage tubes.

Adjustments have been made to the filter system to increase the contact time and improve P removal efficiency (e.g. reducing flow) in the second drainage season, but unfortunately, the performance did not improve. Further investigation is needed to figure out the influential factors which hindered the P absorption.



The upscaled P filter at Inagro showed considerable P removal efficiency during the two drainage seasons of testing. In the second drainage season, with a flow up to 16 m3/day, DRP was reduced from 0.2 ppm to below 0.1 ppm, while achieving an average P removal efficiency of 72.5%.

Unfortunately, the filter faced clogging problems due to the open pond water quality. Despite our attempts to install pre-filters, they were clogged soon after as well due to the large amount of water being treated.

In conclusion, the upscaled P filter certainly has potential for removing P from larger amounts of water in specific cases (namely at the open pond), but the clogging problem will need to be addressed for installations at open waters.