United States
Environmental Protection
Agency
Technology Transfer
Office of Raiaareh '.. -:
arid Development:
Washington DC 20460
.Center for Environmental]
Research Information
'Cincinnati OH 45268
EPA/625/1 -88/022
Design
Manual
Constructed
Wetlands and Aquatic
Plant Systems for
Municipal W^stewater
Treatment
-------
EPA/625/1-88/022
September 1988
Design Manual
Constructed Wetlands and Aquatic
Plant Systems for Municipal
Wastewater Treatment
U.S. Environmental Protection Agency
Office of Research and Development
Center for Environmental Research Information
Cincinnati, OH 45268
-------
Notice
This document has been reviewed in accordance with the U.S. Environmental Protection
Agency's peer and administrative review policies and approved for publication. Mention of trade
names or commercial products does not constitute endorsement or recommendation for use.
This document is not intended to be a guidance or support document for a specific regulatory
program. Guidance documents are available from EPA and must be consulted to address
specific regulatory issues.
-------
Contents
Chapter Pa9e
1 AQUATIC TREATMENT SYSTEMS 1
1.1 Introduction 1
1.1.1 Scope 1
1.1.2 Potential Uses of Natural Systems
1.2 Classification ^
1.2.1 Natural Wetlands 2
1.2.2 Constructed Wetlands 3
1.2.3 Aquatic Plant Systems *
1.3 Natural Wetlands 4
1.4 Constructed Wetlands ^
1.4.1 Free Water Surface Systems (FWS) £
1.4.2 Subsurface Flow Systems (SFS) 5
1.5 Aquatic Plant Systems 5
1.5.1 Floating Plant Systems ^
1.5.2 Submerged Plant Systems 6
1.6 References '
2 ENVIRONMENTAL AND PUBLIC HEALTH CONSIDERATIONS 9
2.1 Introduction 9
2.2 Nitrogen 9
2.3 Phosphorus 10
2.4 Pathogens 10
2.4.1 Parasites 10
2.4.2 Bacteria 10
2.4.3 Viruses 11
2.5 Metals 11
2.6 Trace Organics 11
2.6 References 12
3 DESIGN OF CONSTRUCTED WETLANDS 15
3.1 Types of Constructed Wetlands 15
3.1.1 Free Water Surface Systems with Emergent Plants 15
3.1.2 Subsurface Flow Systems with Emergent Plants 15
3.2 Site Selection 15
3.2.1 Topography 15
3.2.2 Soil Permeability for Free Water Surface Systems 15
3.2.3 Hydrological Factors 16
3.2.4 Water Rights Considerations 16
3.3 Performance Expectations 16
3.3.1 BOD5 Removal in FWS Wetlands 18
3.3.2 BOD5 Removal in SFS Wetlands 19
-------
Contents (continued)
Chapter Page
3.3.3 Suspended Solids Removal 22
3.3.4 Nitrogen Removal 22
3.3.5 Phosphorus Removal 22
3.3.6 Metals Removal 23
3.4 Process Variables 23
3.4.1 Design Objectives 23
3.4.2 BODs Loading Rates 24
3.4.3 Hydraulic Loading Rates 24
3.4.4 Water Depth in FWS Systems 24
3.4.5 Detention Time 25
3.5 Pre-Application Treatment 25
3.6 Vegetation 25
3.6.1 Cattails 25
3.6.2 Bulrushess 25
3.6.3 Reeds 25
3.7 Physical Design Factors 26
3.7.1 System Configurationss 26
3.7.2 Distribution System 26
3.7.3 Outlet Structures 26
3.7.4 Vector Control in Free Water Surface Wetlands 26
3.7.5 Harvesting Vegetation 26
3.8 Case Studies 26
3.8.1 Arcata, California 27
3.8.2 Emmitsburg, Maryland 33
3.8.3 Gustine, California 34
3.8.4 Fabius Coal Preparation Facility 39
3.8.5 Summary 44
3.9 References 45
4 DESIGN OF AQUATIC PLANT SYSTEMS 47
4.1 Background 47
4.1.1 Characteristics of Aquatic Treatment Systems 47
4.1.2 History 47
4.1.3 Climatic Constraints 48
4.2 Vegetation 48
4.2.1 Floating Plants 48
4.2.2 Submerged Plants 51
4.3 Process Design Criteria for Water Hyacinth Systems 52
4.3.1 Organic Loading Rates 52
4.3.2 Hydraulic Loading Rates 52
4.3.3 Water Depth 53
4.3.4 Vegetation Management 53
4.3.5 Mosquitoes and Their Control 54
4.3.6 Suggested Design Parameters 54
4.3.7 Sludge Management 54
4.4 Physical Features of Aquatic Treatment Systems 54
4.4.1 System Configurations 54
4.4.2 Inlet and Outlet Structures 55
4.4.3 Supplemental Aeration 55
4.4.4 Operation and Management of Aeration 55
-------
Figures
Number Page
1-1 Common Aquatic Plants 3
3-1 General Design Schemes for Denitrification 16
3-2 Pilot-Scale Constructed Wetland - Gravel Planted Trench with Brush 18
3-3 Sensitivity of Ce/C0 Ratio to Av 20
3-4 Sensitivity of Ce/C0 Ratio to Temperature 20
3-5 Regression Curve of TKN vs. Detention Time in the Effluent of an
Alternating Typa/Open-Water/Gravel System 23
3-6 Arcata, CA Pilot Marsh System 28
3-7 Arcata, CA Wastewater Treatment Facilities Flow Diagram 31
3-8 Arcata, CA Intermediate FWS System 32
3-9 BOD Performance Data for Emmitsburg, MD SFS 35
3-10 TSS Performance Data for Emmitsburg, MD SFS 35
3-11 BOD5 Performance Data for Gustine, CA Pilot Marsh System 37
3-12 SS Performance Data for Gustine, CA Pilot Marsh System 37
3-13 Gustine, CA Marsh System Flow Schematic 40
3-14 Fabius Coal Facility Site Plan 42
3-15 Fabius Coal Facility Impoundment 1 Wetlands 43
4-1 Morphology of the Hyacinth Plant 50
4-2 Suitable Areas for Hyacinth Systems 50
4-3 Morphology of and Potential Growth Areas for Duckweed Plants 51
4-4 Evolution of Flow pattern Through San Diego, CA Water Hyacinth
Treatment Ponds 56
4-5 Evolution of Pond 3 Flow and Aeration System Configurations at
San Diego,CA 57
4-6 Site Plan for San Diego, CA Aquaculture Pilot Plant 63
4-7 Schematic Diagram of Primary and Secondary Facilities - San Diego, CA
Aquaculture Pilot Plant 64
4-8 Schematic of Hyacinth Pond Step-Feed System With Recycle - San Diego, CA 65
4-9 BOD5 Performance Data for San Diego, CA Pond #3 With 200 Percent Recycle 66
4-10 SS Performance Data for San Diego, CA Pond #3 With 200 Percent Recycle ... 66
4-11 Influent and Effluent BOD, SS, and DO for Step-Feed Hyacinth
Pond - San Diego, CA
4"12 Definition Sketch for the Analysis of a Hyacinth Pond With Step-Feed 68
4 ,.3 and Recycle 70
Analysis of Performance Data for Hyacith Pond 3, San Diego, CA, with
4 *» Step-Feed and Recycle 70
4 15 Hornsby Bend, TX Hyacinth Facility Basin Configuration 72
/,,- Hornsby Bend, TX Hyacinth Facility Pond and Roof Section 72
Iron Bridge, FL Hyacinth Facility Basin Configuration 76
VI
-------
Contents (continued)
Chapter Page
4.5 Performance Expectations 56
4.5.1 Design Equations 56
4.5.2 Nitrogen Removal 56
4.6 Sample Design Problems 58
4.6.1 Sample Problem No. 1 58
4.6.2 Sample Problem No. 2 59
4.7CaseStudies 61
4.7.1 San Diego, California 61
4.7.2 Austin, Texas 69
4.7.3 Orlando, Florida 74
4.7.4 Summary 77
4.8 References 78
APPENDIXA 81
APPENDIXB 83
-------
Tables
Number Page
1-1 Functions of Aquatic Plants In Aquatic Treatment Systems 2
1-2 Percent Removal for Several Pollutants from Secondary Effluent in
Natural Wetlands 4
1-3 Summary of Nutrient Removal from Natural Wetlands 5
1-4 Summary of Nutrient Removal from Constructed Wetlands 5
1-5 Summary of Wastewater Treatment Performance of Aquatic Plant Systems 6
2-1 Pollutants and Pathways of Concern 9
2-2 Trace Organic Removal in Pilot-Scale Hyacinth Basins 12
3-1 Removal Mechanisms in Wetlands for the Contaminants in Wastewater 17
3-2 Performance of Pilot-Scale Constructed Wetland Systems 17
3-3 Predicted vs. Actual Ce/Co Values for Constructed Wetlands 19
3-4 Media Characteristics for Subsurface Flow Systems 21
3-5 Emergent Aquatic Plants for Wastewater Treatment 26
3-6 City of Arcata, CA Wastewater Discharge Requirements 27
3-7 Arcata, CA Pilot Marsh System Hydraulic Loading Ratios and Detention Times . . 29
3-8 Experimental Vegetation and Compartments for Marsh Cells - Arcata, CA 29
3-9 Average Annual BOD5 Concentration (mg/L) Arcata, CA 30
3-10 Arcata, CA Marsh and Wildlife Sanctuary Wastewater Treatment
Plant Performance 33
3-11 Arcata, CA Marsh and Wildlife Sanctuary Project Expenditures 33
3-12 Performance of the Emmitsburg, MD SFS 35
3-13 Determination of Nitrification Component in BOD5Test Gustine, CA 38
3-14 BOD5 and SS Removal Efficiencies As a Function of Detention
Time - Gustine, CA 38
3 15
3_16 Design Criteria for Constructed Wetland at Gustine, CA 39
2_,|7 Initial Operating Schedule of the Gustine, CA Marsh System 41
3_1g Capital Costs for Gustine, CA Marsh Project 41
2_^g Fabius Coal Preparation Facility Marsh System Performance 45
Constructed Wetlands Case Studies Summary 45
4-1
4 2 History of Use of Floating Aquatic Treatment Systems 48
4_2 Performance of Existing Duckweed Systems 52
4_4 Composition of Duckweeds Grown in Wastewater 52
4 c Types of Water Hyacinth Systems 53
4_g Design Criteria for Water Hyacinth Systems 53
4_7 Design Criteria for Effluent Polishing With Duckweed Treatment Systems 54
4_o Recommended Sludge Cleanout Frequency for Water Hyacinth Ponds 54
4_g Nitrogen Removal - Water Hyacinth Tertiary Treatment 58
Design Criteria for Modified Plug-Flow Water Hyacinth Ponds for
4_,|Q Expanded San Diego, CA Aquatic Treatment Facility 67
4 A A Performance Data - Hornsby Bend, TX Hyacinth Facility 73
4_12 Iron Bridge, FL Water Hyacinth System Performance Summary 78
4_13 Iron bridge, FL Water Hyacinth System Monitoring 79
Aquatic Plant Systems Case Studies Summary 80
VII
-------
Acknowledgments
Many individuals contributed to the preparation and review of this manual. Contract
administration was provided by the U.S. Environmental Protection Agency (EPA), Center for
Environmental Research Information (CERI), Cincinnati, Ohio.
Authors:
Ronald W. Crites (Project Manager) - Nolte and Assocoiates, Sacramento, California
Daniel C. Gunther - Nolte and Associates
Andrew P. Kruzic - Nolte and Associates
Jeffrey D. Pelz - Nolte and Associates
George Tchobanoglous - University of California, Davis, California (In-house review)
Contributors and Reviewers:
James F. Kreissl - EPA-Risk Reduction Engineering Laboratory, Cincinnati, Ohio
Sherwood C. Reed - U.S. Army Corps of Engineers, Hanover, New Hampshire
Reviewers:
Lowell L. Leach - EPA-Robert S. Kerr Environmental Research Laboratory, Ada, Oklahoma
John Meagher - EPA-Office of Wetlands Protection, Washington, DC
William Sipple - EPA-Office of Wetlands Protection, Washington, DC
Technical Direction/Coordination:
Denis J. Lussier - EPA-CERI, Cincinnati, Ohio
James E. Smith, Jr. - EPA-CERI, Cincinnati, Ohio
-------
CHAPTER 1
Aquatic Treatment System
1 .1 Introduction
The trend over the past 70 years in the construction
of water pollution control facilities for metropolitan
areas has been toward "concrete and steel"
alternatives. With the advent of higher energy prices
and higher labor costs, these systems have become
significant cost items for the communities that
operate them. For small communities in particular,
this cost represents a higher percentage of the
budget than historically allocated to water pollution
control. Processes that use relatively more land and
are lower in energy use and labor costs are therefore
becoming attractive alternatives for these
communities.
The high cost of some conventional treatment
processes has produced economic pressures and
has caused engineers to search for creative, cost-
effective and environmentally sound ways to control
water pollution.
One technical approach is to construct artificial
ecosystems as a functional part of wastewater
treatment. Wastewater has been treated and reused
successfully as a water and nutrient resource in
agriculture, silviculture, aquaculture, golf course and
green belt irrigation. The conceptual change that has
allowed these innovative processes is to approach
wastewater treatment as "water pollution control" with
the production of useful resources (water and plant
nutrients) rather than as a liability.
The interest in aquatic wastewater treatment systems
can be attributed to three basic factors:
1. Recognition of the natural treatment functions of
aquatic plant systems and wetlands, particularly as
nutrient sinks and buffering zones.
2. In the case of wetlands, emerging or renewed
application of aesthetic, wildlife, and other
incidental environmental benefits associated with
the preservation and enhancement of wetlands.
3. Rapidly escalating costs of construction and
operation associated with conventional treatment
facilities.
7.7.7 Scope
Application of wastewater to wetlands and aquatic
pond systems must be free of unreasonable risks to
public health. Pathogenic organisms may be present
in both wastewaters and sludges and their control is
one of the fundamental reasons for waste
management. Public health considerations of aquatic
plant systems and constructed wetlands are
discussed in Chapter 2.
The portion of this manual concerning constructed
wetlands (Chapter 3) focuses on studies of pilot-
and full-scale systems that have published results.
The general case in favor of constructed wetland
systems is tied to the fact that they can operate in
cold as well as warm climates.
The discussion of aquatic plant systems (Chapter 4)
concentrates on the results with water hyacinth
systems operated in the warm southern regions of the
United States. A few duckweed systems have been
tried either alone or in conjunction with hyacinths.
The projects discussed in this manual reflect this
geographical distribution of project sites and of the
plant species that have been studied extensively.
A list of existing constructed wetlands and aquatic
plant systems is presented in Appendix A.
7.7.2 Potential Uses of Natural Systems
Where natural wetlands are located conveniently to
municipalities, the major cost of implementing a
discharge system is for pumping treatment plant
effluent to the site. Once there, further wastewater
treatment occurs by the application of natural
processes. In some cases, the wetland alternative
can be the least cost advanced wastewater treatment
and disposal alternative. In locations where poorly
drained land that is unsuitable for land application is
available, wetlands can often be constructed
inexpensively with minimal diking.
In considering the application of wastewaters to
wetlands, the relationship between hydrology and
ecosystem characteristics needs to be recognized.
Factors such as source of water, velocity, flow rate,
renewal rate, and frequency of inundation have a
major bearing on the chemical and physical properties
of the wetland substrate. These properties in turn
-------
influence the character and health of the ecosystem,
as reflected by species composition and richness,
primary productivity, organic deposition and flux, and
nutrient cycling (1). In general, water movement
through wetlands tends to have a positive impact on
the ecosystem (2). Rather than wasting water, upland
swamps appear to save water and thus promote
increased regional production indirectly (3).
1.2 Classification
In aquatic systems, wastewater is treated principally
by means of bacterial metabolism and physical
sedimentation, as is the case in conventional
activated sludge and trickling filter systems. The
aquatic plants themselves bring about little actual
treatment of the wastewater (4). Their function is
generally to support components of the aquatic
environment that improve the wastewater treatment
capability and/or reliability of that environment (5).
Some specific functions of aquatic plants in aquatic
treatment systems are summarized in Table 1-1
The morphology of some typical aquatic plants is
shown schematically in Figure 1-1.
Table 1-1. Functions of Aquatic Plants In Aquatic
Treatment Systems (8)
Plant Parts
Function
Roots and/or stems in the water
column
Stems and/or leaves at or
above the water surface
1. Surfaces on which bacteria
grow
2. Media for filtration and
adsorption of solids
1. Attenuate sunlight and thus
can prevent the growth of
algae
2. Reduce the effects of wind
on the water, i.e., the
transfer of gases between
the atmosphere and water
3. Important in the transfer of
gases to and from the
submerged parts of plant.
Wetlands are those areas that are inundated or
saturated by surface or ground water at a frequency
and duration sufficient to maintain saturated
conditions. These can be either preexisting natural
wetlands (e.g. marshes, swamps, bogs, cypress
domes and strands, etc.) or constructed wetland
systems. Constructed systems can range from
creation of a marsh in a natural setting where one did
not permanently exist before to intensive construction
involving earth moving, grading, impermeable barriers
or erection of containers such as tanks or trenches.
The vegetation that is introduced or emerges from
these constructed systems will generally be similar to
that found in the natural wetlands (6).
There are three basic functions of wetlands that make
them potentially attractive for wastewater treatment
(7):
1. Physical entrapment of pollutants through sorption
in the surface soils and organic litter.
2. Utilization and transformation of elements by
microorganisms.
3. Low energy and low maintenance requirements to
attain consistent treatment levels.
Wetlands are comparatively shallow (typically less
than 0.6 m (2 ft)) bodies of slow-moving water in
which dense stands of water tolerant plants such as
cattails, bulrushes, or reeds are grown. In man-
made systems, these bodies are artificially created
and are typically long, narrow trenches or channels
(8).
Three major systems involving wastewater and
wetlands can be observed in the United States (9).
1. Disposal of treated effluent into natural wetlands
2. Use of effluents or partially treated wastewater for
enhancement, restoration, or creation of wetlands
3. Use of constructed wetlands for wastewater
treatment
These three categories provide some degree of
wastewater treatment, either directly or indirectly. In
the United States, however, there are some
constraints on the use of natural wetlands as
functional components of wastewater treatment
systems.
Almost all natural wetlands are waters of the United
States and, as such, a permit is required for any
discharge. The water quality requirements for this
discharge are specified by the applicable federal,
state, and/or local agencies and typically are at least
equal to secondary effluent standards. 10)
On the other hand, constructed wetlands designed
and built for the express purpose of treating municipal
wastewater are not waters of the United States.
There are three categories of aquatic treatment
systems considered in this manual:
1. Natural Wetlands
2. Constructed Wetlands
3. Aquatic Plant Systems
7.2.1 Natural Wetlands
While the interest in wetlands for wastewater
treatment is fairly recent, the term wetlands is also a
relatively new expression, encompassing what for
years have simply been referred to as marshes,
swamps, or bogs. The difference in these wetlands is
related to a large extent to the vegetation which
dominates the area. Grasses or forbs are generally
-------
Figure l-l. Common aquatic plants.
CATTAIL
WATER SUBMERGED DUCK
HYACINTH PLANTS WEED
dominant in marshes, trees and shrubs characterize
swamps, and sedge/peat vegetation occurs in various
bogs.
Natural wetlands are effective as wastewater
treatment processes for a number of reasons. Natural
wetlands support a large and diverse population of
bacteria which grow on the submerged roots and
stems of aquatic plants and are of particular
importance in the removal of BOD5 from wastewater.
In addition, the quiescent water conditions of a
wetland are conducive to the sedimentation of
wastewater solids. Other aspects of wetlands that
facilitate wastewater treatment are the
adsorption/filtration potential of the aquatic plants'
roots and stems, the ion exchange/adsorption
capacity of wetlands' natural sediments, and the
mitigating effect that the plants themselves have on
climatic forces such as wind, sunlight and
temperature (9).
Natural wetland systems are typically characterized
by emergent aquatic vegetation such as cattails
(Typha), rushes (Scirpus), and reeds (Phragmites).
They can also contain some of the floating and
submerged plant species discussed in Chapter 4 as
well as phreatophytes (plants whose roots extend to
the ground-water table or the saturated soil area
immediately above it) (10). Most states (except
Florida, and a few others considering special wetland
standards) make no distinction between the wetland
and the adjacent surface waters and apply the same
requirements to both. Under these conditions,
economics will not favor the utilization of natural
wetlands as a major component in a wastewater
treatment process as the basic treatment must be
provided prior to discharge to the wetland.
Special situations may arise in which natural wetlands
may provide further effluent polishing or, if the
wetland is isolated from other surface waters, more
basic treatment. The use of treated effluent for
enhancement, restoration, or creation of wetlands can
be a very desirable and environmentally compatible
activity (10).
7.2.2 Constructed Wetlands
Studies in the United States have focused on
peatlands, bogs, cypress domes and strands, as well
as cattails, reeds, rushes, and related plants in
wetland settings (6). Constructed wetlands are either
free water surface systems (FWS) with shallow water
depths or subsurface flow systems (SFS) with water
flowing laterally through the sand or gravel. A
constructed wetland involving bulrushes in gravel
filled trenches was developed at the Max Planck
Institute in West Germany. This patented process has
seen limited application to date in the United States.
-------
The constructed wetlands at Santee, California, was
operated in a similar fashion.
1.2.3 Aquatic Plant Systems
Aquatic plant systems are shallow ponds with floating
or submerged aquatic plants. The most thoroughly
studied systems are those which use water hyacinth
or duckweed. These systems include two types
based on the dominant plant types. The first type
uses floating plants and is distinguished by the ability
of these plants to derive their carbon-dioxide and
oxygen needs from the atmosphere directly. The
plants receive their mineral nutrients from the water.
The second type of system consists of submerged
plants and is distinguished by the ability of these
plants to absorb oxygen, carbon-dioxide, and
minerals from the water column. Submerged plants
are relatively easily inhibited by high turbidity in the
water because their photosynthetic parts are below
the water.
1.3 Natural Wetlands
Examples of pollutant removal in natural wetlands
receiving treated wastewater are presented in Table
1-2. The values for percent removal show quite a
range for treatment. This summary table is included
to indicate the general finding for natural wetlands
systems, i.e., that levels of removal for BOD5 and SS
can be high but are not consistently high. Nutrient
removals from several specific natural wetlands
projects are presented in Table 1-3 (11).
Table 1-2. Percent Removal for Several Pollutants from
Secondary Effluent in Natural Wetlands (6)
Pollutant Removal, percent
BOD5 70-96
Suspended Solids 60-90
Nitrogen 40-90
Phosphorus Seasonal
Current experience with wetland systems is generally
limited to the further treatment of secondary effluents
(6). Factors to be considered are potential disruption
of the existing wildlife habitat and ecosystems in a
natural wetland, loss of water via evapotranspiration
for all wetlands in arid climates, the potential for
increased breeding of mosquitos or flies, and the
development of odor. The major benefits that can be
realized from use of wetlands include preservation of
open space, wildlife habitat enhancement, increased
recreation potential, streamflow stabilization and
augmentation in addition to wastewater treatment (6).
1.4 Constructed Wetlands
Constructed wetlands have the positive
characteristics of a natural wetland and can also be
controlled to eliminate the negative aspects of natural
wetlands. The removal efficiency of typical pollutants
are reported in Table 1-4.
Bacteria attached to plant stems and the humic
deposits are the major factor for BOD5 removal. With
respect to phosphorus removal, the contact
opportunities with the soil are limited in most natural
wetland systems (an exception might be peat bogs)
and release of phosphorus has been observed during
the winter in some cases. The surface area for
constructed marshes ranges from 24.6 to 39.6 m2/m3
of applied wastewater per day (23-37 ac/mgd) (6).
The major costs and energy requirements for
constructed wetlands are associated with pre-
application treatment, pumping and transmission to
the site, distribution at the site, minor earthwork, and
land costs. In addition, a constructed system may
require the installation of a barrier layer to limit
percolation to groundwater and additional containment
structures in case of flooding (6).
Possible constraints to the use of constructed
wetlands for wastewater treatment include the
following:
1. Geographical limitations of plant species, as well
as the potential that a newly introduced plant
species will become a nuisance or an agricultural
competitor.
2. Constructed wetlands that discharge to surface
water require 4 to 10 times more land area than a
conventional wastewater treatment facility. Zero-
discharge constructed wetlands require 10 to 100
times the area of conventional wastewater
treatment plants. An example of a zero-discharge
system is the Incline Village Wetlands
Enhancement Facility near Carson City, Nevada.
3. Plant biomass harvesting is constrained by high
plant moisture content and wetland configuration.
4. Some types of constructed wetlands may provide
breeding grounds for disease producing organisms
and insects and may generate odors if not properly
managed.
Constructed wetlands, however, offer the engineer
greater hydraulic control for general use and are not
restricted by many of the environmental concerns and
user conflicts associated with natural wetlands. Unlike
natural wetlands, which are confined by availability
and proximity to the wastewater source, constructed
wetlands can be built anywhere, including lands with
limited alternative uses. They also offer greater
flexibility scope for design and management options
and thus may provide superior performance and
reliability (1).
1.4.1 Free Water Surface Systems (FWS)
These systems typically consist of basins or
channels, with some sort of subsurface barrier to
-------
Table 1-3. Summary of Nutrient Removal from Natural Wetlands
Project
Brillion Marsh, Wl
Houghton Lake, Ml
Wildwood, FL
Concord, MA
Bellaire, Ml
Coots Paradise, Town of Dundas, Ontario, Canada
Whitney Mobile Park, Home Park, FL
Flow, m3/d
757
379
946
2,309
1,1 36d
-
-227
Wetland Type
Marsh
Peatland
Swamp/Marsh
Marsh
Peatland
Marsh
Cypress Dome
Percent
TDPa NH3-N
13
95 71
98
47 58
88
80
91
Reduction
N03-N
51
99°
20
TNb
-
90
64
60-70
89
a Total dissolved phosphorus.
b Total nitrogen.
c Nitrate and nitrite.
d May-November only.
Table 1-4. Summary of Nutrient Removal from Constructed Wetlands
Project
Listowel, Ontario (12)
Santee, CA (10)
Sidney, Australia (13)
Arcata, CA
Emmitsburg, MD
Gustine, CA
Flow, m3/d
17
240
1 1 ,350
132
3,785
Wetland
Tvce
FWSa
SFSb
SFS
FWS
SFS
FWS
BOD5,
Influent
56
118
33
36
62
150
mg/L
Effluent
10
30
4.6
13
18
24
SS, mg/L
Influent
111
57
57
43
30
140
Effluent
a
5.5
4.5
31
8.3
19
Percent Reduction
BOD5
82
75
86
64
71
64
SS
93
90
92
28
73
86
Hydraulic Surface
Loading Rate,
m3ha-d
907
1,543
412
1 Free Water Surface System.
1 Subsurface Flow System.
prevent seepage, soil or another suitable medium to
support the emergent vegetation, and water at a
relatively shallow depth flowing through the unit. The
shallow water depth, low flow velocity, and presence
of the plant stalks and litter regulate water flow and,
especially in long, narrow channels minimize short
circuiting.
Results from Listowel, Ontario are related in Chapter
3 to theoretical results using mathematical modeling
for BOD5 removal. The general result, shown in
Chapter 3, is that Equation 3-5 gives correct order-
of-magnitude predictions of the system response.
For greater accuracy in predicting effluent BOD5
levels for a FWS, system the coefficient of specific
surface for microbial growth must be estimated. This
coefficient is related to the surface area of the
vegetation stems and leaves in the water column.
Predicted results are not extremely sensitive to this
coefficient, as shown in Chapter 3. Water temperature
has a large influence on microbial activity and must
be known rather accurately to predict the extent of
BOD5 degradation in the constructed wetland.
1.4.2 Subsurface Flow Systems (SFS)
These systems are essentially horizontal trickling
filters when they use rock media. They have the
added component of emergent plants with extensive
root systems within the media. Systems using sand
or soil media are also used. Soil media systems
designated as the Root-Zone-Method (RZM) were
developed in West Germany.
A theoretical basis for design of a SFS is shown in
Chapter 3 (Equation 3-7). Unlike the FWS system
equation, in which the specific surface area is
important but not critical, the media porosity is critical
to predicting the required area for a given level of
treatment. Media porosity has a direct mathematical
relationship with the microbial degradation rate
constant.
The general ability of the equations shown in Chapter
3 to predict the extent of BOD5 removal should be
used in conjunction with pilot studies. The
mathematical and theoretical basis is not refined
enough to allow engineering design of a treatment
system from the equations alone.
1.5 Aquatic Plant Systems
1.51 Floating Plant Systems
The water hyacinth Eichhomia crassipes has been
studied extensively for use in improving the
wastewater effluent from oxidation ponds and as the
major component in an integrated, advanced
-------
wastewater treatment system. The major
characteristics of water hyacinths that make them an
attractive biological support media for bacteria are
their extensive root system and rapid growth rate.
The major characteristic that limits their widespread
use is their temperature sensitivity (i.e., they are
rapidly killed by winter frost conditions.) Duckweed
systems have been studied alone and as components
of water hyacinths in polyculture systems.
The major advantage of duckweeds is their lower
sensitivity to cold climates, while their major
disadvantages have been their shallow root systems
and sensitivity to wind. Several projects which have
provided valuable performance data for water
hyacinth and duckweed systems are summarized in
Table 1-5. The Orlando and San Diego projects will
be discussed in more detail in the case studies of
Chapter 4.
1.5.2 Submerged Plant Systems
Submerged plants are either suspended in the water
column or rooted in the bottom sediments. Typically,
their photosynthetic parts are in the water column.
The potential for use of submerged plants for
polishing of effluent seems at least theoretically an
attractive option. The tendency of these plants to be
shaded out by algal growths and to be killed or
severely harmed by anaerobic conditions limits their
practical usefulness.
Table 1-5.
Project
Summary of Wastewater Treatment Performance of Aquatic Plant Systems
BOD5, mg/L SS, mg/L
Flow, m /d Plant Type Influent Effluent Influent Effluent BOD5
_ . _ . .. Hydraulic Surface
Percent Reduction yLogding Rgte
SS
rri/ha-d
Orlando, FL
San Diego, CA
NSTL, MS
Austin, TX
N. Biloxi, MS (Cedar
Lake)
Disney World, FL
30,280
378
8
1,700
49
30
Water
Hyacinth
Water
Hyacinth
Duckweed
and
Penny-wart
Water
Hyacinth
Duckweed
Water
Hyacinth
4.9
160
35
42
30
200
3.1
15
5.3
12
15
26
3.8
120
47.7
40
155
50
3
20
11.5
9
12
14
37
91
85
73
50
87
21
83
76
78
92
72
2,525
590
504
140
700
300
-------
1.6 References
When an NTIS number is cited in a reference, that
reference is available from:
National Technical Information Service
5285 Port Royal Road
Springfield, VA 22161
(703) 487-4650
1. Wile, I., G. Miller, and S. Black. Design and Use
of Artificial Wetlands. In: Ecological
Considerations in Wetland Treatment of Municipal
Wastewaters, Van Nostrand Reinhold Co., NY,
pp. 26-37, 1985.
2. Hantzsche, N.N. Wet/and Systems for
Wastewater Treatment: Engineering Applications.
