Shortcut Navigation:
Kalia YNA 2012


Grade: 9 | State: New York

Using Plant-Based Biofilters to Purify Household Wastewater


This project tested whether “biofilters”—filters based on living plants—could be used to remove common household pollutants from water. I built two biofilters and used them to purify water polluted with laundry detergent to levels at which it could be safely released into nature. I tested the wastewater on live daphnia before and after filtering, as well as with chemical test strips to measure nitrites, nitrates, hardness, chlorine, alkalinity, pH, ammonia, and phosphates. Both methods showed that my filters improved the water quality. The water output from the biofilters sustained the live daphnia as well as the tap water. My experiment confirmed that biofilters could be a viable component of small-scale or domestic water purification. Plant-based biofilters might be used to filter wastewater before returning it safely to the environment, to provide “gray water” for uses other than consumption, or as low-cost prefilters to make filtration of drinking water more efficient. Plant-based biofilters are environmentally conscious and more accessible to areas that cannot afford the industrial technology of municipal filtration.

Background: Scarcity of Clean Water 

Clean water is a critical resource, rapidly becoming scarce because growing populations are consuming more and reducing supply through pollution. Filtering, treating, and otherwise cleaning wastewater to increase supply is necessary, but these methods are expensive and require excessive energy. Modern treatment processes are decidedly not organic, often requiring additional chemicals to be added to the water (e.g., chlorine.) (Faust, 1998) I hypothesized that natural filters utilizing plants and their associated growth media (and microflora) could filter common household pollutants from wastewater. While not yielding potable water, the filtered water might be clean enough to release safely into the environment or reuse (e.g., to flush a toilet or wash a car). A natural filter has many benefits and could be a future alternative for water filtration.

Dangerous Components of Household Wastewater 

I used wastewater containing laundry detergent, a common household pollutant released into the water. It contains many chemicals that are not readily broken down, and is hazardous to the ecosystem because it kills aquatic life, spoils the food chain and compromises human health. For example, the phenols found in detergents are toxins that cause disruptive glandular effects in people and animals, such as gender swaps in fish and death to hypersensitive people. (IPSC Health Safety Guide 2011) Detergent also contains surfactants, builders, bleaches, phosphates and other chemicals, including nitrates, nitrites and enzymes. Surfactants have emulsifying and dispersing properties, and decrease the surface tension of water. Builders, such as sodium tripolyphosphate (STPP), remove the calcium and magnesium ions present in hard water and soil, thereby softening the water and stripping it of minerals. Other common, toxic ingredients of laundry detergent include bleaches, which kill important anionic bacteria, and petroleum distillates (e.g. napthas). In drinking water, phenols quickly caused malignancies in mice and rats. ( 2010) Phosphates in detergent greatly contribute to damaging algae blooms. All of these changes to the balance of natural water systems affect microorganisms, plants, and larger animals such as fish.

Preparation of the Biofilters


The biofilter

Using a 20-gallon aquarium, separated into halves by an acrylic barrier, I built two biofilters, each designed to circulate wastewater repeatedly through common hardy living plants. The water-based filter used a dwarf water lily (Nymphaea) bedded in a mound of sand and floating duckweed (Lemna minor) plants in about five inches of water. A baffle split the water into a U-shaped trough, and a small pump circulated the water from one end of the U to the other. Based on my measurements, this water circulated at about 7.5 liters per hour. The land-based filter used nut grass weeds (Cyperus rotundus) planted in layers of dirt, sand and peat moss, all on top of cheesecloth and wire mesh so that water seeping down through the media would collect below. A second pump circulated this water by pumping the water from the bottom of the filter into a drip line running through the nut grass. Based on my measurements, this water circulated at about 10.6 liters per hour, but perhaps slightly slower when the drip line was lying in the dirt.



The land-based biofilter

I predicted that the land-based biofilter would be more successful since it emulates a municipal filter, with many layers and means of filtration in the various growth media. The drip line and media give the plants and the soil bacteria more time to act on the wastewater. Land-based plants also grow more extensive root systems. (Some roots stretched down to the collection chamber.) I predicted the water-based biofilter would work less well, since the water flowed through the plants more quickly, and the plants were also more directly exposed to the pollutants, having little or no growth medium.

