QUOTES:

Man has a fundamental right to liberty, equality and satisfactory living conditions in an
environment whose quality permits him to live in dignity and well-being. He has solemn
duty to protect and improve the environment for present and future generations............

First principle of the United Nations Conference

On Human Environment, Stockholm, 1992.
(Royston, 1979)

The earth does not belong to man: man belongs to the earth. This we know. All things
are connected like the blood which unites the family. All things are connected. Whatever
befalls the earth befalls the sons of the earth. Man did not weave the web of life; he is
merely a strand in it. Whatever he does to the web, he does it to himself.

Chief Seattle, 1855.

(Royston, 1979)

DECLARATION:
I, Mugera Winnie Nyaguthii, do declare that this report is my original work and to the
best of my knowledge, it has not been submitted for any degree award in any University
or Institution.

Signed ___________________________Date________________________

CERTIFICATION
I have read this report and approve it for examination.

Signed____________________________ Date________________________
Prof.Thumbi

ii

ACKNOWLEDGEMENT
I would like to thank God for giving me the opportunity to undertake this course and for
the grace that enabled this project to come to a completion. I would also like to thank my
family and friends for guiding me and being by my every step of the research.
I am greatly indebted to my mentor,
J.N.Mburu for his guidance, commitment,
patience,believing in me and opening my eyes to the importance of constructed wetlands.

Special thanks to Prof.Thumbi, my Supervisor, for his guidance and positive criticism
through the research process.

To all I say thank you and God bless you.

NyaguthiiMugera

2010.

iii

DEDICATION

To my loving and caring family, Mugera, Wawira, Wambui, Wangeci.

iv

ABSTRACT
The increasing application of Constructed Wetlands for Waste water treatment coupled
with increasingly strict water quality standards is an ever-growing incentive for the
development of better process design tool. The use of constructed wetlands in Kenya is
still limited despite there being a great need for inexpensive and reliable onsite/
decentralized wastewater treatment technology. Indeed, the major challenge facing the
country of Kenya is the need to ensure on-going pollution control of the environment and
water resources, in the face of increased wastewater volumes from domestic, agricultural
and industrial sources. Among the policy and decision makers, there is a strong desire for
the wastewater treatment constructed wetlands technology to be adapted locally with as
much confidence in their operability and pollutant removal levels as with comparable
conventional wastewater treatment technology.
This work evaluates the first order COD kinetics in a pilot scale horizontal subsurface
flow Constructed Wetlands (HSSF-CW) treating domestic wastewater.The HSSF-CW
was situated in JKUAT treatment plant. The first order kinetics evaluated in the study
were associated with area-based, volume-based 1st order plug flow models and tank-inseries design models. The objective of the study was to calibrate these models and
thereafter use them to predict the effluent quality from the HSSF-CW. Further a
comparison of the rate constants calibrated from this wetland with those from literature
was undertaken. Samples were collected and dichromate laboratory tests for COD were
done. Influent COD was found to be between 326-52 mg/l. The effluent COD was found
to be between 30-140 mg/l. Hydraulic Loading Rate was found to between 868m/yr.Average volume-based, area-based and Tank-in series rate constants were found
to be 0.229, 17.016 and 3.3045 respectively. The values for the 3 models were found
to be 1.256,1.177 and 1.064 respectively. The k20 values from the 3models were found to
be 0.0573, 5.789 and 2.683. These constants were lower compared with those from other
studies as reported in the literature were found. This was contrary to the hypothesis that
higher rate constants are expected in the tropics. This could have been caused by the fact
that this study was conducted on an under-developed system that had undergone
renovations four months before the study had commenced and had been compared with
values obtained from mature systems.
v

TABLE OF CONTENTS:

QUOTES:......................................................................................................................... i
DECLARATION: ........................................................................................................... ii
CERTIFICATION .......................................................................................................... ii
ACKNOWLEDGEMENT ............................................................................................. iii
DEDICATION ............................................................................................................... iv
ABSTRACT.................................................................................................................... v
LIST OF TABLES .......................................................................................................... x
LIST OF GRAPHS ......................................................................................................... x
LIST OF PLATES ......................................................................................................... xi
LIST OF ABBREVIATIONS:....................................................................................... xi
CHAPTER 1: ...................................................................................................................... 1
1.0.0 INTRODUCTION ................................................................................................. 1
1.1.0 PROBLEM STATEMENT .................................................................................... 2
1.2.0 PROBLEM JUSTIFICATION............................................................................... 2
1.3.0 RESEARCH OBJECTIVES .................................................................................. 3
1.3.1 General objective ............................................................................................... 3
1.3.2 Specific objectives ............................................................................................. 3
1.4.0 RESEARCH HYPOTHESIS ................................................................................. 3
1.5.0 LIMITATIONS TO THE STUDY ........................................................................ 4
CHAPTER 2: ...................................................................................................................... 5
vi

2.0.0 LITERATURE REVIEW ...................................................................................... 5

2.1.0

CONSTRUCTED

WETLANDS

FOR

DOMESTIC

WASTEWATER

TREATMENT ............................................................................................................ 5
2.1.1 Discovery and evolving of constructed wetlands technology ........................ 5
2.1.2 Classification of constructed wetlands systems ............................................. 6
2.1.3 Design features of a constructed wetland ...................................................... 8
2.1.4 Maintenance of constructed wetlands .......................................................... 11
2.2.0 TREATMENT KINETICS .............................................................................. 11
2.2.1 Plug flow models: ........................................................................................ 13
a)

Area- based plug- flow model................................................................... 13

b)

Volume based plug flow model .............................................................. 13

c)

Tank-in Series models: ............................................................................. 13

2.2.2 Drawbacks of plug flow models .................................................................. 14

2.3.0 Rate constants: Study and literature values...................................................... 15
CHAPTER 3: .................................................................................................................... 16
3.0.0 RESEARCH METHODOLOGY......................................................................... 16
3.1.0 The pilot scale subsurface horizontal flow constructed wetland ..................... 16
3.1.1 Site description: ........................................................................................... 16
3.1.2 Experiments set-up ..................................................................................... 19
3.2.0 SAMPLING ..................................................................................................... 20
3.3.0 WATER QUALITY......................................................................................... 21

vii

3.3.1 TESTS .......................................................................................................... 21

CHAPTER 4: .................................................................................................................... 22
4.0.0 RESULTS AND DISCUSSION .......................................................................... 22
4.1.0 ENVIROMENTAL PARAMETERS RESULTS ............................................ 22
4.2.0 FLOW RATES................................................................................................. 23
4.2.1 Average flow rates, Hydraulic Retention Time and Hydraulic loading rate 23
4.3.0 COD RESULTS:.............................................................................................. 25
4.3.1 COD Removal:............................................................................................. 25
4.4.0 CALIBRATION OF VOLUME-BASED, AREA-BASED PLUG FLOW
MODELS ,TANK-IN-SERIES(TIS) MODEL RATE............................................. 28

4.5.0 K20 AND VALUES ...................................................................................... 30

4.6.0 COMPARISON OF THE KA/V/T

K20 AND VALUES FROM THIS

STUDY AND THOSE OBTAINED FROM ............................................................ 31

4.7.0 INFLUENCE OF ENVIRONMENTAL PARAMETERS .............................. 33

4.7.1 AERATION PROCESS ............................................................................... 33
4.7.2 EFFECTS OF TEMPERATURE ON %COD REMOVAL......................... 34
4.7.3 EFFECTS HYDRAULIC RETENTION TIME(HRT) ON %COD
REMOVAL ........................................................................................................... 35
4.7.4 EFFECTS OF PH ON % COD REMOVAL ................................................ 36
CHAPTER 5 ..................................................................................................................... 37
5.0.0 PREDICTIONS OF EFFLUENT COD USING VOLUME-