In: Ecological Considerations in Wet/and
Treatment of Municipal Wastewaters, Van
Nostrand Reinhold Co., NY, pp. 7-25, 1985.
3. Godfrey, P.J., E.R. Kaynor and S.
Pelczarski. Ecological Considerations in Wet/and
Treatment of Municipal Wastewaters. Van
Nostrand Reinhold Co., NY, 1985.
4. Tchobanoglous, G. Aquatic P/ant Systems for
Wastewater Treatment: Engineering
Considerations. 1987. In: Aquatic Plants for Water
Treatment and Resource Recovery. Magnolia
Publishing, Inc., Orlando, FL, pp. 27-48, 1987.
5. Stowell, R., R. Ludwig, J. Colt, and G.
Tchobanoglous. Toward the Rational Design of
Aquatic Treatment Systems. Presented at the
American Society of Civil Engineers, Spring
Convention, Portland, OR. April 14-18, 1980.
6. Reed, S., R. Bastian, W. Jewell. Engineering
Assessment of Aquaculture Systems for
Wastewater Treatment: An Overview. In:
Aquaculture Systems for Wastewater Treatment:
Seminar Proceedings and Engineering
Assessment. U.S. Environmental Protection
Agency, EPA 430/9-801006, NTIS No. PB 81-
156705, pp, 1-12, 1979.
7. Chan, E., T.A. Bursztynsky, N.N. Hatzsche, and
Y.J. Litwin. 1981. The Use of Wetlands for Water
Pollution Control. U.S. EPA Grant No. R-
806357.
8. Stowell, R., S. Weber, G. Tchobanoglous, B.
Wilson and K. Townzen. Mosquito Considerations
in the Design of Wet/and Systems for the
Treatment of Wastewater. Department of Civil
Engineering, University of California, Davis,
California, and Vector Biology Control Branch,
California State Department of Health Services,
Sacramento, CA, 1982.
9. Reed, S.C., and R.K. Bastian. Wetlands for
Wastewater Treatment: An Engineering
Perspective. In: Ecological Considerations in
Wetlands Treatment of Municipal Wastewaters.
Van Nostrand Reinhold, NY, pp. 444-450, 1985.
10. Reed, S.C., E.J. Middlebrooks, and R.W. Crites.
Natural Systems for Waste Management and
Treatment. McGraw-Hill Book Co., NY, 1987.
11. Hyde, H.C., and R.S. Ross. Technology
Assessment of Wetlands for Municipal
Wastewater Treatment. Municipal Environmental
Research Laboratory, Office of Research and
Development, U.S. Environmental Protection
Agency, EPA600/2-841154, NTIS No. PB 85-
106896, 1984.
12. Herskowitz, J., S. Black, and W. Lewandowski.
Listowel Artrticial Marsh Treatment Project. In:
Aquatic P/ants for Water Treatment and Resource
Recovery. Proceedings of the Conference on
Research and Applications of Aquatic Plants for
Water Treatment and Resource Recovery,
Magnolia Publishing, Inc., Orlando, FL, pp. 247-
261, 1987.
13. Bavor, H.J., D.J. Roser, and S. McKersie.
Nutrient Removal Using Shallow Lagoon-so/id
Matrix Macrophyte Susyems. In: Aquatic Plants
for Water Treatment and Resource Recovery -
Proceedings of the Conference on Research and
Applications of Aquatic Plants for Water
Treatment and Resource Recovery, Magnolia
Publishing, Inc., Orlando, FL, 1987.
-------
-------
CHAPTER 2
Environmental and Public Health Considerations
2.1 Introduction
Protection of public health is the fundamental purpose
of waste treatment. Environmental protection is the
second major purpose. It is the responsibility of the
engineers, scientists and public officials involved to
ensure that waste treatment systems achieve this
goal (1).
Two converging trends encourage engineers to
consider natural processes such as constructed
wetland systems and aquatic plant systems. The first
trend is the ever increasing demand for water at a
time when the least cost water sources have already
been used. The second trend is the increasing
volume of biological and chemical wastes that
potentially enter the surface water system of the
United States from wastewater treatment plants. The
Clean Water Grant Program did much to upgrade
water pollution control facilities so that the United
States population could expect even higher standards
of water quality.
The cost to construct and operate wastewater
treatment facilities that accomplish advanced
treatment in terms of further BOD5 or nitrogen
removal is high compared to the cost of primary and
secondary treatment. The search for a different
approach for polishing effluent and for nutrient
removal has caused renewed interest in land
application and wetlands application of effluent from
conventional wastewater treatment facilities. Systems
that are more "natural" in the sense that they are
influenced more by natural environmental conditions
of temperature, rainfall, sunlight, and wind action are
useful alternatives to conventional systems.
Compared to conventional systems, natural systems
use less electrical energy and require less labor for
operation.
From a public health and environmental health
viewpoint, natural systems have potentially more
points of contact with the environment and with the
public, because of the larger land area involved in the
system. Effluent monitoring is complicated because
indicator organisms (total coliform bacteria counts) do
not clearly indicate the extent of wastewater treatment
(i.e., removal of pathogenic organisms). Any future
application of treated wastewater to constructed
wetland and aquatic plant systems must be free of
unreasonable risks to public health. Public access to
these systems can be controlled with fencing so that
public health issues center on the characteristics of
the effluent and health issues, if any, for the plant
operators.
The principal contaminants of concern in wastewater
fall into the following categories: nitrogen,
phosphorus, pathogenic organisms, heavy metals,
and trace organics. The pathogens include bacteria,
viruses, protozoa and helminths. The heavy metals
include cadmium, copper, chromium, lead, mercury,
selenium, and zinc. Trace organics include highly
stable synthetic compounds (especially chlorinated
hydrocarbons).
The major health concern is possible pollution by
nitrogen, metals, pathogens or organics. These
pollutants and their potential pathways of greatest
concern are summarized in Table 2-1.
Table 2-1. Pollutants and Pathways of Concern
Pollutant Pathway
Nitrogen
Health
Environmental
Phosphorus
Health
Environmental
Pathogens
Health
Environmental
Metals
Health
Environmental
Trace organics
Health
Environmental
Infant water supply
Eutrophication
No direct Impact
Eutrophication
Water supplies, crops, aerosols
Soil accumulation, Infect wildlife
Water supplies, crops, or animals in human
food chain
Long-term soil damage, toxic to plants or
wildlife
Waler supplies, food chain, crops or animals
Soil accumulation
2.2 Nitrogen
Nitrogen is limited in drinking water to protect the
health of infants and may be limited in surface waters
to prevent eutrophication. Nitrogen can be removed in
-------
pond systems by plant or algal uptake, nitrification
and denitrification and loss of ammonia gas to the
atmosphere (evaporative stripping = volatilization).
Nitrogen removal in aquatic plant systems is 26-96
percent, primarily due to nitrification/denitrification
(2,3). In constructed wetlands, nitrogen removal
ranges from 25-85 percent by the same mechanism
(4).
2.3 Phosphorus
Phosphorus removal in wetlands and aquatic plant
systems is not very effective because of the limited
contact opportunities between the wastewater and the
soil. A 28-57 percent phosphorus removal in the
National Space Technology Lab studies with water
hyacinths has been reported (5). The principal
mechanisms for phosphorus removal are plant uptake
or retention in the soil.
2.4 Pathogens
The pathogens of concern in aquatic treatment
systems are parasites, bacteria, and viruses. The
pathways of concern are to the surface waters
receiving discharge from a constructed wetland or
aquatic plant system. Pathways which are generally
not a concern are groundwater contamination and
offsite transmission via aerosols. Groundwater will not
be contaminated in systems that are sealed by an
impervious clay or synthetic material barrier.
Public health effects of wastewater treatment facilities
include the influence on plant workers of aerosols
from pond aerators. Based on several comprehensive
investigations reported, it can be said that people who
have been exposed to aerosolized microorganisms
from wastewater treatment processes generally do
not become infected or ill (6).
2.4.1 Parasites
Research has been conducted on transmission of
parasitic diseases to animals and man by means of
land application of municipal wastewater and sludge
(6). A significant study completed at the San Angelo,
Texas, wastewater irrigation site (7) indicated that
parasites do not increase in cattle grazed or
wastewater irrigated pastures during the period of the
study. These results are similar to those reported
earlier in Poland (8,9) and Australia (10). These
studies, although not on wetlands systems, indicate
that the potential for serious problems does not seem
to be present.
2.4.2 Bacteria
Wildlife may be affected by wetlands systems
because anaerobic muds may contain the causative
organism of avian botulism (Clostridium botulinum).
Control of this wildlife pathogen can be accomplished
largely by multiple dispersion points for FWS
wetlands. This pathogen is not a problem for wild fowl
in SFS wetlands or aquatic plant systems.
The major paths for the transmission of human
disease from wastewater are: direct contact with
applied wastewater, aerosol transport, food chain, and
improperly treated drinking water.
At Santee, California, subsurface flow systems (SFS)
were studied with respect to the contribution of
vegetation to removal of coliform bacteria in
constructed wetlands. Each wetland bed consisted of
a plastic lined (Hypalon, 0.76 mm) excavation, 18.5 m
long x 3.5 m wide x 0.76 m deep (60.7 ft x 11.5 ft x
2.5 ft), containing emergent vegetation growing in
gravel. Influent flow was from primary municipal
wastewater. The hydraulic application rate was 5 cm
(0.2 in)/d and the mean influent total coliform level of
6.75 x 107 MPN/100 ml was reduced to 5.77 x 106
MPN/100 ml (99 percent removal) in the vegetated
bed (7). Hydraulic residence time was 5.5 days. The
population decline of coliforms is due to
sedimentation, filtration, and absorption. Sunlight has
been shown to have a lethal effect on coliforms (11).
In a study of free water surface (FWS) wetlands in
Listowel Ontario, Canada, fecal coliform removal
efficiency was approximately 90 percent when
operated at a 6-7 day residence time (12).
Gearheart et al. found a total coliform removal
efficiency of 93-99 percent during winter and 66-98
percent during summer at 7.5 days retention time in
free water surface wetlands in Arcata, California (13).
Pathogenic bacteria and viruses are removed in
aquatic plant systems by the same mechanisms as in
pond systems. These include predation,
sedimentation, absorption, and die-off from
unfavorable environmental conditions, including UV in
sunlight and temperatures unfavorable for cell
reproduction. In order to quantify the magnitude of the
contribution from the above mechanisms, Gersberg et
al. (14) measured the rate of inactivation of coliform
bacteria in sealed bags with in situ incubation below
the gravel surface of a SFS wetland. The result when
compared to the decay rate through the wetland
system was twice that for the in situ decay rate (i.e.,
without contact with the wetland vegetation). The
difference indicates that half the degradation is due to
vegetation effects including bacterial absorption to
root surfaces and substrate biofilm.
One strong advantage of constructed wetlands over
natural wetlands is that the final effluent can be
chlorinated. Chlorine disinfection of constructed
wetland effluent and aquatic plant systems can
produce waters suitable for unrestricted reuse
applications, since total coliform levels can be
reduced to <2 MPN/100 ml (7). There is a growing
tendency to use chlorine as a disinfectant less often
in cases where the production of trihalomethane
10
-------
(THM) compounds is likely. Disinfection of wetland
effluent with ultraviolet (UV) or ozone are alternatives
that do not produce THMs.
2.4.3 Viruses
Viruses in most treatment systems are more resistant
to inactivation than are bacteria. The removal
efficiency of a SFS system was tested at Santee,
California. An indicator of viral pollution (MS-2
bacteriophages) was reported to be 98.3 percent
removed for a demonstration-scale (800 m2 [8,600
sq ft]) bulrush bed at Santee at a detention time of
5.5 days (7). This involved spiking the influent
wastewater with MS-2 virus and studying
subsequent removal efficiency. MS-2 virus was
chosen because it is an RNA bacteriophage nearly
the same size as enteroviruses and is more resistant
to UV light (15) heat (16) and disinfection (17) than
most enteric viruses.
2.5 Metals
Heavy metals are common environmental pollutants
that are produced as the result of industrial,
commercial and domestic activities. New pretreatment
standards require some industrial discharges, such as
electroplating and metal finishing operations, to limit
heavy metal levels to very low residual concentrations
(18). Studies in New York City show that heavy
metals can be found in municipal wastewater even
when major industrial sources are not part of the
system (19).
Conventional primary and secondary unit processes
at municipal wastewater treatment plants are
inadequate for efficient removal of heavy metals.
Advanced processes including chemical precipitation,
electrolysis, reverse osmosis, and ion exchange are
used for pretreatment of known sources of heavy
metals in industrial wastewater. Use of these
processes to remove low concentrations of heavy
metals in municipal wastewater has the disadvantage
of high capital cost and high operation and
maintenance costs. Additional disadvantages can be
relatively high electrical power costs for electrolysis
and reverse osmosis processes and production of
large amounts of bulky sludges with long settling
times in the chemical precipitation processes.
Since the metal-laden sludges are often disposed of
in land fills, a treatment process that precipitates and
holds heavy metals in the confined area of a
constructed wetland accomplishes the same level of
removal at lower labor and energy costs (i.e., the
heavy metals are returned to the confined
environment of the landfill or the constructed
wetland). The goal of treatment for heavy metals is to
remove the metals from the larger environment and
from the food chain, especially the food chain in river
and ocean waters, The heavy metals are deposited in
landfills or wetlands depending on how they are
removed.
Constructed wetlands (SFS) at Santee, California
received municipal wastewater that was spiked with
the heavy metals copper, zinc and cadmium. At
hydraulic retention times of 5.5 days, removal
efficiencies were 99, 97, and 99 percent respectively
(20). The removal in the constructed wetlands was
attributed to precipitation-adsorption phenomena.
Chemical precipitation is enhanced by wetland
metabolism, especially of algal cells which deplete
dissolved CO2 levels and raise the pH. Metals
removal in MIS wetlands should not be expected to
be significant. In one case, metals removal in a water
hyacinth system was 85 percent for cadmium, 92
percent for mercury, and 60 percent for selenium (6).
2.6 Trace Organics
Municipal and industrial wastewaters contain variable
concentrations of synthetic organic compounds.
During 1960-1970, environmental researchers
became aware of the tendency of some organic
contaminants to resist removal in conventional
wastewater treatment and to persist in the
environment for very long periods. A more disturbing
observation was that persistent, toxic compounds
were found to accumulate in food chains because of
the tendency of the compounds to be fat soluble. A
compound can disappear from solution in an aqueous
system by a number of mechanisms. Among the
mechanisms are: biological, chemical, photochemical
alternatives, and physicochemical processes such as
absorption, sedimentation, and evaporative stripping.
Biological degradation of easily degraded organic
compounds is considered the most important of these
(21).
Evaporative stripping is a major mechanism for land
treatment systems that employ spray irrigation (6);
however, it is not a major mechanism for organic
compound removal from wetlands or aquatic plant
systems. Absorption of trace organics by the organic
matter and clay particles present in the treatment
system is thought to be the primary physicochemical
mechanism for removal of refractory compounds in
wetlands and aquatic plant systems (6). The extent to
which trace organics are removed by a water
hyacinth system is shown in Table 2-2.
11
-------
Table 2-2. Trace Organic Removal in Pilot-Scale Hyacinth
Basins* (6)
Conventration, pg/L
Parameter
Benzene
Toulene
Ethyl benzene
Chloro benzene
Chloroform
Chlorodibromomethane
1,1,1-Trichloroethane
Tetrachloroethylene
Phenol
Butylbenzyl phthalate
Diethyl phthalate
Isophorone
Naphthalane
1 ,4-Dichlorobenzene
Untreated
Wastewater
2.0
6.3
3.3
1.1
4.7
5.7
4.4
4.7
6.2
2.1
0.8
0.3
0.7
1.1
Hyacinth
Effluent
Not Detexcted
Not Detexcted
Not Detexcted
Not Detexcted
0.3
Not Detexcted
Not Detexcted
0.4
1.2
0.4
0.2
0.1
0.1
Not Detexcted
*4.5 day detention time, 76 m3/d flow, 3 sets of 2 basins each in
parallel, plant density 10-25 k/m2 (net weight).
2.6 References
1. Reed, S.C. Health Effects and Land Application of
Wastewater. In: Water Reuse. Ann Arbor Science
Pub., Inc. pp. 753-781, 1982.
2 Gearheart, R.A. et al. Final Report City of Arcata
Marsh Pilot Project. City of Arcata Department of
Public Works. Arcata, CA, 1983.
3. Middlebrooks, E.J. Aquatic P/ant Processes
Assessment. In: Aquaculture Systems for
Wastewater Treatment: an Engineering
Assessment. U.S. Environmental Protection
Agency EPA 430/9-80-006, NTIS No.PB 81-
156705. pp 43-63, 1980.
4. Gersberg, R.M., B.V. Elkins, C.R. Goldman.
Nitrogen Removal in Artificial Wetlands. Water
Res. 17: 1009-1014, 1983.
5. Wolverton, B.C., and R.C. McDonald. Upgrading
Facultative Wastewater Lagoons with Vascular
Aquatic P/ants. JWPCF 51:305-313, 1979.
6. Reed, S.C., E.J. Middlebrooks, and R.W. Crites.
Natural Systems for Waste Management and
Treatment. McGraw-Hill Book Co., New York,
NY, 1987.
7. Gersberg, R.M., R. Brenner, S.F. Lyon, and B.V.
Elkins. Survival of Bacteria and Viruses in
Municipal Wastewaters Applied to Artificial
Wetlands. In: Aquatic Plants for Water Treatment
and Resource Recovery. Magnolia Publishing,
Inc. Orlando, FL. pp. 237-245, 1987.
8. Patyk, S. Worm Eggs in Wroclaw Sewage and on
Meadows and Pastures Irrigated with Municipal
Sewage. Wiad. Parazyt. 4,f 5/6, p. 479-481. In:
Critical Review and Assessment of Polish
Literature on Sewage irrigation, Institute of
Meteorology and Water Management (Wroclaw,
Poland), Technical Interim Report No. 1 on
Project JB-5-532-24, Dec. 1977, Abstract No.
2205. pp. 288-289, 1958.
9. Jankiewicz, L. Survival of Ascaris Eggs On Soils
Irrigated with Communal Sewage, Zesz, nauk,
A.R. - Wroc. Meliorate, XV, No. 90, p. 61-66.
In: Critical Review and Assessment of the Polish
Literature on Sewage irrigation, Institute of
Meteorology and Water Management Wroclaw,
Poland), Technical Interim Report No. 1 on
Project JB-5-532-24, Dec. 1977. Abstract No.
193. pp. 275-276, 1972.
10. Evans, K.J., I.G. Mitchell, and B. Salau. Heavy
Metal Accumulation in Soils Irrigated by Sewage
and Effect in the Plant-animal System.
International Conference on Developments in
Land Methods of Wastewater Treatment and
Utilization. October, Melbourne, Victoria, Australia,
pp. 24/1 -24/14, 1978.
11. A.L.H. Gamerson, and J.R. Saxon. Water Res.
1:279, 1967.
12. Palmateer, G.A., W.L. Kutas, M.J. Walsh, and
J.E. Koellner. Abstracts of the 85th Annual
Meeting of the American Society for Microbiology.
Las Vegas, NV, 1985.
13. Gearheart, R.A., S. Wilber, J. Williams, D. Hull,
N. Hoelper, K. Wells, S. Sandberg, D. Salinger,
D. Hendrix, C. Holm, L. Dillon, J. Morita, P.
Grieshaber, N. Lerner, and B. Finney. City of
Arcata, Marsh Pilot Project, Second Annual
Progress Report. Project No. C-06-2270, State
Water Resources Control Board, Sacramento,
CA, 1981.
14. Weaver, R.W., N.O. Dronen, B.C. Foster, F.C.
Heck, and R.C. Fehrmann. Sewage Disposal on
Agricultural Solids: Chemical and Microbiological
Implications, Vol. II: Microbiological Implications.
Prepared for U.S. Environmental Protection
Agency, RSKERL, Ada. OK, 1978.
15. Kapuscinski, R.B., and R. Mitchell. Sunlight-
induced Mortality of Viruses and Escherichia coli
in Coastal Seawater. Environmental Science
Technology. 1711-6, 1982.
12
-------
16. Burge, W.D., D. Colacicco, and W.N. Cramer.
Criteria for Achieving Pathogen Destruction
During Composting. JWPCF 53:1683-1690,
1981.
17. Havelaar, A.M., and W.M. Hogeboom. A Method
for the Enumeration of Male-Specific
Bacteriophages in Sewage. J. Appl. Bacterial. 56:
439-447, 1984.
18. Federal Register. Electroplating and Metal
Finishing Point Source Categories; Effluent
Limitations, Pretreatment Standards, New Source
Performance Standards. Fed. Reg. 48:137, 1983.
19. Klein, L.A., et al. Sources of Metals in New York
City Wastewater. JWPCF 46: 2653, 1974.
20. Gersberg, R.M., S.R. Lyon, B.V. Elkins, and C.R.
Goldman. The Removal of Heavy Metals by
Artificial Wetlands. In: Proceedings of the Water
Reuse Symposium III. Future of Water Reuse.
AWWA Research Foundation, 1985.
21. Giger, W., and P.V. Roberts. Characterization of
Persistent Organic Carbon. In: Water Pollution
Microbiology, Vol. 2. John Wiley and Sons, NY.
pp 135-175, 1978.
13
-------
-------
CHAPTER 3
Design of Constructed Wetlands
The use of constructed wetlands can be a cost-
effective treatment alternative. Constructing a wetland
where one did not exist before avoids the regulatory
entanglements associated with natural wetlands and
allows design of the wetland for optimum wastewater
treatment. Typically, a constructed wetland should
perform better than a natural wetland of equal area
because the bottom is usually graded and the
hydraulic regime in the system is controlled (1).
In addition to treating municipal wastewaters,
constructed wetlands have been used for a variety of
industrial applications. The fdree water surface (FWS)
wetland is widely used as an inexpensive method of
treating acid mine drainage (1). A FWS wetlands
treatment facility for the Fabius Coal Preparation
Plant, operated by the TVA, is described in the Case
Studies section of this Chapter.
3.1 Types of Constructed Wetlands
Constructed wetlands include FWS, as well as the
more recently developed subsurface flow systems
(SFS). The latter systems involve subsurface flow
through a permeable medium. The "root-zone
method" and "rock-reed-filter" are other names for
these systems that have been used in the literature.
Because emergent aquatic vegetation is used in
these systems they depend on the same basic
microbiological reactions for treatment. The media
type (soil or rock) affects the hydraulics of the
system.
3.7.7 Free Water Surface Systems with Emergent
Plants
A FWS system typically consists of basins or
channels, with a natural or constructed subsurface
barrier of clay or impervious geotechnical material to
prevent seepage, soil or another suitable medium to
support the emergent vegetation, and water at a
relatively shallow depth flowing over the soil surface.
The shallow water depth, low flow velocity, and
presence of the plant stalks and litter regulate water
flow and, especially in long, narrow channels, ensure
plug-flow conditions (1).
3.7.2 Subsurface Flow Systems with Emergent
Plants
A SFS wetland is a constructed wetland consisting of
a trench or bed underlain with an impermeable layer
of clay or synthetic liner. The bed contains media
which will support the growth of emergent vegetation.
The system is built with a slight inclination (1-3
percent) between inlet and outlet. As shown in Figure
3-1, primary or pond effluent is introduced into the
end of the system where it flows into a transverse
channel filled with broken stones.
Alternatively, the inlet channel can be perforated or
gated pipe. From there the wastewater flows
horizontally through the rhizosphere of the wetland
plants. During the passage of the wastewater through
the rhizosphere, the wastewater is treated by
filtration, sorption and precipitation processes in the
soil and by microbiological degradation. The resulting
physical-chemical and biochemical processes
correspond to the mechanical and biological
processes in conventional mechanical treatment
systems including denitrification. The effluent is
collected at the outlet channel which is often filled
with coarse gravel and may be discharged directly
into the receiving water.
Within the class of constructed wetland systems, the
SFS systems studied most completely in the United
States are those with sand or rock media, (e.g.,
Santee, California; Emmitsburg, Maryland).
3.2 Site Selection
3.2.7 Topography
A constructed wetland can be constructed almost
anywhere. The emergent plant species used can
tolerate winter freezing much better than aquatic plant
systems. In Ontario, experimental systems have been
built in heavy clay soils (Listowel) and in an
abandoned mine-tailing basin (Cobalt). Because
grading and excavating represent a major cost factor,
topography is an important consideration in the
selection of an appropriate site.
3.2.2 Soil Permeability for Free Water Surface
Systems
In selecting a site for a free water surface wetland the
underlying soil permeability must be considered. The
most desirable soil permeability is 10~6 to 10~7 m/s
(0.14-0.014 in/hr) (2). Sandy clays and silty clay
loams can be suitable when compacted. Sandy soils
15
-------
Figure 3-1. Typical cross section - SFS system.
SLOTTED PIPE FOR
WASTEWATER
DISTRIBUTION
INLET STONE
DISTRIBUTOR
CATTAILS
'-EFFLUENT OUTLET
HEIGHT VARIABLE
SLOPE 1 %
RHIZOME NETWORK'
WATERTIGHT MEMBRANE
are too permeable to support wetland vegetation
unless there is a restrictive layer in the soil profile that
would result in a perched high ground water table.
Highly permeable soils can be used for small
wastewater flows by forming narrow trenches and
lining the trench walls and bottom with clay or an
artificial liner. In heavy clay soils, additions of peat
moss or top soil will improve soil permeability and
accelerate initial plant growth (3).
3.2.3 Hydrological Factors
The performance of any constructed wetland system
is dependent upon the system hydrology as well as
other factors. Precipitation, infiltration, evapo-
transpiration (ET), hydraulic loading rate, and water
depth can all affect the removal of organics, nutrients,
and trace elements not only by altering the detention
time, but also by either concentrating or diluting the
wastewater. A hydrologic budget should be prepared
to properly design a constructed wetland treatment
system. Changes in the detention time or water
volume can significantly affect the treatment
performance (4).
For a constructed wetland, the water balance can be
expressed as follows:
Qi - Q0 + P - ET = [dV/dt] (3-1)
where,
QJ = influent wastewater flow, volume/time,
QO = effluent wastewater flow, volume/time,
P = precipitation, volume/time
ET = evapotranspiration, volume/time
V = volume of water, and
t = time.
Ground-water inflow and infiltration are excluded
from Equation 3-1 because of the impermeable
barrier.
Historical climatic records can be used to estimate
precipitation and evapotranspiration. Empirical
methods such as the Thornthwaite equation can be
used to estimate evapotranspiration. Pan evaporation
measurements may be useful if the wetlands will
contain a significant percentage of open water areas.
If required, estimates of water losses due to infiltration
can be obtained by conducting infiltration tests such
as outlined in the Design Manual for Land Treatment
Systems (5). Then, if the system operates at a
relatively constant water depth (dV/dt = 0), the effluent
flow rate can be estimated using Equation (3-1) (4).
3.2.4 Water Rights Considerations
In the western states, both riparian and appropriative
water rights may be affected by adopting a
constructed wetlands system. The effects can include
site drainage (quality and quantity), change of location
for surface water discharge, and reduction of the
quantity of a surface water discharge. If an existing
surface discharge is to be affected, replacement of
downstream water rights may be necessary.
3.3 Performance Expectations
Wetland systems can significantly reduce biological
oxygen demand (BOD5), suspended solids (SS), and
nitrogen, as well as metals, trace organics, and
pathogens. The basic treatment mechanisms are
listed in Table 3-1 and include sedimentation,
chemical precipitation and adsorption, and microbial
interactions with BOD5, SS, and nitrogen, as well as
some uptake by the vegetation. The performance of
several pilot-scale wetland systems is summarized
in Table 3-2.
Removal rates for a large-scale pilot study of a SFS
system near Sidney, Australia, have been reported
(6). Trenches were 100 m long x 4 m wide x 0.5 m
16
-------
Table 3-1.
Mechanism
Removal Mechanisms in Wetlands for the Contaminants in Wastewater (from 8)
Contaminant Effected3
Bacteria
Heavy Refractory and
BOD N P Metals Organics Virus
Settleable Colloidal
Solids Solids
Description
Physical
Sedimentation
Filtration
Adsorption
I Gravitational settling of solids (and
constituent contaminants) in
pond/marsh settings.
Particulates filtered mechanically as
water passes through substrate, root
masses, or fish.
Interparticle attractive forces (van der
Waals force).
Chemical
Precipitation
Adsorption
Decomposition
Biological
Bacterial
Metabolism11
Plant Metabolism"
Plant Adsorption
Natural Die-Off
P P Formation of or co-precipitation with
insoluble compounds.
P P S Adsorption on substrate and plant
surfaces.
P P Decomposition or alteration of less
stable compounds by phenomena
such as UV Irradiation, oxidation, and
reduction.
Removal of colloidal solids and
P P P P soluble organics by suspended,
benthic, and plant-supported
bacteria. Bacterial nitnfication/
denitrifi cation.
S S Uptake and metabolism of organics
by plants. Root excretions may be
toxic to organisms of enteric origin.
s s S S Under proper conditions, significant
quantities of these contaminants will
be taken up by plants.
P Natural decay of organisms in an
unfavorable environment.
a P = primary effect; S = secondary effect; I = incremental effect (effect occurring incidental lo removal of another contaminant).
b The term metabolism includes both biosynthesis and catabolic reactions.
Table 3-2. Performance of Pilot-Scale Constructed Wetland Systems (1)
Effluent Concentration, mg/L
Location
Wetland Type
BOD5
SS
NH4
NO,
T N
TP
Listowel, Ontario
Arcata, CA
Santee, CA
Vermontville, Ml
Open water, channel
Open water, channel
Gravel-filled channels
Seepage basin wetland
10
<20
<30
<8
6
<1 0
<5
2
0.2
0.7
<0.2
1.2
8.9
11.6
6.2
0.6
6.1
2.1
' Alum treatment provided prior to the wetland component.
17
-------
deep (328 ft x 13 ft x 1.6 ft) with gravel. Plant species
studied were Myriophyllum aquaticum (parrot feather);
Schoenoplectus validus (bulrush); and Typha
orientails (cumbungi). Secondary effluent was the
system influent flow. Hydraulic loading rate was 264
m3/ha-d (28,225 gpd/ac), and detention time was 9
days. Typical influent and effluent concentrations are
plotted in Figure 3-2. Examination of the plot reveals
that the gravel planted emergent plant systems were
able to remove significant levels of SS, BOD5, and
nitrogen. Phosphorus removal was slight, which is
consistent with the experience of other researchers
with rock and sand based systems.
Figure 3-2. Pilot-scale constructed wetland
planted trench with bulrush (6).
mg/L
60
gravel
Influent
50
40
30
20
10
SS BOD5 TKN NH, TP TOC
Gearheart and Finney (7) concluded from pilot studies
on FSW wetlands in Arcata, CA, that wetlands have
the ability to dampen spikes in effluent characteristics
from an oxidation pond so that the wetland effluent is
more stable and consistent. Constructed wetlands
also reduce SS and fecal coliform levels and bring pH
values to nearly neutral values. The ability of
constructed wetlands to produce a consistent effluent
from a low capital investment with low labor and
energy requirements is a key benefit that is
noteworthy.
