Project overview
The Nine Mile Creek Watershed District and the City of Edina wanted to build an iron-enhanced sand filter (IESF) in Rosland Park to remove phosphorus from stormwater, preventing it from entering Lake Cornelia. IESFs are typically designed to passively filter phosphorus from stormwater; passive filtration requires allowing the runoff to slowly pass through a large volume of iron-enhanced media. However, in Rosland Park, preserving the park’s green space was a high priority. To reduce the filter’s footprint, it was necessary to first identify a filter media material capable of removing phosphorus much more quickly than the material used in a passive IESF. This would allow the runoff to be actively pumped through the system, quickly moving large volumes of water through a much smaller filter area.
To identify a media mix that was up to this challenge, this project’s experimental design called for a four-celled vault. All the runoff was pumped through the first cell, containing an anthracite filter, (cell 1) to remove suspended solids. The water leaving cell 1 was then split between three additional treatment cells. Cell 2’s media mix contained granite sand and iron. Cell 3 contained crushed limestone and iron. Cell 4 contained granite sand and activated alumina/iron oxide composite. The water leaving these three treatment cells was sampled over three seasons and analyzed to compare the phosphorus removal rates of each media mix.
Research questions
- Of the three media mixes tested, which one performs best under conditions of frequent and persistent flow?
- Do seasonal water quality fluctuations or seasonal temperature fluctuations impact phosphorus removal efficiency?
- Of the three media mixes tested, do any remove contaminants other than phosphorus?
- What sort of maintenance and operational procedures are required to maintain treatment effectiveness, and what are the associated costs?
Research findings
Phosphorus removal rates for all three media mixes in this active IESF were approximately half the rates of passive IESFs (approximately 30% removal compared to 60-70% removal). However, a lower phosphorus removal rate might be an acceptable tradeoff in tight spaces where a passive IESF will not fit. The anthracite pretreatment cell (cell 1) effectively prevented the system from clogging and was responsible for about half of the total phosphorus removal through the filtration system. Of the three media mixes tested in cells 2, 3, and 4, cell 3 (crushed limestone and iron) attained the highest median total phosphorus removal rate (29%).
Turbidity and total suspended solids levels peaked in summer and early fall, requiring more frequent backwashing to prevent the anthracite pretreatment filter in cell 1 from clogging. Seasonal variation was not observed to affect phosphorus removal rates.
The anthracite pretreatment cell (cell 1) removed approximately 22% of the nitrogen entering the filter system. After the water left cell 1, cell 2 removed an additional 2%, cell 3 removed an additional 16%, and cell 4 removed none of the nitrogen leaving cell 1.
The estimated annual cost is $9,638 per pound of total phosphorus removed. In the future, the research team expects to modify the design to bring that cost down to approximately $5,400 per pound of total phosphorus removed per year.
What does this mean for Minnesota?
Filter media materials that are capable of rapidly removing phosphorus from stormwater could make it possible to design active iron-enhanced filtration systems that work in tight spaces.
Reports and Presentations
- Final report 2025
- Research brief 2025 .pdf
- Project update 2024 .pdf
- Project Update 2023 .pdf
- Park signage 2022 .pdf