BASED, AREA-

BASED PLUG FLOW MODELS AND TANK-IN SERIES MODELS ..................... 37

viii

5.1.0 Volume-based plug flow model effluent COD predictions ............................ 37

5.2.0 Area-based plug flow model effluent COD predictions .................................. 39
5.3.0 Tank-in Series model effluent COD predictions.............................................. 40
CHAPTER 6 ..................................................................................................................... 43
6.0.0 CONCLUSIONS AND RECOMMENDATIONS .............................................. 43
6.1.0 Conclusions ...................................................................................................... 43
6.2.0 Recommendations ............................................................................................ 43
Bibliography ..................................................................................................................... 45
APPENDIX 1:............................................................................................................... 47
Laboratory tests procedures: ..................................................................................... 47
Open reflux method. ............................................................................................. 47
Dissolved Oxygen - titration method. ................................................................... 48
APPENDIX 2:............................................................................................................... 50
CALIBRATION VOLUME-BASED, AREA-BASED PLUG FLOW MODELS,
TANK-IN-SERIES(TIS) MODEL RATE CONSTANTS....................................... 50
EVALUATION OF K20 AND VALUES .............................................................. 51

APPENDIX 3:............................................................................................................... 53
PLATES:................................................................................................................... 53

ix

LIST OF TABLES
Table 2.1: Rate constants; values from literature.............................................................. 15
Table 4.1: Environmental parameters of the HSSF-CW................................................... 22
Table 4.2: Inflow and Outflow rates ................................................................................. 23
Table 4.3:Average flow rates, Hydraulic Retention Time, Hydraulic Loading Rates ...... 23
Table 4.4: Measured influent and effluent COD............................................................... 23
Table 4.5: Percentage COD removal ................................................................................ 23
Table 4.7: and values ............................................................................................ 23
Table 4.6: Volume-based, Area-based ,Tank-in Series rate constants .............................. 23

Table 4.8: Rate constant values from this study and other studies ................................... 31
The average of kv values was calculated from table 4.6. ................................................. 37
Table 5.1: Predicted effluent COD and Measured effluent COD .................................... 38
Table 5.2: Predicted effluent COD and Measured effluent COD ..................................... 39
Table 5.3: Predicted effluent COD and Measured effluent COD ..................................... 41

LIST OF GRAPHS
Graph 4.1: Influent COD trend ........................................................................................ .23
Graph 4.2:Effluent COD trend..23
Graph 4.3: Influent and Effluent COD comparison ........................................................ .23
Graph 4.4: Influent and Effluent DO comparison.33
Graph 4.5: % COD removal vs Temperature34
Graph 4.6: % COD removal vs HRT.35
Graph 4.7: % COD removal vs pH...36
Graph 5.1: Measured effluent COD vs Predicted effluent COD..38
Graph 5.2: Measured effluent COD vs Predicted effluent COD ...................................... 40
Graph 5.3:Measured effluent COD vs Predicted effluent COD ....................................... 41
Graph a: Ln kv vs Ln (T-20)..51
Graph b: Ln kA vs Ln (T-20) ............................................................................................. 52
Graph c: Ln kT vs Ln (T-20) ............................................................................................. 52

LIST OF PLATES
Plate 3.1: Student collecting samples at the HSSF-CW.................................................... 20
Plate 3.2: Student carrying out tests in the Environmental laboratory.............................. 21
Plate a: Student collecting samples at the influent point of the HSSF-CW ...................... 53
Plate b: Student collecting samples at the effluent point of the HSSF-CW ...................... 54
Plate c: Student testing samples for COD in the Environmental laboratory ..................... 54

LIST OF ABBREVIATIONS:
HSSF-CW =Horizontal Subsurface Flow Constructed Wetland
COD = Chemical Oxygen Demand
DO= Dissolved Oxygen
HRT = Hydraulic Retention Time
CSTR = Continuously Stirred Tank Reactors
Mg/l = Milligrams per liter
TIS = Tank in Series

xi

CHAPTER 1:
1.0.0 INTRODUCTION
Constructed wetlands for wastewater treatment can be defined as a designed and manmade complex of saturated substrate, emergent and submergent vegetation, animal life,
and water that simulate natural wetlands for human use and benefits (Hammer & Bastian,
1989).
Among the treatment wetlands, horizontal subsurface flow (HSSF) constructed wetlands
are a widely applied concept. Pretreated wastewater infiltrates horizontally through the
artificial filter bed, usually consisting of a matrix of sand or gravel and rhizomes. This
matrix is colonized by a layer of attached microorganisms that forms a so-called biofilm.
Purification is achieved by a wide variety of physical, chemical and (micro) biological
processes, like sedimentation, filtration, precipitation, sorption, plant uptake, microbial
decomposition and nitrogen transformations.

Constructed wetlands have proven to be a very effective method for the treatment of
domestic wastewater. For a small community with limited funds for expanding or
updating wastewater treatment plants, constructed wetlands are an attractive option.
Constructed wetlands blend into a natural landscape setting. In addition, wetlands add
aesthetic value, and provide wildlife habitat and recreation opportunities, they are cheap
to construct and maintain no skilled manpower is required, odour- free, free from
mosquitoes and diseases.
A lot of research is therefore required in order to popularize these systems. The first order
kinetics evaluated in the study were associated with area-based, volume-based 1st order
plug flow models and tank-in-series design models
In first order kinetics, the rate of disappearance of reactant is proportional to the amount
reactant present.
The reaction rate, k in waste water treatment is temperature dependent. Generally, as the
temperature increases, so does the rate at which the reaction occurs.

Plug flow models are used to describe reactions in a continuous flowing system. The
operating principle behind plug flow models is, what goes in first into the system is the
1

first to go out. Waste water flowing through the HSSF-CW is modeled as a series of
infinitely thin coherent plugs each with a uniform composition travelling in an axial
direction of the system with each plug having a different composition from the ones
before and after it. As the waste water flows through the system, the residence time of the
plug is a function of its position in the system. Residence time is the average amount of
time a discrete amount of reactant spends inside the system.

1.1.0PROBLEM STATEMENT
Ironically, developing countries with a large burden of untreated wastewaters and
unsatisfactory sanitation conditions have continued to neglect inexpensive wastewater
management solution like the constructed wetlands. One of the contributing factors is the
lack of local research on constructed wetlands in these countries, as evidenced by the
scanty information and data in literature with regard to design parameters. To promote
this optimal alternative technology to conventional waste water treatment, it is important
to derive/evaluate design parameters in the tropical developing countries. This study
evaluates the first order kinetics for the subsurface flow constructed wetland under
tropical climatic conditions.

1.2.0 PROBLEM JUSTIFICATION

1. The existing wastewater treatment plants in the country are becoming
overwhelmed due to increase in population and waste water volumes. Attempts
to implement or upgrade the conventional systems of wastewater treatment are
met by unreasonably high initial costs as well as complex maintenance structures.
2. Need for decentralized wastewater management systems due to mushrooming
townships, remotely located industries and agricultural enterprises. This calls for
an effective and relatively low maintenance method for providing onsite waste
water treatment.

1.3.0 RESEARCH OBJECTIVES

1.3.1 General objective

To evaluate the empirical design models (area-based and volume-based first

order, plug flow models) for COD using a pilot scale subsurface flow constructed
wetland performance in treating domestic waste.

1.3.2 Specific objectives

To determine the organic matter content of both the influent and the effluent in
the HSSF-CW.

To compare the rate constants calibrated from this HSSF-CW study with those
from the literature.

To compare tank-in series (TIS) model for COD with plug- flow models in
describing the HSSF-CW performance.

To determine the influence of environmental conditions i.e., Temperature and

DO (dissolved oxygen) on the treatment performance of a horizontal subsurface
flow constructed wetland.

1.4.0 RESEARCH HYPOTHESIS

Larger constant values are expected in the tropics as compared to those in the temperate
regions. It is therefore hypothesized that the Constructed Wetland systems would be more
efficient in the tropics.

1.5.0 LIMITATIONS TO THE STUDY

Limited financial resources and limited time factor will limit the extensiveness of
the study and therefore only COD will dealt with extensively in the study.

Portable meters (thermometer and pH meter) for monitoring field water

temperature and pH cannot be carried to site as required. This will therefore be
determined in the lab, this way variations may be introduced.