The treatment processes that occur in an artificial
wetland are similar to those that occur in other forms
of land treatment. Removal of settleable organics
occurs primarily as a result of sedimentation.
Removal of colloidal and soluble organics occurs
primarily by aerobic microbial oxidation.
3.3.1 BOD5 Removal in FWS Wetlands
In FWS wetlands, removal of the soluble BOD5 is due
to microbial growth attached to plant roots, stems,
and leaf litter that has fallen into the water. Because
algae are typically not present if plant coverage is
complete, the major sources of oxygen for these
reactions are reaeration at the water surface and
plant translocation of oxygen from the leaves to the
rhizosphere (1).
Specific criteria presented below are suitable for low
to moderate organic loadings. The organic loading
should be distributed over a significant portion of the
area and not applied at a single point. The design
water depth should be 600 mm (24 in) (1) or less to
ensure adequate oxygen distribution, and partial
effluent recirculation might be considered in the
summer months to overcome ET losses and maintain
design flow rates and oxygen levels.
BOD5 removal in a wetland has been described by a
first-order model as follows (1):
[Ce/C0 = exp (-KTt)
(3-2)
where,
Ce = effluent BOD5, mg/L
C0 = influent BOD5, mg/L
KT = temperature-dependent first-order
reaction rateconstant, d"1
t = hydraulic residence time, d
Hydraulic residence time can be represented as:
t = LWd-Q (3-3)
where,
L = length
W = width
d = depth
Q = average flow rate = (flowin + flowout) •*• 2
This equation represents hydraulic residence time for
an unrestricted flow system.
In a FWS wetland, a portion of the available volume
will be occupied by the vegetation, so the actual
detention time will be a function of the porosity (n),
which can be defined as the remaining cross-
sectional area available for flow.
n = Vtf+V
(3-4)
Where Vv and V are volume of voids and total
volume, respectively.
The product (nod) is, in effect, the "equivalent depth"
of flow in the system. The ratio of residence time
from dye studies to theoretical residence time
calculated from the physical dimensions of the
system, should equal the ratio of n«d:d.
18
-------
Combining the relationships in Equations 3-3 and
3-4 with the general model (Equation 3-2) results
in Equation 3-5 (1):
Ce/C0=A exp[(-0.7 KT(AV)175 L W d n)+Q] (3-5)
where,
A = fraction of BOD5 not removed as settleable
solids near headworks of the system (as
decimal fraction)
Av = specific surface area for microbial activity,
m2/m3
L = length of system (parallel to flow path), m
W = width of system, m
d = design depth of system, m
n = porosity of system (as a decimal fraction)
Q = average hydraulic loading on the system,
m3/d
The temperature-dependent rate constant is
calculated from the rate constant for 20°C and the
correction factor of 1.1 (9). The rate constant KT (in
d"1) at water temperature T (°C) can therefore be
defined by Equation 3-6.
KT = K20 (1.1)
(T-20)
(3-6)
where K20 is the rate constant at 20°C.
Other coefficients in Equation 3-5 have been
estimated (1).
A = 0.52
K20 = 0.0057 d"1
A =15.7 m2/m3
n
= 0.75
Typical values used to test the equation against
actual values at Listowel, Ontario, are listed in Table
3-3.
Table 3-3. Predicted vs. Actual Ce/C0 Values for
Constructed Wetlands [Actual Values from
Listowel, Ontario (1 )]
Distance Along
Channel, m
0
67
134
200
267
334 (final effluent)
Predicted
0.52
0.38
0.27
0.20
0.14
0.10
Actual
0.52
0.36
0.41
0.30
0.27
0.17
uvm
Predicted
0.52
0.40
0.31
0.24
0.18
0.14
ler
Actual
0.52
0.40
0.20
0.19
0.17
0.17
For Listowel, Ontario:
T = 17.8°C (summer), 3.0°C (winter)
Q
W =4m
d = 0.14 m (summer), 0.24 m (winter)
= 35 m3/d (summer), 18.0 nWd (winter)
A sample calculation for the above coefficients using
Equation 3-5 yields the following results:
A = 0.52
n
Y
Q
W
d
= 0.0057 d'1
= 15.7 m2/m3
= 0.75
= 17.8°C (summer)
= 34.6 m3/d (summer)
= 4 m
= 0.14 m (summer), 0.24 m (winter)
Predicted Ce/C0 = 0.312 for L = 134 m
Predicted Ce/C0 = 0.187 for L = 267 m
Equation 3-5 is presented here as an example of the
mathematical expression needed for design purposes.
The coefficients have been estimated from the actual
data at Listowel, Ontario. The sensitivity of the
equation to the specific surface area (Av), and the
water temperature (T), were examined.
Figure 3-3 shows the sensitivity of the Ce/C0 ratio to
Av; it indicates that, for values of 12-16 m2/m3
(3.7-4.9 sq ft/cu ft), corresponding to an average
reed stalk diameter of 12-16 mm (0.49-0.66 in),
the Ce/C0 ratio can range from 0.18 to 0.098 at the
end of a 335-m (1,100-ft) wetland channel. For a
stalk diameter of 12 mm (0.5 in) and a vegetation
volume of 5 percent, it was estimated that the
specific surface area is 15.7 m2/m3 (4.8 sq ft/cu ft)
(1). This is a parameter that cannot be measured
directly in an actual wetland environment. It
represents the surface area from all plant litter in the
water column including the reed stems, leaves, and
roots. The sensitivity of the equation to the specific
area coefficient is not high. This means that specific
area can be estimated and predicted results are likely
to match actual results. If the equation proved to be
sensitive to the estimate of specific area, then it
would mean that this coefficient would have to be
known accurately, which is not possible, and the
equation would have been of limited value.
The sensitivity of Equation 3-5 to temperature was
calculated by varying it in the range 5-25°C (41-
77°F) for the predicted Ce/C0 ratio at Listowel. As
shown in Figure 3-4, the degree of treatment at 5°C
(41 °F) is significantly reduced compared to that at
25°C (77°F). This indicates that the equation is
sensitive to temperature and therefore temperature
must be accurately predicted for use in the equation.
As a practical matter, in winter the depth of the
wetland must generally be increased to allow for ice
depth. The increase in detention time in winter from
greater depth has a compensating effect on the
Ce/C0 ratio.
3.3.2 BOD5 Removal in SFS Wetlands
The major oxygen source for the subsurface
components (soil, gravel, rock, and other media, in
19
-------
Figure 3-3. Sensitivity of Ce/C, ratio to Av.
0.1
150 300
Length of Wetlands Channel (meters)
12
13
—t—
14
15
-X-
16
Av
m2/m3
Figure 3-4. Sensitivity of Ce/C0 ratio to temperature.
0.6
0.5
0.4
0.3
0.2
0.1
0
150
300
Length of Wetlands Channel (meters)
10
—»—•
15
-Q-
20
25
Temperature
20
-------
trenches or beds) is the oxygen transmitted by the
vegetation to the root zone. In most cases the
subsurface flow system is designed to maintain flow
below the surface of the bed, so there can be very
little direct atmospheric reaeration (1). The selection
of plant species is therefore an important factor.
Work at the pilot wetlands in Santee, California (10),
indicated that most of the horizontally growing root
mass of cattails was confined to the top 300 mm (12
in) of the profile. The root zone of reeds extended to
more than 600 mm (24 in) and bulrushes to 760 mm
(30 in.). In the cooler climate of western Europe the
effective root zone depth for reeds is also considered
to be 600 mm (24 in) (1). The gravel bed at the
Santee system was 760 mm (30 in) deep, and the
water level was maintained just below the surface
(10). The BOD5 removals observed in the three
parallel bulrush, reed, and cattail units at Santee
reflect the expanded aerobic zone made possible by
the root penetration of the various plants.
Removal of BOD5 in subsurface flow systems can be
described with first-order plug-flow kinetics, as
described in Equation 3-2 for free water surface
systems. Equation 3-2 can be rearranged and used
to estimate the required surface area for a subsurface
flow system. Both forms of the equation are shown
below for convenience.
The bed width is calculated by the following equation.
W = Ac-d
(3-9)
Cross sectional area and bed width are established
by Darcy's law:
Q = ksAsS
(3-10)
Bed cross sectional area and bed width are
independent of temperature (climate) and organic
loading since they are controlled by the hydraulic
characteristics of the media.
The value of KT can be calculated using Equation 3-
6 and a known K20 for subsurface flow wetlands
system. Typical media types including medium to
coarse sands have K20 values of approximately 1.28
d"1. Based on European data and data from Santee,
California, the K20 values presented in Table 3-3B
have been tested for media up to gravelly sand size
at warm temperatures (T>20°C) (1). The combined
effect of large media size (with a resulting small
porosity value) and low temperatures represent a
system that has not been studied and the above
equations may not accurately predict the results.
Expected porosities (n) hydraulic conductivity and K20
are listed in Table 3-4.
[Ce/C0] = exp (-KTt) (3-2)
As= [Q (In C0 - In Ce)] - (KT d n) (3.7)
where,
Ce = effluent BOD5, mg/L
C0 = influent BOD5, mg/L
KT = temperature-dependent first-order
reaction rate constant, d"1
t = hydraulic residence time, d
Q = average flow rate through the system, m3/d
d = depth of submergence, m
= porosity of the bed, as a fraction
A = surface area of the system, m2
The cross sectional area for flow through a
subsurface flow system is calculated according to the
following equation:
(3-8)
where,
Ac = d«W, cross-sectional area of wetland bed,
perpendicular to the direction of flow, m2
d = bed depth, m
R = bed width, m
k =hydraulic conductivity of the medium,
s m3/m2-d
S = slope of the bed, or hydraulic gradient (as a
fraction or decimal)
Table 3-4. Media Characteristics for Subsurface Flow
Systems
Media Type
Medium Sand
Coarse Sand
Gravelly Sand
Max. 10%
Grain
Size, mm
1
2
8
Porosity
(n)
0.42
0.39
0.35
Hydraulic
Conductivity
(ks), m3/m2-d
420
480
500
^20
1.84
1.35
0.86
Sample Design Problem - Subsurface Flow System
Calculate the required area and bed depth for a SFS
system where influent wastewater is from a facultative
lagoon. Assume influent BOD5 to the wetlands will be
130 mg/L. The desired effluent BOD5 is 20 mg/L. The
predominant wetland plant type in surrounding
marshes is cattail. Water temperatures are 6°C
(43°F) in winter and 15°C (59°F) in summer.
Wastewater flow is 950 m3/d (0.25 mgd)
Solution:
1. Choose cattail for this SFS since it is successfully
growing in local wetlands. From the discussion
above, it is known from studies at Santee,
California, that cattail rhizomes penetrate
approximately 0.3 m (1 ft) into the medium. The
bed media depth (d) should therefore be 0.3 m (1
fit).
21
-------
2. The bed slope is based on the site topography.
Most systems have been designed with slope of 1
percent or slightly higher. For this design choose a
slope of 1 percent for ease of construction
(s = 0.01).
3. Reed et al. (1) have indicated the need to check
the value ksS< 8.60. Choose a media of coarse
sand and from Table 3-4, n = 0.39, ks =480 and
K20 = 1.35.
ksS = (480) (0.01) = 4.8 < 8.60
4. Solve for the first-order temperature-dependent
rate constant (KT) using Equation 3-6.
KT = K20 (1.1)
T-20
Winter:
KT = L35
x6-20
Summer:
KT = 1.35
= 0.36
= 0.84
5. Determine the cross section area (Ac) of the bed
with Equation 3-8.
AC = Q^ksS (3-8)
AC = 950-(480)(0.01) = 198 m2
6. Determine the bed width using Equation (3-9).
W = Ac-d (3-9)
w = 198-0.3= 660 m
7. Determine the surface area required with Equation
3-7.
As = [Q (In C0 - In Ce)] ^ (KT d n) (3-7)
Winter:
As = [(950)(4.87-3.00)] •*• [(0.36)(0.3)(0.39)]
= 42,177 m2 = 4.22 ha (10.4 ac)
Summer:
As= 18,076 m2 = 1.81 ha (4.5 ac)
Winter conditions control, so the total bed area must
be 4.22 ha (10.4 ac).
8. Determine the bed length (L) and the detention
time (t) in the system.
L =
(3-11)
L = 42,177-660 = 63.9 m (210 ft)
t = vv-Q = LWdn-Q (3-12)
t = (63.9)(660)(0.3)(0.39) •*• 950
= 5.2 days
9. Divide the required width into individual cells 60 m
wide for better hydraulic control at the inlet zone.
Construct 11 cells, each 60 m x 64 m (197 ft x
210ft).
All 11 cells are required in winter. In summer several
cells could be dried out for regrading or controlled
burning (spring or fall). All cells should remain in
service during winter and summer except for brief
draining for maintenance. The recovery rate for a cell
that has been allowed to go dormant and dries out
fully is slow.
3.3.3 Suspended Solids Removal
Suspended solids removal is very effective in both
types of constructed wetlands, as shown by the data
in Table 3-2 and Figure 3-2. Most of the removal
occurs within the first few meters beyond the inlet,
owing to the quiescent conditions and the shallow
depth of liquid in the system. Controlled dispersion of
the influent flow with proper diffuser pipe design can
help to insure low velocities for solids removal and
even loading of the wetland so that anoxic conditions
are prevented at the upstream end of the channels.
If the water in the wetlands is not shielded from
sunlight by the vegetation, algae could become a
problem. Algae contribute to effluent SS and cause
large diurnal swings in oxygen levels in the water
column.
3.3.4 Nitrogen Removal
Nitrification/denitrification is the major path of nitrogen
removal (1). Removals of 60-86 percent were
reported at Santee (11). It has been shown that
artificial wetlands may be managed so as to fuel the
process of denitrification by using carbon sources
derived from biomass produced within the wetlands
itself (11). Nitrogen (TKN) removals have been
reported that indicate detention times of 5-7 days
will generally produce an effluent with TKN <10 mg/L.
Typical pilot-scale results are shown in Figure 3-5
along with a regression curve superimposed on the
scatter of data (6).
3.3.5 Phosphorus Removal
Phosphorus removal in many wetland systems is not
very effective because of the limited contact
opportunities between the wastewater and the soil.
The exceptions are the submerged bed designs when
proper soils are selected as the medium for the
system. A significant clay content and the presence
22
-------
Figure 3-5. Regression curve of TKN vs. retention time in the effluent of an alternating Typha/open-water/gravel system. The
curve is a logarithmic fit and has a correlation coefficient of 0.70 (6).
Effluent TKN, mg/L
40
30 -
20 -
10
10
15
Detention Time, days
of iron and aluminum will enhance the potential for
phosphorus removal (1). Use of such soils will,
however, reduce the hydraulic capacity and require a
much greater area for treatment.
3.3.6 Metals Removal
There are limited data available on the metal removal
capability of FWS wetlands; because the removal
mechanisms are similar to those described above for
phosphorus, the response is not very effective.
There is greater opportunity for contact and sorption
in SFS systems and metals removal can be very
effective (1). The predominant removal mechanisms
in the artificial wetlands were attributed to
precipitation-absorption phenomena. Precipitation
was enhanced by wetland metabolism which
increased the pH of inflowing acidic waters to near
neutrality. Removal of Cu, Zn, and Cd at the rates of
99, 97, and 99 percent respectively, for a residence
time of 5.5 days in the Santee, California, wetlands
were reported (12). Phosphorus removal and metals
removal will likely be finite due to exhaustion of
exchange sites.
3.4 Process Variables
Constructed wetland systems can be considered
attached growth biological reactors, and their
performance can be described with first-order plug-
flow kinetics. Design guides for BOD5 loading are
presented in this section for both the FWS and SFS
types. These guidelines were derived from a relatively
limited data base, so caution should be used in their
application. A pilot test is strongly recommended for
large-scale projects.
3.4.7 Design Objectives
There are limits to the technology of using
constructed wetlands for high BOD5 wastewater
treatment. Although limited data are available on the
use of wetlands for treating primary effluent,
constructed wetlands have been used in a number of
locations for polishing secondary effluent (4). The
uses for constructed wetlands also include: a) acid
mine drainage treatment, b) stormwater treatment,
and c) enhancement of existing wetlands.
Secondary effluent polishing has been accomplished
in pilot- and full-scale systems at Incline Village's
Carson River wetlands system, and at Arcata,
California (see Section 3.8.1) where polishing
secondary effluent for release into Humboldt Bay was
a lower cost alternative that participation in a regional
deep-ocean outfall pipeline.
The FWS wetland is widely used as an inexpensive
method of treating acid mine drainage. More than 20
such systems were constructed in 1984-1985 in
Pennsylvania, West Virginia, Ohio, and Maryland (1).
Oxygen for oxidation of mine wastes is supplied from
the root zone of the emergent vegetation and from
23
-------
floating algae. Floating algae remove carbon dioxide
from the water column and thereby raise the pH. The
net effect of the pH rise and interaction of metals
from mine waste is physical-chemical precipitation of
metals in the soil and mud of the wetland. The iron
concentration can be reduced from 25-100 mg/L to
less than 2 mg/L in these systems (1).
Enhancement of existing wetlands is a significant
result of constructed wetlands systems. The Incline
Village General Improvement District wetland system
near Minden, Nevada, includes a wetland constructed
adjacent to an existing natural warm springs wetland.
The addition of a more predictable water supply to the
wetland system allowed a larger more stable
population of wetland fowl and desert wildlife to be
established in the vicinity of the wetland.
In the Arcata, California constructed wetland system,
the goal of meeting the NPDES discharge
requirements as well as enhancing the waters of
Humboldt Bay have been generally met. "The
wastewater project will meet a state reclamation
policy in that it reuses wastewater for the creation of
the marsh and the invertebrates of the oxidation
ponds could be used for fish food in the salmon
aquaculture project. The recreation lake water will
continue to provide nutrients for enrichment of the
mudflats of Humboldt Bay and food for juvenile
salmonids planted in the lake as part of the ocean
ranching project" (7).
3.4.2 BOD5 Loading Rates
There are two goals for organic load control in a
constructed wetland system. The first is provision of a
carbon source for denitrifying bacteria. The second is
control of organic loading to prevent overloading of
the oxygen transfer ability of the emergent plants in
the wetland system. If the carbon source is not
available for denitrification, then lower overall nitrogen
removal will result. However, heavy organic loading,
especially if not evenly distributed, will cause plant die
off and odors.
Organic loading in a FWS wetland can be controlled
by step-feed distribution as well as recycle of
wetland discharges. A mass loading rate of about 112
kg BOD5/ha-d (100 Ib/ac-d) is a typical upper
loading rate.
The mathematical justification for this typical loading
capacity of a WS constructed wetland is established
by estimating the oxygen transfer capacity of the
wetland vegetation. Estimation of loading by this
method is a two step process: 1) first, calculate the
required oxygen; then 2) calculate the available
oxygen for the assumed surface area. The following
two equations are used (1):
Required Oxygen = 1.5 BOD5 (3- 13 )
Available Oxygen = (Tr O2) (As) •*• 1,000 (3 - 14)
where,
O, = oxygen required
BOD5 = organic loading, kg/d
Tr O2 = oxygen transfer rate for the vegetation,
20 g/m2'd
As = surface area, m2
As a safety factor, available oxygen should exceed
required oxygen by a factor of 2 (1). Commonly used
emergent plants can transmit 5-45 g O2/d per m2 of
wetland surface (45-400 Ib/ac-d). At a typical
oxygen transfer rate of 20 g/m2-d (180 Ib/ac-d), the
organic loading rate for a wetland should be 133 kg
BOD5/ha-d (118 Ib/ac-d) (1).
3.4.3 Hydraulic Loading Rates
Hydraulic loading rate for FWS systems is closely tied
to the hydrological factors for each wetland and these
factors are specific to the site. Organic loading rate is
also closely tied to the hydraulic loading rate.
Hydraulic loading rates of 150-500 m3/ha-d
(16,000-54,000 gpd/ac), have been reported (13),
and, in general, the site-specific conditions of
weather, soil conditions (permeability especially), and
vegetation type must be considered in establishing
the hydraulic loading rate. From the results obtained
at Listowel, Ontario, it appears that hydraulic loading
rates of 200 m3/ha-d (21,000 gpd/ac) will provide
maximum treatment efficiencies.
The water losses due to evapotrans pi ration can affect
the feasibility of the various wetland designs in arid
climates and their performance during the warm
summer months in all locations. In the western states,
where appropriate laws govern the use of water, it
may be necessary to replace the volume of water lost
to protect the rights of downstream water users.
Evaporative water losses in the summer months
decrease the water volume in the system, and
therefore the concentration of remaining pollutants
tends to increase even though treatment is very
effective on a mass removal basis (1).
In the special case of a wetland that is constructed to
have zero discharge, the hydraulic loading can
become a major concern and the dominant design
consideration. In a zero-discharge wetland, water is
disposed of through the mechanisms of evaporation,
transpiration, and ground-water recharge.
3.4.4 Water Depth in FWS Systems
The water level in the system and the duration of
flooding can be important factors for the selection and
maintenance of wetland vegetation (1). Cattails grow
well in submerged soils and may dominate where
standing water depth is over 150 mm (6 in). Reeds
occur along the shorelines of water bodies where the
water table is below the surface but will also grow in
water deeper than 1.5 m (5 ft). Growth is best in
standing water but the depth seems to have no direct
24
-------
effect. The common reed is a poor competitor and
may give way to other species in nutrient-rich
shallow waters. Bulrushes can tolerate long periods of
soil submergence and occur at water depths of 7.5-
250 mm (0.3-10 in) in California (14). In deeper
water, bulrushes may give way to cattails. Sedges
generally occur along the shore or in shallower water
than bulrushes (1).
3.4.5 Detention Time
Treatment performance in constructed wetlands is a
function of detention time, among other factors.
Ground slope, water depth, vegetation, areal extent,
and geometric shape control the flow velocity and,
thus, the detention time through a wetlands treatment
system (7).
A detention time of 6-7 days has been reported to
be optimal for the treatment of primary and secondary
wastewater (15). Shorter detention times do not
provide adequate time for pollutant degradation to
occur; longer detention times can lead to stagnant,
anaerobic conditions.
Two climatic factors can significantly affect the
detention time at a constant hydraulic loading rate. In
the summertime, evapotranspiration can significantly
increase the detention time, while ice formation in
wintertime can significantly decrease the detention
time. The recommended water depth at Listowel, for
summertime is approximately 100 mm (4 in) and
should be increased to approximately 300 mm (12 in)
in winter if ice formation is expected, to minimize the
effect of climate on the detention time.
Estimating the detention time in wetland systems can
be difficult for several reasons. First, large dead
spaces may exist in the wetlands due to differences
in topography, plant growth, solids sedimentation, and
the degree of flow channelization (i.e. short-
circuiting). Only a fraction of the surface area, in
wetlands, may be available for wastewater flow.
3.5 Pre-Application Treatment
To reduce capital and operating costs, minimal
pretreatment of wastewater prior to discharge to a
wetland is desirable. However, the level of
pretreatment will also influence the quality of the final
marsh effluent, and therefore effluent quality
objectives must be considered (15).
Preceding wetland treatment with a conventional
primary treatment plant is capital intensive and
impractical unless such a facility is already in
existence. Pretreatment with a conventional lagoon is
land consumptive and may generate hydrogen sulfide
in winter and algal problems in warmer weather.
Based on studies at Listowel, some reduction of SS
and BOD5 is desirable to reduce oxygen demand and
prevent sludge accumulations in the upper reaches of
the marsh. Phosphorus reduction by chemical
addition is recommended in the pretreatment step
when phosphorus is required.
3.6 Vegetation
The major benefit of plants is the transferring of
oxygen to the root zone. Their physical presence in
the system (the stalks, roots, and rhizomes) penetrate
the soil or support medium, and transport oxygen
deeper than it would naturally travel by diffusion alone
(1).
Perhaps most important in the FWS wetlands are the
submerged portions of the leaves, stalks, and litter,
which serve as the substrate for attached microbial
growth. It is the responses of this attached biota that
is believed responsible for much of the treatment that
occurs (1).
The emergent plants most frequently found in
wastewater wetlands include cattails, reeds, rushes,
bulrushes and sedges. Information on their
distribution in the United States and some of the
major environmental requirements of each are
provided in Table 3-5 (1).
3.6.7 Cafta;7s
Cattails (Typha spp.) are ubiquitous in distribution,
hardy, capable of thriving under diverse
environmental conditions, and easy to propagate and
thus represent an ideal plant species for constructed
wetlands. They are also capable of producing a large
annual biomass and provide a small potential for N
and P removal, when harvesting is practiced. Cattail
rhizomes planted at approximately 1-m (3.3-ft)
intervals can produce a dense stand within three
months (3).
3.6.2 Bulrushes
Rushes are members of the genus Juncus and are
perennial, grasslike herbs that grow in clumps (5).
Bulrushes (Scirpus spp.) are ubiquitous plants that
grow in a diverse range of inland and coastal waters,
brackish and salt marshes and wetlands. Bulrushes
are capable of growing well in water that is 5 cm (2
in) to 3 m (10 ft) deep. Desirable temperatures are
16-27°C (61-81°F) (1). Bulrushes are found
growing in a pH of 4-9 (16).
3.6.3 Reeds
Reeds (Phagmites communis) are tall annual grasses
with an extensive perennial rhizome. Reeds have
been used in Europe in the root-zone method and
are the most widespread emergent aquatic plant.
Systems utilizing reeds may be more effective in the
transfer of oxygen because the rhizomes penetrate
vertically, and more deeply than cattails (1).
25
-------
Table 3-5. Emergent Aquatic Plants for Wastewater Treatment
Temperature,°C
Common Name
Cattail
Common reed
Rush
Bulrush
Sedge
Scientific Name
Typha spp.
Phragmites communis
Juncus spp.
Scirpus spp.
Carex spp.
Distribution
Throughout the world
Throughout the world
Throughout the world
Throughout the world
Throughout the world
Desirable
10-30
12-23
16-26
18-27
14-32
Seed Germination Tolerance, ppt*
12-24 30
1 0-30 4 5
20
20
Effective pH
Range
4-10
2-8
5-7.5
4-9
5-7.5
*ppt = parts per thousand.
3.7 Physical Design Factors
3.7.7 System Configurations
Studies at Listowel have demonstrated the
importance of a long length-to-width ratio to insure
plug flow hydraulics (3). In the model (Equation 3-5)
plug-flow hydraulics is assumed as the major form
of transport. Internal flow distribution must therefore
be achieved by using high length-to-width ratios or
by internal berming or barriers (3).
3.7.2 Distribution System
The distribution system for a series of channels each
with high length-to-width ratios can be constructed
simply using a manifold pipe and gate valves at the
head of each channel. For each system the influent
flow must have controls to allow distribution to the
preferred channels and wetland segments as well as
an overflow outlet for dispersion of excess flows and
emergency diversions. Distribution of inflow at
multiple points in the wetland is a key requirement for
controlled and efficient operation of the wetland. For
systems with recycle, a pump station with return
pipeline to the distribution system must be
constructed. Alternatively, the plug-flow channels
can be folded back to the inlet to minimize recycle
costs. Flow monitoring is an important component of
the influent distribution system.
3.7.3 Outlet Structures
The configuration of the outlet structure for a
constructed wetland depends on the character of the
receiving water and the number of subunits in the
constructed wetland. The outlet structure for the
surface flow type of wetland is shown in Figure 3-1,
and includes a trench and outlet pipe with adjustable
level for water level control in the wetland. Outlet
structure controls must be able to control depth of
water in the wetlands especially for winter ice
conditions where deeper wetland conditions are
required to maintain treatment levels. Outlet
structures must be constructed to prevent ice
damage and closed control points during freezing
weather.
3.7.4 Vector Control in Free Water Surface
Wetlands
FWS wetlands provide an ideal breeding environment
for many insect pest species, particularly mosquitoes.
In Listowel, population densities of Culex pipiens were
directly related to the presence of high organic
loadings and inversely related to surface water
coverage by dense duckweed (Lemna spp.) growths.
Mosquitoes are not a problem for subsurface flow
wetlands (this is one of the major reasons for using
the subsurface type design).
3.7.5 Harvesting of Vegetation
Generally, harvesting of wetland vegetation is not
necessary especially for subsurface flow systems (2).
For free water surface systems, dry grasses are
sometimes burned off annually to help maintain the
hydraulic profile of the wetland, and avoid build-up
of grassy hillocks, which encourage channelization.
Harvesting of plant biomass is normally not regarded
as a practical method for nutrient removal. For
example, in Listowel, a single, late-season harvest
removed 200 g of plant material (dry weight)/m3 (1.7
lb/l,000 gal) but only 8 percent and 10 percent of the
annual N and P loading to the marsh, respectively (3).
An earlier harvest, prior to translocation of nutrients
by the cattails, or several harvests per season would
be more effective for nutrient removal purposes.
Harvesting may be desirable to reduce the excessive
accumulation of litter that could shorten the life span
of a FWS wetland (3).
3.8 Case Studies
This section provides case study summaries of four
systems (three FWS and one SFS) which are
representative of current knowledge and practice. The
four systems are in Arcata, California; Emmitsburg,
Maryland; Gustine, California; and Jackson County,
Alabama. The Arcata system was chosen because of
the pilot work performed and because one of the
main goals of the project was to enhance the
beneficial uses of the area surface waters. The
Emmitsburg system was chosen because it is a
submerged bed system operating in a relatively cold
winter climate. The Gustine system was used
because of the pilot scale study information and
because it attempts to control the influent quality to
the wetlands system. The Jackson County system
was chosen because it is used in the treatment of
wastewaters associated with mining operations.
26
-------
3.8.7 Arcata, California
The Arcata System is a FWS system that discharges
municipal oxidation pond effluent into a marsh.
3.8.1.1 History
The City of Arcata wastewater treatment plant was
constructed in the 1940s when the first sewers were
installed. At that time, the wastewater received
primary treatment before discharge to Humboldt Bay.
Oxidation ponds were added in 1958 followed by the
addition of chlorination facilities in 1968 and
dechlorination facilities in 1975.
In April 1975, the Comprehensive Basin Plan for the
North Coast Region was adopted by the California
State Water Resources Control Board and was
incorporated into the Bays and Estuaries Policy (16).
The stated policy concerning discharges to Humboldt
Bay was that all wastewater discharges to enclosed
bays and estuaries be "phased out at the earliest
practicable date." The Regional Water Quality Control
Board was empowered to grant exemptions if the
discharger could demonstrate "that the wastewater in
question would consistently be treated and
discharged in such a manner that it would enhance
the quality of receiving waters above that which would
occur in the absence of the discharge."
In 1977, the City of Arcata proposed to the Regional
Water Quality Control Board the use of a wastewater
treatment process consisting of existing primary
sedimentation facilities and 22.3-ha (55-ac)
oxidation pond facilities, and three new constructed
marshes (12.6 ha [31 ac]). The effluent from the
marsh system would flow through a 6.9-ha (17-ac)
recreation lake before being discharged to Humboldt
Bay. The City claimed that the system would protect
all of the existing beneficial uses of Humboldt Bay
and would result in the fuller realization of existing
beneficial uses or in the creation of new beneficial
uses.
The State Water Resources Control Board funded a
three-year pilot study which began in September of
1979. The results of the pilot work were promising
and, in 1983, the Board agreed that the marsh
system would enhance the beneficial uses of
Humboldt Bay for scenic enjoyment and educational
study, and that a full-scale marsh system would
meet the requirements of the Basin Plan. In 1986, the
marsh treatment system was completed and placed in
operation.