Among the chemicals I tested, I predicted the most improved would probably be phosphates, because plants absorb phosphates—they are ingredients in many fertilizers. I predicted that the pH in the water filter would become closer to neutral, because certain water plants can tolerate alkaline environments and continue to lower the pH. Lastly, I predicted that a week might be enough cycles (well over 100 cycles in either filter) to create a noticeable effect in the water purity. After a much longer period in a closed (test) environment, plant deaths and rotting detritus might have negative effects on the water. I predicted that the daphnia, a species that thrives in clean water, would survive longer in the filtered water than in the original wastewater solution.



Using test strips to measure chemical levels

I created a wastewater solution meant to mimic that in a typical washing machine by adding All Stainlifter brand laundry detergent to clean water at a concentration of about 0.0055 ounces per liter (equivalent to following detergent instructions with a 40-gallon washing machine). I tested the effectiveness of my biological filters in two ways: (1) measuring specific chemical levels using test strips (sold for testing aquarium water) and (2) introducing living daphnia, tiny crustaceans that only live in relatively clean water (purchased from Carolina Supply). I created four control solutions by successively diluting my base solution 2:1, 4:1 and 8:1 with clean water, and tested these four solutions with test strips and daphnia.

Using an eyedropper, I placed about six live daphnia in each habitat and monitored their survival over the next day. I chose to filter Solution 2 (the 2:1 concentration) because it was the lowest concentration that killed all the daphnia within 10 hours, while daphnia in Solution 3 survived for more than a day (about the same as in clean water, probably due to suboptimal feeding and water aeration). Solution 2 differed from Solution 3 in the levels of nitrites, chlorine, hardness, phosphate and ammonia (see Chart 1). I computed the volume of each of the filter chambers and added laundry detergent to create the same initial concentration as Solution 2. 



The filtered water with living daphnia

Having created household “wastewater” of a known concentration in each biofilter, I let the filters circulate to see if the plants could remove toxins from the water. I measured the effectiveness of the biofilters by comparing chemical levels measured with test strips and daphnia mortality rates to those of the control solutions, including clean tap water. I took daily samples from each filter for testing. After seven days, I removed filtered water from each biofilter and introduced living daphnia (15 for the land- and 13 for the water-based filter) and charted their survival (see Chart 4).


Both biofilters yielded similarly improved water purity, almost back to the original state of the tap water, based on my tests. Overall, the data from the test strips showed the chemical levels in the biofiltered water to be comparable with or better than control Solution 3—half the detergent concentration initially in the filters. 


The land filter removed the most phosphate, taking it from 5.0 ppm to 0.3 ppm over six days, and ammonia, lowering it from 0.15 ppm to nearly zero over the same period. The land filter also quickly lowered nitrates, nitrites and chlorine from initial levels (of 10, 0.5 and 0.5 ppm, respectively,) to almost negligible levels within the first two days. It also raised alkalinity from 40 to 80 ppm, somewhat below ideal levels of 120–180 ppm for an ecosystem. (Brain 2011) The ideal pH of water is 7, which is perfectly neutral. The land filter came closer, with a slightly acidic final pH of 6.9. The water filter was slightly basic, yielding 7.2. (Small numerical differences in pH are significant because it is a logarithmic scale.)

The water-based filter also lowered phosphate and ammonia gradually to near-zero levels over six days. It took two to three days longer than the land filter to lower nitrite and nitrate levels but did bring them to zero, as per the test strips. The water filter was unable to change the alkalinity from initial levels of 40 ppm. After about two weeks, the lily started to deteriorate in health—it was too large for the aquarium—which likely skewed the results due to the rotting plant material.


Live daphnia introduced into the biofiltered water had survival rates comparable to those in pure water—and better than that in Control Solutions 3 or 4. (See Chart 4.) In Solution 3, daphnia survived for 10 hours but were dead after 24 hours. In the land-based filtered water, only 2 of 15 daphnia died in 24 hours, and in the water-based filtered water, none died. The deaths in the land-based filter water may have been from the brownish peat moss residue or just suboptimal care. The daphnias’ survival rates demonstrated that the biofiltered water was better able to support aquatic life than the original control solutions.