CHAPTER 2:
2.0.0LITERATURE REVIEW

2.1.0CONSTRUCTED

WETLANDS

FOR

DOMESTIC

WASTEWATER

TREATMENT
2.1.1Discovery and evolving of constructed wetlands technology
A lot of curiosity from ecologists over how much the wetlands were valuable arose in the
fifties and sixties. This initiated many studies on the topic. They seemingly stumbled over
the purification capabilities of these wetlands which set off the development of
constructed wetland technologies.
The technology of wastewater treatment by means of constructed wetlands with
horizontal sub-surface flow was pioneered in Germany based on research by the one Dr.
K. Seidel commencing in the 1960s and by Reinhold Kickuth in the 1970s. Her research
seemed heavily criticized since the investigations and calculations were mainly aimed at
nutrient removal through plant uptake which would require a regular harvesting and very
large surface areas.
The growing green awareness in the seventies catalyzed the abandoning of dumping
wastewater in the natural wetlands in favor of constructed wetlands (CWs). Another
positive boost was possibly due to the first energy crisis in 1973. Energy-devouring
technologies suddenly lost their attractiveness to the advantage of low-energy ones.
Natural systems of wastewater treatment are characterized by the use of renewable,
naturally occurring energies such as solar and wind energy, as opposed to conventional
treatment technologies which are dependent on non-renewable fossil fuel energies. The
above mentioned stimuli soon outweighed the classic distrust against new technologies
and, from then on, constructed wetlands development took an exponential growth.
The use of wetlands for wastewater treatment was stimulated by a number of studies in
the early 1970s that demonstrated the ability of natural wetlands to remove suspended

sediments and nutrients, particularly nitrogen and phosphorus, from domestic wastewater
(Mitsch & Gosselink, 1993).
The eighties were once again characterized by precaution and skepticism due to the
discovery of several drawbacks of the technology and failures of most prototypes. Further
research solved most of these problems and led to the maturity of the technology in the
nineties.
Constructed wetlands nowadays have many applications, ranging from the secondary
treatment of domestic, agricultural and industrial wastewaters to the tertiary treatment
and polishing wastewaters treated by means of activated sludge plants and even to the
treatment of storm waters. It has been adopted over the years in various parts of the world
for other various purposes. For example Constructed Wetlands at Milan Army
Ammunition Plant, Milan, Tennessee constructed in the World War II was used to treat
ground water contaminated with residue explosives. This was also used in Iowa Army
Ammunition Plant in Middletown, Iowa for the same purpose. In Fulton Regional Storm
Water Management facility in Edmonton, Alberta, for a storm water management system
(Jones, William.W., 1995).

In the developed countries, wetland treatment system has been identified as a treatment
system that features low operation cost, good reliability and effectiveness, in the face of
increased environmental awareness and enforcement of pollution prevention laws.
Besides, in many situations, a decentralized system of treating sewage wastes with
constructed wetlands, provide not only a more economical and energy efficient means of
achieving treatment objective but also a resource in form of reclaimed water available for
irrigation or creation of wildlife habitats (Campbell & M.H., 1999).

2.1.2 Classification of constructed wetlands systems

The following classification only considers the middle range ecosystems.i.e. the so-called
constructed wetlands, and is based on internationally accepted international water
6

associations scientific and technical reports on constructed wetlands for pollution

control(Kadlec et al.,2000b).
The various types are differentiated by water flow mode and plant species characteristics.
1) Above ground water: free-water-surface (FWS) constructed wetlands
With emergent macrophytes or helophytes, e.g. phragmitesaustralis (common reed),
typhaspp (cattails), scirpusspp, (bulrushes)
with floating-leaved, bottom rooted macrophytes, e.g. nymphaea spp.
(water lilies),nelumbospp,(lotus)
With free-floating macrophytes, e.g.eichhorniacrassipes (water hyacinth), lemna spp.
(duckweed)
With submerged macrophytes, e.g.elodeaspp, (waterweed), myriophyllum spp. (water
milfoil)
With floating mats, e.g.phragimitesaustrils (common weed), typhaspp (cattails), glyceria
maxima (giant sweet grass)
2) Below-ground water: subsurface-flow (SSF) constructed wetlands.
Horizontal-flow systems (HSSF), planted with emergent macrophytes or helophytes, e.g.
phragmitesaustralis (common reed), typhaspp (cattails), scirpusspp, (bulrushes)
Vertical-flow systems(VSSF), planted with emergent macrophysics or helophytes,e.g.
phragmitesaustralis(common reed) , typhaspp(cattails),scirpusspp,(bulrushes)
3)Above ground water flow system consists of a relatively shallow basin (depth between
0.3 and 1.8 meters), isolated from the ground water by means of a plastic liner or by a
local clay layer. Length-width ratios >2 are to be preferred in order to obtain near plugflow conditions. The inlet distribution and effluent abstraction systems should run along
the entire width of the basin to avoid short-circuiting and the existence of dead volumes.
When using free-floating macrophytes, floating barriers are often used to avoid the piling
up of plants in one corner due to wind action.

4)Below-ground water flow Horizontalsystem consists of the shallow(0.5-0.8m

deep)basin, isolated from the groundwater and usually filled with gravel although in
some cases local soils has been used. For the inlet and outlet zone, coarser gravel is
usually applied to allow a better spreading respectively collection of wastewater. In this
case the bed of impermeable material is sloped typically between 0 and 2 percent. It is
during the passage of wastewater through the rhizosphere that it gets cleaned through
microbiological degradation, and physical/ chemical process. The treated wastewater is
evacuated by means of a drainage tube at the bottom of the wetland. An appropriate
choice of filter material (c.q.hydraulic conductivity) and a correct length-width ratio are
indispensible to avoid above-ground water flow, which has detrimental effect on
treatment performance and can cause odour and insect nuisances.
5)Below- ground water Vertical flow systems consists of one or more filter layers of
coarse sand and/gravel with a total depth between 0.6 and 1.0 meter. Wastewater is
preferably spread equally over the top surface, then drains through the filter layers and is
collected at the bottom by means of drainage tubes. Loading often happens intermittently,
i.e. batch wise. Choosing the right filter material is a trade off between high respectively
low hydraulic conductivities.i.e. Less prone to clogging versus a longer hydraulic
retention time. (Risper, 2009)

2.1.3 Design features of a constructed wetland

There are a number of design features that can increase the efficiency ofconstructed
wetlands to trap and retain domestic wastewater pollutants.

Loading Rate: Proper sizing of a constructed wetland in relation to its watershed is

probably the most important factor affecting wetland performance. If the wetland is sized
too small, water flow through the system can be too rapid for effective treatment. Too
little water flowing through can result in stagnant water or temporary dry conditions.
Primary plant productivity and decomposition rates are both higher in flowing water but

high velocities discourage plant growth. Permanently flooded wetlands perform better
than wetlands which dry out seasonally. Therefore, designing the hydraulic loading rate is
critical. Ideally, for optimal performance, the size of the constructed wetland should be
from 1% to 5% of the size of its drainage area. For example, a 25-acre watershed would
require a 1-acre wetland. Designing hydraulic
loading by analyzing existing channel discharge or watershed runoff coefficients is more
precise than the 5% rule above.

Hydraulic Retention Time: Hydraulic retention time refers to the length of time water
remains in the constructed wetland. It is closely related to hydraulic loading rate. The
treatment efficiency of a CW system is usually improved by decreasing the hydraulic
retention time; the longer the hydraulic retention time (HRT), the greater the nutrient
removal (Sakadevan and Bavor, 1999). The most effective HRT ranges from 4 to 15
d(Metcalf and Eddy Inc., 1991; Watson and Hobson, 1989). Gersberg et al. (1989) found
that even a short HRT of 3 to 6 d was effective in removing diseasecausing bacteria and
viruses. Operating with a shorter HRT means smaller land-area requirements.
The longer water remains in the wetland the greater chance ofsedimentation, adsorption,
biotic processing and retention of nutrients. Proper sizing of the wetland is important but
restricting the size of the wetland outlet is also effective. For wetlands with channel flow,
the outlet cross sectional area should be less than 1/3 that of the inlet.

Water Velocity: Peak water velocities through the wetland should not exceed1.5
feet/second. High velocities can wash out rooted vegetation and scour deposited
sediments. Keeps velocities low by regulating hydraulic loading, limiting the gradient
(slope) through the wetland, restricting the outlet size,
Creating sinuous edges and planting persistent emergent vegetation. Ideally, flow
velocities should be less than 0.6 feet/second.

Water Depth: Water depths less than 40 inches result in greater resistance to flow and
shallow depths favor aquatic vegetation. The preferred depth range for emergent plants is
01.0 feet of water; for rooted surface plants, 1.02.0 feet of water; and for rooted
submersed plants, 1.5 6.5 feet of water. Pools deeper
than 40 inches should also be included in the wetland design to maximize sediment
deposition and provide winter fish habitat.