3.8.1.2 Design Objectives
Design objectives for the marsh system at Arcata
were that the marsh effluent meet the NPDES
discharge requirements listed in Table 3-6 as well as
enhance Humboldt Bay waters. To provide for
enhancement of Humboldt Bay waters the marsh
system should also be designed and operated as a
wildlife habitat.
Table 3-6. City of Arcata, CA Wastewater Discharge
Requirements
Constituents*
BOD5 (20°C), mg/L
Suspended Solids, mg/L
Settleable Solids, mL/L
Total Conforms, MPN/100
CI2 Residual, mg/L
Grease and Oil, mg/L
Toxicity Cone., tu
30-Day
Average
30
30
0.1
mL 23
15
1.5
7-Day
Average
45
45
-
-
2.0
Daily
Maximum
60
60
0.2
230
0.1
20
2.5
*6.5 > pH < 8.5 at all times.
3.8.1.3 Pilot Plant Results
a. Pilot Facilities Description
The pilot facilities consisted of 12 experimental cells,
6.1 m wide (20 ft), 61 m long (200 ft), and
approximately 1.2 m deep (4 ft) (17). The 12 cells
consisted of three groups of four cells each (See
Figure 3-6). The depth of water in the cells was set
approximately at either 0.3 or 0.6 m (1 or 2 ft) using
60° V-notch weirs. Seepage from the cells was
prevented by the use of clay in cell bottoms and
berms. Although the cells were initially seeded with
alkali bulrush and hardstem bulrush, the marsh cells
went through a succession of aquatic plants with the
major species at the end of the three-year study
being hardstem bulrush, cattails, water cress, marsh
pennywort, and duckweed.
b. Experimental Design
The first year of the three-year study was devoted to
construction of the experimental facilities and the
establishment of the marsh system. The remaining
two years of the study were spent documenting the
performance of the 12 cells under steady state
operation. The experimental design consisted of
combinations of three hydraulic loading rates (2,400,
1,200, and 600 m3/ha-d) (260,000, 130,000, 65,000
gpd/ac) and two water depths (0.3 and 0.6 m [1 and
2 ft]). This design provided a replicate for each of the
combinations. However, variations in actual weir
heights and measured flow rates from the design
values resulted in variation in the hydraulic loading
rates and the hydraulic detention time between
replicate cells (see Table 3-7). In the second year of
operation the hydraulic loading rate in the first four
cells was reduced from 2,400 to 300 m3/ha-d
(260,000 to 32,500 gpd/ac).
The influent to the marsh system was effluent from
the City's 22.3 ha (55 ac) oxidation pond. The twelve
cells were routinely monitored for influent and effluent
BOD5, SS, total and fecal coliform, organic nitrogen,
ammonia, nitrate, phosphates, metals, pH, DO,
turbidity and toxicity by bioassay. In addition, several
27
-------
Figun 3-6. Arcata, CA pilot marsh system.
JAMES CREEK
MARSH
ENHANCEMENT
PROJECT
JOLLY GIANT C
WASTEWATE
?...-. TREATMENT PL,
PILOT P
ARCATA BAY
'X.^ OXIDATION
^~~-~>_ POND
"\
//S£0\M E NTATION
^/ ( POND
INFLUENT FROM PUMPING DOCK IN OXIDATION POND
f STILLING TOWER DRAIN LINES T ^
t t t
1 1
li
2
~tr
il
-L
i
I
-^x^U...
7~
J
4
-LE'
5
t
'
6
—
/
•z-.
7
8
t
9
t
r~
1O
t
SUMP-
h
t
r~
>
,20'
H
i
12
4
r^
8
28
-------
Table 3-7. Arcata, CA Pilot Marsh System Hydraulic Loading Rates and Detention Times (18)
Marsh Cell Number
1
8
10
12
Flow, m3/d
9/11/80-12/31/81
1/1/82-1 -731 782
Hydraulic Loading, m3/m2-d
9/1/80-12/31/81
I/1/82-1-/31/82
Depth, m
9/1/80-10/31/82
Detention Time, hr
9/1/80-12/31/81
1/1/82-1-/31/82
17.2
3.5
0.24
0.05
0.55
52
257
17.1
1.5
0.24
0.02
0.40
38
411
15.2
1.4
0.19
0.02
0.61
65
697
15.8
1.5
0.22
0.02
0.36
37
369
9.0
9.0
0.12
0.12
0.49
88
88
8.2
8.2
0.11
0.11
0.30
59
59
8.4
8.4
0.11
0.11
0.55
106
106
8.4
8.4
0.11
0.11
0.33
58
58
5.4
5.4
0.07
0.07
0.55
180
160
5.4
5.4
0.07
0.07
0.33
90
90
4.4
4.4
0.06
0.06
0.50
183
183
4.3
4.3
0.06
0.06
0.35
132
132
tracer studies and a disinfection efficiency study were
performed.
Following the first pilot study an additional study was
sponsored by the State Water Resources Control
Board (18). The focus of this study was to determine
the effect of harvesting on performance, wildlife, and
mosquito production and to investigate indicator
organism speciation and removal. Ten of the original
12 cells were restructured and rehabilitated for the
study by harvesting the plants from portions of some
cells and installing baffles in others (see Table 3-8).
The hydraulic loading rate in all cells was maintained
at 700 m3/ha-d (74,000 gpd/ac) and water level was
set at 0.6 m (2 ft), providing a theoretical hydraulic
detention time of 7.5 days. Harvesting was
accomplished by hand, using a weed eater, machetes
and rakes.
Table 3-8. Experimental Vegetation and Compartments for
Marsh Cells - Arcata, CA (32)
Cell 1 Effluent 50 percent of cell harvested3
Cell 2 100 percent of cell harvested
Cell 3 Cell left intact from previous season (Hardstem bulrush)
Cell 4 Alternating 6-m (20-ft) strips harvested starting with
effluent 6m (20 ft)b
Cell 5 Cell left intact from previous season (Cattail)
Cell 6 Marsh cell divided into four 30 m x 6 m (100 ft x 20 ft)
compartments
Cell 7 Marsh cell divided into four 15 m x 6 m (50 ft x 20 ft)
compartments
Cell 8 Marsh cell divided into four 7.5 m x 6 m (25 ft x 20 ft)
compartments
Cell 9 Alternating 15-m (50-H) strips harvested with the last
15 m (50 ft) interval vegetated
Cell 10 Influent 50 percent of cell was harvested
a Cells were harvested In November 1984.
b Cellular compartments were constructed with baffles which allow
flow to transfer to next compartment by use of three V-notch
weirs.
As in the first pilot study, oxidation pond effluent was
the influent to the marsh cells and the marsh cells
were monitored routinely for influent and effluent
BOD5, SS, ammonia, nitrate, phosphates, pH, DO,
turbidity, total and fecal coliforms. Special studies of
indicator organism speciation and mosquito
populations were also performed.
c. Experimental Results
Influent BOD5 during the first pilot study averaged
24.5 mg/L with a standard deviation of 12.3 mg/L.
The average removal percentage for all cells was 46
percent. As expected the lower hydraulic loading
rates produced better effluent quality. The respective
percentage removals for 2,400, 1,200, 600, and 300
m3/ha-d (260,000, 130,000, 65,000, and 32,500
gpd/ac) were 35, 45, 55, and 75 percent (see Table
3-9). Although there were seasonal increases in the
influent BOD5 concentration associated with algal
blooms, variation in BOD5 removal in the marsh
system appeared to be due to factors which are
difficult to quantify, such as the succession of plant
species rather than seasonal factors such as
temperature and sunlight. In general, the marsh
system proved to be effective at all the loading rates
investigated in producing an effluent which would
meet the discharge requirements with the exception
of the first spring and summer of operation.
In interpreting the results with respect to BOD5
removal, the investigators used a mathematical model
developed by Atkinson, in which it is assumed that
soluble substrate is removed by a thin film of
attached microorganisms (17):
-In [Se/SJ = (f h k0) w Z •*• Q
Se = effluent concentration, mg/L
S, = influent concentration, mg/L
f = proportionalityy factor
h = thickness of slime, m
ko = maximum reactor rate, d"1
w = width of section (width of cell), m
Z = filter depth (length of cell), m
Q = volumetric flow rate, m3/d
(3-15)
29
-------
Table 3-9
Average Annual BOD5 Concentration
Arcata, CA (17)
1981
Influent
1*
2
3
4
5
6
7
8
9
10
11
12
Mean
27.3
18.7
18.2
18.1
17.7
16.1
13.8
18.0
23.9
16.7
17.8
14.3
13.6
Std.
Dev.
13.4
8.7
9.3
7.3
6.8
9.1
8.0
7.7
31.3
10.5
11.2
9.6
8.8
1982
Mean
21.9
2.8
7.8
5.7
5.3
10.7
6.4
7.9
6.2
7.1
7.7
4.4
4.7
Std.
Dev.
9.5
3.8
6.2
4.4
5.1
5.4
3.8
4.7
4.0
4.5
5.4
2.8
3.2
(mg/L) -
1981-1982
Mean
25.2
13.9
10.7
14.0
16.8
12.8
13.7
10.4
10.1
Std.
Dev.
12.3
8.2
7.6
8.3
25.8
9.8
10.5
9.1
8.2
* Effluent from Cell 1.
The Arcata investigators determined (f h kj to be
4.95 m/d. They also compared effluent BOD5 to the
BOD6 mass loading and proposed a linear
relationship for determining the areal requirements for
a marsh system treating oxidation pond effluent.
Influent SS averaged 34.9 mg/L with a standard
deviation of 18.9 mg/L. Although there was a larger
average value and range for the influent SS
concentrations as compared to BOD5, the effluent
values were very stable during the study and did not
vary significantly with loading rate. The SS removal
averaged 85 percent for all loading rates with the
highest and lowest average removal rates being 87
percent and 83 percent.
During the first two-year study, no violations of
Arcata's NPDES toxicity standard in either the marsh
influent or effluent from any of the experimental cells
were observed. Total nitrogen removal was measured
during a six-month period and averaged 30 percent
for all cells. Average ammonia removal over the two-
year period varied from 0 to 33 percent for the
various cells. Average phosphorus removal was quite
low, with the highest average removal percentage
being 10 percent.
A disinfection study was conducted in the summer of
1982 to demonstrate that lower SS and pH in the
marsh effluent would result in a lower chlorine
demand. However, it was discovered that the marsh
effluent contained both volatile and non-volatile
compounds which resulted in a higher chlorine
demand compared to the oxidation pond effluent. It
was shown that the volatile compounds, such as
hydrogen sulfide, could be removed by air stripping,
thereby reducing the chlorine demand. It was further
concluded that SS are not a major factor in chlorine
demand in the Arcata system.
The second pilot study was primarily concerned with
the effects of harvesting and baffles on the
performance of marsh systems. In summary, it was
concluded that harvesting resulted in statistically
significant degradation of effluent quality for BOD5
and statistically non-significant degradation for SS.
Baffles did not significantly increase or decrease
effluent BOD5 and SS compared to the control cells
which were neither baffled nor harvested. The results
of the study with respect to the impacts of harvesting
and baffles on the removal of other pollutants were
inconclusive.
No conclusions were drawn concerning the effects of
harvesting and use of baffles on mosquito
populations. The results of the 1985 sampling were
compared with earlier years and it was concluded that
the mosquito population had decreased and in
general, the pilot marsh cells produced approximately
the same densities of mosquitoes as the adjacent
natural marsh system.
Approximately 90 percent of the total coliforms and
more than 95 percent of the fecal coliforms were
removed in the pilot marsh cells. Based on the results
of sampling for 35 species of bacteria, significant
differences in the species composition of the influent
and effluent were not observed.
3.8.1.4 Design Factors
Design of the final treatment system at Arcata was
largely influenced by the existing facilities and by the
results of the first pilot study. It was decided to use
the previously constructed Arcata Marsh and Wildlife
Sanctuary as a final polishing marsh system and to
convert a portion of the existing aerated ponds into an
intermediate marsh system. A flow diagram for the
overall wastewater treatment system is provided in
Figure 3-7.
The primary purpose of the intermediate marsh
system is to remove SS prior to chlorination and
dechlorination. The surface area of the intermediate
marsh system was determined based on a short-
term study, designed to determine the maximum
hydraulic loading rate that met effluent SS standards.
Although a maximum rate was not identified, the
maximum rate used in the study, 12,000 m3/ha-d
(1.28 mgd/ac), provided acceptable effluent SS
levels. The full-scale intermediate marsh system is
16.2 ha (4 ac), which results in design average and
maximum month hydraulic loading rates of 5,400 and
14,000 m3/ha-d (0.58 and 1.5 mgd/ac).
The intermediate marsh system was designed with
several 15-m (50-ft) stretches of open space which
span the full width of the marsh cell (see Figure 3-
8). The purpose of the open space is to provide a
30
-------
Figure 3-7. Arcata, CA wastewater treatment facilities flow diagram.
.ARCATA MARSH AND
WILDLIFE SANCTUARY*,
r CHLORINE
CONTACT
TANK
PRIMARY
CLARIFIERS
INTERMEDIATE
TREATMENT' MARSH
RETURN
PUMP
STATION
ARCATA
BAY
OXIDATION
PONDS
31
-------
Figure 3-8. Arcata, CA intermediate FWS system.
/OPEN WATER
HARDSTEM BULRUSH
(SCIRPUS ACUTUS)
OXIDATION
POND 3
32
-------
habitat for fish which will control the mosquito
population and for wildlife. Another mosquito control
measure was the decision to plant the marsh system
with hardstem bulrush because it allows fish better
access to the planted areas. Routine solids removal
from the intermediate marsh system is planned.
It is anticipated the effluent quality from the
intermediate marsh will meet the discharge
requirements the majority of the year. However, to
enhance the Humboldt Bay waters through the
creation of a wildlife habitat, the City of Arcata is
required to pass a minimum of 8,700 m3 (2.3 Mgal) of
effluent through the Arcata Marsh and Wildlife
Sanctuary daily. The resulting hydraulic loading rate is
700 m3/ha-d (74,000 gpd/ac).
3.8.1.5 Operating Characteristics
One-half of the intermediate marsh system was
planted in March 1986 and the other half during the
summer months of 1987. Currently the water level in
the two intermediate marsh cells is set at 1.1 m (3.5
ft) and the total influent flow is passed evenly through
the cells. Once both cells are fully established the
water level will be reduced to 0.6 m (2 ft). After
chlorination and dechlorination, approximately 11,350
m3/d (3 mgd) of effluent is diverted to the polishing
marsh system at the Arcata Marsh and Wildlife
Sanctuary. Effluent from the polishing marsh is
chlorinated, dechlorinated and discharged to Arcata
Bay. Effluent flows in excess of 1,350 m3/d (3 mgd)
from the oxidation pond are chlorinated, dechlorinated
and discharged directly to Humboldt Bay.
The intermediate marsh system is in the process of
plant establishment and startup. As expected, the
wastewater treatment performance of the one cell
planted in March 1986 was not very good in the first
season, but the operators are confident the
intermediate marsh system will perform as designed
as it matures biologically and when the second cell is
planted and brought into service.
The polishing marsh system has been receiving
effluent from the intermediate marsh system since
June 1986. Influent and effluent BOD5 and SS values
for the first seven months of operation are
summarized in Table 3-10. The polishing marsh has
performed as expected in terms of BOD5 removal but
not as well as hoped in terms of SS removal. The
primary cause for the high effluent SS has been
algae. It is expected that the planting of vegetation
near the effluent collection point will lower the SS
concentrations.
Mosquitoes have not been a problem in either of the
marsh systems. Chlorine usage following the
intermediate marsh system has remained the same
as before the intermediate marsh was constructed.
There was a problem with stickleback fish being
carried over with the polishing marsh effluent into the
Table 3-10.
Arcata, CA Marsh and Wildlife Sanctuary
Wastewater Treatment Plant Performance
Influent, mg/L
Effluent, mg/L
BOD5
SS
BOD,
SS
Aug 1986
Sept
Oct
Nov
Dec
Jan 1987
Feb
Average
34
32
41
46
48
32
20
36.1
49
52
46
39
55
32
27
42.9
8
6
7
21
20
15
19
13.7
17
13
15
42
39
35
58
31.3
chlorine contact tank. This problem was solved by
installing a smaller mesh screen in front of the
effluent pipes and by establishing thick vegetation in
the effluent area.
3.8.1.6 Costs
The construction of both the intermediate and final
wetlands systems at Arcata was totally financed by
the City of Arcata. The total cost of the Arcata Marsh
and Wildlife Sanctuary project (final wetland system)
was $514,600 including planning and environmental
studies and land acquisition. A cost breakdown
summary is provided in Table 3-11.
Table 3-11. Arcata, CA Marsh and Wildlife Sanctuary
Project Expenditures
Item cost, $
Plan of Study 14,000
EIR, Management Plan, and Permits 20,500
Land Acquistion 76,100
Construction 235,000
Accessways (trails, etc.) 19,000
Expenditures from Treatment Plant Modifications 150,000
to Transport and Return Wastewater
Total 514,600
3.8.1.7 Monitoring
In addition to the monitoring required in the discharge
permit, influent and effluent water quality for both of
the marsh systems is monitored weekly for BOD5 and
SS. Monitoring for mosquitoes and vegetation
coverage is on a regular basis.
3.8.2 Emmitsburg, Maryland
3.8.2.1 History
In 1984 the town of Emmitsburg, Maryland was facing
a sewer connection moratorium from the state water
quality regulatory agency because of wastewater
discharge violations. Planning was underway to
construct new treatment facilities but in the interim
33
-------
the town needed to upgrade its existing facilities to
avoid the moratorium. The town decided to use a
SFS constructed wetland system to treat a portion of
its effluent flow. The design and construction was
cooperatively undertaken by the town and the SaLUT
Corporation.
The system was started up in the summer of 1984
and continued in operation until March 1986, at which
time the system did not receive any wastewater for
several days. The resulting stress on the system
eventually caused the death of all the cattails. The
system was reseeded in October 1986.
3.8.2.2 Project Description
The Emmitsburg system is a single basin, 76.3 m
(250 ft) long, 9.2 m (30 ft) wide, and 0.9 m (3 ft)
deep, filled with 0.6 m (2 ft) of crushed rock. Clay
was used in the bottom of the basin to prevent
ground-water contamination. Perforated pipes placed
near the bottom of the basin are used for influent
distribution and effluent collection. The water level
during normal operation is approximately 5 cm (2 in)
below the surface of the gravel. The system was
seeded with 200 broadleaf cattail plants in August
1984 and another 200 plants in July 1985. By March
1986 approximately 35 percent of the basin surface
area was covered by cattails. The planting density
used in this project should have been at least an
order of magnitude higher. Until the plants cover the
entire basin, performance will not be representative of
a SFS system as defined in this manual.
The influent to the Emmitsburg system is trickling
filter effluent. Influent flows have varied between 95
and 132 m3/d (25,000-35,000 gpd), which
corresponds to a surface hydraulic loading rate of
1,420-1,870 m3/ha-d (152,000-200,000 gpd/ac).
Effluent samples are collected and analyzed weekly
for BOD5, SS, total dissolved solids, DO, and pH.
3.8.2.3 Operating Characteristics
The influent BOD5 concentrations to the wetlands
system range between 10 and 180 mg/L while SS
concentrations normally range between 10 and 60
mg/L. Results from two years of operation are
presented in Table 3-12 and Figures 3-9 and 10.
As can been seen in Figures 3-9 and 10, the
performance of the wetlands system has been very
good even with the limited plant coverage. Odors in
the effluent have been an occasional problem but the
frequency of noticeable odors is decreasing as cattail
coverage increases.
3.8.2.4 Costs
The Emmitsburg system was designed by SaLUT,
Inc. and constructed primarily as an in-house project
by the township's Public Works Department.
Engineering and construction costs of the
Emmitsburg system were less than $35,000.
3.8.3 Gustine, California
3.8.3.1 History
The City of Gustine, California is a small agricultural
town on the west side of the San Joaquin Valley. The
city treats approximately 4,500 m3/d (1.2 mgd) of
wastewater of which about one-third originates from
domestic and commercial sources and the remainder
from three dairy products industries. The wastewater
is of high strength, averaging over 1,200 mg/L BOD5,
which reflects the industrial component of the waste.
Until recently, the city's wastewater treatment plant
consisted of 14 oxidation ponds operated in series.
The ponds covered approximately 21.8 ha (54 ac)
and provided about 70 days detention time. Treated
effluent was discharged without disinfection to a small
stream leading to the San Joaquin River.
As with many oxidation pond systems in the United
States, mandatory secondary treatment levels were
not achieved with any consistency. The discharge
regularly exceeded 30 mg/L SS and periodically
exceeded 30 mg/L BOD5.
The city applied for and received federal funding to
analyze alternatives and develop a facilities plan.
Alternatives included the following:
• Oxidation pond treatment followed by land
application (irrigation).
• Oxidation pond treatment followed by reuse in the
form of seasonal flooding of local duck clubs to
attract migrating water fowl.
• Oxidation pond treatment followed by effluent
polishing to meet secondary treatment standards
for river disposal using sand filters, microscreens,
or submerged rock filters.
• Conventional activated sludge treatment to meet
secondary treatment standards for river disposal.
• Oxidation pond pretreatment followed by effluent
polishing in a constructed marsh (using emergent
aquatic vegetation) to meet secondary treatment
standards for river disposal.
From an analysis of the alternatives it was found that
the oxidation pond/constructed marsh was the most
cost-effective solution. The advantages of this
alternative were that suitable land was available, the
treatment method was compatible with the
surrounding area (a lowland area with naturally
occurring aquatic vegetation and virtually no
development) and the consumption of very little
energy.
3.8.3.2 Design Objectives
The primary objective of the Gustine project is to
upgrade the treatment plant effluent quality to meet
30 mg/L BOD5 and SS (30-day average). Mosquito
34
-------
Table 3-12. Performance of the Emmitsburg, MD SFS (19)
Average Flow,
Season m3/d
Fall 1984
Winter 1985
Spring 1985
Summer 1985
Fall 1985
Winter 1988
117
111
130
100
97
106
BUU5,
Influent
29
68
117
87
28
40
mg/L
Effluent
12
29
38
11
7
11
OO. 1
•ng/L
Influent Effluent Effluent DO, mg/L
25
37
37
28
29
25
7
9
13
10
7
4
1.0
0.3
0.0
1.3
2.1
Odor of Effluent
strong
noticeable
occasional
none
occasional
Area Covered
With Cattails, %
<5
-------
control is the second major concern. Mosquito
production has been identified as a major drawback to
the use of an aquatic plant treatment system. The
need to control mosquito production affected the
design of the facility, and will affect operating
procedures as well.
3.8.3.3 Pilot Plant Results
a. Description of the Pilot Facility
A one-year pilot testing program using a 0.4-ha
(1-ac) cattail marsh was initiated to develop design
criteria for a full-scale system. The pilot marsh was
15 m (50 ft) wide by 270 m (875 ft) long. There was a
healthy stand of cattails (Typha spp.) growing on the
site prior to application of wastewater.
Earth berms enclosed the marsh on all sides.
Wastewater was pumped to the marsh through plastic
piping and distributed across the influent end of the
marsh through a manifold with nine valved outlets.
Effluent was collected at the lower end and
discharged with the normal plant effluent.
The condition of the marsh bottom grading was not
known. An average depth between 0.15 and 0.3 m
(0.5 and 1 ft) was estimated but an actual average
depth could not be determined.
b. Experimental Design
The two variables adjusted during the pilot study were
influent source and detention time. Water depth in the
marsh was kept constant so flow rate was used to
vary detention time. Flow rates ranged from 136 to
380 m3/d (36,000-100,800 gpd), corresponding to
actual detention times of 1.3-3.8 days, and hydraulic
surface loading rates of 340-1,000 m3/ha-d
(36,600-108,900 gpd/ac).
Selection of the influent source was based on the
assumption that algae entering the marsh would
penetrate it and would be measured as SS in the final
effluent. There is a visible transition from the first
ponds in the series without significant algal growth, to
ponds (in the latter half of the series) with algal
growth. It is believed that the main mechanism
limiting growth in the first cells is restricted light
penetration due to high turbidity and scum formation.
Five of the 14 oxidation ponds in service at Gustine
were alternately used as an influent source, with
visual observation being the primary method of
selection. The pond farthest downstream without
significant algal growth was selected with the goal of
avoiding high concentrations of algae in the marsh
influent.
Marsh influent and effluent levels of BOD5, SS, pH,
and temperature were measured twice per week. In
addition, two dye tests were performed to measure
the time of flow through the marsh. Routine effluent
total coliform measurements were made, but detailed
bacterial testing was not conducted. (It is assumed
that disinfection of the final effluent will be required.)
Eleven sampling stations were used to monitor
mosquito larva production during the latter portion of
the pilot study.
c. Experimental Results
During a 2-I/2 month "start up and acclimation"
period, both marsh influent and effluent BOD5 and SS
levels were high, at one point exceeding 400 mg/L
and 250 mg/L, respectively. The very high influent
BOD5 was probably due to insufficient pretreatment.
To reduce the influent BOD5 and SS loadings the
source pond was changed from the eighth pond in
the flow sequence to the tenth, farther downstream in
the process. The initial detention time in the marsh
system was 2.1 days.
Influent and effluent levels of BOD5 and SS for the
period December 1982-October 1983 are illustrated
in Figures 3-11 and 3-12. The influent source pond
and marsh detention time for each time period is
shown at the top of the figures. Effluent levels were
generally below 30 mg/L from May through October
1983. The removal of BOD5 was particularly good in
the latter part of the summer and fall, averaging 74
percent. The rise in effluent BOD5 in late July and
early August corresponds with high effluent SS levels.
These increases could be the result of concentration
due to loss of water through evapotranspiration
(calculated to be up to 45 percent in the summer).
SS removal was particularly good following startup,
averaging 80 percent during April, May, and June,
and increasing to 89 percent from July through
August.
There was some suspicion that nitrogenous oxygen
demand was affecting BOD5 test results. In an
evaluation performed at the end of the study
(summarized in Table 3-13) it was found that
nitrification increased the BOD5 readings by up to 16
mg/L. This additional oxygen demand often made the
difference between meeting or exceeding 30 mg/L
effluent BOD5.
The effect of detention time on removal efficiency
was difficult to establish because relatively short
detention times were used. Removal as a function of
detention time is reported in Table 3-14. Because
the data were collected at different times of the year
and at different loading rates, direct conclusions can
not be made. In general it appears that a detention
time of 2.7-3.8 days is the minimum necessary for
adequate treatment during warm weather.
Bacteriological testing consistently showed levels
greater than 2400 MPN/IOO mL. It is assumed that
disinfection of the final effluent will be required.
The marsh was sampled for mosquito larvae by the
Merced County Mosquito Abatement District from
36
-------
Figure 3-11. BOD5 performance data for Gustine, CA pilot marsh system.
Source: Pond 8
Detention Time: 2.1 days
(actual)
Pond 9
3.8 days
BOD5, mg/L
400 r-
300 -
200 -
100
Pond 12
1.3 days
PondS
3.8 days
Jan. 1983
Oct. 1983
Figure 3-12. SS performance data for Gustine, CA pilot marsh system.
Source: Pond 8
Detention Time: 2.1 days
(actual)
Pond 9
3.8 days
Pond 7
2.7 days
Pond€
2.7 day
Pond 12
1.3 days
Pond 8
3.8 days
200 -
100 -
Jan. 1983
Oct. 1983
37
-------
Table 3-13. Determination of Nitrification Component in BOD5 Test - Gustine, CA (201)
Test Date Station3
10/13/83" Influent
1+00
2 + 00
3 + 00
4 + 00
5 + 00
6 + 00
7 + 00
8 + 00
Effluent
10/6/83c Effluent
11/3/83c Effluent
Water Temp.°C Standard
19 251
93
50
48
21
22
13
22
14
30
20 33
20
BOD5, mg/L
With Nitrification
Inhibitor
244
81
48
39
11
20
10
19
7
25
17
14
Difference
7
12
2
9
10
2
3
3
7
5
16
6
a Stationing measured from effluent end of marsh, each station 30.5 m (100 ft) apart.
b Tests performed by UC Davis Environmental Engineering Laboratory.
c Tests performed by California Water Lab, Modesto, CA.
Table 3-14. BOD5 and SS Removal Efficiencies As a
Function of Detention Time - Gustine, CA (20)
Actual
Detention
Time, d
1.3
2.1
2.7
3.8
Removal Efficiency*, %
BOD5
49
48
74
68
S S
61
28
89
80
Time Period
3/10-4/4
1 2/23-3/9
7/7-10/13
4/12-7/6
* Removal efficiencies based on average influent and effluent
concentrations over the time period covered.
June 10 through October 20, 1983. The average
number of mosquito larvae at each of the 11 sampling
stations ranged from 3.0 to 7.8 larva per dip. Both
Culex pipiens and Culex tarsalis larva were found in
about equal numbers. Based on experience and data
of the abatement district, it was concluded that a
marsh of the type tested may represent a mosquito
breeding source which must be subject to control
measures.
3.8.3.4. Design Factors
Based on research conducted at the University of
California, Davis (21) it has been shown that low
winter temperatures will control system sizing. Longer
detention time is required for BOD5 removal due to
reduced biological activity.
It is expected that the lowest water temperatures will
occur in January or February and be about 5°C
(40°F). At this temperature and a BOD5 loading rate
of 112 kg/ha-d (100 Ib/ac-d), using data from
U.C.D. research, it appears that a maximum detention
time of about 11 days is required. At the design flow
of 3,785 m3/d (1 mgd), a marsh area of about 9.3 ha
(23 ac) at a water depth of 0.45 m (1.5 ft) would be
required. The corresponding hydraulic loading is 407
m3/ha-d (43,500 gpd/ac). Detention time can
probably be reduced to as little as 4 days during the
warmest summer months.
The primary method of varying detention time will be
depth control. The ability to control water depth, and
to drain cells completely, is necessary to facilitate
harvesting of the plants, and other maintenance
activities. Such operational flexibility is a key design
factor.
Annual burning of the dormant marsh vegetation will
be practiced to enhance treatment by maintaining
plug flow characteristics, and to control mosquito
production. Individual cell width is limited to 12-15 m
(40-50 ft) by the access requirements. Levees
separating cells must be capable of accommodating
service vehicles.
Stocking of the mosquito larva eating fish, Gambusia
affinis, is the primary method of mosquito control.
High mass loading leads to low oxygen levels
produced by increased biological activity. Low oxygen
levels inhibit the movement of fish so loading rates
should be maintained between 112 and 168 kg/ha-d
(100 and 150 Ib/ac-d). This influent is to be
distributed to avoid organic "hot-spots." Levee side
slopes are steep, and vegetation is managed to allow
penetration of fish throughout the system. Bottom
slopes are designed to facilitate rapid draining of the
cells, if required, to interrupt the reproductive cycle of
the mosquitoes.
The above factors were used to develop the design
criteria summarized in Table 3-15.