How the Biofilters Cleaned the Water 

An industrial water treatment plant uses mechanical, biological and chemical processes to filter and purify wastewater. Screening, a grit chamber and sedimentation separate larger particles from the water by mechanical means. A grit chamber is a dense material that slows the flow of the water so that finer solids are removed. A sedimentation tank, or clarifier, rotates the water slowly so that heavier sediment sinks and oil rises. “Activated sludge” and aeration provide biological filtration. (Faust 1998) Activated sludge is an oxygenated environment that encourages the growth of saprotrophic bacteria—bacteria that break down organic matter—and other organisms that metabolize pollutants. (Mountain Empire Community College 2010) Finally, flocculation (chemicals that precipitate colloidal pollutants), chlorination and disinfection provide chemical purification. (Faust 1998) My biological, plant-based filters may have filtered the water through analogous processes. I believe that the water was filtered by three methods, (1) mechanically, by the soil, sand, and peat moss, (2) biologically, by activated bacteria in the water and media, and (3) by the plants themselves. 

Mechanical Filtration

In the land filter, water passed through soil, sand and peat moss. When polluted water goes through soil, oils, heavy metals and excess nutrients are filtered out mechanically and by soil-borne organisms that absorb or metabolize them. Forcing the wastewater to seep slowly through the media gave these organisms time to break down or absorb various pollutants. The sand may have also acted as a grit chamber, taking solids out of the water. In addition to filtering and/or absorbing pollutants, peat moss is known to soften water chemically because it binds calcium and magnesium ions, and releases tannic and gallic acids into the water. These acids target bicarbonates in water and reduce the carbonate hardness and pH. ( 2011) This leads me to believe that peat moss was the main agent in neutralizing pH and hardness. Although the water filter had a sand bedding for the water lily, it probably provided less mechanical filtration. 

Biological Filtration

My biofilters likely provided biological filtration from bacteria living in the water and the growth media. The water was aerated as it dripped from the pump, and nutrients for bacteria were provided in the top soil and from the detergent, which contains compounds that bacteria consume, such as phosphates, nitrates and nitrites. When some of the duckweed died, its plant matter may have provided food for the bacteria. These bacteria likely played a major role in removing toxins from the water. Activated sludge filters can oxidize carbonaceous matter, turn ammonium and nitrogen into biological materials, remove phosphates, and absorb gases such as carbon dioxide, ammonia and nitrogen. (Mountain Empire Community College 2010) The chemical strips showed that the concentrations of some of these substances declined throughout the seven-day period.

Filtration by the plants

Finally, I believe the plants in both biofilters played a great role in removing phosphates, nitrates, nitrites and ammonia. Many substances harmful to people and animals are conducive to plant growth. Plants require ammonia, phosphates and nitrates, and most synthetic fertilizers (as well as laundry detergent) contain these chemicals. However, high concentrations of these chemicals can cause algae or other plant species to “bloom,” disrupting the environmental balance, and sufficiently high concentrations can kill the same plants. Duckweed, nut grass and water lilies are known to absorb these substances, and lily pads are planted in many ponds to control algal blooms. ( 2011) Duckweed expands and spreads as it gains phosphates, as do water lilies. In addition to absorbing chemicals useful to them, certain plants can “lock up” harmful substances such as lead, zinc and cadmium, preventing them from harming other species or getting into the groundwater. Duckweed has been proven to be an incredible biofilter in absorbing not only phosphorus, but also dangerous heavy metals. It was tested by a team of Israeli scientists to clean wastewater from a nuclear power plant, which after passing through the biofilter was 99% clean. (Cafe 2011)


My experiment showed that plant-based, biological filters can effectively remove pollutants from household wastewater. While my experiment was limited in size and duration (and also because of the difficulty of obtaining living plants during the winter), it removed measurable amounts of key pollutants. In addition, living specimens verified that the water quality did improve. I researched examples of successful, larger biofilter experiments. Plant-based filters are economical, accessible in less-developed countries, ecologically safe and produce none of the noise, odor and unsightliness associated with water-treatment plants. (Logson 2002) Further research could test the effectiveness of more mature plants, different plant species and other combinations of cycling water through multiple filters and larger-scale filtration.