Maximize Edge: Sinuous edges between the terrestrial and aquatic zones provide more
resistance to flow and more edge habitat for plants and animals. Round wetlands have the
least amount of edge per given surface area.

Minimize Edge Slope: The terrestrial-aquatic boundary should have a very gradual slope.
This allows for the establishment of a continuum of emergent species and reduces the
erosive effects of waves hitting a sharp shoreline boundary.

Persistent Emergent Vegetation: Persistent emergent vegetation has stems which persist
even after the growing season. This provides year-round resistance to water flow.
Persistent emergent plants include: cattail (Typha spp.),iris (Iris pseudacorusor I.
versicolor), rush (Juncus spp.), cordgrass (Spartina
spp.), reedgrass (Calamagrostis spp.), sawgrass (Cladiumjamaicense), andswitch-grass
(Panicumvirgatum).

Woody

plants

such

as

alder

(Alnus

spp.),buttonbush

(Cephalanthusoccidentalis), black willow (Salix nigra), and othersare useful edge species
with persistent stems. Aquatic bed or submergent vegetation removes nutrients seasonally
but does not offer significant frictional resistance to suspended sediments.

Pre-sedimentation

Basin:

Many

constructed

wetland

designs

incorporate

presedimentationbasin to trap sediments and large particulates before they enter the
wetland. This can extend the life of the constructed wetland and ultimately enhance
treatment efficiency.
10

2.1.4 Maintenance of constructed wetlands

Constructed wetland planning should not overlook the need for long-term maintenance.
Additional vegetation planting may be required to speed plant coverage, replace damaged
plants or to try more suitable varieties. Perimeter fencing may be required if livestock
grazing is anticipated to be a problem.
Maintenance may be needed to control the spread of undesired plant species such as
purple loosestrife.
Inlets and outlets can become blocked with debris which will require periodic removal.
Inlet and outlet structures should be inspected weekly and especially following big storm
events. Most importantly, if the wetland functions well as a sediment and nutrient trap, it
may eventually require dredging to remove accumulated materials. Thus, vehicular
access to the site must be provided for maintenance vehicles and possibly dredging
equipment (Jones, William.W., 1995)
2.2.0TREATMENT KINETICS
The state-of-the-art in constructed treatment wetlands modeling consists of first-order
equations which in case of constant conditions (e.g. influent, flow and concentrations)
and an ideal plug-flow behavior predict an exponential profile between inlet and
outlet(D., Rousseau, 2005).
The wetland is designed as a channel with a limited width and can therefore be calculated
as a plug flow reactor where the wastewater moves as a front in one direction (along the
channel). Most systems in the US and Europe, whose design is based on kinetics, use a
first order plug flow model ((EPA), 1993).

The following first-order reaction equation (IWA, 2000)is used to describe pollutant
removal:

Ce =Exp (-kv x HRT) (1)

Co

11

Where: Ce is the constituent concentration in the effluent (mg/l)

Co is the constituent concentration in the influent (mg/l)
kv is the volumetric decay rate (d-1)
The values of kv are obtained through regression of ln (Ce/Co) versus HRT.
Some reported kinetic data (IWA, 2000)(Kadlec, H, & Knight, 1996)are based on area:

K =kv x h..(2)
Where k is the area-based decay rate constant,

is

the porosity or space available for

water to flow through the wetland and h is the water depth(Reed et al., 1995);(IWA,







2000). Equation (1) can be modified by substituting kv from equation (2) giving;

Where HLR is hydraulic loading rate.

  


The influent mass loading rates of constituentsis calculated as:

 

Where Q is the inflow rate and A is the surface area of the system
The removal efficiencies of (%) of pollutants were calculated as:



and the mass removal rate = IL x

  
%


12

2.2.1 Plug flow models:

The following design models were used during the research:(M.F.Dahab & Wenxin
(Emy) Liu, 2009)
a) Area- based plug- flow model

   

 !  "# $!"#%

Cout = effluent concentration (mg/l)

Cin = influent concentration (mg/l)
t = hydraulic retention time(d)

kT = area- based temperature dependent rate constant (d-1)

= the temperature coefficient (Dimensionless)
T = temperature at different sampling time 0C
b) Volume based plug flow model



'
 &


Cout= effluent concentration (mg/l)

Cin = influent concentration (mg/l)
q = hydraulic loading rate (m/yr)

kT = volume- based temperature dependent rate constant (d-1)

y = fractional distance through the wetland

c)

Tank-in Series models:



)
*
()
) + ! ,

With this method, the wetland maybe conceptually partitioned into a number of equal
sized sections

(n), each of which is presumed to be completely mixed,

where:

13

n = number of tanks
t = HRT in the nth CSTR reactor

The tanks-in series (TIS) model is commonly used for modeling of pollutant removal in
ponds and wetlands (Kadlec, H, & Knight, 1996). The model represents a series of
continuously stirred tank reactors (CSTRs) where a substance is removed in each tank
according to first-order kinetics. The number
Of tanks, N represents the degree of mixing. High value of N means a small degree of
dispersion,
i.e., a plug flow reactor (PFR), while N = 1 defines completely mixed reactor.

2.2.2 Drawbacks of plug flow models

According to(D., Rousseau, 2005), after a research on Model based design for a pilotscale constructed reed bed belonging to Aqua fin NV and located in Aartselaar, Belgium;

Calibration of the parameters k, C*, and is mostly done on the basis of inlet
and outlet concentrations, and not on the basis of transect data, although the latter
are to be preferred for calibration purposes. Because these parameters lump a
large number of other characteristics representing the complex web of
interactions in constructed treatment wetland as well as external influences like
weather conditions, a large variability can be observed in reported k, C*, and
values.

The equations are based on the assumptions of plug flow and steady-state
conditions. However, small scale wastewater treatment plants under which most
treatment wetlands can be ranged are subject to large influent variations whereas
the larger ones are subject to hydrological influences, thus causing in both cases
non-steady state conditions.

14

Short-circuiting and dead zones are common phenomena in constructed treatment

wetlands causing non-ideal plug-flow conditions, thus jeopardizing the use of the
first-order model.

The so-called rate constants do not seem to be constant at all but dependent on
the influent concentrations, the HLR and the water depth.

2.3.0 Rate constants: Study and literature values

Table 2.1: Rate constants; values from literature
STUDIES

1st

VOLUME-BASED
ORDER MODEL
kv

k20(d-1)

1st

AREA-BASED

TIS MODEL

ORDER MODEL

kA(m/yr)

k20(d1

kT

k20(d-1)

Nebraska(2009)

--

0.256

1.011

24.6

--

1.008

--

0.353

1.015

Axler, et al(2000)

--

--

--

19.5

--

1.071

--

--

--

IWA(2000)

--

--

--

21.9-113

--

--

--

--

--

Kadlec&

--

--

--

180

--

--

--

--

--

Reed, et al(1995)

0.68

1.104

1.06

--

--

--

--

--

--

European

--

0.1

--

--

--

--

--

--

--

USEPA(1988)

--

0.86

1.06

--

--

--

--

--

--

Jing et al

0.40

--

--

--

4.79

--

--

--

--

Knight(1996)

Cooper(1990)

15

CHAPTER 3:
3.0.0 RESEARCH METHODOLOGY
3.1.0 The pilot scale subsurface horizontal flow constructed wetland
3.1.1 Site description:
The pilot scale constructed wetland is sited within Jomo Kenyatta University of
Agriculture and Technology (JKUAT) sewage treatment works premises.(JKUAT is
located in Juja town, 40 Km North East of Nairobi City on Nairobi-Thika road at latitude
o

1 10S and longitude 37 E, and at an altitude of 1463m above sea level).

JKUAT Sewage works treats domestic wastewater by use of a series of 5 wastewater

stabilization ponds: two primary facultative ponds, two secondary facultative ponds and
one maturation pond. The final effluent is discharged to a near-by river via a natural
water way.

A schematic diagram of the wetland set-up at the JKUAT sewage works is shown in
figure below.