3.8.3.5 Description and Operating Characteristics
of the Treatment System
A schematic of the completed marsh treatment
system is shown in Figure 3-13. Pretreatment is
accomplished in up to 11 of the existing oxidation
38
-------
Table 3-15. Design Criteria for Constructed Wetland at
Gustine, CA (22)
Item Value
Effluent Criteria, mg/L
BOD5,
s s
Design Flow, m3/d
Area, ha
Surface Hydraulic Loading, ma/ha-d
Depth, m
Detention Time, d
Inlets
Outlets
30
30
3,785
9.7
380
0.1-0.45
4-1 1
Head end of channels,
and one-third point
Adjustable weirs
ponds operated in series. Following the ponds, 24
marsh cells, each about 0.4 ha (1 ac) in size, operate
in parallel. The operator can draw wastewater from
any one of the last seven oxidation ponds. This
method of operation allows the operator to control the
detention time in the ponds from 28 to 54 days, and
to adjust the degree of pretreatment. The operator
can thus avoid applying heavy concentrations of algae
which develop in the latter ponds throughout the
summer.
Pond effluent flow is split into six parts in a
distribution structure and each portion of the flow is
directed to a group of four marsh cells. Each of the
24 cells is 11.6 m (38 ft) wide, 337 m (1,107 ft) long
and has an adjustable water depth of 10-45 cm (4-
18 in). Levees, 3 m (10 ft) wide, separate the cells
from one another.
Flow is introduced across the width of the marsh cells
at the head end and also at a point one third of the
total length from the head end. The initial flow split
will be 67 percent at the head, and 33 percent at the
one-third point. Overloading of the inlet zone of the
cell is thus avoided. The reverse arrangement (33
percent at the head and 67 percent at the one-third
point) can also be used with flow from the first third
used to dilute the flow applied to the second two
thirds.
Effluent from each cell flows over an adjustable weir
used to control water depth in the cell. The effluent is
then pumped to a disinfection process prior to
discharge.
Hydraulic detention time is controlled by varying the
number of cells in service, and by varying the water
depth in each cell. The operator is offered great
flexibility in attaining the desired detention time in the
marsh which varies from about four days in the
summer to 11 days in the winter. This operational
flexibility allows the cells to be taken out of service
sequentially each summer for vegetation management
and other maintenance requirements.
Up to 12 cells may be taken out of service at any one
time. The initial operating schedule listing the number
of cells and the hydraulic detention time for each
month is presented in Table 3-16.
In September 1986, six of the marsh cells were
seeded with locally acquired hardstem bulrush
rhizomes and six cells were seeded with locally
acquired cattail rhizomes. The specified minimum
planting density was 11 rhizomes/m2 (1/sq ft for
bulrush rhizomes and 5 rhizomes/m2 (0.45/sq ft) (18
in center-to-center) for cattail rhizomes. For both
species, rhizomes were distributed in the marsh cells
at a density greater than specified using a manure
spreader and then were disked into the soil with a
disk-harrow. The contractor was responsible for
assuring that the rhizomes germinated and that a
healthy crop was established. However the early
winter rainfall which the contractor had planned on for
germination was very low in 1986 and the rhizomes
did not germinate.
The marsh cells were replanted in June 1987. Both
bulrush and cattail rhizomes were purchased from a
nursery in Michigan and were planted using a tomato
planter.
Several methods of vegetation management are
under consideration, including mechanical harvesting
and burning. Most management techniques require
that the cell be taken out of service and allowed to
dry. A ramp was constructed into each cell to allow
access of harvesting equipment, should the
mechanical harvest option be selected.
Six marsh cells have been planted in bulrush. Stands
of bulrush are generally less dense than stands of
cattail and therefore allow greater movement of
mosquito fish. If treatment in the bulrush cells is
comparable to that in the cattail cells, replanting of
the marsh in bulrush may be considered.
3.8.3.6 Costs
Bids for the Gustine treatment plant improvements
were received in August 1985. Approximate costs for
the marsh system portion of the project were
extracted from the lump sum bid and are summarized
in Table 3-17.
City-owned land for the Gustine project was
available. Land requirements for this system were 9.7
ha (24 ac) net for the area actually planted and about
14.5 ha (36 ac) gross for the whole marsh system,
including all interior cell divider levees and an outer
flood protection levee.
3.8.4 Fabius Coal Preparation Facility
3.8.4.1 History
The Tennessee Valley Authority (TVA) processed
coal at the Fabius Coal Preparation Plant, located in
39
-------
Figure 3-13. Gustine, CA marsh system flow schematic.
o
0.
o
o
_s>X
o
20.
7
I
Ifc
I
K
O
at
5
o
40
-------
Table 3-16. Initial Operating Schedule of the Gustine, CA
Marsh System (23)
Month No. Cellis in Service, Hydraulic Detentron Time, d
January
February
March
April
May
June
July
August
September
October
November
December
24
24
20
16
16
12
12
12
16
20
24
24
11
11
10
8
6
5
4
4
6
8
10
11
Table 3-17. Capital Costs for Gustine,
(23)
CA Marsh Project
Item
Pond effluent piping3
Earthwork"
Flow distribution structure0
Flow distribution piping in marshd
Marsh cell water level control structure6
Marsh effluent collection piping'
Planting0
Paving11
Total
cost, $
(August 1985)
192,000
200,000
16,000
205,000
27,000
83,000
69,000
90,000
882,000
a Includes 790 m (2,600 H) Of 53-cm (21-in) PVC gravity piping.
five manholes, seven pond outlet control pipes with wooden
access platformss.
b Total earthwork volume, approx. 334,000 m3 (45,000 cu yd).
Cost includes clearing and grubbing, extra effort to work in area
of very shallow ground water and to construct a 2-m (6.5-H)
high outer levee to enclose the marsh area and protect it from the
100-year flood.
0 A concrete structure with V-notch wears, grating, access stairs,
and handrail.
d Approximately 850 m (2,800 H) of 20-cm (8-in) PVC gravity
sewer pipe, 760 m (2,500 H) of 20-cm (8-in) gated aluminum
pipe, and wooden support structures with concrete base slabs for
the gated pipe installed at the one-third of length point.
e Small concrete structures in each cell with weir board guides and
60-mm (0.24-in) mesh stainless steel screen.
f Approximately 460 m (1,500 ft) of 10-38 cm (4-5 in) PVC
gravity sewer pipe plus manholes.
e Based on mechanical planting of bulrush and cattail rhizomes on
45 and 90-cm (18 and 36-in) grid, respectively. Total
bulrush area of about 2.4 ha (6 ac); 7.2 ha (18 ac) for cattails.
h Aggregate base paving of the outer levee and selected inner
levees of the marsh area.
Jackson County, Alabama, from 1971 to 1979. In
1979 the facility was closed and in 1984 reclamation
efforts were started. One of the main reclamation
efforts was to involve the two coal refuse (coal slurry)
disposal ponds which have a combined water surface
area of 17 ha (42.5 ac) (see Figure 3-14). The coal
slurry water stored in the ponds is treated and
discharged but until recently seepage from the toe of
the impoundment dam was not treated. Seepage
flows range from 45 to 150 m3/d (11,900-39,600
gpd) and contain high concentrations of iron and
manganese. DO is less than 2.0 mg/L, SS exceed 98
mg/L, and pH averages 6.0.
In April 1985, it was decided to experiment with a
marsh/pond wetland system for treating the two large
and several small seeps (seepage flows) emanating
from the impoundment dam.
3.8.4.2 Project Description
A series of four wetland areas were created in June
1985 by clearing 1.2 ha (3 ac) of woodlands,
constructing four dikes with overflow spillways in the
drainage path of the combined seepage flows, and
transplanting a number of wetlands plant species in
the diked areas (see Figure 3-15). Transplanted
species, selected from nearby acid seeps, included
bulrush (Scirpus), rush (Juncus), spikerush
(Eleocharis), cattails (Typha), and scouring rush
(Equisetum).
The water surface area of all the wetlands areas is
approximately 0.6 ha (1.48 ac), and water depth
varies from 0 to 1.5 m (0-5 ft) in the larger ponds.
Based on seepage flows of 45-150 m3/d (8-27
gpm), the surface hydraulic loading rate on the
system is 75-250 m /ha-d (8,000-26,700 gpd/ac).
After the clearing and dike construction were
completed, it was discovered that there are additional
seeps entering the wetlands system. One large and
one small seep emerge at Pond 3 and two smaller
seeps emerge at Pond 4. The four large wetland
areas were stocked with mosquitofish and fathead
minnows.
In July 1985, water quality sampling of the two major
influents, the final effluent and at four locations in the
wetlands system was initiated. Samples were
collected biweekly and analyzed for pH, redox
potential (Eh), DO, total iron, total manganese, and
total SS.
An experiment to determine the capacity of the
wetlands system for treating coal slurry water in
addition to the seepage was performed in December
1985.
During a four-week period, supernatant from the
coal slurry ponds was fed into the wetlands system at
a flowrate of 110-220 m3/d (29,000-58,000 gpd).
Effluent quality dropped below discharge
requirements at the next sampling and the experiment
was stopped two weeks later. Starting in May
experiments with treating supernatant flows were
resumed but at much lower flow rates (5.4 m3/d
[1,400 gpd)
41
-------
RECLAIMED
PREPARATION PLA
SITE
RECLAIMED
IMPOUNDMENT-
IMPOUNDMENT
WETLANDS
EARTH DAM
^IMPOUNDMENT 1
WETLANDS
\
TO DRY CREEK
-------
Figure 3-15. Fabius Coal Facility Impoundment 1 wetlands.
SEEP 2
jfX ^—SANDBAG
DIKES
SEEP AREA
EEP 1
AG DIKE
•u/
POND 1
CEMENT BAG SPILLWAY
POND 2
\ POND 3
BOUNDARY OF
CLEARED
FINAL DISCHARGE
MAJOR TYPES OF
PLANTED VEGETATION
CATTAIL
BULLRUSH
100
SCALE
43
-------
3.8.4.3 Operating Characteristics
The range and average water quality values for the
influent, pond 1 effluent and final effluent with the
wetlands system during a 12-month period treating
coal slurry seepage and supernatant are summarized
in Table 3-18. The average values include results
during the four-week period when a relatively large
flow of coal slurry supernatant was treated. Wetlands
effluent water quality during periods when only
seepage was being treated was always better than
the discharge requirements. Compared with the
alternative, chemical treatment, the wetlands system
is much simpler in operation and maintenance and
would appear to be more stable in terms of effluent
quality.
3.8.4.4 Costs
The construction of the wetlands system at Fabius
was performed by TVA personnel and cost
approximately $28,000.
3.8.5 Summary
Although the four case studies presented in this
chapter cover only portions of the range of possible
applications of constructed wetlands, they represent
four different approaches to the use of a constructed
wetlands system for wastewater treatment. A
comparison of the four systems is difficult but a
summary of each system's design and operating
characteristics and costs is provided in Table 3-19.
Constructed wetlands systems offer several potential
advantages as a wastewater treatment process.
These potential advantages include simple operation
and maintenance, process stability under varying
environmental conditions, lower construction and
operating costs, and in the case of free water surface
systems, the possibility to create a wildlife habitat.
The potential problems with free water surface
constructed wetlands include mosquitoes. Startup
problems in establishing the desired aquatic plant
species can be a problem with FWS and SFS
wetlands alike.
44
-------
Table 3-18. Fabius Coal Preparation Facility Marsh System Performance (24)
Period
7/85-9/85
10185-12/85
1/86-3/86
4/86-6/86
7/86-9/86
10/86-12/86
Pond 4 Effluent
Flow, m3/d
52.3
53.4
91.6
62.7
26.7
107.4
Inf.
6.0
6.0
4.7
6.3
pH
Eff1
6.4
-
6.5
6.1
6.4
6.4
DO, mg/L
Eff4 Inf.
6.6 0
6.5 0
6.5
6.3
7.0
6.6
Eff1
6.2
10.9
10.9
5.3
8.2
Eff4
-
7.2
11.2
7.4
7.3
9.7
Fe, mg/L
Inf.
80
97
-
59
40
Eff1
2.6
9.3
3.5
14.7
4.3
Eff4
0.64
0.79
0.94
0.71
0.70
0.63
Mn, mg/L
Inf.
8.7
9.9
-
-
18
8.6
Eff1
1.4
-
11.2
3.1
2.6
6.2
Eff4
0.43
0.48
5.9
2.1
1.1
1.6
SS, mg/L
Inf.
95
74
-
155
48
Eff1
8.9
33
19.3
46.7
18.3
Eff4
2.2
4.0
2.8
4.7
3.0
2.0
Table 3-19. Constructed Wetlands Case Studies Summary
Type
Aquatic Plants
System Type
Influent
Special Design Features
Design Flow, m3/d
Wetlands Surface Area, ha
Influent/Effluent BOD5 mg/L
Influent/Effluent SS, mg/L
Hydraulic Surface Loading, m3/ha-d
Capital Cost, $/m3/d
Capital Cost, $/ha
Arcata, CA
Hardstem bulrush
Cattails
Free Water Surface
Oxidation Pond Effluent
Intermediate Wetlands
System Wildlife
Sanctuary
11,150
12.6
36.1/13.7
42.9/31 .3
907
45
41 ,000
Emmitsburg, MD
Cattails
Subsurface Flow
Tricking Filter Effluent
130
0.07
61.5/18.0
30.2/8.3
1,540
- 264b
- 495,000b
Gustine, CA
Cattails
Hardstem bulrush
Free Water Surface
Oxidation Pond Effluent
Variable Source of
Influent, Step Feed
3785
9.3
-150/- 24a
~140/- 19a
412
232
94,000
Fabius Coal Facility
Bulrush
Cattails
Rush
Spikerush
Free Waler Surface
Coal Slurry Pond
Seepage
227
0.6
374
a Pilot plant results.
b Costs are not representative of full-scale system costs.
3.9 References
When an NTIS number is cited in a reference, that
reference is available from:
National Technical Information Service
5285 Port Royal Road
Springfield, VA 22161
(703) 487-4650
1. Reed, S.C., E.J. Middlebrooks, and R.W. Crites.
Natural Systems for Waste Management and
Treatment. McGraw-Hill Book Co. NY, 1987.
2. Hyde, H.C., and R.S. Ross. Technology
Assessment of Wetlands for Municipal
Wastewater Treatment. U.S. Environmental
Protection Agency, EPA/600/2-84-154, NTIS
No. PB 85-106896, 1984.
3. Miller, I.W.G., and S. Black. Design and Use of
Artificial Wetlands. In: Ecological Considerations
in Wetland Treatment of Municipal Wastewaters,
Van Nostrand Reinhold Co., NY, pp. 26-37,
1985.
4. Zirschky, J. Basic Design Rational for Artificial
Wetlands. ERM-Southeast, Inc, prepared for
USEPA RSKERL, Ada, OK, 1986.
5. Process Design Manual: Land Treatment of
Municipal Wastewater. U.S. Environmental
Protection Agency, Cincinnati, Ohio. EPA-
625/1-81/013, 1981.
6. Bavor, H.J., D.J. Roser, and S. McKersie.
Nutrient Removal Using Shallow Lagoon-Solid
Matrix Macrophyte Systems. In: Reddy, K.R. and
W.H. Smith (Eds). Aquatic Plants for Water
Treatment and Resource Recovery. Magnolia
Publishing Inc. pp. 227-235, 1987.
7. Gearheart, R.A., and B.A. Finney. Utilization of
Wetlands for Reliable Low-Cost Wastewater
Treatment - A Pilot Project. Paper Presented to
IV World Congress on Water Resources, at
Buenos Aires, Argentina, September 5-9, 1982.
45
-------
8. Stowell, R., G. Tchobanoglous, J. Colt, and A.
Knight. The Use of Aquatic Plants and Animals for
the Treatment of Wastewater. Departments of
Civil Engineering and Land, Air, and Water
Resources, University of California, Davis, pp.
639-645, September 1979.
9. Tchobanoglous, G., and G. Gulp. Aquaculture
Systems for Wastewater Treatment: An
Engineering Assessment, U.S. Environmental
Protection Agency, Office of Water Program
Operations, Washington, D.C. EPA/430/9-80-
007, NTIS No. PB 81-156689, pp. 13-42, 1980.
10. Knight, R.L. Wetlands - A Natural Land
Treatment Alternative. Proceedings of the
Conference: Reuse and the Protection of
Florida's Waters, Sept. 17, 1984.
11. Gersberg, R.M., B.V. Elkins, C.R. Goldman.
Nitrogen Removal in Artificial Wetlands. Water
Res. 17:1009-1014, 1983.
12. Gersberg, R.M., S.R. Lyon, BV. Elkins, and C.R.
Goldman. The Removal of Heavy Metals by
Artificial Wetlands. In: Proceedings of the Water
Reuse Symposium III, San Diego, CA. AWWA
Research Foundation, Denver, CO, 1985.
13. Hantzsche, N.N. Wet/and Systems for
Wastewater Treatment: Engineering Applications.
In: Ecological Considerations in Wet/and
Treatment of Municipal Wastewaters, Van
Nostrand Reinhold Co. NY, pp. 7-25, 1985.
14. Wolverton, B.C. Artificial Marshes for Wastewater
Treatment. In: Aquatic Plants for Water
Treatment and Resource Recovery, Magnolia
Publishing, Inc., Orlando, FL, 1987.
15. Stephenson, M. et. al. The Use and Potential of
Aquatic Species for Wastewater Treatment.
Appendix A. The Environmental Requirements of
Aquatic P/ants. SWRCB Publication No. 65.
Sacramento, CA. October 1980.
16. California State Water Resources Control Board,
CRWQCB, North Coast Region, Water Quality
Control Plan, Klamath River Basin I-A,
September, 1975.
17. Gearheart, R., et al., Final Report, City of Arcata
Marsh Pilot Project, Volume 1, Effluent Quality
Results - System Design and Management,
Project No. C-06-2270, April, 1983.
18. Gearheart, R., et al., Final Report, City of Arcata
Marsh Pilot Project, Wet/and Bacteria Speciation
and Harvesting Effects on Effluent Quality, Project
No. 3-154-500-0, January, 1986.
19. Thiesen, A., and C.D. Martin. Municipal
Wastewater Purification in a Vegetative Filter Bed
in Emmitsburg, Mary/and. In: Aquatic P/ants for
Water Treatment and Resource Recovery,
Magnolia Publishing, Inc., Orlando, FL, pp. 295-
298, 1987.
20. Nolte and Associates, Marsh System Pilot Study
Report, City of Gustine, California, EPA Project
No. C-06-2824-010, November, 1983.
21. Stowell, R. et a I., Mosquito Considerations in the
Design of Wet/and Systems for the Treatment of
Wastewater, Dept. of Civil Engineering, U.C.
Davis, Davis, California, and Vector Biology and
Control Branch, California State Dept. of Health
Services, Sacramento, California, December
1982.
22. Crites, R., and T. Mingee, Economics of Aquatic
Wastewater Treatment Systems, In: Aquatic
P/ants for Wafer Treatment and Resource
Recovery, Magnolia Publishing, Inc., Orlando, FL,
1987.
23. Nolte and Associates, Operation and
Maintenance Manual, City of Gustine Wastewater
Treatment Facility Improvements, November
1986.
24. Brodie, G., et al., Treatment of Acid Drainage
from Coal Facilities with Man-Made Wetlands,
In: Aquatic P/ants for Water Treatment and
Resource Recovery, Magnolia Publishing, Inc.,
Orlando, FL, 1987.
46
-------
CHAPTER 4
Design of Aqua tic Plant Systems
4.1 Background
Aquatic plant systems are engineered and
constructed systems that use aquatic plants in the
treatment of industrial or domestic wastewater. They
are designed to achieve a specific wastewater
treatment goal. Aquatic plant systems can be divided
into two categories:
1. Systems with floating aquatic plants such as water
hyacinth, duckweed, pennywort; and
2. Systems with submerged aquatic plants such as
waterweed, water milfoil, and watercress
Until recently, most of the floating aquatic plant
systems have been water hyacinth systems. Use of
water hyacinth for wastewater treatment in the United'
States can be traced back to field-scale experiments
in Texas and laboratory research by NASA
researchers at the Bay St. Louis Experimental Station
in Mississippi carried out in the early 1970s. Water
hyacinths have been used in a variety of experimental
and full-scale systems for treating various quality
wastewaters.
However, use of water hyacinth has been limited, in
geographic location, to warm weather regions
because of the sensitivity of water hyacinth to
freezing conditions. Water hyacinth systems have
been most often used for either removing algae from
oxidation pond effluents or for nutrient removal
following secondary treatment. Since a conference on
aquaculture systems for wastewater treatment at the
University of California's Davis campus in September
1979, additional data have been accumulated on the
use of aquatic plants in wastewater treatment (1,2).
Since 1970, aquatic treatment systems have been
used successfully in a variety of treatment
applications including secondary, advanced
secondary, and tertiary treatment. Most of the
performance data reported in the literature for these
systems have, however, been observational rather
than quantitative. Hydraulic detention time, hydraulic
loading rate, and organic loading rate are the most
common parameters used and needed to size aquatic
plant systems.
4.7.7 Characteristics of Aquatic Treatment
Systems
Aquatic treatment systems consist of one or more
shallow ponds in which one or more species of water
tolerant vascular plants such as water hyacinths or
duckweed are grown (3). The shallower depths and
the presence of aquatic macrophytes in place of
algae are the major differences between aquatic
treatment systems and stabilization ponds. The
presence of plants is of great practical significance
because the effluent from aquatic systems is of
higher quality than the effluent from stabilization pond
systems for equivalent or shorter detention times.
This is true, particularly when the systems are
situated after conventional pond systems which
provide greater than primary treatment.
In aquatic systems, wastewater is treated principally
by bacterial metabolism and physical sedimentation,
as is the case in conventional trickling filter systems.
The aquatic plants themselves, bring about very little
actual treatment of the wastewater (3). Their function
is to provide components of the aquatic environment
that improve the wastewater treatment capability
and/or reliability of that environment (4).
4.7.2 History
General reviews of the use of water hyacinth and
several other aquatic plants for wastewater treatment,
including duckweed, are presented elsewhere (2,5-
7). The locations of several pilot- and full-scale
tests which were significant in developing methods
and performance data for aquatic treatment systems
are presented in Table 4-1. The systems in
Mississippi and Texas (presented in Table 4-1)
received facultative pond effluent as their influent
source. Primary effluent was the influent wastewater
for the systems at Walt Disney World, San Diego, and
Hercules. All other locations in Table 4-1 received
secondary effluent into the floating aquatic plant
system.
Research at Walt Disney World in Florida (8) has,
since 1978, been directed toward an integrated
system including: 1) aquaculture treatment of
wastewater to meet federal, state and local standards,
2) biomass management for achieving optimum
yields, and 3) anaerobic digestion of harvested
47
-------
Table 4-1. History of Use
Location
University of Florida
Bay St. Louis, MSa
Lucedale, MS
Orange Grove, MS
Cedar Lake (Biloxi), MSb
Williamson Creek, TX
Austin-Hornsby, TX
Alamo-San Juan, TX
San Benito, TX
Rio Hondo, TX
Lakeland, FL
Waft Disney World, FL
Coral Springs, FL
Orlando, FL
University of California Davis,
Hercules, CA
Roseville, CA
San Diego, CA
of Floating Aquatic Treatment
Scale
Experimental
Full
Full
Full
Full
Pilot
Pilot/Full
Full
Full
Full
Full
Pilot
Full
Full
CA Experimental
Full
Pilot
Pilot
Systems (Water Hyacinths,
Objective
Research
Secondary
Secondary
Secondary
Secondary
Secondary
Secondary
Secondary
Secondary
Secondary
Tertiary
Secondary
Tertiary
Tertiary
Research
Advanced Secondary
Nitrification
Advanced Secondary
except as noted)
Date
1964-1974
1976
1970s
1970s
1979
1975
1970s
1970s
1976
1970s
1977
1978
1978
1985
1978-1983
1980-1981
1981
1981
Status
Completed
Ongoing
Abandoned
Abandoned
Ongoing
Abandoned
Ongoing
Abandoned
Ongoing
Abandoned
Ongoing
Completed
Abandoned
Ongoing
Completed
Abandoned
Abandoned
Ongoing
' Frost killed hyacinths; pennywort and duckweed are now used.
' Duckweed.
aquatic vegetation to produce methane for energy
recovery.
4.7.3 Climatic Constraints
The water hyacinth systems that are currently used to
treat wastewater in the United States are located in
the warm temperate climates of the southern states.
The optimum water temperature for water hyacinth
growth is 21-30°C (70-96°F). Air temperatures of
-3°C (-26°F) for 12 hours will destroy the leaves
and exposure at -5°C (-23°F) for 48 hours will kill
the plants. If a water hyacinth system were to be
used in a colder climate, it would be necessary to
house the system in a greenhouse and maintain the
temperature in the optimum range (9). Based upon
the limited data available, it would be uneconomical to
attempt to develop a water hyacinth wastewater
treatment system in cold regions (9). Duckweed
(Lemna spp.) is more cold tolerant than water
hyacinths and can be grown practically at
temperatures as low as 7°C (45°F) (10).
4.2 Vegetation
Aquatic plants have the same basic nutritional
requirements as plants growing on land and are
influenced by many of the same environmental
factors. The treatment responses in an aquatic plant
system are due to the presence of the plants in the
water system altering the physical environment of the
systems (11). Water hyacinth plant roots act as a
living substrate for attached microbial organisms,
which provide a significant degree of treatment (11).
4.2.7 Floating Plants
Floating plants have their photosynthetic parts at or
just above the water surface with roots extending
down into the water column. In photosynthesis,
floating aquatic plants use atmospheric oxygen and
carbon dioxide. Nutrients are taken up from the water
column through the roots. These roots are an
excellent medium for the filtration/adsorption of
suspended solids and growth of bacteria. Root
development is a function of nutrient availability in the
water and nutrient demand (i.e., growth rate) of the
plant. Thus, in practice, the density and depth of
treatment medium (i.e., plants roots) will be affected
by wastewater quality/pretreatment and factors
affecting plant growth rate such as temperature and
harvesting.
With floating plants, the penetration of sunlight into
water is reduced and the transfer of gas between
water and atmosphere is restricted. As a
consequence, floating plants tend to keep the
wastewater nearly free of algae and anaerobic or
nearly so, depending on design parameters such as
BOD5 loading rate, detention time, and the species
and coverage density of floating plants selected for
use (4). An observation of interest is that some
molecular oxygen produced by photosynthetic tissue
is translocated to the roots and may keep root zone
48
-------
microorganisms metabolizing aerobically, though the
surrounding water is anaerobic/anoxic (4).
4.2.1.1 Water Hyacinths
Water hyacinth (Eichhornia crassipes) is a perennial,
freshwater aquatic vascular plant with rounded,
upright, shiny green leaves and spikes of lavender
flowers (11) (See Figure 4-1). The petioles of the
plant are spongy with many air spaces and contribute
to the buoyancy of the hyacinth plant. When grown in
wastewater, individual plants range from 0.5 to 1.2 m
(20 to 47 in) from the top of the flower to the root tips
(11). The plants spread laterally until the water
surface is covered and then the vertical growth
increases. Hyacinths are very productive
photosynthetic plants. Their rapid growth is a serious
nuisance problem in many slow flowing southern
waterways. These same attributes become an
advantage when used in a wastewater treatment
system.
In the United States, this plant is widely distributed in
Alabama, California, Florida, Mississippi, Louisiana,
and Texas (5). After years of using expensive
physical and chemical control measures, the water
hyacinth problem has been generally reduced to
manageable levels through the use of the hyacinth
weevil (Neochetina eichhomiae and N. bruchi) and
hyacinth mite (Orthogalumna terebrantis). Both of
these biological control agents were imported from
South America, the native location of the water
hyacinth. The mite probably was introduced
accidentaly with the original water hyacinth plants at
the Cotton States Centennial Exposition in New
Orleans, Louisiana, in 1884 (12,13). These biological
control agents have reduced water hyacinth
populations to manageable levels so the hyacinth is
no longer considered a major concern for
maintenance of open waterways.
Water hyacinth is a rapid growing aquatic macrophyte
and is ranked eighth among the world's top 10 weeds
in growth rate (5). It reproduces primarily by
vegetative propagation, but seeds may be a major
source of reinfestation once the parent plants have
been removed. Water hyacinth also develops a large
canopy, which may provide a good competitive edge
over other floating aquatic plants growing in the same
system. Growth of water hyacinth is influenced by: 1)
efficiency of the plant to use solar energy, 2) nutrient
composition of the water, 3) cultural methods, and 4)
environmental factors (5).
Plant growth is described in two ways. The first is to
report the percentage of pond surface covered over a
period. The second more useful method is to report
the plant density in units of wet plant mass per unit of
surface area. Under normal conditions, loosely
packed water hyacinth can cover the water surface at
relatively low plant densities (10 kg/m2 [20 Ib/sq ft]
wet weight). It can reach a maximum density of 50
kg/m (100 Ib/sq ft) wet weight (5), before growth
ceases.
As in other biological processes, growth rates in
water hyacinth systems depend on temperature. Both
air and water temperature are important in assessing
plant vitality. Water hyacinths are reported to survive
24 hour exposure at temperatures of 0.5 to -5°C (33
to 31 °F) but die at temperatures of -6 to -7°C (21
to 19°F) and cannot become established in regions
where winter temperatures average 1 °C (34°F) (14).
Growth is rapid at 20-30°C (68-96°F) and nearly
stops at 8-15°C (46-59°F) (14). Suitable areas for
growing water hyacinths include the southern portions
of California, Arizona, Texas, Mississippi, Alabama,
Georgia, and Florida. Areas within the continental
United States where cultivation of water hyacinth
systems is possible are shown in Figure 4-2.
Hyacinth systems can be used to advantage in
correcting algal bloom problems in oxidation ponds.
Use of hyacinths in summer only is a technically
feasible solution for some rural systems that
experience discharge problems with high SS (from
algae).
4.2.1.2 Pennywort
Pennywort (Hydrocotyle umbellata) is not a free
floating plant; it tends to intertwine and grows
horizontally, and at high densities the plants tend to
grow vertically. Unlike water hyacinth, photosynthetic
leaf area of pennywort is small, and at dense plant
stands, yields are significantly reduced as a result of
self shading (15). Pennywort exhibits mean growth
rates greater than 10 g/m-d (73 Ib/ft-d) in central
Florida (15). Although rates of N and P uptake by
water hyacinth drop sharply during the winter, nutrient
uptake by pennywort is approximately the same
during both warm and cool seasons. Nitrogen and
phosphorus uptake during the winter months is
greater for pennywort than for water hyacinth (16).
Although annual biomass yields of pennywort are
lower than water hyacinth, it offers a great potential
as a cool season plant that can be integrated into
water hyacinth/water lettuce biomass production
systems (15).
4.2.1.3 Duckweed
Duckweeds are small, green freshwater plants with
leaflike frond from one to a few millimeters in width.
Lemna and Spirodela have a short root usually less
than 10 mm (0.4 in) in length (Figure 4-3).
Duckweed such as Lemna spp., Spirodela spp., and
Wolffia spp., have all been tested for pollutant
removal or used in wastewater treatment systems
(11). Potential growth areas for duckweed in the
United States are shown in Figure 4-3 for various
seasons of the year.
49
-------
Figure 4-1. Morphology of the hyacinth plant [Reprinted with permission of McGraw-Hill Book Company (11)).
Leaves
Roots
Stolon
Figure 4-2. Suitable areas for hyacinth systems (10).
