United States
                     Environmental Protection
Robert S. Kerr Environmental
Research Laboratory
Ada OK 74820
                     Research and Development
EPA-600/S2-83-067  Oct. 1983
v>ERA         Project  Summary

                    Wastewater Treatment with
                    Plants  in  Nutrient  Films
                    W. J. Jewell, J. J. Madras, W. W. Clarkson. H. DeLancey-Pompe, and R. M.
                      The nutrient film technique (NFT) is
                    a unique modification of a hydroponic
                    plant growth system which utilizes
                    plants growing on an impermeable
                    surface. A thin film of water flowing
                    through the extensive root system
                    provides nutrients for plants and
                    associated microbial growth. Root
                    masses up to 15 cm thick or more have
                    been  obtained. This self-generating
                    plant system could be used as a fitter to
                    immobilize and use the gross and trace
                    organics in wastewater. The goal of this
                    study was to determine the economic.
                    technical, and  practical feasibility of
                    using plants grown in the NFT system
                    as pollution control systems.
                      NFT systems appear capable of
                    providing secondary quality treatment
                    with  some nutrient  removal on a
                    relatively small  area  compared to
                    overland flow systems. At loading rates
                    of 10 cm per day the effluent quality
                    with primary settled sewage was often
                    less than 10 mg/l for suspended solids
                    and biochemical oxygen demand. The
                    influent sewage temperature was 9C.
                    Estimated area needs of an NFT system
                    designed for  BOD and SS removal
                    appear to be approximately 3 hectares
                    for a community of 10,000 people,
                    whereas up to 10 times this amount
                    may be needed to provide nutrient
                    '   This EPA-sponsored study was par-
                    tially supported by the Office of Water
                    Resources and Technology.  U.S. De-
                    partment of the Interior.
                      This Project Summary was developed
                    for EPA's Robert S. Kerr Environmental
                    Research Laboratory, Ada, OK. to
                    announce key findings of the research
                    project that is fully documented in a
separate report of the same title (see
Project Report ordering information at

  The use of plants in wastewater
treatment has most often been limited to
slow rate land treatment and the use of
nuisance plants  of limited value in
hydroponic systems. A simple advanced
hydroponic system referred to as the
"Nutrient Film Technique" (NFT) enables
all species of plants to be considered for
use in water pollution control systems.
  There is increasing interest in the use
of low-cost natural systems that have
existing or built-in pollution control
mechanisms. This  project focuses on a
new plant production system that could
lead to new ways of treating wastewater
and water supplies with a solar-powered
pollution control system that could have a
number  of useful and valuable  by-
products. The main goal of this study was
  determine the feasibility of using plants
  as pollutant-concentrating, pollutant
  assimilating, and nutrient-recycling
  facilities in a unique hydroponic system
  utilizing the Nutrient Film Technique.
  The five specific objectives were to:
  1.  define advantages of using plants
     in  the Nutrient Film Technique
     system over conventional systems
     to collect, concentrate, and assimi-
     late pollutants;
  2.  identify research and development
     needs to support a  long-term
     program to define the full potential
     of the Nutrient Film Technique
     system as a pollutant management/
     resource recovery system;
  3.  define the engineering  require-

      ments to establish a total plant-
      based pollutant control system.
      This will include area requirements,
      hardware, plants and plant produc-
      tion system;
  4.  gather  baseline  data on system
      performance using energy and
      mass balance to support econo-
      metric data; and
  5.  determine the feasibility  of re-
      covering  energy, food,  and nu-
      trients from the plant material.
  Design parameters  were developed
based on hydraulics  and removal  of
nutrients, organics, trace organics and
cadmium. Plant selection and  manage-
ment methods were investigated and are
described. Finally,  the potential evapo-
transpiration (ET) and actual ET and
recovery of ET are discussed.