16

5
7

6
8

22

Figure 3.1: Schematic diagram of the pilot scale experiment set-up at the JKUAT
sewage treatment works
1. Raw sewage
4. Primary facultative pond
6. Pilot scale constructed wetland

2. Bar screens

3. Grit chamber

5. Secondary facultative pond

7. Maturation pond

8. Treated effluent

17

7.5m

(Vegetation and gravel)

A (vegetation and gra
3m

CELL A
B (gravel)
(Gravel)
3m

3m

End

CELL B

Effluent
C (pond)

(Pond)

3m

3m
CELL C

(Vegetation and gravel)

3m
CELL D

3m

Figure 3.2: Plan layout of the pilot scale HSSF-CW.

Vegetation

Influent
Effluent

Inlet structure

Outlet
structure
Plastered
reinforced
concrete base (Slope at
1%)

Gravel
material

Wastewater
level

18

3.1.2 Experiments set-up

Controlled quantities of wastewater and rain were the only inputs into the pilot HSSFCW system.
2

The wetland consists of four cells set in parallel, each 22.5m .

The wetland cells are 7.5m long and 3m wide. They have vertical sides and a bottom
horizontal slope of 1%. Cells A, B, and D were filled with gravel to a depth of 0.6m. The
gravel ranges in size from 9-37mm, with a porosity of 45 % and a hydraulic conductivity
3

Ks = 4050 m /m .d
The macrophyte Cyperus papyrus was introduced into cells A and D, using clumps at a
spacing of 0.75m by 0.75m. While cell B is unvegetated and cell C is a pond with no
gravel or vegetation.
The wastewater level in the wetland cells is maintained at an average depth of 0.55m
within the gravel and regulated by outlet pipes positioned at a height 0.54m from the
wetland floor.

Cells A was used in the measurements of the various parameters required for this
study.
The wetland receives a continuous feed of primary effluent from a primary facultative
pond at the JKUAT sewage works. This influent to the pilot wetland is tapped from the
primary facultative pond effluent stream at a manhole with the aid of a sluice valve. The
desired flow rate is maintained manually by regulating the sluice valve.

19

3.2.0 SAMPLING

Sample collection was carried out between 6:30am and 8:00am on a weekly basis
(on Wednesdays or Thursdays) during the study. Samples of the effluent and
influent flow were collected at their respective outflow and inflow points, into
clean sampling bottles.

Care was taken not to collect deleterious materials {which would have otherwise
altered lab results} into the bottles.

Safety gear such as an overall and safety gloves were worn to prevent contact
with the wastewater.

The period between sampling and carrying out laboratory tests was maintained
below the allowable six hours; in so doing, preservation procedures were
unnecessary saving on both time and resources.

Plate 3.1: Student collecting samples at the HSSF-CW

20

3.3.0 WATER QUALITY

3.3.1 TESTS

This included both on-site monitoring and laboratory tests

Temperature: A Mercury thermometer was used on both the influent and

effluent samples collected.

pH:A pH meter was used to monitor the pH of both the influent and effluent of
the samples collected.

Flow rates: volumetric method was used to measure the flow rate of both
the influent and effluent.

DO: Dissolved Oxygen titration method was carried out in the lab because a DO
meter was not available in the laboratory. See appendix 1 for procedure.

Chemical Oxygen Demand (COD): Open reflux method was used to measure
the COD in both the influent and effluent. See appendix 1 for procedure.

Plate 3.2: Student carrying out tests in the Environmental laboratory

21

CHAPTER 4:
4.0.0 RESULTS AND DISCUSSION
4.1.0 ENVIROMENTAL PARAMETERS RESULTS
Environmental parameters were monitored to determine their influence on COD removal.
The following environmental parameters were obtained:
Table 4.1: Environmental parameters of the HSSF-CW

DO

DO

Date

Temp,T( c)

pH

(Influent, mg/l)

(Effluent, mg/l)

4/11/2009

23.5

7.3

0.38

0.42

5/11/2009

23

7.2

0.4

0.46

11/11/2009

21.5

7.0

0.48

0.68

12/11/2009

23

7.3

0.46

0.52

18/11/2009

23

7.2

0.38

0.4

19/11/2009

23

7.1

0.4

0.46

26/11/2009

21

6.9

0.44

0.56

27/01/2009

23

7.0

0.36

0.42

2/02/2010

22

7.2

0.4

0.44

3/02/2010

23

7.0

0.34

0.38

9/02/2010

23

7.3

0.42

0.48

22

4.2.0 FLOW RATES

The following flow rates were obtained using the volumetric method from the
HSSF-CW:
Table 4.2: Inflow and Outflow rates
Date
4/11/2009

INFLOW (m3/d)

OTUFLOW (m3/d)
7.39

13.13

5/11/2009

0.69

0.52

11/11/2009

1.04

2.25

12/11/2009

3.89

4.18

18/11/2009

0.95

1.67

19/11/2009

2.16

1.67

26/11/2009

1.91

1.86

27/01/2009

2.25

2.23

2/02/2010

1.77

1.71

3/02/2010

1.89

1.75

9/02/2010

3.55

2.59

4.2.1 Average flow rates, Hydraulic Retention Time and Hydraulic loading rate
Using table 4.2 the equations shown below were used to compute the Average
flow rates, Hydraulic Retention Time (HRT) and Hydraulic Loading Rate with the
help Ms Excel spreadsheet software:
./01230456712809, ;2/


$;< + ;0%
2

where ;<: inEluent Elow rate and ;0: efEluent Elow rate

MNO12P5<QR080S8<6ST<U0, T 
 12.375ma

MNO12P5<Qf62O<S3R280, g 

V7$W%
where V7: useable wetland water volume
;2/

;<
.7

W: porosity  0.45

where .7: wetland surface area  22.5U"

23

The results obtained were as shown below:

Table 4.3:Average flow rates, Hydraulic Retention Time, Hydraulic Loading Rates

Date

Average flow rates

(Qav) (m3/d)

Hydraulic
Retention Time
(HRT) d

Hydraulic Loading
Rate (m/yr)

4/11/2009

10.26

0.54

119.84

5/11/2009

0.60

9.21

8.41

11/11/2009

1.64

2.48

36.44

12/11/2009

4.03

1.33

67.74

18/11/2009

1.31

3.33

27.10

19/11/2009

1.91

2.91

27.10

26/11/2009

1.89

3.00

30.11

27/01/2009

2.24

2.50

36.20

2/02/2010

1.74

3.25

27.79

3/02/2010

1.82

3.18

28.37

9/02/2010

3.07

2.15

41.95

24

4.3.0 COD RESULTS:

The following COD results were obtained using the standard laboratory test:
Table 4.4: Measured influent and effluent COD
Date

COD in (mg/l)

COD out (mg/l)

4/11/2009

326

140

5/11/2009

94

52

11/11/2009

156

114

12/11/2009

110

84

18/11/2009

84

72

19/11/2009

124

118

26/11/2009

116

94

27/01/2009

160

122

2/02/2010

52

30

3/02/2010

142

110

9/02/2010

104

88

4.3.1 COD Removal:

From Table 4.4 the following trends were be observed:
Graph 4.1: Influent COD trend

CODin(mg/l)

9/2/2010

3/2/2010

2/2/2010

27/01/2010

26/01/2010

19/11/2009

18/11/2009

12/11/2009

11/11/2009

5/11/2009

4/11/2009

400
300
200
100
0

CODin(mg/l)

25

The influent COD range of this system was found to be between 52 -326mg/l. The
highest COD value in the waste water could have been associated with technical
failure of the pond system (on this particular day) that pre-treats the waste water
before it gets to the constructed wetland. No particular trend was observed during
the sampling days.

Graph 4.2:Effluent COD trend

CODout(mg/l)

9/2/2010

3/2/2010

2/2/2010

27/01/2010

26/01/2010

19/11/2009

18/11/2009

12/11/2009

11/11/2009

5/11/2009

4/11/2009

150
100
50
0

CODout(mg/l)

The effluent COD range was found to be between 140-50mg/l which means that
the waste water effluent had not reached the ministry of water requirements
(which is >50mg/l) and therefore required further treatment before being released
into any river. No particular trend was observed during the sampling days.