Water Surface
Rhizome
50
-------
Figure 4-3. Morphology of and potential growth areas for duckweed plants (7).
Fronds
Water
Surface
Roots
MORPHOLOGY
LEGEND
Growth is likely during
all 12 months of the year
Growth is likely during
9 months of the year
Growth is likely during
6 months of the year
Duckweeds are the smallest and the simplest of the
flowering plants and have one of the fastest
reproduction rates. A small cell in the frond divides
and produces a new frond; each frond is capable of
producing at least 10-20 more during its life cycle
(9). Lemna sp. grown in wastewater effluent (at 27°C
[81 °F]) doubles in frond numbers, and therefore in
area covered, every four days. It is believed that
duckweed can grow 30 percent faster than water
hyacinths. The plant is essentially all metabolically
active cells with very little structural fiber (11).
Performance of some existing duckweed systems is
summarized in Table 4-2. Duckweed systems are
developed by following the conventional design
procedures for faculaltive lagoons. Effluent from a
duckweed covered system should exceed
performance of conventional facultative lagoons for
BOD5, SS and nitrogen removal (11). The effluent
from a duckweed system is likely to be anaerobic and
post aeration may be necessary. The advantage of
duckweed systems over a similar additional facultative
lagoon cell is lower algae concentrations in the
effluent. This is due to extensive shading of the water
by the mat of surface duckweed.
Duckweed, like hyacinths, contains about 95 percent
water; the composition of the plant tissue is given in
Table 4-3 (4). Duckweed contains at least twice as
much protein, fat, nitrogen, and phosphorus as
hyacinth. The value of duckweed as a food source for
a variety of birds and animals has been confirmed by
several nutritional studies (17).
Small floating plants, particularly duckweed, are
sensitive to wind and may be blown in drifts to the
leeward side of the pond. Redistribution of the plants
requires manual labor. If drifts are not redistributed,
decreased treatment efficiency may result due to
incomplete coverage of the pond surface. Also piles
of decomposing plants can result in the production of
odors.
4.2.2 Submerged Plants
Submerged plants are either suspended in the water
column or rooted in the bottom sediments. Typically,
their photosynthetic parts are in the water column, but
certain vascular species may grow to where their
photosynthetic parts are at or just below the water
surface.
The potential for use of submerged aquatic plants for
treatment of primary or secondary effluent is severely
limited by their tendency to be shaded out by algae
and their sensitivity to anaerobic conditions. The
mechanism by which submerged plants are able to
remove ammonia from the water column is related to
their photosynthetic processes which remove carbon
dioxide from the water (unlike hyacinths) thus raising
the pH and driving ammonia to the gaseous form
which can diffuse into the atmosphere. Ammonia gas
is the most toxic form of nitrogen for fish. This
51
-------
Table 4-2. Performance of Existing Duckweed Systems (7)
Location
Biloxi, MS
Collins, MS
Sleep Eye. MN (Del Monte)
Wilton, AR
NSTL, MS
Influent
Facultative Pond Effluent3
Facultative Pond Effluent
Facultative Pond Effluent
Facultative Pond Effluent3
Package Plant Effluent
BOD5
Influent
30
33
420
35.5
mg/L
Effluent
15
13
18
6.5
3.0
TSS,
Influent
155
36
364
47.7
mg/L
Effluent
12
13
34
7.4
11.5
Depth, m
2.4
0.4
1.5
2.7
0.4
Detention Time,
days
21
7
70
0.7
8
3 Partially aerated.
b Theoretical hydraulic detention time for duckweed cell only.
Table 4-3. Composition of Duckweeds Grown in
Wastewater (after 11)
Percent of Dry Weight
Constituent
Range
Average
Crude protein
Fat
Fiber
Ash
Carbohydrate
TKN
Phosphorus, as P
32.7-44.7
3.0-6.7
7.3-13.5
12.0-20.3
4.59-7.15
0.5-0.7
38.7
4.9
9.4
15.0
35.0
5.91
0.6
mechanism is of some concern for healthy
populations of mosquito fish which are generally
encouraged in aquatic treatment ponds for mosquito
control.
At night these plants respire (i.e., use oxygen) in
competition with the mosquito fish. It is generally
considered that this category of plant will not have
widespread usage in aquatic systems because of its
pronounced diurnal effect on the aquatic environment,
tendency to be shaded out by nuisance algae, and
sensitivity to anaerobic conditions (2). No pilot- or
full-scale systems using submerged plants are
known and therefore none are reported in this
manual.
4.3 Process Design Criteria for Water
Hyacinth Systems
Water hyacinth systems represent the majority of
aquatic plant systems that have been constructed.
Organic loading is a key parameter in the design and
operation of water hyacinth systems. Three types of
hyacinth systems can be described based on the
level of DO and the method of aerating the pond.
Aerobic hyacinth systems without supplemental
aeration will produce secondary treatment or nutrient
(nitrogen) removal depending on the organic loading
rate. This type of system is most common of the
hyacinth systems already constructed. Its advantages
include few mosquitoes or odors.
For a system location in which no mosquitoes or
odors can be tolerated, an aerobic system with
supplemental aeration is required. The added
advantage of such a system is that, with aeration,
higher organic loading is possible and reduced land
area is required. The characteristics of these two
systems are summarized in Tables 4-4 and 4-5.
The third configuration for a hyacinth system is to
operate it under high organic loading. The purpose is
to achieve secondary treatment, and these systems
are capable of producing consistent treatment without
aeration under high organic loading. Disadvantages
include increased mosquito populations and potential
for odors. The early systems at Disney World were
this type (called facultative/anaerobic in this manual).
Facultative/anaerobic systems are not commonly
being designed any more because it has been
recognized that organic loading rates of up to 100
kg/ha-d (89 Ib/ac-d) produce consistent results
without the disadvantages of high loading.
4.3.1 Organic Loading Rates
BOD5 loading rates for water hyacinth systems can
range from 10 to 300 kg/ha-d (9-268 Ib/ac-d) (see
Table 4-5). For primary effluent loading at Disney
World, Florida, the basins were loaded at 55-440
kg/ha-d (50-400 Ib/ac-d), without significant odor
problems except at the higher loadings. Average
loadings on the entire system without aeration should
not exceed 100 kg/ha-d (89 Ib/ac-d).
4.3.2 Hydraulic Loading Rate
Hydraulic loading rate, expressed in units of m3/ha-
d, is the volume of wastewater applied per day
divided by the surface area of the aquatic system.
The hydraulic loading rates applied to water hyacinth
facilities have varied from 240 to 3,570 m3/ha-d
(25,650-381,650 gpd/ac) when treating domestic
wastewaters (9). For secondary treatment objectives
(BODj and SS<30 mg/L), the hydraulic loading rate
is typically between 200 and 600 m3/ha-d (21,600-
64,600 gpd/ac). For advanced secondary treatment
with supplemental aeration, hydraulic loading rates of
1,000 m3/ha-d (107,000 gpd/ac) have been used
successfully. However, organic loading rates will
generally control hydraulic loading.
52
-------
Table 4-4. Types of Water Hyacinth Systems
Type Purpose
Typical BOD5
Loading, kg/ha-d
Advantages
Disadvantages
Aerobic Non-aerated Secondary Treatment
Aerobic Non-aerated
Aerobic Aerated
Nutrient Removal
40-80
10-40
Secondary Treatment 150-300
Facultative/Anaerobic* Secondary Treatment 220-400
Limited mosquitoes;
limited odors
Limited mosquitoes;
limited odors
nutrient removal
No mosquitos; no odors;
higher organic loading rates;
reduced land area
Higher organic loading rates;
reduced land area
More land area required;
harvesting may be more
difficult (depends on pond
configuration)
More land area required;
harvesting may be more
difficult (depends on pond
configuration)
Additional harvesting required;
supplemental power required
Increased mosquito
population; potential for odors
a Only suitable where odors and mosquitoes may not be a problem.
Table 4-5. Design Criteria for Water Hyacinth Systems
Type of Water Hyacinth System
Factor
Aerobic Non-aerated
Aerobic Non-aerated
Aerobic Aerated
Influent Wastewater
Influent BOD5, mg/L
BOD5 Loading, kg/ha-d
Expected Effluent, mg/L
BOD5
SS
TN
Water Depth, m
Detention Time, days
Hydsraulic Loading, m3/ha-d
Harvest Schedule
Screened or Settled
130-180
40-80
<30
<30
415
0.5-0.8
10-36
> 200
Annualy
Secondary
30
10-40
<10
<10
<5
0.6-0.9
6-18
< 800
Twice per Month
Screened or Settled
130-180
150-300
<15
<15
<15
0.9-1.4
4-8
550-1,000
Monthly
4.3.3 Water Depth
The recommended depth of hyacinth ponds is 0.4-
1.8 m (1.2-6 ft) with the majority of investigators
recommending a depth of 10.9 m (3 ft) (9). The
critical concern is to provide adequate depth for the
root system to penetrate through the majority of the
liquid flowing through the hyacinth pond. A greater
depth is sometimes recommended for the final cell in
a series of hyacinth ponds since the hyacinth roots
will be longer when fewer nutrients are present in the
water (11). Recommended depth from the San Diego
project (with aeration) is 1.07-1.37 m (3.5-4.5 ft)
(18). For duckweed systems, operating depths of
1.5-2.5 m (58.2 ft) have been used.
4.3.4 Vegetation Management
The literature on water hyacinths as a wastewater
treatment process contains considerable speculation
on the use of the water hyacinth upon harvesting (8).
Composting, anaerobic digestion for the production of
methane, and the fermentation of the sugars into
alcohol are techniques proposed as a means to
partially recover the costs of wastewater treatment.
These digestion techniques may have use in a large
scale operation; however, it is unlikely that a typical
small wastewater treatment production system will
approach the economic break-even point from
methane gas production.
The need for harvesting depends on the water quality
objectives for the project, the growth rates of the
plants, and the effects of predators such as weevils.
Harvesting of aquatic plants is needed to maintain a
crop with high metabolic uptake of nutrients. For
example, frequent harvesting of hyacinths (every
three to four weeks) is practiced in Florida to achieve
nutrient removal. Nitrogen and phophorus removal by
the plants is achieved only with frequent harvesting.
In areas where weevils pose a threat to healthy
hyacinth populations, selective harvesting can
theoretically be used to keep the plants from being
infected. The State of Texas recommends an annual
draining and cleaning of each basin instead of regular
plant harvesting (11). Duckweed harvesting for
nutrient removal may require frequencies of at least
one week during warm periods.
53
-------
The harvested plants are typically dried and landfilled.
The drying process may be a source of significant
odors. At Kissimmee, Florida, the hyacinths are
vermicomposted. Ground duckweed can be used as
animal feed without air drying.
4.3.5 Mosquitoes and Their Control
The objective of mosquito control is to suppress the
mosquito population below the threshold level
required for disease transmission or nuisance
tolerance level. Strategies that can be used to control
mosquito populations include (from [3]):
1. Stocking ponds with mosquito fish (Gambusia
afinis).
2. More effective pretreatment to reduce the total
organic loading on the aquatic system to help
maintain aerobic conditions.
3. Step feed of influent waste stream with recycle.
4. More frequent harvesting.
5. Application of man-made control agents.
6. Diffusion of oxygen (with aeration equipment).
Effective mosquito control is based on two very
difficult operational parameters: the maintenance of
DO at 1 mg/L and the frequent harvesting and
thinning of the water hyacinths. Supplemental
aeration has been employed at San Diego to maintain
this goal.
In many parts of the United States, the growth of
mosquitoes in aquatic treatment systems may be the
critical factor in determining whether the use of such
systems will be allowed (3). Fish used for control of
mosquitoes (typically Gambusia affinis) will die in
anaerobic conditions caused by organically
overloaded ponds. In addition to inhibited fish
populations, mosquitoes may develop in dense
hyacinth systems when plants have been allowed to
grow tightly together. Pockets of water form within the
plant body and are accessible to the mosquito larvae
but not the fish.
4.3.6 Suggested Design Parameters
Design parameters used to size aquatic systems
include hydraulic detention time, organic loading rate,
and less frequently nitrogen loading rate. Design
parameters based on the required level of treatment
have been summarized in Table 4-5. Design criteria
for effluent polishing using duckweed in facultative
ponds are summarized in Table 4-6.
4.3.7 Sludge Management
Sludge consists of both wastewater solids and plant
detritus. It must eventually be removed from aquatic
plant system ponds. The quantities of sludge that
Table 4-6. Design Criteria for Effluent Polishing With
Duckweed Treatment Systems
Factor Secondary Treatment
Wastewater Input
BOD5 Loading, kg/ha-d
Hydraulic Loading, m3/ha-d
Water Depth, m
Hydraulic Detention Time, days
Water Temperature, °C
Harvest Schedule
Facultative Pond Effluent
22-28
<50
1.5-2.0
15-25
>7
Monthly
accumulate were estimated at Williamson Creek,
Texas, to be 1.5 to 8 x 10~4 m3 of sludge/m3 of
wastewater treated (150-800 gal/mgd) (1). This
compares to 1.8 x 10"3 m3 of sludge/m3 of
wastewater treated (1,800 gal/mgd) for conventional
primary stabilization ponds. Generally the rate of
sludge accumulation in a pond containing water
hyacinths is a function of the pretreatment provided.
Very little information is available regarding sludge
accumulation for ponds with aquatic plants other than
water hyacinths. A cleaning frequency for hyacinth
ponds based on the degree of treatment and the
frequency of plant harvesting has been suggested
(11). Suggested cleaning frequencies are shown in
Table 4-7.
Table 4-7. Recommended Sludge Cleanout Frequency for
Water Hyacinth Ponds (11)
Pond Type Cleaning Frequency
Primary Cells in Shallow High-
Rate Systems
Secondary Cells
Tertiary Cells
Deep Secondary Cells
(regularly harvested)
Secondary Cells (irregularly
harvested)
Systems Used Only Seasonally
Annual
2-3 years
2-3 years
5 years
Annual
Annual
4.4 Physical Features of Aquatic
Treatment Systems
4.4.7 System Configurations
Most of the early hyacinth systems involved
rectangular basins operated in series similar to
stabilization ponds. Long, narrow channels were used
in the Disney World research in Florida.
The San Diego Aquaculture Project, as an example,
is a pilot-scale water hyacinth project for treatment
of primary effluent to secondary effluent quality. The
current configuration of this system has evolved to
solve earlier problems with hydrogen sulfide odors,
and presence of mosquito larvae in the ponds. The
solution to the above two major problems is reflected
54
-------
in the unique system configuration and influent
distribution system.
Early operating experience at the San Diego facility
indicated that hydrogen sulfide odors and mosquito
larvae were a problem. Because of the urban setting,
this pilot plant had stringent requirements of no odors
and no mosquito larvae being allowed. Initial solutions
included lower organic loading and recycle flow from
the effluent end of the pond to the front end to dilute
the influent flow and distribute the organic load more
completely throughout the pond. This solution was
only partly successful and organic loading rates had
to remain low to prevent anaerobic conditions at the
head end of the pond.
A series of tests of BOD5 concentration along the
length of the pond indicated that most of the BOD5
removal occurred in the first 15 m (50 ft) of the 120-
m (400-ft) pond. The most recent system
configuration includes recycle of effluent and step
feed of influent at eight locations approximately 15 m
(50 ft) apart along the length of the pond.
Supplemental aeration has also become a regular part
of the pond configuration.
The evolution of the distribution system experiments
which have resulted in the choice of the step feed
with recycle is shown in Figure 4-4. Step feed and
recycle should be used as operational tools to control
organic loading in the pond. With these tools, and a
pond with high length-to-width ratios (>10:1), the
operator can control the treatment process for best
performance. Pattern C represents the current San
Diego system and pattern D represents the
wraparound pattern planned for future facilities. The
wraparound design shortens recycle lines and step
feed lines.
4.4.2 Inlet and Outlet Structures
Shallow, rectangular basins with a high length to
width ratio are usually designed for aquatic treatment
systems to reduce the potential for short circuiting
and to simplify harvesting operations. The use of
baffles and influent distribution manifolds helps to
maximize the retention time. Influent manifolds and
multiple inlet (step feed) systems can also be used
effectively for recycling treated effluent to reduce the
influent concentrations of wastewater constituents. An
effluent manifold across the basin will maintain a low
velocity into the manifold which serves to maintain
quiescent conditions near the outlet. If variable
operating depths are planned, it should be possible to
remove effluent at a depth of 0.3 m (1 ft) below the
most shallow operating depth.
4.4.3 Supplemental Aeration
The need for aeration is derived from the strict need
for mosquito control and odor control. Aeration of the
ponds helps maintain DO> 1 mg/L for the mosquitoe
fish in the system and minimizes H2S gas production.
A successful configuration of the aeration system at
San Diego uses fine bubble diffusers. Fine bubble
diffusers produced DO levels 0.5-1.0 mg/L higher
than coarse bubble diffusers in similarly configured
ponds with the same BOD5 loading rate and total air
flow (Figure 4-5). The capacity of the aeration
system was equal to two times the BOD5 load. [A
method for sizing the aeration system is shown in
Sample Problem No. 2.] During daylight hours an
automated system will cycle on and off as required to
maintain DO 1 mg/L When water hyacinths are
actively photosynthesizing, they transport oxygen to
their roots and at the same time to the microbes
attached to the roots. This process lowers the
supplemental aeration equipment and associated
energy costs for aeration.
One method of supplemental aeration cited is spray
irrigation. In this method, wastewater is recycled
through spray heads onto the hyacinths. This
technique is also often cited for frost control in
climates where winter temperatures are marginal for
hyacinths. A spray recycle system in which more
temperature tolerant plants are used in order to
function as a living trickling filter has also been
proposed (19).
The San Diego project experimented with spray
irrigation for supplemental aeration and found both
effective aeration and lower mosquito larvae
populations (18). The lower vector population was
probably due to the disruption of the water/air
interface by the simulated rain effect during the
nighttime when the mosquitoes are actively breeding
(19).
An important negative effect of spraying was related
to hyacinth plant health. The plants in the San Diego
project which were sprayed began to show stress and
yellowing while plants just outside the spray area
flourished (16). Plant health was improved by limiting
spray period to 12 evening hours. Although spray
irrigation effectively raises dissolved oxygen levels,
and limits mosquito larvae populations, concerns over
increased TDS due to excessive evaporation and cost
of pumping tend to minimize the value of this
approach (18).
4.4.4 Operation and Maintenance of Aeration
DO should be measured at least twice per day. The
goal should be to maintain an average DO
concentration of 1-2 mg/L along the pond length. If
the DO falls below 1 mg/L, additional aeration should
be added, or the influent flow should be reduced, until
the pond recovers. This operation can be automated
and optimized with DO probes. Fine bubble diffusers
can develop a thick biological slime buildup, after
several months of operation, especially if aeration is
intermittent (18). Monthly inpond cleaning with a
coarse brush can be effective in at least temporarily
controlling the biological slime growth.
55
-------
Figure 4-4. Evolution of flow pattern through San Diego, CA water hyacinth treatment ponds: a)origianl plug-flow, b) plug-
flow with recycle, c) step-feed with recycle, d) step-feed with recycle in wraparound pond (18).
Influent
Effluent
Recycle
Step Feed
Recycle
T
Influent
Recycle
Effluent
I
4.5 Performance Expectations
4.5.1 Design Equations
The San Diego water hyacinth project examined a
series of time sequenced profile tests with various
recycle flows. The tests were done to determine 1)
the maximum allowable organic loading rate and 2)
the optimum recycle ratios (18). Based on the results
of the testing program, it was concluded in the San
Diego Aquaculture Project that the modified step feed
system could be modeled as a series of continuous
flow stirred tank reactors (18). [This flow diagram is
shown in Figures 4-8 and 4-13.1
4.5.1.1 BOD5 Removal
The steady state materials balance for the first
reactor in the series of eight reactors assuming first
order BOD5 removal kinetics is (18):
accumulation = inflow - outflow + generation
(4-1)
0=Qr(C8)+0.125 Q(c0) - (Qr+0.125 Q)(d) -
(4-2)
Where,
Q = recycle flow, m3/d
Cg = BOD5 concentration in effluent from
reactor number 8 in series, mg/L
0.125Q = inflow to each individual cell (Q-8),
C0 = BOD5 concentration in influent, mg/L
C° = BOD5 concentration in effluent from
reactor number 1 in series, mg/L
V1
= First order reaction rate constant at
temperature, T, d-t
= Volume of first reactor in series, m3
The estimated kT value to be used in Equation 4.2 is
1.95 d-t at 20°C (68°F). An important aspect of the
recycle system as shown in Figure 4-8a is that the
recycle ratio is 16:1 for the first reactor in series and
23:1 for the last reactor in series. If the recycle flow
had been mixed directly with the influent before being
applied to the pond, the recycle ratio would have
been 2:1. The difference between these two modes
of operation is significant with respect to the
performance of the pond.
4.5.1.2 Temperature Effect
Based on the results of the daily testing program, the
value of the temperature coefficient, 0, in the
following equation is estimated to be about 1 .06 (18).
kT =
20
n
-------
Figure 4-5. Evolution of Pond 3 flow and aeration system configurations at San Diego, CA: a) plug-flow with air diffusion
tubing (Hinde), b) step feed with recycle and coarse bubble aeration system, c) stepfeed with recycle and fine
bubble aeration system (18).
a. In operation May 84 - April 86
Aeration lines
./^typical) 8 at Z0C /I6
,..„ Aeration lines
- IOO' hi | I08' 4-32'+— IOO'
Influent header
Influent box
Pre -aeration
manhole
b. In operation May 86-October 87
Aeration lines Air supply line Stadium aeration
/ (typical) / /
:
?"
(
V
c
, j
—
8 -
— |
7 -1
6
/
/'"'
1
J
— 1
5 -1
4
*]
*J
^
*]
-
2
1
-J
I
->j
J
X
To drain
T"
.«w^
•influent
Cell number
Recycle
To pond 4
c. In operation November 87-Present
go g g
C7-
fine bubble
aerators
/ 1.5 ft2at 10'OC
Stadium aerator
lii
D Q D Q |O Q
Recycle
T
To pond 4
Air supply
Influent
57
-------
review of existing water hyacinth treatment systems,
Weber concluded that nitrification followed by
denitrification was the principal nitrogen removal
mechanism (20). Only when water hyacinth systems
received low nitrogen loadings and significant
harvesting was conducted did plant uptake become
the principal nitrogen removal mechanism (20).
The nitrogen removal for 54 data points from case
studies including locations at Coral Springs, FL;
Williamson Creek, MS; and University of Florida, FL
have been summarized (21). The results of these
studies are presented in Table 4-8 as the percent of
expected nitrogen removal for a particular surface
loading rate.
Table 4-8. Nitrogen Removal - Water Hyacinth Tertiary
Treatment (21)
Hydraulic Loading, m /ha-d
9,350
4,675
2,340
1,560
1,170
< 5935
Total Nitrogen Reduction, %
10-35
20-55
37-75
50-90
65-90
70-90
4.53 Phosphorus Removal
Phosphorus removal from aquatic macrophyte
systems is due to plant uptake, microbial
immobilization into detritus plant tissue, retention by
the underlying sediments, and precipitation in the
water column. Since P is retained by the system, the
ultimate removal from the system is achieved by
harvesting the plants and dredging the sediments
(22).
Phosphorus uptake in Florida marshes averaged 11
percent with signs of net export of phosphorus from
the marsh in the winter (23). Reddy and Tucker (24)
studied the productivity of water hyacinths grown with
various nitrogen sources. The optimum N/P ratio in
the water medium should be 2.3-5 to achieve
maximum biomass yields (24). This ideal range can
be used to estimate whether nitrogen or phosphorus
is limiting to water hyacinths in a particular pond
environment.
Phosphorus removal by precipitation/adsorption in
aquatic systems not involving plant harvesting was
found to be approximately 2 kg P/ha-d (1.8 Ib/ac-d)
in one study (4). Phosphorus may be removed prior
to applying wastewater to an aquatic system by a
chemical addition and precipitation reaction.
Precipitation may be the cost effective method of
removal depending on the degree of phosphorus
removal required.
4.6 Sample Design Problems
The following two sample design problems indicate
how the design criteria in Tables 4-4 and 4-7 can
be applied. Example No. 1 is also used to show the
use of Table 4-8 in estimating nitrogen removal.
4.6.7 Sample Problem No. 1:
Design a hyacinth system to produce secondary
effluent with screened, raw municipal wastewater as
influent.
Assume:
Design flow rate = 730 m3/d;
BOD5 = 240 mg/L
SS = 250 mg/L
TN = 20 mg/L
TP = 10 mg/L
critical winter temperature > 20°C.
Effluent requirements:
BOD5 <30 mg/L
SS < 30 mg/L.
Solution:
1. Determine BOD5 loading:
(240 mg/L) (730 m3/d) (103 L/m3) (1 kg/106 mg)
= 175kg/d
2. Determine basin surface areas required based on
criteria in Table 4-4:
50 kg/ha-d BOD5 for entire area
100 kg/ha-d BOD5 for first cell
Total area required = (175 kg/d) + (50 kg/ha-d)
= 3.5 ha
Area of primary cells = (175) •*• (100)
= 1.75 ha
3. Use two primary cells, each 0.88 ha in area; with
L:W =3:1, the dimensions at the water surface will
be:
Area of primary cells = L+ W = (L)(L/3) = L2 + 3
(0.88 ha) (10,000 m2/ha) = L2- 3
8,800 m2 = L2-3
L = 163 m
W = 163+3 = 54 m
4. Divide the remaining required area (3.5 ha - 1.75
ha = 1.75 ha) into two sets of two basins (four cells
- 0.44 ha each) to produce a total system with
two parallel sets with three basins each.
58
-------
2 .
Area of final cells = L + W = (L)(L/3) = L" + 3
(0.44 ha) (10,000 rrf/ha) = L" + 3
4,400 m2 = L2+3
L= 115 m
W = 115 m+3 = 38 m
5. Allow 0.5 m for sludge storage and assume a 1.2
m "effective" water depth for treatment; total pond
depth = 1.7 m. Use 3:1 sideslopes, and use the
equation below (approximate volume of a frustum)
to determine the treatment volume.
V = [(L) (W) + (L - 2sd) (W - 2sd)
+ 4 (L - sd) (W - sd)] d+6
Where,
v = volume of pond or cell, m3
L = length of pond or cell at water surface, m
W = width of pond or cell at water surface, m
s = slope factor (e.g., 3:1 slope, s = 3)
d = depth of pond, m
Primary cells:
V=[(163) (54) + (163-20301.2) (54-20301.2)
+4 (163 - 2-1.2) (54 - 221.2)] 1.2+6
v = 9,848 m3
Final cells:
V=[(115) (38) + (115 - 2Z3-1.2) (38 - 2-3-1.2)
+ 4(115 - 2 x 1.2) (38 - 2-1.2)] 1.2+6
V = 4,745 m3
6. Determine the hydraulic detention time in the
"effective" treatment zone:
Primary cells:
t = (2) (9,848 m3) + (730 m3/d) = 27 days
Final cells:
t = (2) (4,745 m3) + (730 m3/d) = 26 days
Total detention time =27 + 26
= 53 days > 40 days, OK
7. Check hydraulic loading:
(730 m3/d) + (3.5 ha) = 209 m3/ha-d
>200 m3/ha-d, OK
8. Estimate nitrogen removal with Table 4-8 to be
sure sufficient nitrogen is present to sustain growth
in the final cells and to determine harvest
frequency. Hydraulic loading is:
209 m3/ha-d
From Table 4-8 essentially 90 percent removal is
predicted at a hydraulic loading <935 m3/ha-d.
Since the hydraulic loading for this example is 250
m3/ha-d it is reasonable to expect 5 mg/L of
nitrogen in the final effluent or less. Because the
nitrogen will not be at optimum growth levels in this
system an annual harvest is suggested. An influent
flow diffuser in each of the primary cells is
recommended to properly distribute the untreated
influent.
4.6.2 Sample Problem No. 2:
Design an aerated hyacinth system with recycle to
produce secondary effluent on a site with limited
available area.
Assume:
Design flow = 730 m3/d
BOD5 = 240 mg/L
SS = 250 mg/L
TN = 20 mg/L
TP = 10 mg/L
winter water temperature = 20°C.
Effluent requirements:
BOD5 <30 mg/L
SS<30 mg/L
Assume 80 percent plant coverage is maintained on
the basins and routine monthly harvests are included.
Solution:
1. Since the site area is limited, space is not available
for preliminary treatment in a pond unit. Use Imhoff
tanks for primary treatment and supplemental
diffused aeration in the hyacinth ponds to minimize
area requirements. The Imhoff tank has the added
advantage for this relatively small flow in that
separate sludge digestion is not required.
2. Design the Imhoff tank.
Typical criteria:
Sedimentation detention time = 2 hr
Surface loading = 24 m3/m2-d
Overflow weir loading = 600 m3/m-d
Surface area for scum = 20% of total surface
Sludge digestion volume = 0.1 m3/capita for the
population served, or
about 33% of total
tank volume
Minimum sedimentation area needed
= (760 m3/d) + (24 m3/m2-d) = 31.7 m2
Surface area needed for scum
= (0.20) (31.7 m2) = 6.3 m2
Total surface area needed
= sedimentation area + scum area
=31.7 + 6.3 = 38 m2
59
-------
A typical tank might be 8 m long and 5 m wide. In
this case the central sedimentation chamber might
be 4 m wide with open channels on each side,
about 0.5 m wide, for scum accumulation and gas
venting. The slotted, sloping bottom (bottom walls
sloped at 5:4) would have to be about 3 m deep to
provide the necessary 2-hour detention time. The
total depth of the hopper bottomed tank might be
6-7 m including an allowance for freeboard and
the sludge digestion volume.
A properly maintained Imhoff tank can achieve
about 47 percent BOD5 removal and up to 60
percent SS removal. Assuming no nitrogen or
phosphorus losses the primary effluent for this
example would be:
BOD5 = (240 mg/L) (0.53) = 127 mg/L
SS = (250 mg/L) (0.40) = 100 mg/L
TN = 25 mg/L
TP = 15 mg/L
3. The BOD5 loading on the hyacinth basins would
be:
(127 mg/L) (730 m8/d) (103 L/m3) (1 kg/106 mg)
= 92.7 kg/d
4. Determine the basin volume using Equation 4-2.
Assume a recycle ratio of 2:1 as in the San Diego
case. Also, design the system with step feed at
eight points as shown in Figure 4-5. In order to
solve Equation 4-2, the concentration of the
effluent from the eight sections of the basin can be
estimated from the recycle ratio as shown in Figure
4-13 in the case studies.
0=Qr(C8)+0.125 Q(C0) - (Qr+O.I25 Q)(d) -
Where,
Q
(4-2)
= recycle flow, m/d
= 2Q = 2(730) = 1,460 m3/d
C8 = BOD5 concentration in effluent from
reactor number 8 in series, mg/L
= C0-8 = 127 + 23 = 5.52 mg/L
0.125Q = inflow to each individual cell (Q-8),
m3/d
= 730-8 = 91.25 m3/d
C = BOD5 concentration in influent, mg/L
° = Imhoff tank effluent = 127 mg/L
C| = BOD5 concentration in effluent from
reactor number 1 in series, mg/L
= C0+16 = 127+16 = 7.94mg/L
kT = first order reaction rate constant at
temperature, T, d-t
= 1.95 d'1 at 20°C
V1 = volume of first reactor in series, m3
= total Volume + 8, m3
0 = (1,460) (5.52) + (91.25) (127)
- (1,460 + 91.25) - (1.95) (7.94)
0 = 8,059 + 11,589 - 12,317 - 15.5 V!