NFT Basic Concept
  The NFT version of hydroponics utilizes
plants grown directly on an impermeable
surface to which a thin film of water is
continuously applied (see Figure 1). The
root  production on this impermeable
surface will result in a large mass of roots
and  accumulated  matter with a large
surface  area. Root masses  have been
observed to accumulate up to 30 cm deep,
separate from the stalk and fruit. Virtually
all plants tested have been found to grow
well  in this system. The hypothesis here
is that  these  large masses of self-
generating root systems can be used as
living filters.  Plant top growth  also
provides nutrient uptake, shade for
       Grow Block
protection from algae, and water removal
in the form of transpiration. Sludge that
would settle in  the root filter would be
held in place by the roots, and the filter
itself would gradually expand as the
sludge accumulated and occupied more
space.  After removal of organics and
suspended solids, the remaining refrac-
tory soluble organics,  nutrients, and
remaining toxic elements would continue
to pass through  the fine root filter. Since
high flow  rates  are possible  in the
absence of  large amounts of suspended
and organic matter, subsequent loading
rates would be related to the sorption
rates of critical wastewater constituents
such as nutrients. This general interaction
of pollutants, plants and water lead to the
following hypothesized wastewater treat-
ment system.
  The  hypothesized  system would be
composed of three distinct sections, with
plant characteristics dependent upon the
pollutant removals required:
  1.  Roughing or preliminary treatment:
      plant species with   large root
     systems capable of surviving and
     growing  in  a grossly polluted
     condition. Large sludge accumula-
     tions, anaerobic conditions, and
     trace metal precipitation and en-
     trapment  would characterize this
     section.  A  large  portion of the
      BOD5 and SS would be removed in
     this section.
  2.  Nutrient conversion and recovery: an
      active nutrient uptake, high bio-
      mass and/or food production
  Nutrient System
 (Continuous Flow)
                  ~     s  _-<    f    S "*^y^^ f   f   *^

                     r    I     *    /    f    /"~J^ Capillary Pad
                   Purified * Water         \     T
Figure 1.   The nutrient film technique variation of hydroponic plant production systems.

      section would follow the  first
      section. The major interaction here
      would be nutrient conversions, but
      suspended solids and trace organic
      removals would improve.
  3.  Water polishing: the third section
      would be a polishing section that
      would  necessarily have nutrient-
      limited plant production, depending
      on the  required  effluent water
  A schematic of the three-plant series
wastewater treatment system, showing
the major pollution control functions and
the by-products produced in each section,
is presented in Figure 2.
  Obviously,  the three  modules in the
NFT treatment system can also be  used
separately for different levels of treatment
of varying input water quality.

Conclusions and
  The Nutrient Film Technique (NFT) has
been shown to be a viable alternative for
domestic sewage treatment in this 3-year
multidisciplinary effort. Pilot scale units
up to 36 meters long have been operated
continuously  with domestic  sewage  at
flow rates of up to 11,000 l/d (3000 gpd)
in New York and New Hampshire. Data
presented  here should  be considered
conservative since most experiments
were conducted under "worst case"
climate and temperature conditions. No
attempt was  made to control the plant
environments during most of the testing.
  The general approach of this study was
to choose one plant species to  demon-
strate the concept, and then to test a wide
range of species under  conditions that
would lead to the definition of  process
controlling parameters. Reed  canary
grass was selected  as  the main  test
species that could accomplish all phases
of treatment (i.e.,  roughing treatment
through  nutrient polishing). Wetland
plants and commercially valuable plants,
such as ornamental roses, were ultimately
tested. Reed canary grass grown in the
NFT system from seed and from trans-
planted sod resulted in the production of
better than secondary effluent quality at
an application of 10 cm/day of settled
domestic sewage and synthetic waste-
water applications throughout the year.
Application rates at 20 cm/day  were
found to result in the  destruction of the
reed canary grass. Attempts to optimize
multiple species systems that could exist
in low dissolved oxygen environments
eliminated the 20 cm/d limitation.
  Example data for a 36-meter long unit
for  synthetic  wastewater are shown in
Figure 3.  Note that the higher loading


Lagoon Preapplication Treatment
Equalization &
Emergency Storage

Preliminary Treatment


Plant Series 1

Plant Products
Stable ^
X Sludge &
Material \
	 Plant Series II
Suspended Solids Nitrogen and
and Organic Phosphorus and
Removal Potassium Removal
Plant Products
Plant Series III
Polishing of
Soluble Organics
Suspended Solids.
Nutrients and
Toxic Elements
Water Rec.
Figure 2.    Schematic diagram showing three-stage hypothesized NFTtreatment system for domestic wastewater with functions for each stage and
            fate of all materials.
rates tested are equivalent to a system
area required to treat sewage generated by
10,000 people to secondary level of less
than 2 ha.
  The suspended solids removal capability
was one of the most efficient character-
istics  of the NFT. Even  at high loading
rates the turbidity of the effluents were
low and the suspended  solids less than
  After some of the removal mechanisms
were defined, a small optimized unit was
constructed and operated for a short time
on  raw domestic sewage (unsettled
sewage). Optimized operation  included
rapid batch addition of  the sewage
followed by slow withdrawal and a brief
resting phase to encourage aeration. The
following  represents  typical  results
obtained at a loading rate of 30 cm/d
(equivalent to a land area requirement of
1.3 ha to treat the sewage from 10,000
Total COD, mg/l
TSS, mg/l
TKN. mg/l
Removal Mechanisms