Comparing the influent COD and effluent COD the following was observed:

26

Graph 4.3: Influent and Effluent COD comparison

350
300
250
200
150
100
50
0

CODin(mg/l)

9/2/2010

3/2/2010

2/2/2010

27/01/2010

26/01/2010

19/11/2009

18/11/2009

12/11/2009

11/11/2009

5/11/2009

4/11/2009

CODout(mg/l)

From the above compar

comparisons,
isons, there was a significant amount of COD removal on
each of the days sampling was done proving that the HSSF
HSSF-CW
CW was efficient in
COD removal.

The removal efficiencies were as shown:

Table 4.5: Percentage COD removal

Date

COD in (mg/l)

COD out (mg/l)

4/11/2009

326

140

%
EFFICIENCY
57.06

5/11/2009

94

52

44.68

11/11/2009

156

114

26.92

12/11/2009

110

84

23.64

18/11/2009

84

72

14.29

19/11/2009

124

118

4.84

26/11/2009

116

94

18.97

27/01/2009

160

122

23.75

2/02/2010

52

30

42.31

3/02/2010

142

110

22.54

9/02/2010

104

88

15.38

27

4.4.0 CALIBRATION OF VOLUME-BASED, AREA-BASED PLUG FLOW

MODELS ,TANK-IN-SERIES(TIS) MODEL RATE CONSTANTS
The volume- based area-based plug flow models, Tank-in-Series (TIS) model
were calibrated with the help of Ms Excel spreadsheet software.
ijkl  mnopjkqrjslpj:

k, 

tn  mnopjkqrjslpj:

k,

un)  ()  vt(olpj:

k,
()

()

w

() i , (i)




 '
&

.(ii)

y (iii)

)xu ,

7z010: {|  /65PU0}290O80U~0128P10O0~0SO0S81280Q6S982S8$O %

{  2102}290O80U~0128P10O0~0SO0S81280Q6S982S8$O %

{ !  T2S{  <S901<0980U~0128P10O0~0SO0S81280Q6S982S8$O )
8  MNO12P5<QR080S8<6ST<U0 $O%

 8z080U~0128P10Q6044<Q<0S8 $<U0S9<6S5099%
g  MNO12P5<Qf62O<S3T<U0 $UN1%

N  412Q8<6S25O<982SQ08z16P3z8z070852SO

S  SPU}016482S{9  3 <S8z<9~218<QP521M 

  0445P0S8Q6SQ0S8128<6S $U35%

  <S45P0S8Q6SQ0S8128<6S $U35%

(See appendix 2 for a step by step evaluation of the models.)

The rate constants obtained were as follows:

28

Table 4.6: Volume-based, Area-based ,Tank-in Series rate constants

Volume-based

rate

Area-based

Date

constant (kV)

constant (kA)

4/11/2009

1.5573

101.2926

5/11/2009

0.0643

4.9789

11/11/2009

0.1265

11.4302

12/11/2009

0.2022

18.2681

18/11/2009

0.0462

4.1771

19/11/2009

0.0171

1.3440

26/11/2009

0.0701

6.3313

27/01/2009

0.1086

9.8148

2/02/2010

0.1692

15.2877

3/02/2010

0.0802

7.2445

9/02/2010

0.0776

7.0072

rate

Tank-in series rate

constant (kit)
5.9857
4.4231
3.1053
2.9286
2.5000
2.1525
2.7021
2.9344
4.2000
2.8727
2.5455

29

4.5.0 K20 AND VALUES

To obtain k20 and values, the following equation was used and regression done
with the help of Ms Excel spreadsheet software.

$/i/u%   $u%.(iv)

7z010:  8z080U~0128P10O0~0SO0S81280Q6S982S8$$O %
{"#  8z01280Q6S982S828280U~0128P106420

T  80U~0128P1028O<44010S892U~5<S38<U0$6 %

(See appendix 2 for a step by step evaluation of the models)

The {"# and values obtained are as shown:
Table 4.7: and values
Constants


Volume-based
rate constant(kava)
0.0573

Area-based rate
constant(kA)
5.789

Tank-in-series
rate constant(kit)
2.683

1.256

1.177

1.064

From table 4.7it was observed that the values obtained were greater than 1. This

Significance of k20 and values obtained:

proved that the organic removal in the HSSF-CW is actually microbial. If the

value was= 1, the process usually is physical (Filtration, sedimentation).It was

further observed (using equation iv) that:

On doubling/increasing the value, a higher kV/A/T value was obtained.

Reducing the value significantly (0.12.values) a lower kV/A/T value was
obtained.

On doubling/increasing the K20 value, a higher kV/A/T value was obtained.

On halving the value, a lower kV/A/T value was obtained.

When the temperature value was increased, a higher kV/A/T value was
obtained and vies-versa.
This means that the temperature effects were very significant in the COD
removal process. (i.e. the higher the temperature the more the rate of
purification and vies-versa)

30

4.6.0 COMPARISON OF THE KA/V/T

K20 AND VALUES FROM THIS

STUDY AND THOSE OBTAINED FROM LITERUAURE

The comparison was as shown:

Table 1.8: Rate constant values from this study and other studies
VOLUME-BASED

STUDIES

st

1 ORDER MODEL
kv

k20(d1

This

0.22

0.057

1.25

study(2010)

Nebraska(200

--

0.256

1.01

9)
Axler,

--

--

--

TIS MODEL

ORDER MODEL
kA(m/y

k20(

kT

-1

k20(

-1

r)

d )

17.016

5.78

1.17

3.304

2.68

1.06

--

1.00

--

0.35

1.01

--

--

--

24.6

1
et

1st

AREA-BASED

d )

8
19.5

--

al(2000)

1.07
1

IWA(2000)

--

--

--

21.9-

--

--

--

--

--

113
Kadlec&

--

--

--

180

--

--

--

--

--

0.68

1.104

1.06

--

--

--

--

--

--

--

0.1

--

--

--

--

--

--

--

--

0.86

1.06

--

--

--

--

--

--

0.40

--

--

--

4.79

--

--

--

--

Knight(1996)
Reed,

et

al(1995)
European
Cooper(1990)
USEPA(1988
)
Jing et al

31

Lower kv, kA and their corresponding k20 values was obtained in this study as
compared to other studies. This was contrary to the hypothesis that a larger
kvalues would be expected in the tropics.
This could have been due to:

The HSSF-CW being used for the study had undergone

renovations a few months before the study began and therefore
the system was not fully developed. Usually a minimum period of
SIX months of operation is recommended. During the course of
the research, the system was about FOUR months old.
The other studies on the other hand could have been carried out
on mature systems. Mature systems have a well developed root
system which provides sufficient room for biofill attachment.
Colonies are much larger and the biofilm is mature.

The other studies could have also been carried out during the
summer period when microbial activities are much more
enhanced. (The temperatures for this study ranged between 2024oC. Summer temperatures on the other hand are much higher
than this.)

The rate constants obtained were however comparable with values from other
studies.

There was not much data available for comparison on studies done for Tank-in
Series models. The little data obtained however showed that the k value obtained
for TIS model in this study was higher. This is not very conclusive as more
studies are needed for concrete conclusions to be made.

32

4.7.0 INFLUENCE OF ENVIRONMENTAL PARAMETERS

4.7.1 AERATION PROCESS
COD removal from waste water is aerobic. The higher the oxygen levels, the
higher the degradation rates. Plants add oxygen into the system during the
purification process.
A comparison of the influent DO and effluent DO was as shown:

Graph 4.4:: Influent and Effluent DO comparison

0.7
0.6
DO (mg/l)

0.5
0.4
0.3
INFLUENT DO(mg/l)

0.2

EFFLUENT DO(mg/l)

0.1
9/2/2010

2/2/2010

26/01/2010

18/11/2009

11/11/2009

4/11/2009

Dates

From the above comparisons, there was a signi

significant
ficant amount of oxygen addition
on each of the days sampling was done proving that the system was very
effective.

33

4.7.2 EFFECTS OF TEMPERATURE ON %COD REMOVAL

A scatter plot of %COD removal against temperature was as shown:

Graph 4.5: % COD removal vs Temperature

%COD REMOVAL vs
TEMPERATURE
80.00
%COD REMOVAL

70.00
60.00
50.00
40.00
30.00

Series1

20.00

Linear (Series1)

10.00
0.00
20

21

22

23

24

Temperature oC
R = 0.106

The correlation between %COD removal and Temperature was found to be

10.6%. The temperature range was between 20-24oC. Temperature data was not
enough to reveal a trend.