V, = 7,331 - 15.5 = 473 m3
Total Basin Volume = 8 (V^ = 8 (473) = 3,784 m3
5. Calculate the number of ponds required. Refer to
Table 4-5 for pond dimensions. Length, width and
depth should be 122 m x 8.5 m x 1 m (400 ft x 28
ft x 3.3 ft), respectively. These pond dimensions
result in a volume of 745 m3 (1 96,000 gal). For the
total basin volume required, 3,784 nr (1 Mgal),
five ponds will be needed.
Assume that the required oxygen is double the
organic loading, the air contains about 0.28 kg/m3
oxygen, and the aeration efficiency (E) in the
shallow basins is about 8 percent.
Total air required
= [2 (BOD5, mg/L) (Q, L/d) (10'6 mg/kg)]
+ [E (0.28 kg/ma)]
= [(2) (127) (730 x 103)(10-6] + [(0.08) (0.28)]
= 8,260m3/d = 5.73m3/min
From Table 4-5, maximum air flow per aerator is
0.028 m3/min. Since there are five ponds, the
number of aerators required per pond is:
Number Aerators
= (5.73 m3/min) + [(5 ponds) (0.028 m3/min)]
= 40.9 aerators/pond
Divide the aerators into eight sections within the
ponds as illustrated in Figure 4-5C. Each aerator
should have a surface area of 0.14 m2 (1.5 sq ft).
6. An inlet step feed system is essential for the ponds
to ensure uniform distribution of influent. The use
of Gambusia fish or other biological or chemical
agents are necessary for mosquito control. Plants
should be harvested about every three to four
weeks. No more than 20 percent of the plant cover
removed at any one time.
7. The treatment system designed in this example will
provide better performance than the system
developed in Example No. 1 on one-half to one-
third of the land area. The major reasons are the
use of the Imhoff tank for primary treatment, step
feed, and aeration. In locations where land is
limited or very expensive this approach to
treatment might be cost effective when secondary
level treatment is required.
The added cost of aeration equipment may be cost
effective where land area is at a premium. An
aerated hyacinth system becomes a hybrid system
60
-------
that is more complex than a natural aquatic system
as described in this manual and less complex than
conventional treatment with trickling filters or
rotating biological contactors.
4.7 Case Studies
The purpose of this section is to provide a view of the
state of the art in the design and operation of aquatic
plant systems by providing case study summaries of
three systems which are representative of current
knowledge and practice. The three systems are in
San Diego, California; Austin, Texas; and Orlando,
Florida. The San Diego pilot scale water hyacinth
system was chosen as a case study because it
attempts to treat primary effluent to secondary
effluent quality. The Austin water hyacinth system
was chosen because it utilizes a cover for frost
protection. The Orlando water hyacinth system was
chosen because it attempts to remove BOD5, SS,
nitrogen and phophorus from an effluent that has
undergone advanced secondary treatment.
4.7.7 San Diego, California
4.7.1.1 History
The City of San Diego depends on imported water for
at least 90 percent of its water supply. Recognizing
that its available water supply sources will fall short of
the projected needs by the year 2000, San Diego has
been working since the 1950s on ways to meet future
water demands. Early attempts at using secondary
treated wastewater for irrigation and distilling ocean
water into a potable supply were unsuccessful.
In 1964, San Diego began work on reclaiming water
by using reverse osmosis (RO) to remove salt.
Primary treated effluent was passed through RO units
for use in low pressure steam boilers. This 76-m3/d
(20,000-gpd) pilot plant successfully produced high
purity boiler feedwater. In a cooperative program
sponsored by the California Department of Water
Resources, it was found that the RO units at San
Diego also removed viruses. Based on that finding,
the use of reclaimed water to meet the City's water
requirements was first given serious consideration.
In 1974, the RO pilot plant was moved from Point
Loma to a site adjacent to the Jack Murphy Stadium.
At this location, the objective was to reclaim water for
irrigation of the stadium sod farm (25).
The demonstration wastewater reuse project known
as Aqua I, was operated from September 1981 to
June 1986. The complete pilot plant included the
following treatment processes: secondary treatment
with hyacinths, lime stabilization, ultrafiltration,
pressure sand filtration, reverse osmosis, carbon
adsorption, ozone and ultraviolet light disinfection,
and digestion of harvested hyacinths for methane
production. The capacity of the Aqua I treatment
facility was 114 m3/d (30,000 gpd).
The current San Diego pilot plant, described in this
case study summary, is an extension and expansion
of the previous facility. An advisory board was formed
to advise the San Diego researchers, to review
results, and to make recommendations for operation
of the pilot facility.
4.7.1.2 Project Description
Conceptually, the overall water reclamation program
in San Diego can be divided into four parts: aquatic
wastewater treatment using water hyacinths;
advanced water treatment using reverse osmosis to
produce raw potable water; anaerobic digestion of
hyacinth biomass and wastewater sludge to produce
methane; and a health effects study to compare the
risks of using reclaimed water to those using the
current water supply. The aquatic treatment portion of
the project consists of four separate phases and will
be completed in 1989.
Phase 1, completed in 1984, included the design and
construction of two alternative 1,890-m3/d (0.5-
mgd) primary facilities and four alternative 380-m3/d
(0.11-mgd) secondary facilities, including six water
hyacinth treatment ponds.
Phase 2, completed in 1986, included operation and
evaluation of the pilot plant under alternative
treatment schemes.
Phase 3, to be completed in 1989, will include
construction of the Water Reclamation Plant using the
aquatic treatment scheme selected on the basis of
the results from Phase 2. The pilot plant from Phase
2 will be scaled up to 3,785 m3/d (1 mgd) capacity,
with a 1,890-m3/d (0.5-mgd) advanced treatment
system added to reduce salt concentrations and to
further remove pollutants. An anaerobic reactor will be
included to produce methane gas from the water
hyacinths.
Phase 4 will include operation and evaluation of the
3,785-m3/d (1-mgd) facility.
4.7.1.3 Pilot Plant Results
The overall goal of the pilot program was to
demonstrate an innovative/alternative water
reclamation process with cost-effective recovery of
energy. The program was intended to provide a firm
basis for preparing the engineering design of a
large-scale system. Note that the original funding for
the program covered only the demonstration of
wastewater treatment using aquatic plants. Under the
original grant, the objective of the aquatic treatment
system was to meet 30 mg/L each for BOD5 and SS.
With an additional grant for the evaluation of
advanced treatment and health effects, the objectives
for the pilot aquatic treatment system became: 1) to
supply suitable water to the advanced water treatment
system for further processing, and 2) to supply
hyacinth biomass and wastewater sludge to the
61
-------
anaerobic digester for energy recovery through
methane production. The goals of the program
remained the same. Utilization of a natural biosystem,
coupled with low energy systems and energy
recovery has been the goal. The additional goal is
reclamation of water for useful purposes such as
irrigation and raw potable water supplies.
Phase II Studies (Early Developments)
The two primary and four secondary treatment
processes were operated in various combinations to
form seven different treatment trains for comparison
and evaluation of overall system efficiencies (see
Figures 4-6 and 4-7). Primary facilities consisted of
a sedimentation basin and a rotary disk filter, each
1,890 m3/d (0.5 mgd) in capacity. Secondary facilities
consisted of a pulsed bed filter (PBF), a sludge
blanket/fixed film reactor (SB/FF), a hybrid rock filter
(HRF), and water hyacinth ponds. The six water
hyacinth ponds were each 8.5 m x 126 m long x 1.2
m deep (28 ft x 416 ft x 4 ft). The ponds were
constructed with earthen berms and a clay lining. The
hyacinth ponds were operated in three sets of two
ponds each, with piping and slide gate arrangements
permitting operation in parallel or in series, and at
varying depths. The hyacinth ponds were utilized both
as a secondary treatment process and as a polishing
treatment following the other secondary treatment
processes.
The complete treatment trains evaluated in the Phase
1 portion of the program included:
a. Primary Sedimentation Basin - Hybrid Rock Filter
- Hyacinth Ponds.
b. Rotary Disk Filter - Sludge Blanket Fixed Film
Reactor - Hyacinth Ponds.
c. Primary Sedimentation Basin - Sludge Blanket
Fixed Film Reactor - Hyacinth Ponds.
d. Primary Sedimentation Basin - Pulsed Bed Filter
- Hyacinth Ponds.
e. Primary Sedimentation Basin - Hyacinth Ponds.
f. Rotary Disk Filter - Hyacinth Ponds.
g. Rotary Disk Filter - Pulsed Bed Filter - Hyacinth
Ponds.
Data collection for the treatment trains began in
September 1984 and continued through September
1985. Data collected for each treatment train included
BOD5, SS, nutrients and the concentrations of sulfur
compounds throughout the system. BODs and SS
concentrations were measured to determine if the
final effluent met secondary treatment discharge
standards. Nutrient data, including measurement of
the various forms of nitrogen and phosphorus, were
included for evaluation of microbial processes and
nutrient uptake by the water hyacinths. Measurement
of sulfur compounds was included because of the
potential for formation of hydrogen sulfide and odor
problems. Based on a detailed analysis of the
performance data for the alternative process trains
cited above, it was concluded that the most cost-
effective system was train F (rotary disk filter-
hyacinth ponds) (27). The flow sheet involving the
hybrid rock filter was rejected because of clogging.
The anaerobic reactor was rejected because of odor
generation.
Experiments conducted since September 1985 were
with the selected process flowsheet. To determine
how BOD5 and SS were removed in the pond, profile
tests along the length of the ponds were undertaken
in the fall of 1985. From profile testing, it was found
that most of the treatment for BOD5 and SS occurred
in the first 50 ft of the hyacinth ponds. Based on this
finding, flow was introduced in intervals or "steps"
along the entire length of the ponds. This presents
organic overloading of the head-end of the pond.
Phase // Studies (Step-Feed Hyacinth Ponds)
Based on the findings from the profile testing
program, Ponds 3 and 5 were modified to test the
effect of effluent recirculation and of step feeding the
influent at several locations along the length of the
ponds. Each pond was divided into eight cells, each
15.2 m (50 ft) long, with influent fed at the front end
of each cell (see Figure 4-8). The existing
recirculation system was utilized. The bulk of the
recycle flow was pumped to the cascade aerators.
The remaining portion of the recycle flow was
returned to the aeration manholes where it was
combined with the influent and reintroduced to the
ponds through the influent step feed piping. An
aeration system covering the entire length of each
pond was constructed using PVC pipe with holes
drilled at 0.3-m (1-ft) intervals. Aeration was also
provided to the aeration manhole. Installation of the
aeration system was necessary to overcome the
problems associated with the presence of high sulfate
levels in the influent wastewater.
The step-feed system was put into operation in
March 1986. The system was monitored to
determine: 1) the treatment capability of the pond and
individual cells, 2) air requirements for different
influent feed rates, 3) the maximum feed rate at
which a DO concentration of 1 mg/L could be
maintained, 4) the effect of recirculation on DO
concentrations and pond chemistry, and 5) the effect
of total coverage by the aeration system.
Performance Data for Step-Feed Hyacinth Ponds
Design criteria for the planned expansion of the water
hyacinth system, as well as criteria for the original
system, are summarized in Table 4-9. Performance
data for the various treatment processes that
62
-------
Figure 4-6. Site plan for San Diego, CA aquaculture pilot plant (26).
1 1
1
1
1
1
1
1
1
1
1
2 '
|| j
i
r 1
I
1
1
1
I
!
i
i
K
o
z
I
sfls
-ill!
63
-------
Figure 4-7. Schematic diagram of primary and secondary facilities - San Diego, CA aquaculture pilot plant (25).
64
-------
Figure 4-6. Schematic of hyacinth pond step-feed system with recycle - San Diego, CA (26).
Influent Flow to each seqment =Q -r 8
Q 4 /
i
t
»»•
• 1
1
L
Recycle = Qr
comprise the aquatic treatment system are presented
in Figures 4-9 and 4-10, for December, 1982
through October, 1983. As shown, the aquatic pond
effluent BOD5 values were consistently well below 30
mg/L regardless of the significant variation in the
influent BOD5 (125 to 375 mg/L) with the exception of
a single value, all of the SS values were also below
30 mg/L.
Profile testing for BOD5, SS, and DO was conducted
to determine treatment efficiencies throughout the
pond. Time sequenced samples were taken at the
influent of each cell and approximately 3 m (10 ft)
before the influent of the next cell. Results of a typical
profile test for BOD5, SS and DO are shown on
Figure 4-11. The profiles show the results of dilution
from the recycle flow at the head of the pond.
Loading appears to be consistent throughout each
cell, with adequate treatment achieved throughout the
entire pond. DO showed the greatest variation, with a
general decline throughout the last four cells to about
1 mg/L.
Harvesting and Hyacinth Productivity
Harvesting was accomplished at the pilot facilities
primarily to provide open water surfaces and low plant
densities to allow for more effective control of
mosquito larvae by mosquito fish. Hyacinths were
removed from the ponds during harvesting using a
truck mounted, articulating clam shell bucket and
loaded into an emptiable box also attached to the
truck.
Hyacinth productivity for the first year of operation
was significantly higher than reported at similar
facilities in Florida. Average productivity for the
second year, 67 dry metric tons/ha-yr (30 t/ac-yr),
was typical of other installations. The lower
productivity during the second year was probably the
result of maintaining a lower plant density in the
ponds by systematic harvesting. There was no
attempt to correlate hyacinth productivity with
treatment performance since the main purpose of the
harvesting was to provide for mosquito control.
Ocfors and Ocfor Control
The design of the pilot plant included provisions to
control odors from the various treatment processes.
The odor control provisions included: 1) enclosing the
primary settling basin and the rotary disk filter in a
separate building and routing exhaust air from the
building through a carbon adsorption unit; 2)
precipitating sulfides with ferric chloride in the
anaerobic filter sludge blanket fixed-film reactor
(SBFFR) and hybrid rock filter (HRF); and 3) providing
carbon canisters to adsorb hydrogen sulfide and other
odors at each of the aeration manholes. Aeration
manholes, located downstream from each of the three
secondary processes, contain aerators to increase
the DO concentration of the processed wastewater
before being introduced to the ponds.
Odor control measures provided for the primary
facilities successfully prevented odors in the vicinity
of the pilot plant. Carbon canisters at the SBFFR and
aeration manholes also controlled odors, except for
an incident at the SBFFR when the carbon canister
became depleted and had to be replaced. Several
incidents of odor were associated with the HRF. Most
of the problems occurred during the first few months
of operation. The unit was taken out of service in
June 1984, one month after startup, because of the
odor problems. When ferric chloride was added to the
HRF influent in late June, odors became less intense
and less frequent. However, isolated incidents of odor
were reported, mostly when ponding occurred on the
surface of the HRF as a result of clogging of the
medium.
The most serious odor problems were associated with
the hyacinth ponds. Hydrogen sulfide odors were
noticed at the effluent boxes and at the aeration
tubing. The principal cause of the odors is the
reduction of the sulfates in the wastewater to
65
-------
Figure 4-9. BOD5 performance data for San Diego, CA Pond #3 with 200 percent recycle (27).
BOD5, mg/L
250
200 -
150 tc.
100 -
6/86
1/87
6/87
Figure 4-10. SS performance data for San Diego, CA Pond #3 with 200 percent recycle (27).
Effluent
6/86
1/87
6/87
66
-------
Table 4-9.
Item
Design Criteria for Modified Plug-Flow Water
Hyacinth Ponds for Expanded San Diego, CA
Aquatic Treatment Facility (18)
Original
Expanded
Pond Configuration
Cross-Section Trapezoidal
Flow Scheme Plug-Flow
Pond Dimensions
Max. length, m 122
Base width, m 3.55
Side slopes 2:1
Max. height, m 1.22
Top width at 1.5 m
Surface Area, ha
at 1.07 m 0.097
at 1.22 m
at 1.37 m
Process Design and Operation
BOD5 loading
(BOD5/COD = 0.45). kg/ha-d 123b
Influent flow per pond, ma/d
Operating depth, m 98
Recycle ratio Variable
Max. airflow per aerator11, m3/min 0
DO in pond, mg/L
Expected effluent, mg/L (% of time)
BOD5
SS
Trapezoidal
Step-Feed
with
Recycle3
122
3.66
2:1
1.52
9.76
0.097
0.105
0.113
359C
313
1.37C
2:1
0.028
1.2
120 (90)
< 10 (50)
< 25 (90)
1 1 (50)
a Wrap around pond design is to be used (see Figure 4-4d).
b Based on an assumed BOD5/COD ratio of about 0.7.
c Tentative based on completion of depth tests.
a Aeration system (see Figure 4-4c).
hydrogen sulfide under the anaerobic conditions in
the bottom sludge deposits. The solution to the odor
problem was to change the method of operating the
ponds, as discussed previously, and to raise the DO
concentration sufficiently to satisfy the oxygen
requirements of the wastewater and produce a DO
residual of at least 1 mg/L in the remainder of the
ponds.
Vectors and Vector Control
The primary objective of the vector control program
was to evaluate the mosquito breeding potential of
the hyacinth ponds and to identify effective measures
for controlling mosquito populations. Observations
were made of the changing populations of larval
mosquitoes and mosquito fish (Gambusia), adult
mosquitoes, midges, and invertebrate mosquito
predators. Incidental observations concerning the
ecology of the ponds were also made.
Mosquitoes were controlled adequately when a
sufficient population of Gambusia was maintained in
the ponds. However, the low DO levels throughout
the ponds during much of the testing period, together
with low water temperatures in the winter, significantly
reduced the fish populations and required use of
other mosquito control measures. Two man-made
agents (BTI (bacillus thurengensis israulis) and
Golden Bear Oil 1111) were used successfully, but
continued applications were necessary.
Performance Summary
Based on the performance of Pond 3 it was
concluded that a step feed system with recirculation
greatly increases the treatment capacity of the pond.
Introduction of influent at 15.2-m (50-ft) intervals
resulted in a nearly uniform loading distribution and
effective treatment throughout the pond, with effluent
BOD5 and SS concentrations well below the limits for
secondary treatment. However, continuous aeration
throughout the pond is required to maintain aerobic
conditions so as to eliminate the development of
odors. Air requirements are proportional to the pond
BOD5 loading, with approximately 2.5 standard L of
air required per second to treat 1 kg of BOD5 (2.4
scfm/lb). Recirculation provided initial dilution of the
incoming wastewater and helped in distributing the
loading throughout the pond. At higher recirculation
rates, effluent turbidity increased. High turbidity can
cause excessive chlorine demand and thus increase
the cost of chlorination. However, SS levels were
generally within the limits of secondary treatment
standards, even at recirculation ratios as high as 51.
4.7.1.4 Design Factors
Process design factors for the hybrid aquatic system
used at San Diego involve consideration of: 1)
pollutant surface loading rates, 2) operating water
depths, 3) process kinetics, and 4) temperature
effects. The aquatic system is considered to be a
hybrid because of the need to aerate due to the
specific characteristics of the local wastewater.
Although the final design factors have not been
selected for a 3,785-m3/d (1-mgd) facility, the
values given below are consistent with the findings to
date.
Pollutant Surface Loading Rates
A commonly used loading parameter for aquatic plant
systems is based on surface area and is expressed
as mass CBOD5/area-d. Loading rates, based on
using a step-feed system with recycle and
supplemental aeration, of 200-250 kg CBOD5/ha-d
(180-225 Ib/ac-d) are recommended by the San
Diego researchers.
Operating Water Depths
The San Diego researchers believe that operating
water depths for aquatic systems are extremely
important with respect to process performance and in
defining the hydraulic detention time and the mixing
conditions within the pond system. Recommended
operating water depth for a hybrid step-feed water
hyacinth system with aeration is 0.9-1.2 m (30-42
in).
67
-------
Figure 4-11. Influent and effluent BOD, SS, and DO for step-feed hyacinth pond - San Diego, CA (26).
120
100
ao
60
40
20
0
120
too
80
60
4O
2O
Effluent BOO
\
SS
Effluent SS
:\
\
^
Effluent DO
N
'
Influent DO
J L
i 1 I I
Influent
Box
Cell 1
Sample Location
Cell 8 EMIuent
Box
68
-------
Hydraulic Surface Loading Rates
During the Phase II pilot studies, the hydraulic surface
loading rate was held at 0.058 m /m2-d (62,000
gpd/ac). The resultant hydraulic detention time was
21 days.
Process Kinetics
The San Diego step-feed hyacinth system with
recycle has been modelled, as shown in Figure 4-
12, as a series of CFSTRs (continuous-flow
stirred-tank reactors) (28). Using the cascade flow
model, the treatment performance of the pond system
was described adequately assuming first-order
kinetics (see Figure 4-13). The corresponding first
order reaction rate constant k was found to be about
1.95 d-t.
Temperature Effects
The performance of all aquatic treatment systems is
temperature dependent. Based on both experimental
studies and an analysis of data presented in the
literature, it appears that a modified van't Hoff-
Arrhenius temperature relationship can be used to
estimate the effect of temperature on wastewater
treatment using aquatic systems. Based on
experimental studies with water hyacinth and
emergent plant systems, it is estimated that the value
of the temperature coefficient is about 1.09.
4.7.1.5 Operating Characteristics
The performance of the hybrid step-feed hyacinth
system with recycle and supplementary aeration has
proven to be stable with respect to the effluent quality
(see Figures 4-9 and 10). Even before the step-
feed system was developed, the effluent quality was
good (both BOD5 and SS less than 30 mg/L)
regardless of the condition of the pond (whether the
DO levels were low or nonexistent and the pond was
odorous).
There are two constraints to the operation of a
hyacinth system in San Diego that dictate operating
practices that may not be factors in other locations. A
hyacinth system must be odorless and free of
mosquitoes. These constraints are backed by
requirements for a minimum dissolved oxygen of 1
mg/L in the ponds and zero mosquito larvae per dip.
The mosquito requirement can only be met by
maintaining a large and healthy mosquito fish
population in the ponds and low plant densities so
that mosquito fish have access to breeding locations.
It has been recommended that the DO in the ponds
be maintained above 1 mg/L for the mosquito fish.
The climate in San Diego is such that cold weather
stress on the hyacinths was not a factor in effluent
quality. Another problem common to the southern
United States, introduced weevil and mite species for
biological control of hyacinths, has not been a
problem in San Diego primarily because these
species have not been introduced in the area.
4.7.1.6 Costs
In addition to the costs associated with conventional
hyacinth pond construction, costs for hybrid step-
feed hyacinth facility include the capital and O&M
costs for the step feed distribution piping,
recirculation pumps and piping, and a complete in-
pond aeration system. The costs of all these features
were included in the cost analysis developed for the
hyacinth ponds.
Based on an applied wastewater BOD5 concentration
of 175 mg/L, and a pond loading rate of 225 kg
BOD5/ha-d (200 Ib/ac-d), a 3,785-m3/d (1-mgd)
facility would require a pond surface area of 2.9 ha
(7.3 ac). Capital costs for the ponds would be
approximately $2.18 million with an annual O&M cost
of $494,000 (mid 1986 dollars). Anaerobic digestion
of the harvested hyacinths has the potential to
generate methane having an energy equivalent of
about 2 billion BTU/yr. Use of this energy to generate
electricity could significantly reduce the outside power
costs for the entire treatment facility.
4.7.2 Austin, Texas
4.7.2.1 History
The State of Texas has been gathering information on
the use of water hyacinths to improve the quality of
stabilization pond effluent since 1970. Field-, pilot-,
and full-scale studies of hyacinth systems have
taken place at various locations, including the City of
Austin. The use of a water hyacinth system in
wastewater treatment has been shown to be feasible
but winter freezing is a recurring problem.
The city's Hornsby Bend Sludge Treatment Facility
receives excess activated sludge from area
wastewater treatment plants. It was placed into
operation in the 1950s and is undergoing a major
expansion and renovation program.
Original plant design called for supernatant from three
sludge holding lagoons to be passed through a
chlorine contact basin and then discharged to the
Colorado River. The quality of the treated supernatant
was not meeting the discharge requirements
established for the facility.
Water hyacinths were introduced into the 1.2-ha
(3-ac) chlorine contact basin in 1977 and they
served as a seasonal upgrade to the treatment
process for several years. Basin configuration was
not well suited to hyacinth treatment and the
hyacinths were usually damaged by freezing
conditions each winter. A greenhouse structure was
proposed to protect the hyacinth and offer year round
treatment.
69
-------
Figure 4-12. Definition sketch for the analysis of a hyacinth pond with step-feed and recycle (28).
. Influent
. Flow to each segment = Q •*• 8
I6M I7M I8'l I9 = l 20=l 2I--I 22=l 23'l
Nominal recycle ratio for each reactor
based on an overall recycle ratio of 2:1
Effluent
Figure 4-13. Analysis of performance data for hyacinth pond 3, San Diego, CA, with step-feed and recycle (18).
o>
E
cf
o
c
cu
o
o
o
May 16, 1986
Q= 227 mVd (60,000 got/d)
Qr = 464m3/d (122,600 ga!/d)
Temp=20°C
Computed curve
Influent
CBOD
Influent
box
.Effluent
;Q._
ss
/Effluent
Cell no.
Typical Removal Curves For BOD And SS Along Pond Length For Pond 3
When Operated In The Step-Feed Mode With Recycle.
70
-------
A new water hyacinth facility with three basins
covered with a permanent greenhouse structure was
included in the Hornsby Bend expansion and
renovation plans. The Hornsby Bend Hyacinth Facility
(HBHF) qualified for funding under the EPA
Construction Grants Program as an innovative
wastewater treatment process. It was the first
hyacinth facility with a permanent greenhouse
structure funded by the USEPA.
4.7.2.2 Design Objectives
The design objective is to provide year-round
upgrading of sludge lagoon supernatant quality to
meet 30 mg/L BOD5, and 90 mg/L SS discharge
limits.
4.7.2.3 Design Factors
Basin Design
The Texas Department of Health specifies a
maximum design surface hydraulic loading rate of
1,870 m3/ha-d (0.2 mgd/ac) for water hyacinth
wastewater treatment basins (29). Loading rates up to
4,680 m3/ha-d (0.5 mgd/ac) are permitted on a
case-by-case basis. Austin's Hornsby Bend
Hyacinth Facility (HBHF) incorporates three basins
that have a total surface area of 1.6 ha (4.0 ac) when
filled to their 17,000-m3 (4.5-Mgal) capacity. The
basins were designed to receive a maximum daily
flow of 7,570 m3/d (2 mgd); equivalent to a surface
loading rate of 4,680 m3/ha-d (0.5 mgd/ac).
The center basin has an area of 0.64 ha (1.6 ac) and
the two outer basins have areas of 0.48 ha (1.2 ac)
each (see Figure 4-14). All three basins are 265 m
(870 ft) long. The center basin is 24.2 m (80 ft) wide
and the outer basins are 18.1 m (60 ft) wide. Basin
depths vary from 0.9 m (3 ft) at the upstream end to
1.5 m (5 ft) at the downstream end. The middle basin
receives roof runoff during storms. The investigators
at HBHF believe temperature changes due to runoff
entering the basin have caused stress to some
species stocked in the pond, and plans are underway
to divert roof drainage out of the facility.
Influent flow to the basins is distributed uniformly
across the width of each basin at the upstream end
via a 30-cm (12-in) diameter perforated pipe. Two
secondary distribution pipes at 63.9 m (210 ft) and
127.8 m (419 ft) downstream of the primary inlet in
each basin are available for experimental step
application of influent.
Maintenance of the basins will include harvesting of
plants and removal of detritus accumulation. The
basin slope facilitates cleaning. A drain valve at the
bottom of the outlet structure is separate from the
adjustable telescoping valve used to set water depth.
Capacity of the facility is adequate to treat design
flows with one of the three basins out of service for
cleaning. Berms separating the basins accommodate
a 3-m (10-ft) wide unsurfaced roadway used during
harvesting.
Mosquito control was a major consideration in the
design of the basins. The primary method of control
is stocking of predators of mosquito larva and adults.
These include mosquito fish (Gambusia affinis), grass
shrimp (Palemonetes kadiakensis), and leopard, tree,
and cricket frogs. Eight open areas are incorporated
into the design of each basin to maintain oxygen
levels adequate for survival of the Gambusia and
grass shrimp. The openings consist of either a 55.7-
m2 (600-sq ft) or a 74.2-m2 (800-sq ft) area
protected from water hyacinth intrusion by chain-link
fence fabric. Light is allowed to penetrate the water
surface and sponser growth of algae on the gravel
lined bottom of the aerators. The open areas help
insure mosquito fish survival. After leaving the
hyacinth facility, the polished secondary effluent
passes over a two step cascade aerator with a total
drop of 3.4 m (11 ft). DO concentration of the effluent
has exceeded 5 mg/L at the discharge.
Greenhouse Design
A 2-ha (5-ac) greenhouse structure covers the
three hyacinth basins to prevent winter freezing of the
plants. The three bays of the concrete and steel
structure are completely enclosed in clear, reinforced
fiberglass decking with a light transmission value of
65 percent. Light transmission of the fiberglass is
critical to plant growth so the performance of the
panels is being monitored over time. A section view
of the greenhouse structure is provided in Figure 4-
15.
Sidewalls are 3.4 m (11 ft) high to permit
maneuvering of maintenance vehicles and equipment.
Seven overhead doors at each end of the building
originally provided access for both personnel and
equipment. Separate personnel doors were recently
added. Moveable barriers are placed across open
doorways to exclude snakes and other predators of
organisms stocked for mosquito control.
The barriers will also prevent the return of Nutria (a
large pond dwelling rodent) which inhabited the facility
for several months but have since moved out. Both
the doors and roof ridge vents running the entire
length of the building provide ventilation. The ridge
vents are screened to reduce immigration of adult
mosquitoes.
4.7.2.4 Operating Characteristics
Biological stability of the water hyacinth basins is the
prime requirement for successful wastewater
treatment. As a result of maintenance work on one of
the sludge lagoons that feed the hyacinth basins,
influent loading levels were erratic and, subsequently,
so were treatment levels during the first six months of
operation. Influent and effluent BOD5, SS, NH3-N,
and N03-N values are presented in Table 4-10 for
71
-------
Figure 4-14. Hornsby Bend, TX hyacinth facility basin configuration (30).
NATURAL. ASAATOft.
(TYff)
274m. (1OOFO
fai*.rr*)
4=1-
1ft
r~-
PLAN VIEW
Figure 4-15. Hornsby Bend, TX hyacinth facility pond and roof section (30).
62'-6"
.
FIBEROLA53 CORRUGATED
ROOF PANELS
CLEAR
FIBERGLAS
fiOINC PANELS.
WATER DEPTH
VARIES FROM
3' TO 5'
^CONCRETE COLUMNS
2S O.C.
72
-------
Table 4-10.
Date
Performance Data - Hornsby Bend, TX Hyacinth Facility' (31)
nj-j BOD5, mg/L TSS, mg/L
VSS. mg/L
NH3-N, mg/L
Influent
Effluent
Influent
Effluent
Influent
Effluent
Influent
Effluent
Influent
' Monthly average of approximately 12 samples (composites) per month.