  This study was able to define some of
the general pollutant removal mechan-
isms. By accumulating large quantities
of biomass in the form of fine roots, the
possibility of removing pollutants is
greatly enhanced. The solids entrapped in
the roots provide the largest capability of
      400 -
      300 -
      200 -
      700 -
                                                         Condition 3   Condition 4
                                         Figure 3.    Influent and effluent chemical oxygen demand for the  Cornell NFT treating
                                                    synthetic wastewater in spring, 1981. Conditions correspond to areal loadings of
                                                    6.9. 10.2, 20.3. and 40.6 centimeters per day, respectively, for a 36 m long unit.
this treatment system. Measurements of
the solids within the roots indicated that
over 1,000 gm/m2 of  entrapped solids
were accumulated. These solids represent
significant potential for manipulation of
pollutant  cycles. If, for  example, this
biological organic material can be utilized
for  pollutant removal,  then the solids
retention  time becomes important. In
such operations, reasonably high loading
rates result in solids retention times of
greater than 100 days. This indicates that
the process could be stable and provides
an efficient treatment system.
  Organic  removals  are  limited  by the
level  of aeration. The capillary  mat of
dense root systems  may significantly
increase aeration. The use of plants such
as cattails that translocate  air to their
roots  could also be a significant factor.
Additional studies are needed  in this
  The fine suspended solids were found
to biocoagulate in the system, and high

clarity effluent was achieved under high
loading rates.
  A  major goal of this study was to
manipulate the nitrogen cycle using both
plants  and microbial interactions via
nitrrfication-denitrification.   Nitrification
was rarely observed during the study and
produced only 2 or 3 mg/l of nitrate-
nitrogen  under the lowest loading
condition during the warmer time periods.
This lack of nitrification  resulted in
limited capability for control  of the
nitrogen cycle with the NFT.
  Phosphorous  removal appeared to be
limited by the plant nutrition requirements.
Due to the stress condition of most of the
testing that occurred, phosphorous
removal  rates  were  low  in the NFT
  The energy that is contained in domestic
sewage is likely to provide enough heat to
maintain plant cultures in the NFT in most
areas of the  U.S. where low cost solar
greenhouses  can be utilized. In some
locations the greenhouse  cover of the
NFT system may be minimized due to this
available energy.
  The kinetics of pollutant removal was
examined using several different ap-
proaches. Due  to the large number of
variables,  no  comprehensive model was
established  that  would  predict the
process efficiency. The following  rates
indicate the range of observed nutrient
and organic removals that were obtained
in the system:
                         Removal Rate
Parameter                   kg/ha-d
  Both  heavy metal and organic toxic
materials were examined in this system.
Cadmium was added to the synthetic
wastewater tested. Trace organics were
added to the domestic sewage in the New
Hampshire test facility. It is proposed that
the removal mechanisms for the metals
and the organics were the extensive root
surface area and the large accumulated
biomass. Between 91 and 98 percent of
the following trace organics were removed:
chloroform, tetrachlorethylene,  benzene,
toluene, trichloroethylene, xylene, bro-
moform, M-nitrotoluene, PCB 1248.

Kinetic Analysis

  An attempt was  made to define the
relationship of hydraulic loading rates
and hydraulic retention time within the
root zone. Due to the large number of
factors that affected the plants and the
flow through the system,  only general
comments can be made.
  The most promising empirical approach
provided by the kinetic analysis of data
indicated that the loading  rate/effluent
quality relationships would hold over a
limited range for domestic sewage. These
relationships will be of limited value in
designing the  NFT system for other
  This study attempted to conduct
parallel testing with  actual domestic
sewage and a synthetic sewage. Parallel
comparison  of these systems indicated
that it was possible to simulate the
domestic sewage systems with the syn-
thetic wastewaters.  Since  the synthetic
sewage contained  soluble substrates
only, it was not possible tosimulate solids
behavior or interactions with  the syn-
thetic sewage.

Evapotranspiration Water

  The water balance was established for
a number of bench scale and pilot tests.
Although in many  circumstances the
expected loss of water through evapo-
transpiration was equal  to  literature
values of approximately 50,000 l/ha-day,
several test conditions resulted in water
losses up to 100,000  l/ha-day.  These
exceptionally high values of evapotrans-
piration represent a potential source of
high quality water.
  The energy and process requirements to
recover evapotranspired water from the
greenhouse were  examined. Energy
costs of ET recovery are extremely high
and prohibit this alternative unless a low
cost energy source is available.  The
potential alternative that was identified
was the use of the temperature differen-
tial between the influent wastewater and
the greenhouse air.
            NFT Plants for Sewage