Data Tables

Land-based filter

Date Nitrate Nitrite Hardness Chlorine Alkalinity pH Ammonia Phosphate
Control 0 0 25 0 40 7.0 0 0
3/14/11 10 0.5 25 0.5 40 6.8 0.15 5
3/15/11 5 0.5 75 0.5 40 6.8 0.15 5
3/16/11 0 0 75 0 40 6.9 0.1 3
3/17/11 0 0 75 0 40 6.9 0.1 2
3/18/11 0 0 75 0 80 6.9 0.1 1
3/19/11 0 0 75 0 80 6.9 0 1
3/20/11 0 0 75 0 80 6.9 0 0.3

Water-based filter

Date Nitrate Nitrite Hardness Chlorine Alkalinity pH Ammonia Phosphate
Control 0 0 25 0 40 7.0 0 0
3/14/11 10 0.5 75 0.5 40 6.2 0.15 5
3/15/11 10 0.5 75 0 40 6.4 0.15 4
3/16/11 10 0.5 75 0 40 6.4 0.15 2
3/17/11 5 0.5 75 0 40 6.2 0.15 2
3/18/11 5 0.5 75 0 40 6.8 0.15 1
3/19/11 0 0 75 0 40 6.8 0 1
3/20/11 0 0 75 0 40 7.2 0 0.5


Control Solutions

  Pure Water Solution 1 Solution 2 Solution 3 Solution 4
Nitrate 0 0 0 0 0
Nitrite 0 0.5 0.5 0 0
Hardness 25 25 25 0 0
Chlorine 0 0.25 0.25 0 0
Alkalinity 40 300 180 0 0
pH 7.0 7.5 6.2 6.8 6.8
Ammonia 0 2 0.125 0 0
Phosphate 0 5 1.5 1 0.5


  Next morning That night Starting population After 12 hours After 24 hours
Solution 1 0 0 6 0% 0%
Solution 2 0 0 5 0% 0%
Solution 3 2 0 5 40% 0%
Solution 4 2 2 6 33% 33%
Land filtration 13 13 15 87% 87%
Water filtration 13 13 13 100% 100%
Control–pure water 6 6 8 75% 75%


“Activated Sludge.” Disinfection of Water. Retrieved from the World Wide Web on 15 February 2011.

“Are Leading Brand Laundry Detergents Environmentally Friendly?” Laundry Alternative. Retrieved from the World Wide Web on 20 February 2011.

Brain, Marshall. “How Sewer and Septic Systems Work.” Howstuffworks. Retrieved from the World Wide Web on 1 March 2011.

Faust, D. Samuel. Chemistry of Water Treatment. Washington, D.C.: Lewis Publishers (CRC Press), 1998.

Fuertes, James H. Water Filtration Works. New York: J. Wiley and Sons, 1901.

“How Surfactants Become Bodily Toxins.” Brighthub. Retrieved from the World Wide Web on 13 March 2011.

Logson, Gary.Water Filtration Practice, Including Slow Sand Filters and Pre-Coat Filtration. Boston: Taylor and Francis Group, 2002.

Marti, Isabel. “Spectrophotometric determination of low levels of anionic surfactants in water by solvent extraction in a flow injection system.” RSC Publishing. Retrieved from the World Wide Web on 13 February 2011.

“Phenol Health and Safety Guide.” IPSC Health and Safety Guide. Retrieved from the World Wide Web on 1 March 2011.

“Toxic Dangers of Typical Laundry.” Retrieved from the World Wide Web on 14 March 2011.

“Using Peat Moss to Lower Water Hardness and Ph Levels in Ponds.” Pet Retrieved from the World Wide Web on 1 March 2011.

Vigil, Kenneth. Clean Water. New York: Lewis Publishers (CRC Press), 2003.

“Water Plants as Biofilters to Polluting Heavy Metals and Radioactive Materials.” Ideas Galore. Retrieved from the World Wide Web on 14 February 2011.

Winter, Ruth. A Consumer’s Dictionary of Household, Yard and Office Chemicals: Complete Information about Harmful and Desirable Chemicals Found in Everyday Home Products, Yard Poisons, and Office Polluters. Miami: Macmillan Ltd, 2001.

Zoller, Uri. Handbook of Detergents. Boca Raton, Florida: Taylor and Francis Group, 2009.

American Museum of Natural History

Central Park West at 79th Street
New York, NY 10024-5192
Phone: 212-769-5100

Open daily from 10 am - 5:45 pm
except on Thanksgiving and Christmas
Maps and Directions