Effluent
9/87
10/87
11/87
12/87
1/88
2/88
3/88
4/88
5/88
5/88
7/88
8.4
8.3
8.3
8.2
8.1
8.1
8.1
7.9
7.9
8.0
8.1
7.1
7.8
7.8
7.7
7.6
7.7
7.6
7.6
7.4
7.7
7.7
97
39
153
106
79
84
357
143
156
99
30
11
9
14
18
45
139
34
30
28
140
120
245
142
127
84
155
162
121
117
132
31
19
21
24
17
36
41
47
26
30
19
90
169
240
111
96
71
91
160
68
79
104
28
22
17
14
16
12
37
49
8
23
12
22.9
26.5
26.1
41.9
121.1
95.6
77.6
76.8
43.5
47.0
24.7
38.6
43.0
39.3
39.1
31.0
36.4
42.0
42.5
21.9
33.9
37.4
1987 and 1988. The primary method for assuring
relatively constant loading rates in the future will be
maintaining a constant influent flow rate. The HBHF
operates under 30 mg/L BOD5 and 90 mg/L SS
discharge requirements, with a maximum 30-day
average flow of 7,570 m3/d (2 mgd). By June, 1989
no discharge of pond effluent to the Colorado River
will be permitted. Plans for future disposal include
using HBHF effluent to irrigate approximately 80 ha
(200 ac) of agricultural land near the facility. When
the facility was placed into operation in February of
1986, the basin effluent BOD5 concentration was at
or near 10 mg/L.
Mosquito control measures have been effective. In
addition to the species stocked, dragonflies have
inhabited the facility. Dragonfly larva feed on
mosquito larva and the adults prey on the mosquito
adults. There was a noticeable increase in the
mosquito population, believed to be due to
immigration of adults, when the weather became
cooler.
DO concentration within the natural aerators has been
measured as high as 5 mg/L. Small plants and debris
are removed daily from each aerator to maintain a
constant light source for the oxygen producing algae
attached to the rock at the bottom.
No harvesting was necessary during the first five
months of operation but was needed constantly
during July and August. Harvesting was much less
frequent during the following winter. A modified tractor
mounted backhoe is used to remove hyacinth from a
1.2-1.8 m (4-6 ft) strip along the perimeter of each
basin. In addition to acting as temporary aerators, the
cleared areas facilitate movement of mosquito larvae
predators. Harvested plant material is first dried on an
asphalt pad, then mixed with thickened waste
activated sludge and recycled by the city's
Department of Parks and Recreation. The recycling
program was implemented in January of 1987.
Based on operating experience at other hyacinth
facilities in the Austin area it is known that humus
accumulation will occur at a relatively fast rate, and
that most accumulation will take place near the inlet
end of the basin. It is hoped that partial draw-down
of a basin will be adequate for cleaning, without
requiring restocking of plants and other organisms.
There are several unanswered questions regarding
the operating characteristics of the HBHF under
extreme weather conditions. Of primary concern is
survival of the hyacinth plants during very cold
weather. Outside temperatures during the 1985-86
winter were mild and did not provide an indication of
the greenhouse's ability to retain heat entering the
system. Outside air temperatures rose to above 37°C
(98°F) during the summer of 1986 without causing
heat stress damage to the plants within the
greenhouse; inside temperatures were approximately
55°C (131 °F) on those days. Another concern is the
potential decreased light transmissivity of the
fiberglass over time. Deterioration of the fiberglass,
and algal growth brought on by condensation on the
inside surfaces may inhibit light transmission.
Roadways inside the facility have been subject to
moisture and are deteriorating. Condensation dripping
onto the road surfaces and capillary rise weaken the
road structure. Installation of a permanent road
surface on the berms is planned.
4.7.2.5 Costs
Total engineering design and construction cost of the
facility was estimated to be $1,200,000. A more
detailed accounting is not yet available.
4.7.2.6 Monitoring Programs
Under the requirements of the HBHF discharge
permit and the need to evaluate other aspects of
basin performance, influent and effluent levels of the
following contaminants are being monitored: BOD5,
SS, VSS, NHs-N, N03-N, and TP. Research work
during 1986 included efforts to develop a
73
-------
mathematical model of BOD5, SS and nutrient
removal, and study of the nitrification process within
the hyacinth system. A consistent effort to monitor
and maintain the biological structures of the treatment
system will be necessary during the formative stages
of its unique ecosystem. Biological maturity and
stability will not occur overnight, but it is ultimately
essential to dependable treatment performance.
4.7.3 Orlando, Florida
4.7.3.1 History
The City of Orlando's Iron Bridge Wastewater
Treatment Facility (IBWTF) was constructed in 1979
to provide regional wastewater treatment. It was
designed to achieve tertiary treatment standards
using primary clarification and RBCs for
carbonaceous BOD5 removal and nitrification,
submerged RBCs for denitrification, chemical addition
and sedimentation facilities for phosphorus removal,
and rapid sand filters for final polishing. The discharge
permit for the plant required an effluent of 5 mg/L
BOD5, 5 mg/L SS, 3 mg/L TN, and 1 mg/L TP and
allowed a maximum discharge of 90,000 m3/d (24
mgd).
By 1982, flows to the plant were increasing but the
city was faced with meeting the existing waste load
allocation for the Iron Bridge discharge to the St.
Johns River. The city of Orlando started looking for
ways to achieve higher levels of treatment for a
portion of the total flow. One proposal was to use a
water hyacinth system to treat 30,000 m3/d (8 mgd)
to achieve an effluent quality of 2.5 mg/L BOD5, 2.5
mg/L SS, 1.5 mg/L TN, and 0.5 mg/L TP. This level
of treatment would allow for a maximum influent flow
of 106,000 m3/d (28 mgd). The city of Orlando
decided in 1983 to test the feasibility of this proposal
by building and operating a pilot hyacinth facility.
Based on the results of the pilot study it was decided
to build the full scale water hyacinth system. The
full-scale system was completed in the summer of
1985 and has been in operation since.
4.7.3.2 Design Objective
The major design objective for the water hyacinth
system at IBWTF was to treat a portion of the total
plant flow to a higher effluent quality to allow for an
increased effluent discharge flow without violating the
waste load allocation for the discharge. The specific
goal was to remove 50 percent of the major pollutants
in 350 L/s (8 mgd) of effluent, allowing for an increase
of 175 Us (4 mgd) in discharge flow.
4.7.3.3 Pilot Plant Results
Pilot Facilities Description
The hyacinth pilot facility consisted of five ponds built
in series, each 5.2 m x 9.8 m (17 ft x 32 ft) providing
required pond surface area was determined using a
computer model (HYADEM) developed by Amasek,
Incorporated after assuming a wet crop density of
12.2 kg/m2 (2.5 Ib/sq ft) and an influent flow of 54.5
m3/d (14,400 gpd). The pond depth was set at 0.6 m
(2 ft) resulting in a nominal hydraulic detention time of
2.8 days. The nominal surface loading rate was 2,240
m3/ha-d (0.24 mgd/ac).
Experimental Design
The stated goals of the pilot study (25) were:
1. Demonstrate the ability of the hyacinth system to
achieve the desired effluent concentrations on an
average monthly basis with nitrogen being the
major concern.
2. Demonstrate the ability of the hyacinth system to
perform during the winter months.
3. Demonstrate the ability of the hyacinth system to
recover following a freezing event.
4. Determine the need for micronutrient addition.
5. Determine the applicability and degree of reliability
of Amasek design and operational models.
6. Reveal specific operational adjustments required.
The pilot system was operated under steady state
conditions. Influent and effluent samples were
analyzed twice weekly during November and
December 1983 and daily during the period from
January 1 to March 15, 1984 for BOD5, SS, TN, and
TP. Additionally there was periodic determination of
standing crop densities, total crop biomass, and
micronutrient concentrations in the influent and
effluent.
Experimental Results
The five ponds were stocked with water hyacinth in
September 1983. Problems with influent quality
control made it difficult to evaluate pilot plant
performance for the following three months. Plant
growth during this adjustment period was below
expected rates. Factors which may have caused poor
growth were a possible micronutrient deficiency and
activity of the hyacinth weevil (Neochetina
eichhorniae).
By December the wet standing hyacinth crop had
increased from 455 kg (1,000 Ib) to 1,650 kg (3,636
Ib), approximately 6.5 kg/m2 (1.34 Ib/sq ft). On
December 25 and 26 a freeze occurred that produced
a noticeable effect on the plants but did not kill them.
Treatment efficiencies decreased in January. A
meaningful evaluation of the effect of the freeze was
not possible due to the instability of the system.
Actual loading rates were not as had been planned.
a total pond area of 253 m (2,720 sq ft). The Flow was reduced to 21.2 m /d (5,600 gpd) during
74
-------
the second week in January 1984 to accommodate
the higher nitrogen loading. Initially, iron, potassium
and phosphorous were added as a micronutrient
supplement to the influent. In January, zinc, copper,
manganese, molybdenum, boron, and sulfur were
added to the supplement program, and the last two
ponds were covered with a portable greenhouse
structure in order to assess their performance during
freeze events.
Pollutant removal from February 15 to March 15 was
stable and the system did not have any major
operating problems. Removal of BOD5, SS, TN, and
TP during this one-month period averaged 60, 43,
70, and 65 percent respectively.
In a Amasek report assessing the performance of the
pilot facility, it was concluded that covering a water
hyacinth system for freeze protection at the Iron
Bridge plant was not cost effective considering "the
ability of hyacinths to recover from even severe
Florida freeze events, and considering some of the
negative features associated with a covered system
(32).
4.7.3.4 Design Factors
The areal requirements and standing crop density of
the system were determined using the same
computer model that had been used in the design of
the pilot hyacinth system. The premise of the model
is that nutrient removal is tied directly to plant growth.
Plant growth is modelled using Monod kinetics and
the van't Hoff-Arrhenius temperature relationship,
and apparently assuming that growth is occurring in a
reactor with a constant concentration of the limiting
nutrient. Growth rate is then related to plant density
and surface area coverage, and the average daily rate
of nutrient uptake is calculated. The calculation of
effluent nutrient content is made with the following
relationship:
Cn =
- Nu -
where,
Cn = effluent nutrient concentration
C| = influent nutrient concentration
QI = daily flow in
QO = daily flow out
Nu = daily mass nutrient removal by plant uptake
N| = daily mass nutrient removal by incidental
processes.
In general, most researchers have concluded that
nitrogen removal is by nitrification/denitrification with
only incidental removal by the plant biomass.
Results of the pilot-scale system were used to
determine the necessary constants for the growth
relationships.
The system consists of two ponds each having a
surface area of 6 ha (15 ac) and hyacinth digesting
facilities (see Figure 4-16). Each pond is further
divided into five basins 67 m long x 183 m wide (220
ft x 600 ft) using berms.
Weirs are located at six points in the dividing berms
to distribute flow evenly across the full width of the
berms to prevent short-circuiting. AWT effluent is
fed to both ponds through an influent manifold. The
west pond has an influent line from the secondary
facilities in addition to the AWT influent line.
Supplementary nutrient addition is provided by
chemical dosing and mixing facilities, and chemical
feed pipes to the influent lines and to the weirs in
each dividing berm. Pond depth is 0.9 m (3 ft)
resulting in a hydraulic detention time of
approximately 3.5 days.
4.7.3.5 Operating Characteristics
The Iron Bridge hyacinth facility was initially stocked
with water hyacinth in late 1984. Until July 1985 the
system was operated in a start-up mode. During this
time the system met nutrient removal requirements.
In July 1985 Amasek took over operation of the
system. In a report to the City of Orlando (33),
Amasek summarized the process problems
encountered from July 1985 to February 1986:
1. At the time Amasek took over operations, the crop
had developed extensive weevil populations and
there was considerable encroachment of alligator-
weed.
2. Amasek attempted to improve crop viability by
selective harvesting. Growth of the remaining crop,
however, was not as projected, and extensive
algae development resulted in violation of SS
limits.
3. As weevil populations developed, a spraying
program (Sevin) was initiated. Also new hyacinth
stock was brought in to enhance crop
development.
4. Improved crop viability was noted as a result of the
spraying. However, crop growth was inconsistent,
and coverage was not being achieved as designed.
This resulted in a continuation of algae
development and solids violations. However,
adequate nutrient removal continued.
5. By January, 1986, it had become evident that the
crop was experiencing serious growth problems.
Nutrient removal was still being observed, although
there was a considerable decline in the rate of
removal.
6. Several potential causes for the growth problems
were identified during a series of meetings with the
city. These were as follows:
75
-------
Figure 4-16. Iron Bridge, FL hyacinth facility basin configuration.
,
1
s
id
u
tj
1
;?
S
T
U
\i
F
_j
5!
j.
f
2
I
nj
z
5
-t
-t
-t
1A 1
-f
-f
fl
J \
\ r
U
-{
-f
-f
1 I
-f
-f
-f
^""-DISTPJI
h- weiR
U i r /
H
I 2A
h
"JJ* ~lrt
|— fVsMrv
n
j N.
\ c
u
h-
H
r-
I 2 "«r
r-
H
h-
-(
-4
-{
'
-f
-£
-t
-t
-{
^
-«
-f
-«
h -t
Y- -(
h -f
I 3A I
I- -f
c
h -f
I- -f
n
j ^
^ /•
U
h -f
h -4
h -f
3 1
h 220' -*
)- -(
)- -f
)- -4
1- -{
I- -{
I 4A 1
H -t
"^DIVIDER.
)_ &CPM ^ £
H -(
n
J V
•^ r
U
- -f
-(
h- -(
4 1
-f
1- -f
1- -f
h ~~!
— -*- 1.
1 5A |
|_ _^L
1- — f
- —- f
A J
J\ ;
W s
If Nj
h- —j
h- — !
}- —
1 6
H —
H — Is
h- —
> AWT AND £
/ INFLUENT 1]
'
fL
rf
« 4 .
P 0
., cl
bl
/«x
?pr^"^
3 " ' fe
Ej]-5
4
\ AWT INFLU
DISTRIBUTIC
L
Key to Numbered Structures:
1. Methane Generation Tank
2. Equalization lank
3. Compressed Methane Gas Storage Tanks
4. Equipment Wash Pad
6. Covered Maintenance Work Area
6. Maintenance Storage Room
7. Control Room
8. Heater Equipment Room
9. Chemical Feed System
10. Parking Area
76
-------
a. Metal toxicity, with aluminum as the primary
suspect.
b. Biological interferences or competition,
principally from the algae populations.
c. Macronutrient deficiencies, with phosphorus the
principle concern.
d. Micronutrient deficiencies.
7. In mid-January, 1986, the system was shut down
in an attempt to restore crop health and to facilitate
solids control. The east pond was fertilized to bring
levels of nitrogen, phosphorus, iron and calcium to
excess concentrations. A series of experiments
were established to test impacts of various
additives. Extensive testing of plant and water
quality was conducted to identify toxic or deficient
levels.
8. By late January plant morphology indicated a very
serious growth problem, and the standing crop
began to decline significantly. The east pond
showed no response to the added nutrients,
indicating that either a toxic influence or a
micronutrient deficiency was the problem.
9. In February, 1986, flow to the west side was
reinstated, and an improvement in crop health was
noted almost immediately. This verified the
suspicion that there were no chronic toxic
influences from the Iron Bridge effluent. This had
been noted also within the set of contained
experiments. A micronutrient deficiency therefore
became the principal suspect. In evaluating this
concern more closely, the Iron Bridge plants and
water were compared to plants and water in
Amasek's other systems.
In summary, the growth problems have been
assessed as follows:
A molybdenum deficiency has developed as a result
of: 1) precipitation and filtration of aluminum
molybdate prior to discharge to the hyacinth lagoons,
2) interference of molybdenum uptake by sulfates
which are put into the system as ferrous sulfate, and
3) low sediment pH and poor system buffering
because of low alkalinity which inhibits molybdenum
uptake.
To correct the hyacinth growth problems at the Iron
Bridge hyacinth facility, molybdenum and boron are
added as part of the supplementation program, ferric
chloride is used instead of ferrous sulfate, and lime or
soda ash is added to increase influent alkalinity to 60
mg/L as CaC03.
From February to May 1986, the hyacinth system was
operated in a start-up mode to establish a healthy
crop. Starting in June the west pond was operated as
designed except that the influent nitrogen levels were
approximately 13 mg/L rather than 3 mg/L. In
September the east pond was also placed in service.
Influent and effluent concentrations of BOD5, SS, TN,
and TP for six months of relatively steady operation
(June to November) are presented in Table 4-11.
During this steady operation period, the hyacinth
system did not meet its treatment goals for either
BOD5 or SS. The BOD5 and SS concentrations were
reduced on average from 4.87 and 3.84 mg/L in the
influent to 3.11 and 3.62 mg/L in the effluent. In terms
of mass of nitrogen removed, the system did achieve
the removal rates predicted. Effluent phosphorus
levels were always below the design goal of 0.5 mg/L,
although it was necessary to add supplemental
phosphorus to the influent to assure phosphorus was
not limiting plant growth.
4.7.3.6 Costs
The construction costs of the hyacinth system at the
Iron Bridge plant were $1,200,000 for the hyacinth
digester, and $2,000,000 for the basins and piping.
Operation and maintenance is performed under
contract by the Amasek Corp. for a yearly fee of
$550,000 which covers all O&M costs associated
with the hyacinth system, such as pumping and
sludge disposal.
4.7.3.7 Monitoring Programs
Amasek is performing extensive monitoring of the
hyacinth facility as part of their O&M contract with the
city of Orlando. A summary of the monitored
parameters and frequency of monitoring is provided in
Table 4-12. In addition to the influent and effluent
water quality parameters, standing crop biomass is
monitored to allow for control of harvesting
operations. Monitoring of hyacinth predators and
micronutrient contents in the influent is also
performed to assure the hyacinths remain healthy.
4.7.4 Summary
The three case studies provided in this chapter
represent a broad range of the potential uses of
aquatic plant systems. A comparison of the three
systems is difficult but a summary of each system's
design and operating characteristics and costs is
provided in Table 4-13.
What is clear from these case studies is that aquatic
plant systems can be designed and operated to
accomplish a variety of wastewater treatment tasks,
but the designs and the operation are not always
simple. Hyacinth systems are susceptible to cold
weather and particularly in the southern states, can
be affected by biological controls introduced to help
control water hyacinths in the natural environment.
Concerns of health agencies for mosquitoes can play
a very big factor in the design and operation of
77
-------
Table 4-11. Iron bridge, FL Water Hyacinth System Performance Summary
Date
6/86
7/86
8/86
9/86
10/86
Average
Wastewater -
Flow, m3/d
16,680b
17,450b
16,850b
32,500C
31,190C
23,250
BOD5,
Influent
3.24
4.12
3.33
6.16
4.43
4.87
mg/L
Effluent
4.58
1.73
3.70
2.66
3.11
3.11
SS, mg/L
Influent
3.06
3.85
3.58
5.23
2.70
3.84
Effluent
6.31
1.86
4.28
2.91
3.56
3.62
TN, mg/L
Influent
12.52
12.44
12.77
12.66
14.49
13.00
Effluent
8.09
8.06
7.62
7.96
9.66
8.16
TPa
Influent
0.37
0.33
0.55
0.75
0.89
0.61
Effluent
0.24
0.11
0.19
0.15
0.22
0.22
' Phosphorus is added to the hyacinth system Influent as a nutrient supplement.
^ West hyacinth pond in operation.
: Both hyacinths ponds in operation.
1 Both ponds in operation for portions 01 the period
aquatic plant systems. Finally although water hyacinth
systems may be useful in nutrient removal, there are
limits to the treatment capacity and dependability of
hyacinth systems in terms of meeting very low
effluent values.
4.8 References
When an NTIS number is cited in a reference, that
reference is available from:
National Technical Information Service
5285 Port Royal Road
Springfield, VA 22161
(703) 487-4650
1. Aquaculture Systems for Wastewater Treatment:
Seminar Proceedings and Engineering
Assessment. U.S. Environmental Protection
Agency, EPA-430/9-80-006, NTIS No. PB
81-156705, 1980.
2. Reddy, K.R., and W.H. Smith (Eds.). Aquatic
Plants for Water Treatment and Resource
Recovery. Magnolia Publishing Inc., 1987.
3. Tchobanoglous G. Aquatic P/ant Systems for
Wastewater Treatment: Engineering
Considerations. 1987. In: Aquatic Plants for Water
Treatment and Resource Recovery. Magnolia
Publishing Inc., Orlando, FL, pp. 27-48, 1987.
4. Stowell, R., R. Ludwig, J. Colt, and G.
Tchobanoglous. Toward the Rational Design of
Aquatic Treatment Systems. Presented at the
American Society of Civil Engineers, Spring
Convention, Portland OR. April 14-18, 1980.
5. Reddy, K.R., and D.L. Sutton. Waterhyacinths for
Water Quality improvement and Biomass
Production. J. Environ. Qual. 13:1-8, 1984.
6. Reddy, K.R., and W.F. DeBusk. Nutrient Removal
Potential of Selected Aquatic Macrophytes. J.
Environ. Qual. 14:459-462, 1985.
7. Zirschky, J.O., and S.C. Reed. The Use of
Duckweed for Wastewater Treatment. JWPCF
60:1253-1258, 1988.
8. Hayes, T.D., H.R. Isaacson, K.R. Reddy, D.P.
Chynoweth, and R. Biljetina. Water Hyacinth
Systems for Water Treatment. In: Reddy, K.R.
and W.H. Smith (Eds). 1987. Aquatic Plants for
Water Treatment and Resource Recovery, pp.
121-139, 1987.
9. Reed, S.C., and R.K. Bastian. Aquaculture
Systems for Wastewater Treatment: An
Engineering Assessment. U.S. EPA Office of
Water Program Operations, EPA 430/9-80-007,
1980.
10. Leslie. M. 1983. Water Hyacinth Wastewater
Treatment Systems: Opportunifies and
Constraints in Cooler Climates. U.S.
Environmental Protection Agency, EPA/600/2-
83-095, Washington D.C.
11 Reed, S. C., E.J. Middlebrooks, and R.W. Crites.
Natural Systems for Waste Management and
Treatment. McGraw-Hill Book Co. NY, 1987.
12. Wunderlich, W.E. The Use of Machinery in the
Control of Aquatic Vegetation. Hyacinth Contr. J.
6:22-24, 1967.
13. Klorer, J. The Waterhyacinth Problem. Assoc.
Eng. Soc. 42:33-48, 1909.
14. Stephenson, M., G. Turner, P. Pope, J. Colt, A.
Knight, and G. Tchobanoglous. Publication No.
65, The Use and Potential of Aquatic Species for
Wastewater Treatment. Pub. by California State
Water Resources Control Board, 1980.
78
-------
Table 4-12. Iron Bridge, FL Water Hyacinth System
Monitoring
Parameter
Frequency
Influent Flow
Air Temperature
Wastewater Temperature
PH
Conductivity
DO
Rainfall
Wind Velocity
Wind Direction
Chlorine
TKN
NH4-N
NO3-N
N02-N
TN
OP
TP
BOD5
TSS
TDS
Na
K
Fe
Ca
Mn
Mg
Mn
B
Zn
Cu
Mb
Cr
Al
Pb
Hg
Ni
Cd
S04
Plant Constituents
Harvested Biomass
Stocked Biomass
Standing Crop Biomass
Weevils
Sameodes
Mosquitoes
Encroaching Vegetation
Root Macroinverterbrates
Fungal Isolates
Daily
5 days/week
5 days/week
5 days/week
5 days/week
5 days/week
Daily
5 days/week
5 days/week
Twice/week
Twice/week
Twice/week
Twice/week
Twice/week
Twice/week
Twice/week
Twice/week
Twice/week
Twice/week
Once/week
Once/week
Once/week
Once/week
Once/week
Once/week
Once/week
Once/week
Once/week
Once/week
Once/week
Once/week
Once/week
Once/week
Once/week
Once/week
Once/week
Once/week
Once/week
As Needed
As Needed
As Needed
Once/week
Once/week
Once/week
Once/week
Daly
As Needed
As Needed
15. Reddy, K.R., and W.F. DeBusk. Growth
Characteristics of Aquatic Macrophytes Cultured
in Nutrient-enriched Water: I. Water Hyacinth,
Water Lettuce, and Pennywort. Econ. Bot.
38:225-235, 1984.
16. DeBusk, T.A., and K.R. Reddy. Wastewater
Treatment Using Floating Aquatic Macrophytes:
Containment Removal Processes and
Management Strategies. In: Aquatic Plants for
Water Treatment and Resource Recovery.
Magnolia Publishing inc., Orlando, FL, pp. 643-
656, 1987.
17. Hillman, W.S., and D.C. Cully. The Use of
Duckweed. American Scientist 66:442-451,
1978.
18.Tchobanoglous, G., F. Maitski, K. Thomas, and
T.H. Chadwick. Evolution and Performance of City
of San Diego Pilot Scale Aquatic Wastewater
Treatment System Using Water Hyacinths.
Presented at the 60th Annual Conference of the
Water Pollution Control Federation, Philadelphia,
PA. October 5-8, 1987.
19. Dinges, R. Personal Communication. 1988.
20. Weber, A.S., and G. Tchobanoglous. Rational
Design Parameters for Ammonia Conversion in
Water Hyacinth Treatment Systems. JWPCF
57:316-323, 1985.
21 Gee and Jenson. Water Hyacinth Wastewater
Treatment Design Manual for NASA/National
Space Technology Laboratories, NSTL Station,
MS, 1980.
22. Reddy, K.R. Nutrient Transformations in Aquatic
Macrophyte Filters Used for Water Purification. In:
Water Reuse Symposium III, pp. 660-669, 1985.
23. Mitsch, W.J. Waterhyacinth (Eichhornia crassipes)
Nutrient Uptake and Metabolism in a North
Central Florida Marsh. Arch. Hydrobiol. 81:188-
210,1977.
24. Reddy, K.R. and J.C. Tucker. Productivity and
Nutrient Uptake of Water Hyacinth, Echhornia
Crassipes /. Effect of Nitrogen Source. Econ. Bot.
37:237-247, 1983.
25. Curran, G.M., S.B. Pearson, S.A. Curtin, and T.H.
Chadwick, San Diego's Aquatic Treatment
Program with Total Resource Recovery, in K.R.
Reddy and W.H. Smith, Aquatic Plants for Water
Treatment and Resource Recovery, Orlando,
Florida, March, 1987.
26. Black & Veatch, Engineers-Architects, Interim
Progress Report San Diego Aquaculture Project,
Kansas City, Ml, October, 1986.
27. Martinson, S. Personal Communication, California
State Water Resources Control Board,
Sacramento, CA, February, 1987.
28. Tchobanoglous, G. Aquatic Plant Systems for
Wastewater Treatment: Engineering
Considerations in K.R. Reddy and W.H. Smith,
Aquatic Plants for Water Treatment and Resource
Recovery, Orlando, Florida, March, 1987.
79
-------
Table 4-13. Aquatic Plant Systems Case Studies Summary
Item
Aquatic Plants
Preapplication Treatment
Special Design Features
Design Max. Flow, m3/d
Pond Surface Area, ha
Influent/Effluent BOD, mg/L
Influent/Effluent SS, mg/L
Influent/Effluent TN, mg/L
Hydraulic Surface Loading,
ma/ha-d
Capital Cost, $/ma-d
Yearly O&M Cost, $/m3'd
Capital Cost, $/ha
San Diego, CA
Water Hyacinths
Primary
Supplemental Aeration
3.79
0.65
~ 130/-9.5
-107/-10
23a/9a
583
580b
132b
340,000
Austin, TX
Water Hyacinths
Ponds
Covered System
7,570
1.6
131/17.6
142/11.3
55/1 2a
4,730
158
741 ,000
Orlando, FL
Water Hyacinths
AWT
Supplemental Nutrient Addition
30,300
12.1
4.9/31
3.8/36
13.0/8.2
2,500
66
18
165,000
NH4 + NO3-N.
Demonstration facility.
29. Design Criteria for Sewerage Systems, Texas
Department of Health, 1981.
30. Doersam, J., Use of Water Hyacinth for the
Polishing of Secondary Effluent at the City of
Austin Hyacinth Greenhouse Facility in K.R.
Reddy and W.H. Smith, Aquatic Plants for Water
Treatment and Resource Recovery, Orlando,
Florida, March, 1987.
31. Doersam, J. Personal Communication. August
1988.
32. Amasek, Inc., Assessment of Iron Bridge Water
Hyacinth Pilot Study, March 1984.
33. Amasek, Inc., Assessment of Winter Time
Nutrient Removal Performance of Five Water
Hyacinth Based Wastewater Treatment Systems
in Florida, March 1986.
80
-------
APPENDIX A
This Appendix contains lists of municipal and selected
industrial facilities that employ, or have employed,
constructed wetlands and aquatic plant systems. The
lists indicate the state and city where projects have
been identified. In some cases, the projects will have
been abandoned. These lists are included so that the
manual user can identify nearby projects and visit
them if so desired.
1. Constructed Wetlands
Alabama
Theodore, Jackson County
Russelville
Stevenson
Sand Mountain
Arizona
Lakeside
Camp Verde
Showlow
California
Arcata
Gustine
Hay ward
Laguna Miguel
Las Gallinas
Martinez
Santee
Idaho
Idaho City
Iowa
Granger
Norwalk
Riverside
Kansas
Saint Paul
Kentucky
Benton
Hardin
Pembroke
Louisiana
Benton
Carville
Haughton
Sibley
Maryland
Anne Arundel County
Emmitsburg
Glen Burnie
Massachusetts
Spenser
Michigan
Vermontville
Boscommon
Mississippi
Collins
Vay St. Louis (NASA/NSTL)
New Jersey
Avalon
Bernard
Beverly
Hightstown
Washington Township
Nebraska
Kimball
Nevada
Incline Village
Oregon
Cannon Beach
Pennsylvania
Elverton
Iselin
Lake Winola
81
-------
South Dakota
Hitchcock
Spenser
Wosley
Tennessee
Gatlinburg
Kingston
Tellico
Virginia
Monterey
In Other Countries
Richmond Australia
Mannersdorf, Austria
Listowel, Canada
Port Perry, Canada
Ringsted, Denmark
Rodekro (Jutland), Denmark
Othfresen, Germany
Windelsbleiche, Germany
Coromandel Township, New Zealand
Whangarei, New Zealand
2. Aquatic Plant Systems
Alabama
Enterprise
Arkansas
Wilton
Florida
Jupiter
Kissimmee
Melbourne
Orlando
Minnesota
Sleepy Eye
Mississippi
Bay St. Louis
North Biloxi
North Dakota
Devil's Lake
Texas
Austin
Baytow
San Benito
Virginia
Craig-NC
Washington
Stewart Park
82
-------
Multiply
m3/d
g/m2"d
kg/ha-d
kg/m2
m3/ha-d
m3/m2-d
m
m2
ha
m3
APPENDIX B
CONVERSION FACTORS
by
264
8.92
0.892
0.2
106.9
25
3.28
10.76
2.47
264.2
To Get
gpd
Ib/ac-d
Ib/ac-d
Ib/sq ft
gpd/ac
gpd/sq ft
ft
sqft
ac
gal
*u.s. o.P.o. 1988-548-158:87ol3
83
-------
------- |