              Although reed canary  grass was
            examined in the majority of tests in this
            study, a  wide  range of plants was
            cultured and propagated to examine their
            viability in the NFT system when applied
            to wastewater treatment. The following
            summarizes those plants that grew well
            and those that were  less acceptable
            under adverse conditions:
                                                    that Flourished
                  with Marginal Growth
                                                     Japanese millet
                                                     Napier grass
                      Bristly sedge
                      Reed canary grass
                      Soft rush
  The above list of plants includes some
that  have high  monetary value.  Reed
canary grass, when grown under relatively
low  nutrient conditions, resulted in
biomass with a total nitrogen content of 5
percent with most values greater than 4
percent. This indicates  a  total  protein
content of greater than 30 percent of the
total dry weight in many cases and a plant
that would have a significant animal food
value. Plants with ornamental value such
as shrubs,  trees,  and roses would
represent plants with significant com-
mercial value. Other plants that showed
promise in this study were several food
plants that would be useful for propagation
purposes. Certain plants with a high
value,  such  as berry plants, could be
cultured  in  this system. The potential
carry-over of toxic materials would be a
concern with food products.
  The growth of plants in sewage, where
the organic and suspended solids loading
rates are high, results  in an optimum
condition for inhibition of plant yield and
pest invasions. The application of hearty
plants  such  as  cattails  in the roughing
sections of the NFT eliminated most of the
plant pathogen problems that were
observed; however,  fungus invasions,
insect attack on many of the plants, and
other  problems of greenhouse  plant
production were common. All the prob-
lems were controlled when advice from
specialists was sought and implemented.
The  destruction  of  a total treatment
system by pests or  by contaminants in
wastewater is  a concern that needs
careful evaluation  prior to  full scale

Feasibility Considerations
  This study outlines a new approach to
the use of plants as solar-powered water
and  wastewater treatment devices.
  The  area requirements and necessity
for a greenhouse cover will vary depending
on the location of the system.  An NFT
secondary treatment system is shown in
Figure 4. The  area  needs appear to be
approximately  2.7  ha (6.6 acres) for a
wastewater  flow from  a population  of
10,000 people. This would be divided into

two sections. The first section would be a
roughing section of approximately one
hectare. It  would  be  subdivided into
relatively small sections so  that the
wastewater  could be rapidly introduced,
allowed to remain in a quiescent state for
approximately 1 hour, and then removed
slowly so that  laminar  flow conditions
would occur. The second section would
be a  polishing  and nutrient removal
section  of  about  2 ha. The capital
investment of such  a system appears to
be attractive, and energy requirements
would be  low  since  only low  head
pumping would  be  required in such  a
system. Although a detailed economic
analysis was not completed in this study,
it appears  that  the capital  investment
would be less  than conventional  sec-
ondary treatment alternatives.
  The application of the  NFT for nutrient
removal is subject to the limitations of all
plant systems.  Even with maximum
growth rates the nutrient removal rates
are relatively low in comparison to high
rate  nutrient manipulation processes
such  as  microbial nitrification-denitri-
fication. However, partial nutrient polishing
could be achieved in NFT units of 10 ha or
larger for flow rates produced by 10,000
  The estimates for  tertiary treatment or
water reclamation with the NFT indicate
area  requirements that may not be
competitive with a conventional  unit
process unless the plant  products have a
commercial  value. If plant management
techniques are achieved that allow plant
products to  enter commercial markets,
the use of the NFT as a water  reclamation
facility shows significant promise.


* * t

Roughing NFT 1
1 i i

0 ha') '
: : i


       Sludge to
                                      *    t   t  t  J
                                                       Low Quality Biomass
                                                                Effluent Air
                                                               Scrubbed for
                                                               Odor Removal
                                     NFT for
                                   Treatment and
                                Nutrient Conversion
                                       2 ha
                                (Area Available for
                            Commercial Plant Production
                               Approximately 1 ha)
           -^  Useful Plant
           {Ornamental, Chemical,
               or Animal Feed)
Figure 4.  Schematic of NFT treatment facility capable of treating domestic sewage from
          10,000 people.
   W. J. Jewell, J. J. Madras, W. W. Clarkson, H. DeLancey-Pompe, and P. M. Kabrick
    are with Cornell University, Ithaca, NY 14853.
   William R. Duffer andJ. L. Wltherow are the EPA Project Officers (see below).
   The complete report, entitled "Wastewater Treatment with Plants in Nutrient
    Films," (Order No. PB 83-247 494; Cost: $44.50, subject to change) will be
    available only from:
          National Technical Information Service
          5285 Port Royal Road
          Springfield, VA 22161
          Telephone: 703-487-4650
   The EPA Project Officers can be contacted at:
          Robert S. Kerr Environmental Research Laboratory
          U.S. Environmental Protection Agency
          Ada. OK 74820

United States
Environmental Protection
Center for Environmental Research
Cincinnati OH 45268
Official Business
Penalty for Private Use $300
                                                                                                  * U.S. GOVERNMENT PRINTING OFFICE: 1983-759-102/0770