34

4.7.3 EFFECTS HYDRAULIC RETENTION TIME (HRT) ON %COD

REMOVAL

A scatter plot of %COD removal against HRT was as shown:

Graph 4.6: % COD removal vs HRT

% COD REMOVAL

% COD REMOVAL VS HRT

50.00
45.00
40.00
35.00
30.00
25.00
20.00
15.00
10.00
5.00
0.00
0.0000

Series1
Linear (Series1)

5.0000
HRT,t (d-1)

10.0000
R = 0.345

The correlation between %COD removal and HRT was found to be 34.5%. This
was fairly correlated.
It was further observed, that an increase Hydraulic Retention Time (HRT) led to
an increase in %COD removal.

35

4.7.4 EFFECTS OF PH ON % COD REMOVAL

pH an important factor that affects the microbiological processes. The optimal
range fluctuates for the different processes but generally varies between 7.0 and
8.5. This treatment system buffers itself.
A scatter plot of %COD removal against pH was as shown:

Graph 4.7: % COD removal vs pH

%COD REMOVAL vs pH
80.00
70.00
% COD REMOVAL

60.00
50.00
40.00
Series1
30.00

Linear (Series1)

20.00
10.00
0.00
6.8

6.9

7.0

7.1

7.2

7.3

7.4

pH

R = 0.157

The correlation between %COD removal and pH was found to be 15.7%. It was
further observed, that the pH levels were circum-neutral.

36

CHAPTER 5
5.0.0 PREDICTIONS OF EFFLUENT COD USING VOLUMEBASED, AREA-BASED PLUG FLOW MODELS AND TANK-IN
SERIES MODELS

The volume-based, Area-based plug flow models and Tank-in series models were
used to predict effluent COD using the obtained kv, kA and kT values.

5.1.0Volume-based plug flow model effluent COD predictions

The average of kv values was calculated from table 4.6.
Equation(i) was used to predict effluent COD (mg/l)
N/B: The average kv was kept constant, Cin (mg/l) and HRT (t) were varied
accordingly.

The predicted effluent COD (mg/l) values from the volume -based plug flow
model are as shown below and compared with the measured effluent COD (mg/l)
values:

37

Table 5.1: Predicted effluent COD and Measured effluent COD

Date

Predicted effluent
COD (mg/l)
287.90

4/11/2009

Measured effluent
COD (mg/l)
140

5/11/2009

11.41

52

11/11/2009

88.43

114

12/11/2009

81.05

84

18/11/2009

39.15

72

19/11/2009

63.72

118

26/11/2009

58.35

94

27/01/2009

90.35

122

2/02/2010

24.70

30

3/02/2010

68.49

110

9/02/2010

63.51

88

Graph 5.1: Measured effluent COD vs. Predicted effluent COD

MEASURED EFFLUENT COD (mg/l)

MEASURED EFFLUENT COD (mg/l) vs PREDICTED

EFFLUENT COD (mg/l)
200
150
100
50
R = 0.496

0
0

50

100

150

200

250

300

350

PREDICTED EFFLUENT COD (mg/l)

Using Ms Excel software, the correlation between the predicted and measured
effluent COD values using the volume-based plug flow model was found to be
49.6%

38

5.2.0Area-based plug flow model effluent COD predictions

The average of kA values was calculated from table 4.6.
Equation(ii) was used to predict Cout (mg/l)
N/B: The average kA was kept constant;Cin (mg/l) and Hydraulic Loading Time
(q) were varied accordingly. y=1
The predicted effluent COD (mg/l) values from the area -based plug flow model
are as shown below and compared with the measured effluent COD (mg/l) values:

Table 5.2: Predicted effluent COD and Measured effluent COD

4/11/2009

Predicted effluent
COD (mg/l)
282.85

Measured effluent
COD (mg/l)
140

5/11/2009

12.43

52

11/11/2009

97.80

114

12/11/2009

85.57

84

18/11/2009

44.83

72

19/11/2009

66.18

118

26/11/2009

65.92

94

27/01/2009

99.99

122

2/02/2010

28.19

30

3/02/2010

77.95

110

9/02/2010

69.32

88

Date

39

Graph5.2: Measured effluent COD vs. Predicted effluent COD

MEASURED EFFLUENT COD (mg/l)

MEASURED EFFLUENT COD (mg/l) vs PREDICTED

EFFLUENT COD (mg/l)
200
150
100
50
R = 0.529

0
0

50

100

150

200

250

300

PREDICTED EFFLUENT COD (mg/l)

Using Ms Excel software, the correlation between the predicted and measured
effluent COD values using the Area-based plug flow model was found to be
52.9%

5.3.0Tank-in Series model effluent COD predictions

The average of kT values was calculated from table 4.6.
Equation(iii) was used to predict Cout (mg/l)
N/B: The average kT was kept constant,Cin (mg/l) and HRT (t) were varied
accordingly. n=3

The predicted effluent COD (mg/l) values from the Tank-in Series model are as
shown below and compared with the measured effluent COD (mg/l) values:

40

Table 5.3: Predicted effluent COD and Measured effluent COD

4/11/2009

Predicted effluent
COD (mg/l)
189.00

Measured effluent
COD (mg/l)
140

5/11/2009

1.12

52

11/11/2009

22.38

114

12/11/2009

33.28

84

18/11/2009

7.64

72

19/11/2009

14.03

118

26/11/2009

12.50

94

27/01/2009

22.50

122

2/02/2010

4.93

30

3/02/2010

13.91

110

9/02/2010

18.11

88

Date

Graph 5.3:Measured effluent COD vs Predicted effluent COD

MEASURED EFFLUENT COD (mg/l)

MEASURED EFFLUENT COD (mg/l) vs PREDICTED

EFFLUENT COD (mg/l)
200
150
100
50
R = 0.311

0
0

50

100

150

200

PREDICTED EFFLUENT COD (mg/l)

Using Ms Excel software, the correlation between the predicted and measured
effluent COD values using the Tank-in series model was found to be 31.1%

41

Significance of effluent COD predictions made from the 3 models

This showed that the Area-based 1st order plug flow model produced the best
predictions (the predicted effluent COD had a correlation of 52.9% with the
measured effluent COD). This could only mean one thing, that Hydraulic Loading
Rate (q) is a very important factor in the performance of the HSSF-CW. The
wetland surface area and inflow rates should be well taken care of during the
design and construction of the HSSF-CW.
It was further observed that the Measured effluent COD was generally higher than
the Predicted effluent COD. This meant that the models predicted cleaner effluent.
This could be disadvantageous during design of the systems using the models as
the probability of under designing would be high.

42

CHAPTER 6
6.0.0 CONCLUSIONS AND RECOMMENDATIONS
6.1.0 Conclusions
From the results and subsequent discussion, the following has been concluded.

The horizontal subsurface flow constructed wetlands have great

potential in secondary treatment of domestic wastewater because a
relatively significant removal of COD was observed during the study.
The k values obtained were also comparable to those found in
literature.

All the three models produced good predictions.

The Area-based first order plug flow model however, produced the
best effluent COD predictions.

The Area-based and Volume-based plug flow models were easy to

calibrate, unlike Tank-in series model where value of n had to be
found first.

6.2.0 Recommendations

Further research on the same is required in the tropics but this time on a
mature system. Annual evaluation of the tropics shows that they have a
higher energy turnover all season. The HSSF-CW should therefore be
very effective in the tropics as the purification process is temperature
dependent.

Further research should be done on the same but this time:

Using Higher Hydraulic Loading Rates than this study. (They
were between 120-8 m/yr for this study).

Using longer study duration.

Using different vegetation (Cyperus papyrus were used for this
study)
43


Covering many seasons (dry- wet seasons) where a wide range

of temperature would be used.

More studies should be done on the Tank-in series model. The 1st order
plug-flow models have a limitation as they ignore the mixing element in
the system assuming that what goes in first into the system is the first to
go out. Short-circuiting and dead zones are common phenomena in
constructed treatment wetlands causing non-ideal plug-flow conditions,
thus jeopardizing the use of the first-order model.

44

Bibliography
1. (EPA), U. S. (1993). Subsurface flow constructed wetland for wastewater
treatment:A technology assessment.
2. Bruggen, J. (20008). Wetlands for Water Quality Module. Delft, Netherlands.
3. Campbell, C., & M.H., O. (1999). Constructed wetlands in the sustainable
landscape. New York: J. Wiley.
4. Connors, K. A. (1991). Chemical kinetics ,the study of reaction rates in solutions.
Michigan: VCH .
5. D, R., D, G., PA, V., & N, D. P. (2002). Short- term behaviour of constructed
reed-beds: pilot plant experiments under different temperature conditions. Eigth
International Conference on Wetland systems for water pollution, (pp. 5-19).
Arusha, Tanzania.
6. D., Rousseau. (2005). Performance of constructed wetlands: Model based
evaluation and impact of operation and maintenance. Ghent, Belgium: Ghent
university.
7. Hammer, & Donald. (1992). Creating Freshwater Wetlands. Michigan: Ann
Arbor.
8. Hammer, D., & Bastian, R. (1989). Wetland Ecosystems: Natural water purifiers.
Michigan: Lewis publishers.
9. IWA. (2000). Activated Sludge Models.
10. Jones, William.W. (1995). Design features of constructed wetland for non-point
source treatment. Bloomington,Indiana: Indiana University.
11. Kadlec, H, R., & Knight, R. L. (1996). Treatment Wetlands. Michigan: Ann
Arbor.
12. M.F.Dahab, P., & Wenxin (Emy) Liu, P. (2009). Eavaluation o fFirst Order
Kinetics in Subsurface Flow Constructed Wetlands Treatment. Lincoln,NE:
University of Nebraska,Iowa Department of Natural Resourses.

45

13. Marble, & D, A. (1992). A Guide to Wetland Functional Design. Michigan: Ann
Arbor.
14. Mitsch, W., & Gosselink, J. (1993). Wetlands. Newyork.
15. Persson, J. (2001). How hydraulic conditions enhance water quality in Wetlands.
16. Risper, O. (2009). Evaluation of the performance of Horizontal Subsurface Flow
performing secondary treatment of domestic wastewater. Nairobi.
17. Schmidt, & D., L. (1998). The Engineering of chemical reactions. New York:
Oxford University .
18. Thunhorst, & A., G. (1993). Wetland Planting Guide for the Northeastern United
States-Plants for Wetland creation,Restoration and Enhancement. Maryland:
Environmental Concern Inc.

46

APPENDIX 1

Laboratory tests procedures:

Open reflux method.
Introduction: it is used as a measure of the oxygen equivalent of the organic matter
content of a s ample that is susceptible to oxidation by strong chemical oxidant. The test
is usually for monitoring and control of wastewater processes.
Procedure:
1. 50ml of the sample is placed in 500ml refluxing flask.
2. 1gm of mercuric sulphate, several glass beads are added. 5ml of
H2SO4 (concentrated sulphuric acid) was added slowly while mixing
to dissolve mercuric acid.
3. The sample is mixed while cooling to avoid possible losses of
volatile materials.
4. 25ml of 0.0471M k2cr2o7 solution is added and mixed thoroughly.
5. The samples are then transferred to the liebig condenser apparatus.
The remaining sulphuric acid reagent (70ml) is then added through
the open end of the condenser.
6. The samples are continuously stirred while adding the acid reagent.
7.

The reflux mixture is mixed thoroughly by applying heat to prevent

local heating of the flask bottom and a possible blow out of the flask
contents.

8. The open end of the condenser is covered with a foil to prevent

foreign materials from entering the refluxing mixture.
9. Refluxing is then carried out for 2hours.
10. The condenser is thereafter cooled and washed down using distilled
water.
11. The reflux is then disconnected and diluted to about twice its original
volume with distilled water.
12. This mixture is then cooled to room temperature (due to the
exothermic nature of the reaction). The excess k2cr2o7 is then titrated
using FAS using 0.1-0.15ml (2 to 3 drops) ferroin indicator.

47

13. The same volume of ferroin indicator is used for all titrations.
14. The end point of the reaction is taken as the first sharp change in
colour, from blue-green to reddish brown.
15. Similarly a blank of equal volume to sample, distilled water is
titrated against FAS.

Calculations:
COD = ((A-B) x m x 8000)
ml of sample
Where: A = ml , FAS used for blank water
B = ml, FAS used for sample water
m = molarity of the FAS

Dissolved Oxygen-titration method.

Introduction: It is required for supporting fish and aquatic life in water. When it reaches
the level below 2mg/l some aquatic life die.
Reagents:
1. Manganous sulphate solution. Mnso4.7h20.
2. Alkaline azide iodide solution.
3. Sulphuric acid solution.
4. Starch indicator.
5. Standard sodium thiosulphate solution, 0.025N

Procedure
1. 200ml of the sample is collected in a beaker.
2. 1ml of manganous sulphate is added to the sample while stirring. A
pipette was used during the experiment.

48

3. 1ml of alkaline azide iodide is then added to the sample whilst stirring.
Some precipitates form.
4. 1ml of concentrated sulphuric acid is then added to the mixture. The
sulphuric acid dissolves the precipitates.
5.

5 drops of starch indicator are added to the mixture. At that point the
mixture turns blue-black colour.

6. The mixture is then

solution.

titrated against standard sodium thiosulphate

Calculations:
The 200ml of solution taken for titration corresponds to 200ml of the
original sample, because 1ml of 0.025N sodium thiosulphate solution titrant
is equivalent to 0.2mg DO, each millilitre of sodium thiosulphate titrant is
equivalent to 1mg/ltr DO. When volume equal to 200ml of the original
sample is titrated therefore the DO concentration is ml of titrant used under
the above conditions.

49

APPENDIX 2:
CALIBRATION VOLUME-BASED, AREA-BASED PLUG FLOW MODELS,TANKIN-SERIES(TIS) MODEL RATE CONSTANTS

The step by step evaluation was shown:

ijkl  mnop jkqrjslpj:

k,  () i , (i)

tn  mnop jkqrjslpj:

k,


fS *
-  {| 8fS0

1

 {|  fS *
2{<S3{| 8z09P}0Q8:
8


S816OPQ<S3fS6S}68z9<O09:

()

S816OPQ<S3fS6S}68z9<O09:

2{<S3{ 8z09P}0Q8:

un)  ()  vt(o lpj:

k,
()

w

<SO<S38z016686S}68z9<O09:
2{<S3{ ! 8z09P}0Q8:
{! 




 '
&


{
-   NfS0

g
g

 {  fS *
N


fS *
)

.(ii)

y (iii)

)xu ,


S
*
-*

S + {! 8

S  *


- $S + { ! 8%


50

EVALUATION OF K20 AND VALUES

The step by step evaluation was shown:

$/i/u%   $u%.(iv)

Introducing natural log (Ln) on both sides of equation ( iv):

fS$/i/u%  fS + $u  %fS.(v)

Equation (v) was compared with $'  l

+ % where the Ln of k values from all the
models were plotted in the y-axis, T-20 valuesin the x-axis, Ln as the slope of the
graph and ) as the y-intercept.
The graphs are shown below:
Graph a: Ln kvvs Ln (T-20)

Ln kv vs Ln (T-20)
1.0000
0.0000

Ln kv

-1.0000

4
Series1

-2.0000

Linear (Series1)

-3.0000
-4.0000
-5.0000

y = 0.228x - 2.860
Ln (T-20)

Using equationN  0.228  2.860;

  .  . 
 .  . 

51

Graph b:Ln kA vs Ln (T-20)

Ln kA vs Ln (T-20)
5.0000

Ln kA

4.0000
3.0000
2.0000

Series1

1.0000

Linear (Series1)
y = 0.163x + 1.756

0.0000
0

Ln (T-20)
Using equationN  0.163 + 1.756;
  .  .

 .  . 

Graph c:Ln kT vs Ln (T-20)

Ln kT vs Ln(T-20)
2.0000

Ln kT

1.5000
1.0000
Series1
0.5000

Linear (Series1)
y = 0.062x + 0.987

0.0000
0

Ln (T-20)
Using equationN  0.062 + 0.987;
  .  .

 .  . 

52

APPENDIX 3:
PLATES:

Platea: Student collecting samples at the influent point of the HSSF-CW

53

Plate b:Student collecting samples at the effluent point of the HSSF-CW

Plate c: Student testing samples for COD in the Environmental laboratory

54